Surgical guidelines article_2

CHAPTER 1

INTRODUCTIONM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia

Randall Chesnut, M.D.Department of Neurological Surgery,University of WashingtonSchool of Medicine,Harborview Medical Center,Seattle, Washington

Jamshid Ghajar, M.D., Ph.D.Department of Neurological Surgery,Weil Cornell Medical College ofCornell University,New York, New York

David Gordon, M.D.Department of Neurological Surgery,Montefiore Medical Center,Bronx, New York

Roger Hartl, M.D.Department of Neurological Surgery,Weil Cornell Medical College ofCornell University,New York, New York

David W. Newell, M.D.Department of Neurological Surgery,Swedish Medical Center,Seattle, Washington

Franco Servadei, M.D.Department of Neurological Surgery,M. Bufalini Hospital,Cesena, Italy

Beverly C. Walters, M.D., M.Sc.Department of Neurological Surgery,New York UniversitySchool of Medicine,New York, New York

Jack E. Wilberger, M.D.Department of Neurological Surgery,Allegheny General Hospital,Pittsburgh, Pennsylvania

Reprints requests:Jamshid Ghajar, M.D., Ph.D.,Brain Trauma Foundation,523 East 72nd Street,New York, NY 10021.Email: [email protected]

Neurosurgery 58:S2-1-S2-3, 2006 DOI: 10.1227/01.NEU.0000210361.83548.D0 www.neurosurgery-online.com

Traumatic brain injury (TBI) affects up to2% of the population per year, and con-stitutes the major cause of death and

severe disability among young people. By far,the most important complication of TBI is thedevelopment of an intracranial hematoma,which complicates 25 to 45% of severe TBIcases, 3 to 12% of moderate TBI cases, andapproximately 1 in 500 patients with mild TBI(20). Without effective surgical management,an intracranial hematoma may transform anotherwise benign clinical course with the ex-pectation of recovery to a situation in whichdeath or permanent vegetative survival willoccur. Moreover, prolonged delay in the diag-nosis or evacuation of an intracranial hema-toma may produce a similar result.

As many as 100,000 patients per year mayrequire surgical management for a posttrau-matic intracranial hematoma in the UnitedStates alone. For these reasons, the impact thatneurosurgeons can have on the care of suchpatients is enormous, and perhaps, more than inany other area of emergency medicine, the ag-gressiveness and rapidity with which care isprovided for an intracranial hematoma will de-termine the outcome (8). Picard et al. (13) haveshown that craniotomy for evacuation of anacute epidural hematoma is one of the mostcost-effective of all surgical procedures. For thisparticular subgroup, which may represent up to5% of patients with severe and moderate TBI,the quality of outcome has been shown to varydramatically among different hospitals with dif-ferent levels of commitment to acute neuro-trauma care (2, 6, 11). It is for this reason, morethan any other, that neurosurgical consultationin the Emergency Room should be promptlyavailable and is a mandated requirement forLevel I certification of Trauma centers (1).

Although there is evidence that posttraumaticintracranial mass lesions have been removedsurgically up to 4000 years ago by the Egyptiansand Meso-Americans, it was not until a series ofpublications emerged in the late 1960s that itbecame generally accepted that excellent resultscould be achieved with craniotomy for removal

of extradural hematomas (9). For acute subduralhematomas and intraparenchymal lesions, suchas contusions and traumatic intracerebral hema-tomas, the outcome has historically been muchworse, because up to 60% of patients with acutesubdural hematomas will die or remain severelydisabled (10).

During the early 1970s, a series of publica-tions from the Medical College of Virginia dem-onstrated that wide decompressive craniotomywith duraplasty was one of the most effectiveforms of therapy for raised intracranial pressurein patients with mass lesions (4). Subsequently,most neurosurgical centers with an interest inTBI have also applied the same craniotomy tech-nique to patients with intraparenchymal contu-sions, with improvements in outcome. How-ever, there is also widespread dissent andcontroversy regarding the surgical managementof intraparenchymal lesions, with some neuro-surgeons maintaining that aggressive surgicalintervention, although able to preserve life, willresult in a very poor quality of life for survivors(7, 15).

With our increasing understanding of thepathomechanisms in severe and moderate TBIhave come changes in our approach to man-agement of patients with posttraumatic intra-cranial mass lesions. For example, it is nowwell accepted that most intraparenchymalmass lesions (contusions and intracerebral he-matomas) will enlarge with time, necessitatingserial computed tomographic scanning, andusually intracranial pressure monitoring dur-ing the first few days (12, 17). Similarly, thepropensity of patients with posttraumatic co-agulation disorders to develop intraparenchy-mal bleeding that is more severe is now wellaccepted, and has led to management of coag-ulation disorders in the head injured popula-tion that is much more aggressive (18).

In turn, these practices led to an increase inthe performance of craniotomy, both for evac-uation of intraparenchymal mass lesions andas a decompressive measure. Recently, severalpublications have shown that, within the con-text of modern aggressive neuro-intensive

NEUROSURGERY VOLUME 58 | NUMBER 3 | MARCH 2006 SUPPLEMENT | S2-1

care therapy, decompressive craniotomy is an effective meansof controlling raised intracranial pressure after severe TBI,especially in those patients with intraparenchymal lesions (14,19). Many neurosurgeons, however, are reluctant to imple-ment such aggressive surgical techniques in patients withraised intracranial pressure after severe TBI, claiming thatimproved quality of life has never been conclusively demon-strated.

It is, therefore, the overall aim of these Guidelines is topresent rigorous literature-based recommendations for thesurgical management of patients with posttraumatic intracra-nial mass lesions. We have chosen to focus on those acutemass lesions that develop within 10 days of injury and, thus,we have chosen not to cover chronic subdural hematoma,subdural hygroma, and posttraumatic hydrocephalus, whichusually are delayed. Similarly, we have chosen to focus onclosed TBI in general because a comprehensive set of manage-ment guidelines for patients with penetrating TBI has alreadybeen formulated (3).

Compared with the Guidelines for The Management of SevereTraumatic Brain Injury (5), the literature regarding surgical man-agement after TBI suffers from extensive limitations, in bothquality and scope. Most notably, although our group reviewedmore than 700 manuscripts for the preparations of these Guide-lines, there are no controlled clinical trials in the literature tosupport different forms of surgical management, or to supportsurgical versus conservative therapy. Consistent with these lim-itations, we have been unable to formulate recommendations atthe standard level requiring Class I evidence.

As in all other areas of �evidence-based medicine,� theseGuidelines have been formulated strictly in accordance withexternally imposed constraints. Only clinical human-basedliterature has been reviewed. Only literature from 1975through 2001 has been reviewed. Mainly literature in English,with far fewer articles in other languages, was reviewed. Forthese reasons, the reader must clearly understand that thescope and level of magnitude of the recommendations madehere are distilled from the available literature and interpretedaccording to the rules of �evidence-based medicine� (16).

An important aspect of this document is, therefore, to alsoformulate critical questions that need to be resolved by futureclinical trials or prospective cohort studies, to determine themost effective forms of therapy for the future. As with theother guidelines in severe TBI, therefore, this is a document inevolution, and frequent revisions will be made to keep upwith the evolving state of knowledge in this area.

These Guidelines have been organized on the basis of the tra-ditional literature-based classification of posttraumatic mass le-sions: namely, epidural hematoma, acute subdural hematoma,intraparenchymal lesions (contusion and intracerebral hema-toma), acute posterior fossa mass lesions, and depressed frac-tures of the cranium. We recognize, however, that, for mostpatients with severe TBI, and for some patients with moderateTBI, more than one of these acute posttraumatic mass lesionsmay coexist at the same time. For example, the majority ofpatients with acute subdural hematomas will also demonstrate

concomitant intraparenchymal contusions on their computed to-mographic scan. In some patients, there may be multiple sites inwhich intraparenchymal mass lesions occur—e.g., bifrontal con-tusions, bitemporal contusions, or temporal and frontal lesions.For high-volume lesions (�50 cm3), management decisions areeasier, and generally are in favor of surgery. Low-volume lesions(�25 cm3) are usually not operated on, however, for lesionsbetween high and low volumes, the decisions may be very dif-ficult and associated factors, e.g., shift, cisterns, and GlasgowComa Scale, become especially important.

Within the literature, the terms surgical decompression,decompressive craniectomy, evacuation, and internal decom-pression are often used interchangeably. This aspect is clari-fied as much as possible in the individual sections.

We describe methods for posttraumatic mass volume mea-surement in Appendix 1 and simple methods and definitions ofmidline shift, subarachnoid hemorrhage, and status of basalcisterns in Appendix 2.

REFERENCES

1. ACS-COT: American College of Surgeons-Committee on Trauma: Resourcesfor Optimal Care of the Injured Patient. Chicago, American College of Sur-geons, 1999.

2. Alberico A, Ward J, Choi S, Marmarou A, Young H: Outcome after severehead injury. Relationship to mass lesions, diffuse injury, and ICP course inpediatric and adult patients. J Neurosurg 67:648–656, 1987.

3. Anonymous: Management and prognosis of penetrating brain injury.J Trauma 51[Suppl]: S1–S49, 2001.

4. Becker D, Miller J, Ward J, Greenberg R, Young H, Sakalas R: The outcomefrom severe head injury with early diagnosis and intensive management.J Neurosurg 47:491–502, 1977.

5. Brain Trauma Foundation, American Association of Neurological Surgeons,Joint Section on Neurotrauma and Critical Care: Guidelines for the manage-ment of severe traumatic brain injury. J Neurotrauma 17:457–554, 2000.

6. Bricolo A, Pasut L: Extradural hematoma: Toward zero mortality. A pro-spective study. Neurosurgery 14:8–12, 1984.

7. Clark K, Nash TM, Hutchison GC: The failure of circumferential craniotomyin acute traumatic cerebral swelling. J Neurosurg 29:367–371, 1968.

8. Consensus conference: Rehabilitation of persons with traumatic brain in-jury. NIH Consensus Development Panel on Rehabilitation of Persons withTraumatic Brain Injury. JAMA 282:974–983, 1999.

9. Jamieson KG, Yelland JD: Extradural hematoma: Report of 167 cases.J Neurosurg 29:13–23, 1968.

10. Jamieson KG, Yelland JD: Surgically treated subdural hematomas.J Neurosurg 37:137–149, 1972.

11. Klauber MR, Marshall LF, Luerssen TG, Frankowski R, Tabaddor K,Eisenberg HM: Determinants of head injury mortality: Importance of thelow risk patient. Neurosurgery 24:31–36, 1989.

12. Lobato RD, Gomez PA, Alday R, Rivas JJ, Dominguez J, Cabrera A,Turanzas FS, Benitez A, Rivero B: Sequential computerized tomographychanges and related final outcome in severe head injury patients. ActaNeurochir (Wien) 139:385–391, 1997.

13. Picard J, Bailey S, Sanderson H, Reese M, Garfeld JS: Steps towards costbenefit analysis of regional neurosurgical care. BMJ 301:629–635, 1990.

14. Polin R, Shaffrey M, Bogaev C, Tisdale N, Germanson T, Bocchicchio B, JaneJ: Decompressive bifrontal craniectomy in the treatment of severe refractoryposttraumatic cerebral edema. Neurosurgery 41:84–92, 1997.

15. Ransohoff J, Benjamin MV, Gage EL Jr, Epstein F: Hemicraniectomy in themanagement of acute subdural hematoma. J Neurosurg 34:70–76, 1971.

16. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS: Evidencebased medicine. What it is and what it isn’t. BMJ 312:71–72, 1996.

SURGICAL MANAGEMENT OF TBI AUTHOR GROUP

S2-2 | VOLUME 58 | NUMBER 3 | MARCH 2006 SUPPLEMENT www.neurosurgery-online.com

17. Soloniuk D, Pitts LH, Lovely M, Bartkowski H: Traumatic intracerebralhematomas: Timing of appearance and indications for operative removal.J Trauma 26:787–794, 1986.

18. Stein SC, Young GS, Talucci RC, Greenbaum BH, Ross SE: Delayed braininjury after head trauma: Significance of coagulopathy. Neurosurgery 30:160–165, 1992.

19. Taylor A, Butt W, Rosenfeld J, Shann F, Ditchfield M, Lewis E, Klug G,Wallace D, Henning R, Tibballs J: A randomized trial of very early decom-pressive craniectomy in children with traumatic brain injury and sustainedintracranial hypertension. Childs Nerv Syst 17:154–162, 2001.

20. Thurman D, Guerrero J: Trends in hospitalization associated with traumaticbrain injury. JAMA 282:954–957, 1999.

INTRODUCTION


CHAPTER 2

METHODOLOGYM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia










Neurosurgery 58:S2-4-S2-6, 2006 DOI: 10.1227/01.NEU.0000210362.83548.0B www.neurosurgery-online.com

The concept of formulating a set of rec-ommendations for patient managementhas its basis in the advent of scientific

investigation in clinical medicine. Along withthis came the realization that deriving recom-mendations from our own experience as clini-cians, although tempting, has limitations re-lated to the fact that humans remember thebest or the worst of their experiences, andcannot easily objectify their results. What clin-ical experience does do for us is to help withthe formulation of hypotheses that can betested through human clinical trials experi-mentation. Rather than seeing clinical experi-ence and clinical trials as dichotomous, polarsources of information regarding outcomes ofpatient care, it is probably more useful tothink of the gamut of clinical testing along acontinuum. On one end, we have the more-subjective impressions of care delivery thatare attractive because they are grounded inpersonal experience in the real world of caringfor patients. This is expressed well by one ofthe fathers of clinical epidemiology, AlvanFeinstein. Dr. Feinstein, in his seminal work,Clinical Judgment (2), writes: “. . .any decentdoctor reflects on alternatives, is aware of un-certainties, modifies judgments on the basis ofaccumulated evidence, balances risks of vari-ous kinds, considers the potential conse-quences of his or her diagnoses or treatments,and synthesizes all of this in making a rea-soned decision that he or she decrees right forthe patient.” He further goes on to underscorethe value of clinical experience when hewrites, “In caring for patients, clinicians con-stantly perform experiments. During a singleweek of active practice, a busy clinician con-ducts more experiments than most of his lab-oratory colleagues do in a year.”

This philosophy underscores and supportsthe notion that we can make recommenda-tions from the wealth of clinical experienceprovided by direct, day-to-day clinical care.However, a close look at the process of trueexperimentation yields a long list of potentialerrors that can be made in the gathering and

interpretation of data, even in highly con-trolled circumstances. Sources of error havebeen classified into two broad categories: sys-tematic error (bias) and random error. In gen-eral, the former source of error is controlled bycareful study design, and the latter is dealtwith by randomization. Without these protec-tors against error, the chances of making mis-takes in deriving recommendations from prac-tice are great. For this reason, the notion ofcategorizing recommendations into groups as-sociated with various strengths of clinicalstudies on the basis of scientific rigor wasintroduced in the last decade-and-a-half of the20th century.

In 1990, the Institute of Medicine published alandmark work entitled Clinical Practice Guide-lines: Directions for a New Program (1). The insti-gating factors for this publication were the ob-servation of wide variability of practicethroughout American healthcare and the devel-opment of mechanisms for establishing qualityassurance and review within Medicare, alongwith the increasing use of technology and con-comitant costs of care delivery. In this docu-ment, the authors made recommendations re-garding attributes defining the validity ofrecommendations. These include the following:1. The available scientific literature should besearched using appropriate and comprehensivesearch terminology. 2. A thorough review of thescientific literature should precede guideline de-velopment. 3. The evidence should be evaluatedand weighted, reflecting the scientific validity ofthe methodology used to generate the evidence.4. There should be a link between the availableevidence and the recommendations, with thestrength of the evidence being reflected in thestrength of the recommendations, reflecting sci-entific certainty (or lack thereof). 5. Empiricalevidence should take precedence over expertjudgment in the development of guidelines. 6.Expert judgment should be used to evaluate thequality of the literature and to formulate guide-lines when the evidence is weak or nonexistent.7. Guideline development should be a multidis-


ciplinary process, involving key groups affected by the recom-mendations.

SEARCH OF THE LITERATURE

The literature is generally searched using the National Li-brary of Medicine computerized database, either through in-dividual search engines in academic medical centers orthrough the internet. This, by definition, will limit the scope ofthe search to 1966 and forward. In addition, if modern neuro-imaging techniques are pertinent, as in this document, thesearch should be limited to the years since the modality wasavailable. Therefore, our searches were limited to the period of1975 to the present. Because guidelines are applicable to pa-tient care, the literature is limited to human studies in the areaof traumatic brain injury and surgical management, alongwith the imaging characteristics in patients with surgicallyamenable lesions. In addition, further limitations in the searchwere imposed by the fact that most of the literature is inEnglish, and limited ability existed for reading the usefularticles that appear in different languages. Therefore, mostsearches were limited to the English language. Appropriatesearch terms were chosen, as demonstrated in each of theindividual Guideline sections.

EVALUATION AND WEIGHTING OF THELITERATURE

The journal articles found have been carefully read andevaluated, including an assessment of the methodology usedin the studies. This not only includes the establishment of theclinical question addressed (e.g., therapeutic effectiveness, di-agnostic tests, prognostic studies, etc.) and type of study (ran-domized controlled trial, case-control study, case series, etc.),but also the quality of the study with respect to potentialerrors in design, execution, or conclusions reached. Therefore,studies that might, on the surface, represent evidence support-ing one level of recommendation, may instead be flawedenough to be devalued to support a recommendation of lesserstrength. The quality of the literature was evaluated in thisway according to well-established criteria (3). All articles werecross-reviewed and disagreements were resolved by consen-sus.

LINK BETWEEN EVIDENCE ANDGUIDELINES

The general concept of relating strength of recommenda-tions to strength of evidence reflecting varying degrees ofclinical certainty was formalized into a scheme that has beenfollowed by medical societies, including organized neurosur-gery, from the inception of the Guideline development process.Despite problems with the strict application of this paradigm(some of which are displayed and discussed in this supple-ment), the scheme has the benefit of using scientific evidencerather than expert opinion for the substrate of the recommen-dations, although expert opinion is used to formulate therecommendations themselves, as well as to make judgmentsregarding the quality of the evidence. The evidence-basedscheme used in these and all Guidelines regarding therapeuticeffectiveness endorsed by the American Association of Neu-rological Surgeons and the Congress of Neurological Surgeonsbegins with classification of the literature into three categoriesof evidence, as outlined next and in Table 1.

The classification of evidence into these three categoriesleads to the formulation of recommendations called Standards,Guidelines, and Options. Class I evidence is used to supporttreatment recommendations of the strongest type, practiceStandards, reflecting a high degree of clinical certainty. Class IIevidence is used to support Guidelines, reflecting a moderatedegree of clinical certainty. Class III evidence supports practiceOptions reflecting unclear clinical certainty. This terminologywas developed to indicate, in normal vocabulary, the strengthof the recommendations on the basis of strong to weak med-ical evidence. In neurosurgery, this scheme has been used toformulate Guidelines, rather than a scheme that uses letters ornumbers that have no grounding in language and are, there-fore, more easily misinterpreted. The link between scientificevidence and recommendations has been highlighted in theseGuidelines by presenting those studies in the scientific founda-tion that support the stated recommendation in boldface type.

Recommendations are not necessarily weak in cases inwhich the evidence is weak. Examples of this appear in manyof the Guidelines developed in neurosurgery to date, and au-thors and readers alike feel frustrated at the impotency of therecommendations when the evidence is weak, especially whenthe logic of the recommendation and all of the evidence sup-porting it, however weak, support the recommendation. Oneexample of this, among many that can be found in this sup-

TABLE 1. Classification of Evidence on Therapeutic Effectiveness

Class I Evidence from one or more well-designed, randomized, controlled clinical trials, including overviews of such trialsClass II Evidence from one or more well-designed comparative clinical studies, such as nonrandomized cohort studies, case-control

studies, and other comparable studiesClass III Evidence from case series, comparative studies with historical controls, case reports, and expert opinion

METHODOLOGY


plement, is in the recommendation regarding timing of evac-uation of epidural hematomas. According to the paradigmembraced and used in this set of Guidelines, case series indi-cating that patients who have a Glasgow Coma Scale score of8 or below with evidence of a “blown pupil” and who areoperated on early achieve better outcomes can only support apractice “Option.” However, no competent neurosurgeonwould allow a patient in this clinical scenario to be neglectedwhen the need for surgical relief of brain compression is soclear. It is fairly certain to say that there will never be arandomized controlled trial for this circumstance, and, thus,never a practice “Standard.” However, a patient databasecould be used to generate a case-control study, thus, yieldinga higher recommendation. However, there is no evidence thatwaiting to operate on such a patient is beneficial, and, there-fore, an “Option” to delay surgical evacuation will also prob-ably never be promulgated. If, indeed, such a recommenda-tion were put forward, it would never be accepted by theprofession, and rightfully so.

EXPERT JUDGMENT AND EMPIRICALEVIDENCE

There are two ways in which expert judgment comes intoGuideline development. The most common use of expert opin-ion is in developing recommendations for practice. This hasbeen a usual method in the past (as well as the present, in theform of textbook chapters), but has more recently given way tomore formalized approaches embraced by evidence-basedmedicine methodology, such as that used in this supplement.However, even in evidence-based methodology, expert opin-ion is used to evaluate the literature as well as to frame theconcepts and wording of the recommendations. In addition, ifthe evidence is weak and conflicting, expert opinion is used toderive recommendations. This use is unavoidable, but theexpert opinion is guided by the evidence published in theliterature, rather than from personal experience alone.

OPTION LEVEL RECOMMENDATION

All of the recommendations in the Surgical Management ofTraumatic Brain Injury are at the option level, supported only

by Class III scientific evidence. Unfortunately, option canmean choice and, as discussed above, it would be unethicalnot to operate on surgical mass lesions in salvageable patients.Currently, all neurological guidelines approved by the Amer-ican Association of Neurological Surgeons and the Congressof Neurological Surgeons use option recommendations forClass III evidence. In the future, this may change because thesesurgical guidelines put maximized strain on using such ter-minology.

Given that the recommendations in these surgical guide-lines are at the option level (Class III evidence) we state therecommendation at the beginning of the chapter on each topic,and refer the reader to this methodology section for an expla-nation of the level of evidence.

MULTIDISCIPLINARY PROCESS

In all Guidelines published under the auspices of the BrainTrauma Foundation and the American Association of Neuro-logical Surgeons, other professional organizations were in-volved in either developing the Guidelines or reviewed andapproved them. In these Surgical Management of TraumaticBrain Injury Guidelines, however, only neurosurgeons wereinvolved. These neurosurgeons represent a wide range oforganizations. There were representatives from the AmericanAssociation of Neurological Surgeons, the Congress of Neu-rological Surgeons, the European Brain Injury Consortium, theAmerican College of Surgeons (Committee of Trauma) and theWorld Federation of Neurological Surgeons (Neurotraumasection) involved in the development of these Surgical Man-agement of Traumatic Brain Injury Guidelines.

REFERENCES

1. Committee to Advise the Public Health Service on Clinical Practice Guide-lines (Institute of Medicine): Clinical Practice Guidelines: Directions for a NewProgram. Washington, D.C., National Academy Press, 1990.

2. Feinstein A: Clinical Judgement. Baltimore, Williams & Wilkins, 1967.3. Walters B: Clinical Practice Parameter Development, in Bean J (ed): Neuro-

surgery in Transition. Baltimore, Williams & Wilkins, 1998, pp 99–111.



CHAPTER 3

SURGICAL MANAGEMENT OF ACUTE

EPIDURAL HEMATOMASM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia










RECOMMENDATIONS(see Methodology)

Indications for Surgery● An epidural hematoma (EDH) greater than 30 cm3 should be surgically evacuated

regardless of the patient’s Glasgow Coma Scale (GCS) score.● An EDH less than 30 cm3 and with less than a 15-mm thickness and with less than

a 5-mm midline shift (MLS) in patients with a GCS score greater than 8 without focaldeficit can be managed nonoperatively with serial computed tomographic (CT)scanning and close neurological observation in a neurosurgical center.

Timing● It is strongly recommended that patients with an acute EDH in coma (GCS score �

9) with anisocoria undergo surgical evacuation as soon as possible.Methods● There are insufficient data to support one surgical treatment method. However,

craniotomy provides a more complete evacuation of the hematoma.

KEY WORDS: Coma, Computed tomographic parameters, Craniotomy, Epidural, Head injury, Hematoma,Surgical technique, Timing of surgery, Traumatic brain injury

Neurosurgery 58:S2-7-S2-15, 2006 DOI: 10.1227/01.NEU.0000210363.91172.A8 www.neurosurgery-online.com

OVERVIEW

Incidence

Since the introduction of CT scanning as theimaging study of choice to detect intracraniallesions after trauma, the incidence of surgicaland nonsurgical EDH among traumatic braininjury (TBI) patients has been reported to be inthe range of 2.7 to 4% (8, 11, 25, 41). Amongpatients in coma, up to 9% harbored an EDHrequiring craniotomy (10, 35). The peak inci-dence of EDH is in the second decade, and themean age of patients with EDH is between 20and 30 years of age ( 3, 8, 9, 13, 16–18, 20, 22,26, 29, 32, 37, 39). EDH are a rare entity inpatients older than 50 to 60 years of age. Inpediatric patients, the mean age of patientsharboring EDH is between 6 and 10 years (21,34), and EDH is less frequent in very youngchildren and neonates (27, 30).

Pathogenesis

Traffic-related accidents, falls, and assaultsaccount for 53% (range, 30–73%), 30% (range,

7–52%), and 8% (range, 1–19%), respectively,of all EDH (3, 8, 20, 22, 26, 27, 36, 40). Inpediatric patients, falls are the leading causeof EDH in 49% of cases (range, 25–59%) andtraffic-related accidents are responsible for34% (range, 25–41%) of all EDH (21, 25–27, 30,34). EDH can result from injury to the middlemeningeal artery, the middle meningeal vein,the diploic veins, or the venous sinuses. His-torically, bleeding from the middle meningealartery has been considered the main sourcefor EDH. In a recent report on EDH in 102pediatric patients and 387 adults, arterialbleeding was identified as the source of theEDH in 36% of the adults and only in 18% ofthe children (27). In 31% of the pediatric pa-tients, a bleeding source could not be identi-fied and venous bleeding accounted for ap-proximately 32% of EDH in both age groups.

Location

In surgical series, EDH are more frequentlylocated in the temporoparietal and temporal re-gions as compared with other locations (3, 6, 25,


27, 29, 32). In 2 to 5% of patients, bilateral EDH are found (11, 18,40), and there seems to be a slight predominance of right-sidedEDH over left-sided lesions (6, 40).

Clinical Presentation

In patients with EDH, 22 to 56% are comatose on admissionor immediately before surgery (3, 17, 20, 22, 25, 32). Theclassically described “lucid interval,” i.e., a patient who isinitially unconscious, then wakes up and secondarily deterio-rates, was observed in a total of 456 of 963 patients (47%)undergoing surgery for EDH in seven studies (3, 8, 18, 22, 28,31, 39). Between 12 and 42% of patients remained consciousthroughout the time between trauma and surgery (3, 8, 17, 22).Pupillary abnormalities are observed in between 18 and 44%of patients, and up to 27% (3–27%) of patients are neurologi-cally intact. Other presenting symptoms include focal deficits,such as hemiparesis, decerebration, and seizures. Early sei-zures are noted in 8% of pediatric patients presenting withEDH (21).

Mortality

The mortality in patients in all age groups and GCS scoresundergoing surgery for evacuation of EDH is approximately10% (range, 7–12.5%) (7, 8, 14, 17, 18, 20, 22, 28, 31, 32).Mortality in comparable pediatric case series is approximately5% (25, 30).

Determinants of Outcome in Patients UndergoingSurgical Evacuation of an EDH

GCS, age, pupillary abnormalities, associated intracraniallesions, time between neurological deterioration and surgery,and intracranial pressure (ICP) have been identified as impor-tant factors determining outcome from EDH.

Age and GCS

The influence of age on outcome in the subgroup of patientswith EDH is not as pronounced as it is in TBI patients overall.Three studies using multiple regression analysis found thatGCS was a better predictor of outcome than age in patientsundergoing surgery for EDH (20, 22, 38). In a retrospectiveanalysis of 98 patients of all age groups with EDH undergoingcraniotomy, van den Brink et al. (39) investigated determi-nants of outcome at 6 months. They identified GCS, age, andthe CT diagnosis of subarachnoid hemorrhage as significantfactors correlated with outcome, using multivariate analysis.Admission GCS or GCS before surgery is the single mostimportant predictor of outcome in patients with EDH under-going surgery (3, 10, 20, 22, 24, 38, 39). In three studies usingmultivariate analysis in a total of 284 patients, the admissionGCS score was identified as the most significant factor deter-mining outcome at 6 months (20, 38, 39). In one study with 200patients undergoing craniotomy, admission and preoperativeGCS both correlated with functional outcome at 1 year (22).Gennarelli et al. (10) analyzed the relationship between type oflesion, GCS score on admission, and 3-months outcome in

1107 comatose patients with TBI. The highest mortality wasfound in patients with a subdural hemorrhage and a GCSbetween 3 and 5 (74%). Patients with an EDH and a GCS of 3to 5 had a mortality of 36%, and patients with an EDH and aGCS of 6 to 8 had a mortality of only 9%.

Pupils

Pupillary abnormalities, such as pupillary asymmetry orfixed and dilated pupils occur in approximately 20 to 30% ofpatients with EDH undergoing surgery (3, 16, 20, 40) and in62% of patients who are comatose on admission (32). Onestudy showed that ipsilateral mydriasis was not associatedwith adverse outcome and was reversible when operated onwithin 70 minutes after pupillary dilation (6). Bilateral mydri-asis, however, is associated with a high mortality (3, 6, 8, 24,32, 39). Mydriasis contralateral to the hematoma is also asso-ciated with high mortality (24, 32). Van den Brink et al. (39), ina multivariate model evaluating the relative prognostic valueof predictive parameters, found that, in patients in all agegroups and GCS scores, pupillary abnormalities were signifi-cantly related to unfavorable outcome. Adverse outcome wasobserved in 30% of normal pupillary responses, in 35% ofunilateral fixed pupils, and in 50% of bilateral fixed pupils.Bricolo and Pasut (3) achieved a good outcome in 100% ofpatients who presented with anisocoria and in 90% of patientswho presented with anisocoria and hemiparesis. The onlypatient with bilateral mydriasis in their case series died.

Associated Lesions

Associated intracranial lesions are found in between 30 and50% of adult patients with surgically evacuated EDH (3, 8, 13,16, 20, 22, 23, 27, 29, 31, 32, 35). These are predominantlycontusions and intracerebral hemorrhage followed by sub-dural hematoma (SDH) and diffuse brain swelling (8, 16, 29,32, 35). The incidence of associated lesions is less in the pedi-atric age group (25, 27, 30). SDH and/or parenchymal lesionsin association with EDH lower the chance of a good outcome.In two studies with a total of 315 patients operated on forevacuation of an EDH, the frequency of associated intracraniallesions was 33% (20, 22). In both studies, a significant relation-ship was found between the presence of other lesions inaddition to the EDH and an adverse outcome. Lee et al. (22)identified associated brain lesions as one of four independentpredictors of unfavorable outcome after surgery for EDH andthis has been confirmed by several others (8, 13, 16, 23, 32).Cranial fractures are present in between 70 and 95% of cases(15, 17, 20, 25, 30, 37). The impact of fractures on outcome iscontroversial. Kuday et al. (20) observed a significant relation-ship between cranial factures and adverse outcome in 115patients undergoing surgery for EDH. Lee et al. (22) did notsee this relationship in a series of 200 patients managed sim-ilarly, and Rivas et al. (32) actually reported a significantlylower mortality rate in patients with cranial fractures. Signif-icant extracranial injury is present in 7 to 23% of patientsoperated on for an EDH (8, 13, 17, 24, 27). Lobato et al. (24)



found extracranial injury in 20% of their patients, and themortality rate in this subgroup was lower than the overallmortality (7.6% versus 28%). No data were found on theassociation of hypotension and outcome in patients with anacute EDH.

ICP

There is only one study available in which postoperativeICP and its relationship to outcome 6 months after trauma wasstudied. Lobato et al. (24) monitored ICP in 54 (83%) of 64comatose patients after removal of an EDH. Elevated ICP (�15mm Hg) was found in 67% of cases, and ICP greater than 35mm Hg was significantly associated with a higher mortality.

PROCESS

A MEDLINE computer search using the following key-words for the years 1975 to 2001 was performed: “traumaticbrain injury” or “head injury” and “epidural” or “extradural”and “hematoma” or “hematoma” or “hemorrhage.” Thesearch was narrowed by including the keywords “surgicaltreatment” or “surgery” or “operation” or “craniotomy” or“craniectomy” or “craniostomy” or “burr holes” and exclud-ing “spinal.” These searches combined yielded 168 articles.The reference lists of these publications were reviewed and anadditional 22 articles were selected for analysis. Case reports,publications in books, and publications regarding penetratingbrain injuries on spinal EDH and on exploratory burr holeswithout a preoperative CT scan were not included. Articleswere excluded if the diagnosis of EDH was not based on CTscanning, or if subgroups of patients who did not undergo CTscanning were not clearly identified. Publications with fewerthan 10 patients or publications that did not include informa-tion on outcome were excluded. Of these 190 articles, 18 wereselected for analysis.

SCIENTIFIC FOUNDATION

Indication for Surgery

The decision to operate on an acute EDH is based on thepatient’s GCS score, pupillary exam, comorbidities, CT find-ings, age, and, in delayed decisions, the patient’s ICP. Neuro-logical deterioration over time is also an important factorinfluencing the decision to operate. Trauma patients present-ing to the emergency room with altered mental status, pupil-lary asymmetry, and abnormal flexion or extension are at highrisk for either an SDH and/or EDH compressing the brain andbrainstem.

CT Characteristics and Outcome

CT is the imaging study of choice for the diagnosis of anEDH. CT scanning is recommended in patients at risk forharboring an acute EDH. It allows not only diagnosis of theprimary lesion but also identification of additional featuresthat affect outcome, such as MLS, traumatic subarachnoid

hemorrhage, obliteration of the basal cisterns, thickness of theblood clot, and hematoma volume.

In a series of 200 patients who were treated surgically forEDH, Lee et al. (22) found that a hematoma volume greaterthan 50 cm3 was significantly related to higher mortality andunfavorable functional outcome. Unfavorable functional re-covery was observed in 6.2% of patients with a hematomavolume less than 50 cm3, and in 24% of patients with ahematoma volume greater than 50 cm3. Mixed density of theblood clot, indicating acute bleeding, was observed in 32% oftheir patients and correlated with unfavorable outcome butnot with mortality. Patients with an MLS greater than 10 mmshowed a higher mortality and more unfavorable outcomewhen compared with those with less displacement. Partial ortotal obliteration of the basal cisterns was observed in 59% oftheir patients and correlated with both mortality and func-tional outcome. Multivariate analysis identified only hema-toma volume as an independent predictor of unfavorableoutcome.

In contrast, in 98 patients with EDH who underwent sur-gery, van den Brink et al. (39) found that the status of the basalcisterns, MLS, and hematoma volume were not related tooutcome. The authors only identified the presence of trau-matic subarachnoid hemorrhage to be significantly associatedwith unfavorable outcome. Patients with favorable outcomehad a hematoma volume of 56 � 30 cm3 and, with unfavorableoutcome, the hematoma volume was 77 � 63 cm3, but thisdifference was not significant.

Rivas et al. (32) found that hematoma volume and severityof MLS were related to preoperative coma in patients withEDH. In comatose patients, a hematoma volume greater than150 cm3 and an MLS greater than 12 mm were associated withincreased mortality. Mixed-density blood clots were observedin 62% of their patients and were related to poor outcome.Location of the lesion did not influence outcome. Seelig et al.(35) did not find a relationship between location of blood clot,MLS, and outcome in 51 comatose patients undergoing sur-gery for EDH.

In summary, most authors could not detect a relationshipbetween blood clot location and outcome. However, it is likelythat hematoma volume, MLS, mixed density of the blood clot,and traumatic subarachnoid hemorrhage are related to out-come, but more studies are needed to clarify this issue.

Surgery and Nonoperative Treatment

Prospective, randomized trials comparing surgical treat-ment with nonoperative management are not available. Somestudies compared patients who were treated either surgicallyor nonoperatively, and used logistic regression analysis andmultivariate analysis models to determine factors that wereassociated with either treatment (36, 37). Some investigatorslooked at patient series that were initially all treated nonop-eratively and analyzed the factors associated with subsequent,delayed surgery (2, 7, 19, 37). There are no studies on nonop-erative treatment of comatose patients with EDH.

ACUTE EPIDURAL HEMATOMAS


What are the factors leading to surgery? The followingstudies compared characteristics between patients who weretreated with either surgery or were managed nonoperatively.The value of such analyses is doubtful because it merelydocuments the criteria used to select patients for surgery.

Servadei et al. (36) conducted a prospective study including158 consecutive patients with GCS 14 and 15 with EDH whowere admitted to three neurosurgical units. A treatment pro-tocol was not defined for these hospitals. One hundred-sixteenpatients underwent surgery and 42 patients were managednonoperatively. Ninety-three percent of patients with an MLSgreater than 5 mm, and 91% of patients with a hematomathickness greater than 15 mm underwent surgery. A logisticregression analysis identified hematoma thickness and MLS asthe factors associated with surgery. Location and the presenceof associated lesions did not reach significance. Outcome wasgood in all patients. Similar results were obtained in 33 pedi-atric patients, 20 of whom were treated surgically (1). Bothgroups did not differ in terms of age, GCS, and outcome.Multivariate logistic regression analysis revealed that MLS,hematoma thickness, and volume, as well as temporal locationof the blood clot were related to surgery. Hematoma volumeand MLS were 41 cm3 and 8 cm3, and 4 mm and 0.5 mm, forthe surgical and nonsurgical groups, respectively.

A review of 30 patients who were treated with craniotomyand 18 patients treated nonoperatively revealed that patientsmanaged with surgery had lower GCS scores, were morelikely to present with pupillary abnormalities and hemipare-sis, and had larger blood clots and more MLS (12). Temporallocation of the hematoma and the presence and location of afracture were not related to surgery.

Factors Determining Delayed Surgery

Bezircioglu et al. (2) conducted a prospective study on thenonoperative management of 80 patients with EDH and GCSscores between 9 and 15. Patients with a GCS score greaterthan 8, an EDH volume less than 30 ml, a hematoma thicknessless than 2 cm, and without neurological deficit were treatednonoperatively. Five patients deteriorated and underwent cra-niotomy. One of these patients died, the others had goodoutcomes. The only factor significantly associated with de-layed surgery was a temporal location of the hematoma,which was observed in all five surgical patients but only in24% of the 75 patients treated without operation.

In a study of 74 patients with initially asymptomatic EDHmanaged nonoperatively, 14 required delayed surgery be-cause of neurological deterioration or increase in the size ofthe hematoma (5). The authors found that a hematoma volumegreater than 30 cm3, a hematoma thickness greater than 15mm, and an MLS greater than 5 mm were significantly morefrequent in patients requiring surgery. A hematoma volumegreater than 30 cm3, an EDH thickness greater than 15 mm,and an MLS greater than 5 mm were observed in 5%, 27%, and28% of patients managed without surgery, and in 57%, 71%,and 79% of patients who had surgery, respectively. The au-

thors used the “ellipsoid” or “ABC/2” method to estimate thevolume of EDHs (see Appendix I). Hematoma location, bonefractures, and time-to-initial CT scan were not related to out-come. In a small study on 22 patients, 7 of whom required latersurgery, a time interval of less than 6 hours after injury to thefirst CT scan and a cranial fracture that crossed a major vesselwere significantly related to surgery (19).

Studies Describing Successful Nonsurgical Management

Bullock et al. (4) treated 12 of 123 patients presenting withEDH nonsurgically. All patients were conscious (GCS 12–15)with a hematoma volume between 12 and 38 cm3 (mean, 26.8cm3) and an MLS less than 10 mm on the initial CT scan. Noneof the hematomas were in the temporal region. All patientsmade a good outcome. Cucciniello reported on 57 patientswith EDH who were treated nonoperatively (9). Initial GCSwas between 10 and 15. Five hematomas were in the temporalregion. The maximum hematoma thickness ranged between 6and 12 mm. Only one patient had an MLS. All patients madea good recovery.

Timing of Surgery

Time between Injury and Surgery

The effect of surgical timing on outcome from EDH isrelevant for a subgroup of patients in whom the EDH causescompression of brain structures that, with time, could causepoor outcome. This subgroup is usually categorized as havingpupillary abnormalities and/or a GCS score less than 9(coma). Generally, studies of EDH reveal that only 21 to 34%of patients present to the hospital with a GCS score less than8 or 9 (3, 20, 22, 25). Studies do not find a relationship betweensurgical timing and outcome if patients of all GCS scores areincluded. In 200 patients with EDH that were surgically evac-uated, Lee et al. (22) failed to demonstrate a significant rela-tionship between surgery within 4 hours of trauma or surgerywithin 2 hours of admission and outcome, using multivariateanalysis. However, a significant correlation was observed be-tween the duration of brain herniation, as evidenced by aniso-coria, and the outcome. The time lapse between the onset ofpupillary abnormalities and surgery is related to outcome.Cohen et al. (6) studied 21 patients with EDH and GCS scoreless than 9 who underwent surgery. Ten of these patientsdeveloped anisocoria after admission. All 5 patients withanisocoria for longer than 70 minutes before surgical evacua-tion of the EDH died. Patients with anisocoria for shorter than70 minutes achieved a good outcome. Haselsberger et al. (13)studied 60 patients with EDH, and 34 patients developedcoma before surgery. They found that patients treated within2 hours after loss of consciousness exhibited a mortality rate of17% and good recovery in 67%, compared with a mortalityrate of 56% and good results in 13% in patients operated onlater. Sakas et al. (33) found that all patients with either SDHor EDH with fixed and dilated pupils for longer than 6 hourdied.



Patient Transfer and Timing of Surgery

The question of whether a patient with acute EDH shouldbe treated at the nearest hospital or transferred to a specializedtrauma center has been debated but poorly documented instudies. This is an important timing issue and is significant inthe group of patients who are deteriorating. Another issue isthe surgical evacuation of EDH by nonneurosurgeons withsubsequent transfer to a neurosurgical center. Obviously,these studies are uncontrolled with regard to the efficacy ofsurgery and the type of patients included in both arms. In theabove timing of surgery, the group of patients in a coma andwith pupillary abnormalities can be expected to do worse thelonger the interval to evacuation of the EDH. Thus, because ofthe delay, transferred patients would have longer intervaltimes to surgery.

Wester (40) studied 83 patients with acute EDH that under-went craniotomy. Twenty-eight patients were transferredfrom other hospitals and 11 of these underwent emergencysurgery at those outside institutions. Patients who underwentsurgery outside the parent institution by nonneurosurgeonshad a significantly worse outcome at 3 months as comparedwith patients who were directly admitted to the study hospi-tal. This was mainly attributed to the technical inadequacy ofthe primary operation at the outside institution. The authorsinterpreted this as support for the strategy of directly trans-ferring patients to an adequate trauma center, but they did notcontrol for other confounding variables, such as admissionGCS and pupillary exam.

Another study analyzed 107 patients operated on for EDH(3). The majority (67%) of these patients were transferred fromoutlying hospitals. The authors noted that only 6% of thedirect admissions experienced a poor outcome, as comparedwith 18% of patients who were transferred after undergoingCT scanning at an outside institution. This difference failed toreach statistical significance. Poon and Li (31) studied 71 pa-tients with EDH managed surgically primarily at a neurosur-gical hospital and 33 patients transferred from an outsideinstitution. Time delay from neurological deterioration to sur-gery was 0.7 � 1 hour and 3.2 � 0.5 hours for the direct versusindirect transferred group, respectively. Six-months outcomewas significantly better in patients who were directly admittedwith a minimal delay from deterioration in neurological examto surgery.

SUMMARY

In patients with an acute EDH, clot thickness, hematomavolume, and MLS on the preoperative CT scan are related tooutcome. In studies analyzing CT parameters that may bepredictive for delayed surgery in patients undergoing initialnonoperative management, a hematoma volume greater than30 cm3, an MLS greater than 5 mm, and a clot thickness greaterthan 15 mm on the initial CT scan emerged as significant.Therefore, patients who were not comatose, without focalneurological deficits, and with an acute EDH with a thickness

of less than 15 mm, an MLS less than 5 mm, and a hematomavolume less than 30 cm3 may be managed nonoperativelywith serial CT scanning and close neurological evaluation in aneurosurgical center (see Appendix II for measurement tech-niques). The first follow-up CT scan in nonoperative patientsshould be obtained within 6 to 8 hours after TBI. Temporallocation of an EDH is associated with failure of nonoperativemanagement and should lower the threshold for surgery. Nostudies are available comparing operative and nonoperativemanagement in comatose patients. The literature supports thetheory that patients with a GCS less than 9 and an EDHgreater than 30 cm3 should undergo surgical evacuation of thelesion. Combined with the above recommendation, it followsthat all patients, regardless of GCS, should undergo surgery ifthe volume of their EDH exceeds 30 cm3. Patients with anEDH less than 30 should be considered for surgery but may bemanaged successfully without surgery in selected cases.

Time from neurological deterioration, as defined by onset ofcoma, pupillary abnormalities, or neurological deterioration tosurgery, is more important than time between trauma andsurgery. In these patients, surgical evacuation should be per-formed as soon as possible because every hour delay in sur-gery is associated with progressively worse outcome.

KEY ISSUES FOR FUTURE INVESTIGATION

•Effect of transfer versus direct admission to a traumacenter on timing of surgery and outcome from EDH.

•Identification of subgroups that do not benefit from sur-gery: old patients with low GCS scores, pupillary abnormali-ties, and associated intracerebral lesions.

•Surgical technique.

REFERENCES

1. Bejjani G, Donahue D, Rusin J, Broemeling L: Radiological and clinicalcriteria for the management of epidural hematomas in children. PediatrNeurosurg 25:302–308, 1996.

2. Bezircioglu H, Ersahin Y, Demircivi F, Yurt I, Donertas K, Tektas S:Nonoperative treatment of acute extradural hematomas: Analysis of 80cases. J Trauma 41:696–698, 1996.

3. Bricolo A, Pasut L: Extradural hematoma: toward zero mortality. A pro-spective study. Neurosurgery 14:8–12, 1984.

4. Bullock R, Smith R, van Dellen JR: Nonoperative management of extraduralhematoma. Neurosurgery 16:602–606, 1985.

5. Chen T, Wong C, Chang C, Lui T, Cheng W, Tsai M, Lin T: The expectanttreatment of �asymptomatic� supratentorial epidural hematomas. Neurosur-gery 32: 176–179, 1993.

6. Cohen J, Montero A, Israel Z: Prognosis and clinical relevance of anisocoria-craniotomy latency for epidural hematoma in comatose patients. J Trauma41:120–122, 1996.

7. Cook R, Dorsch N, Fearnside M, Chaseling R: Outcome prediction in extra-dural haematomas. Acta Neurochir (Wien) 95:90–94, 1988.

8. Cordobes F, Lobato R, Rivas J, Munoz M, Chillon D, Portillo J, Lamas E:Observations on 82 patients with extradural hematoma. Comparison ofresults before and after the advent of computerized tomography.J Neurosurg 54:179–186, 1981.

9. Cucciniello B, Martellotta N, Nigro D, Citro E: Conservative management ofextradural haematomas. Acta Neurochir (Wien) 120:47–52, 1993.



10. Gennarelli T, Spielman G, Langfitt T, Gildenberg P, Harrington T, Jane J,Marshall L, Miller J, Pitts L: Influence of the type of intracranial lesion onoutcome from severe head injury. J Neurosurg 56:26–32, 1982.

11. Gupta S, Tandon S, Mohanty S, Asthana S, Sharma S: Bilateral traumaticextradural haematomas: Report of 12 cases with a review of the literature.Clin Neurol Neurosurg 94:127–131, 1992.

12. Hamilton M, Wallace C: Nonoperative management of acute epidural he-matoma diagnosed by CT: The neuroradiologist’s role. AJNRAm J Neuroradiol 13:853–859, 1992.

13. Haselsberger K, Pucher R, Auer L: Prognosis after acute subdural or epi-dural haemorrhage. Acta Neurochir (Wien) 90:111–116, 1988.

14. Heinzelmann M, Platz A, Imhof H: Outcome after acute extraduralhaematoma, influence of additional injuries and neurological complicationsin the ICU. Injury 27:345–349, 1996.

15. Hunt J, Hill D, Besser M, West R, Roncal S: Outcome of patients withneurotrauma: The effect of a regionalized trauma system. Aust N Z J Surg65:83–86, 1995.

16. Jamjoom A: The influence of concomitant intradural pathology on thepresentation and outcome of patients with acute traumatic extraduralhaematoma. Acta Neurochir (Wien) 115:86–89, 1992.

17. Jamjoom A: The difference in the outcome of surgery for traumatic extra-dural hematoma between patients who are admitted directly to the neuro-surgical unit and those referred from another hospital. Neurosurg Rev20:227–230, 1997.

18. Jones N, Molloy C, Kloeden C, North J, Simpson D: Extradural haematoma:Trends in outcome over 35 years. Br J Neurosurg 7:465–471, 1993.

19. Knuckey N, Gelbard S, Epstein M: The management of �asymptomatic�epidural hematomas. A prospective study. J Neurosurg 70:392–396, 1989.

20. Kuday C, Uzan M, Hanci M: Statistical analysis of the factors affecting theoutcome of extradural haematomas: 115 cases. Acta Neurochir (Wien) 131:203–206, 1994.

21. Lahat E, Sheinman G, Feldman Z, Barzilai A, Harel R, Barzilay Z, Paret G:Metabolic and clinical markers of prognosis in the era of CT imaging inchildren with acute epidural hematomas. Pediatr Neurosurg 33:70–75, 2000.

22. Lee E, Hung Y, Wang L, Chung K, Chen H: Factors influencing the func-tional outcome of patients with acute epidural hematomas: Analysis of 200patients undergoing surgery. J Trauma 45:946–952, 1998.

23. Lobato R, Cordobes F, Rivas J, de la Fuente M, Montero, A, Barcena A, PerezC, Cabrera A, Lamas E: Outcome from severe head injury related to the typeof intracranial lesion. A computerized tomography study. J Neurosurg59:762–774, 1983.

24. Lobato R, Rivas J, Cordobes F, Alted E, Perez C, Sarabia R, Cabrera A, DiezI, Gomez P, Lamas E: Acute epidural hematoma: An analysis of factorsinfluencing the outcome of patients undergoing surgery in coma. JNeurosurg 68:48–57, 1988.

25. Maggi G, Aliberti F, Petrone G, Ruggiero C: Extradural hematomas inchildren. J Neurosurg Sci 42:95–99, 1998.

26. Meier U, Heinitz A, Kintzel D: Surgical outcome after severe craniocerebraltrauma in childhood and adulthood. A comparative study [in German].Unfallchirurg 97:406–409, 1994.

27. Mohanty A, Kolluri V, Subbakrishna D, Satish S, Mouli B, Das B: Prognosisof extradural haematomas in children. Pediatr Neurosurg 23:57–63, 1995.

28. Otsuka S, Nakatsu S, Matsumoto S, Sato S, Motozaki T, Ban S, Yamamoto T:Study on cases with posterior fossa epidural hematoma—Clinical featuresand indications for operation. Neurol Med Chir (Tokyo) 30:24–28, 1990.

29. Paterniti S, Fiore P, Macri E, Marra G, Cambria M, Falcone F, Cambria S:Extradural haematoma. Report of 37 consecutive cases with survival. ActaNeurochir (Wien) 131:207–210, 1994.

30. Pillay R, Peter J: Extradural haematomas in children. S Afr Med J 85:672–674, 1995.

31. Poon W, Li A: Comparison of management outcome of primary and sec-ondary referred patients with traumatic extradural haematoma in a neuro-surgical unit. Injury 22:323–325, 1991.

32. Rivas J, Lobato R, Sarabia R, Cordobes F, Cabrera A, Gomez P: Extraduralhematoma: Analysis of factors influencing the courses of 161 patients.Neurosurgery 23:44–51, 1988.

33. Sakas D, Bullock M, Teasdale G: One-year outcome following craniotomyfor traumatic hematoma in patients with fixed dilated pupils. J Neurosurg82:961–965, 1995.

34. Schutzman S, Barnes P, Mantello M, Scott R: Epidural hematomas in chil-dren. Ann Emerg Med 22:535–541, 1993.

35. Seelig J, Marshall L, Toutant S, Toole B, Klauber M, Bowers S, Varnell J:Traumatic acute epidural hematoma: Unrecognized high lethality in coma-tose patients. Neurosurgery 15:617–620, 1984.

36. Servadei F, Faccani G, Roccella P, Seracchioli A, Godano U, Ghadirpour R,Naddeo M, Piazza G, Carrieri P, Taggi F, Pagni CA: Asymptomatic extra-dural haematomas. Results of a multicenter study of 158 cases in minor headinjury. Acta Neurochir (Wien) 96:39–45, 1989.

37. Sullivan T, Jarvik J, Cohen W: Follow-up of conservatively managed epi-dural hematomas: Implications for timing of repeat CT. AJNRAm J Neuroradiol 20:107–113, 1999.

38. Uzan M, Yentur E, Hanci M, Kaynar MY, Kafadar A, Sarioglu AC, Bahar M,Kuday C: Is it possible to recover from uncal herniation? Analysis of 71 headinjured cases. J Neurosurg Sci 42:89–94, 1998.

39. van den Brink WA, Zwienenberg M, Zandee SM, van der Meer L, Maas AI,Avezaat CJ: The prognostic importance of the volume of traumatic epiduraland subdural haematomas revisited. Acta Neurochir (Wien) 141:509–514,1999.

40. Wester K: Decompressive surgery for �pure� epidural hematomas: Doesneurosurgical expertise improve the outcome? Neurosurgery 44:495–500,1999.

41. Wu J, Hsu C, Liao S, Wong Y: Surgical outcome of traumatic intracranialhematoma at a regional hospital in Taiwan. J Trauma 47:39–43, 1999.



EVID

ENTI

AR

YTA

BLE

TAB

LE1.

Surg

ical

man

agem

ent

ofac

ute

epid

ural

hem

atom

asa

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Bej

jani

etal

.(1

)33

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

Dis

char

geR

etro

spec

tive

anal

ysis

offa

ctor

saf

fect

ing

the

deci

sion

toop

erat

ein

33pe

diat

ric

patie

nts

with

EDH

.13

patie

nts

unde

rwen

top

erat

ive

trea

tmen

t.

Mas

sef

fect

,te

mpo

ral

loca

tion

ofbl

ood

clot

,M

LS,

thic

knes

sof

clot

,an

dvo

lum

ew

ere

inde

pend

ently

rela

ted

tosu

rger

y.O

ther

clin

ical

fact

ors,

such

asag

e,G

CS,

and

asso

ciat

edfr

actu

res

wer

eno

t.O

utco

me

was

good

in�

90%

ofpa

tient

s.

EDH

Ope

rati

vetr

eatm

ent

Non

oper

ativ

etr

eatm

ent

Thic

knes

s(m

m)

209

Vol

ume

(cm

3 )41

8

MLS

(mm

)4

0.5

Bez

irci

oglu

etal

.(2

)80

IIIG

CS

�8

Surg

ery

and

nons

urgi

cal

GO

Sat

2m

oPr

ospe

ctiv

est

udy

onno

nope

rativ

em

anag

emen

tof

patie

nts

with

EDH

.Pa

tient

sw

itha

EDH

volu

me

�30

ml,

thic

knes

s�

2cm

,G

CS

�8,

none

urol

ogic

alde

ficit,

and

adm

itted

with

in24

hof

TBI

wer

etr

eate

dno

nope

rativ

ely.

Of

80pa

tient

sw

how

ere

trea

ted

nono

pera

tivel

y,5

dete

rior

ated

and

need

edsu

rger

y.O

nedi

ed,

and

4ha

da

good

outc

ome.

Tem

pora

llo

beED

Hw

asre

late

dto

dela

yed

surg

ery.

Bri

colo

and

Pasu

t(3

)10

7III

All

GC

SA

llsu

rger

yG

OS

at6

mo

Neu

rolo

gica

lsi

gns

onad

mis

sion

No.

ofpa

tien

tsG

ood

outc

ome

(%)

Non

e17

100

One

dila

ted

pupi

l8

100

Hem

ipar

esis

only

2391

Hem

ipar

esis

and

one

dila

ted

pupi

l10

90

Dec

ortic

atio

n9

44

Dec

ereb

ratio

n4

25

Bot

hpu

pils

fixed

10

GC

Son

adm

issi

onN

o.of

pati

ents

GR

/MD

(%)

SD/V

S(%

)D

ead

(%)

3–4

425

2550

5–7

3278

139

8–1

571

973

0

Bul

lock

etal

.(4

)12

IIIG

CS

12–1

5N

onsu

rgic

alD

isch

arge

Ret

rosp

ectiv

ean

alys

isof

the

nono

pera

tive

trea

tmen

tof

12pa

tient

sw

ithED

H.

All

patie

nts

mad

ea

good

reco

very

.N

ote

mpo

ral

EDH

patie

nts

wer

ein

clud

ed,

and

basa

lci

ster

nsw

ere

open

oron

lypa

rtia

llyef

face

din

all

patie

nts.

MLS

was

�10

mm

inal

lpa

tient

s.M

ean

volu

me

ofth

eED

Hw

as26

.8cm

3 .

Che

net

al.

(5)

74III

GC

S�

12Su

rger

yan

dno

nsur

gica

lG

OS

at1

mo

Ret

rosp

ectiv

ean

alys

isof

111

of46

5pa

tient

sw

ithED

Hw

houn

derw

ent

initi

alno

nope

rativ

etr

eatm

ent.

All

nono

pera

tive

case

sha

da

GC

S�

12,

and

14pa

tient

sun

derw

ent

dela

yed

surg

ery

beca

use

ofne

urod

eter

iora

tion

orin

crea

sein

the

size

ofth

eED

H.

37pa

tient

sw

ere

excl

uded

beca

use

ofin

com

plet

eC

Tda

ta.

All

patie

nts

achi

eved

ago

odG

OS.

Patie

nts

requ

irin

gde

laye

dsu

rger

ypr

esen

ted

mor

efr

eque

ntly

with

anED

Hvo

lum

e�

30cm

3 ,cl

otth

ickn

ess

�15

mm

,or

MLS

�5

mm

onth

ein

itial

CT

scan

.

Ope

rati

vetr

eatm

ent

Non

oper

ativ

etr

eatm

ent

Pva

lue

No.

ofpa

tient

s14

97

Tem

pora

l/fr

onto

tem

pora

llo

catio

nof

EDH

42.8

%36

.6%

n.s.

EDH

�30

cm3

83

P�

0.00

1

MLS

�5

mm

1117

P�

0.00

1

EDH

�15

mm

thic

knes

s10

16P

�0.

01

(Tab

leco

ntin

ues)



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Coh

enet

al.

(6)

21III

GC

S�

8A

llsu

rger

yN

otdo

cum

ente

dPr

ospe

ctiv

eda

taco

llect

ion

on21

adul

tpa

tient

sad

mitt

eddu

ring

3yr

with

acut

eED

Han

dG

CS

�8,

who

unde

rwen

tsu

rger

y.10

patie

nts

deve

lope

dne

wan

isoc

oria

afte

rad

mis

sion

.O

nly

14pa

tient

sha

dC

Tsc

ans.

3pa

tient

sha

dem

erge

ncy

cran

ioto

my

with

out

radi

ogra

phic

stud

ies.

Ani

socr

iafo

rm

ore

than

70m

inw

asas

soci

ated

with

100%

mor

talit

y,an

dfo

rle

ssth

an70

min

with

GO

S4

and

5.

Tim

efr

oman

isoc

oria

tocr

anio

tom

yN

o.of

pati

ents

GR

/MD

(%)

Dea

d(%

)

�71

min

510

00

�89

min

50

100

Cuc

cini

ello

etal

.(9

)57

IIIG

CS

10–1

5N

onsu

rgic

alN

otdo

cum

ente

dR

etro

spec

tive

anal

ysis

ofa

case

seri

esof

57pa

tient

sou

tof

144

with

EDH

who

wer

em

anag

edno

nope

rativ

ely.

All

patie

nts

had

good

outc

ome.

Onl

yon

epa

tient

dem

onst

rate

dM

LSan

dm

axim

umhe

mat

oma

thic

knes

sw

asbe

twee

n6

and

12m

m.

Ham

ilton

and

Wal

lace

(12)

48III

All

GC

SSu

rger

yan

dno

nsur

gica

lD

isch

arge

Ret

rosp

ectiv

ean

alys

isof

patie

nts

who

unde

rwen

tsu

rgic

alan

dno

nope

rativ

etr

eatm

ent

for

EDH

.

Patie

nts

who

unde

rwen

tsu

rgic

altr

eatm

ent

had

low

erG

CS

scor

es,

mor

epu

pilla

ryab

norm

aliti

es,

larg

erED

H,

grea

ter

MLS

,an

dw

ere

mor

elik

ely

tosh

owun

cal

hern

iatio

non

CT.

Has

elsb

erge

ret

al.

(13)

60III

All

GC

SA

llsu

rger

yN

otdo

cum

ente

dA

revi

ewof

171

patie

nts

who

pres

ente

dw

ithei

ther

subd

ural

(111

patie

nts)

orep

idur

al(6

0pa

tient

s)he

mat

oma.

The

influ

ence

ofsu

rgic

altim

ing

onm

orta

lity

and

func

tiona

lre

cove

ryw

asan

alyz

ed.

Patie

nts

with

anac

ute

subd

ural

orep

idur

alhe

mat

oma

had

alo

wer

mor

talit

yan

dim

prov

edfu

nctio

nal

reco

very

whe

nop

erat

edon

�2

haf

ter

onse

tof

com

a.

Tim

efr

omco

ma

onse

tto

surg

ery

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS

(%)

D(%

)

�2

h18

6716

17

�2

h16

1331

56

Knu

ckey

etal

.(1

9)22

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

Not

docu

men

ted

Ret

rosp

ectiv

ean

alys

isof

22pa

tient

sof

all

age

grou

psw

how

ere

initi

ally

trea

ted

nono

pera

tivel

y.7

patie

nts

need

edsu

bseq

uent

cran

ioto

my.

Earl

yC

Tsc

an�

6h

afte

rtr

aum

aan

dcr

ania

lfr

actu

res

cros

sing

big

vess

els

orth

em

iddl

em

enin

geal

arte

ryw

ere

asso

ciat

edw

ithde

teri

orat

ion.

Initi

alG

CS,

age,

and

size

ofth

eED

Hw

ere

not.

Kud

ayet

al.

(20)

115

IIIA

llG

CS

All

surg

ery

GO

Sat

6m

oR

etro

spec

tive

anal

ysis

ofpr

ospe

ctiv

ely

colle

cted

data

on11

5pa

tient

sof

all

age

grou

psun

derg

oing

surg

ery

for

�sig

nsof

hern

iatio

n.�

GC

Sis

the

para

met

erco

rrel

atin

gm

ost

clos

ely

with

outc

ome.

The

pres

ence

ofa

cran

ial

frac

ture

and

“lon

g”in

terv

albe

twee

non

set

ofsy

mpt

oms

and

inte

rven

tion

also

affe

ctou

tcom

e.

Lee

etal

.(2

2)20

0III

All

GC

SA

llsu

rger

yG

OS

at1

yrA

retr

ospe

ctiv

est

udy

of20

0pa

tient

sw

ithep

idur

alhe

mat

omas

requ

irin

gop

erat

ion.

Ana

lysi

sof

fact

ors

lead

ing

toa

poor

outc

ome.

Func

tiona

lou

tcom

eco

rrel

ated

with

decr

ease

dle

ngth

ofhe

rnia

tion

time,

nopo

stur

ing,

norm

alpu

pils

,an

da

high

GC

Ssc

ore.

Unf

avor

able

radi

olog

ical

sign

sw

ere

asso

ciat

edbr

ain

inju

ry,

MLS

�1

cm,

oblit

erat

edba

sal

cist

erns

,an

dsi

zean

dde

nsity

ofth

ebl

ood

clot

.

Inte

rval

betw

een

anis

ocor

iaan

dde

com

pres

sion

(h)

No.

ofpa

tien

tsSD

/VS/

D(%

)

No

126

7.9

�1.

518

22.2

1.5–

2.5

2725

.9

2.5–

3.5

1330

.8

3.5–

4.5

1353

.8

�4.

53

66.7

Hem

atom

avo

lum

e(m

l)N

o.of

pati

ents

SD/V

S/D

(%)

�50

816.

2

51–1

0081

14.8

101–

150

3138

.7

�15

07

71.4

MLS

(mm

)N

o.of

pati

ents

SD/V

S/D

(%)

�5

859.

4

6–1

069

8.7

11–1

535

37.1

�15

1163

.6

(Tab

leco

ntin

ues)



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Poon

and

Li(3

1)10

4III

All

GC

SA

llsu

rger

yG

OS

at6

mo

Apr

ospe

ctiv

est

udy

of10

4pa

tient

sof

all

age

grou

psw

ithis

olat

edED

H.

One

-thi

rdof

the

patie

nts

wer

etr

ansf

erre

dfr

oman

outs

ide

hosp

ital.

Asi

gnifi

cant

incr

ease

inm

orta

lity

and

mor

bidi

tyw

asre

late

dto

incr

ease

dtim

efr

omne

urol

ogic

alde

teri

orat

ion

tosu

rger

y.

No.

ofpa

tien

ts

Tim

ebe

twee

nde

teri

orat

ion

ofco

nsci

ous

leve

lan

dsu

rger

y(h

)

D(%

)V

S/SD

/MD

(%)

GR

(%)

710.

7�

14

1086

333.

2�

0.5

2427

49

Saka

set

al.

(33)

11III

“Com

atos

e”A

llsu

rger

yG

OS

at1

yrA

naly

sis

of40

seve

reTB

Ipa

tient

sw

houn

derw

ent

cran

ioto

my

afte

rde

velo

ping

bila

tera

lfix

edan

ddi

late

dpu

pils

.

Patie

nts

with

SDH

had

asi

gnifi

cant

incr

ease

inm

orta

lity

(64%

)co

mpa

red

with

thos

ew

ithED

H(1

8%).

Patie

nts

who

unde

rwen

tde

laye

d(�

3h)

surg

ery

had

aw

orse

outc

ome.

Tim

efr

ompu

pilla

ryno

nrea

ctiv

ity

tosu

rger

y

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS/

D(%

)

�3

h20

3030

40

�3

h16

2512

63

Seel

iget

al.

(35)

51III

“Com

atos

e”A

llsu

rger

yG

OS

at3

mo–

2yr

Pros

pect

ive

mul

ticen

ter

data

base

anal

yzed

retr

ospe

ctiv

ely,

look

ing

atth

efa

ctor

sin

fluen

cing

outc

ome

afte

rsu

rger

yfo

rED

H.Th

em

ost

pow

erfu

lin

dica

tor

ofou

tcom

ew

asm

otor

scor

ebe

fore

surg

ery.

Mot

orsc

ore

No.

ofpa

tien

tsG

R/M

D/S

D(%

)V

S/D

(%)

1–3

1932

68

4–

632

6931

Serv

adei

etal

.(3

6)15

8III

GC

S14

/15

Surg

ery

and

nons

urgi

cal

GO

Sat

6m

oPr

ospe

ctiv

est

udy

of15

8co

nsec

utiv

epa

tient

sw

ithm

inor

TBI

(GC

S14

/15)

adm

itted

to3

neur

osur

gica

lun

its.

No

trea

tmen

tpr

otoc

olw

assp

ecifi

edan

da

logi

stic

regr

essi

onan

alys

isw

asco

nduc

ted

toid

entif

yth

efa

ctor

sle

adin

gto

surg

ery.

116

patie

nts

unde

rwen

tsu

rger

yan

d42

patie

nts

wer

em

anag

edno

nope

rativ

ely.

Of

all

fact

ors

anal

yzed

,on

lyth

ickn

ess

ofhe

mat

oma

and

MLS

wer

ere

late

dto

the

deci

sion

toop

erat

e.O

utco

me

was

the

sam

ein

both

grou

ps.

MLS

(mm

)N

o.of

pati

ents

Cra

niot

omy

(%)

Non

oper

ativ

em

anag

emen

t(%

)

0–5

130

7030

6–1

019

89.5

10.5

�10

810

00

van

den

Bri

nket

al.

(39)

98III

All

GC

SA

llsu

rger

yG

OS

at6

mo

Ret

rosp

ectiv

ean

alys

isof

CT

para

met

ers

and

outc

ome

in98

patie

nts

ofal

lag

egr

oups

with

acut

eED

H.

Vol

ume

ofth

eED

Hdi

dno

tco

rrel

ate

with

outc

ome.

Suba

rach

noid

bloo

dan

dpu

pilla

rydy

sfun

ctio

n,ag

e�

20yr

,an

dG

CS

onad

mis

sion

wer

epa

ram

eter

sco

rrel

atin

gw

ithou

tcom

e.

Age

(yr)

SD/V

S/D

(%)

�20

6

21–

4017

41–

6031

�60

50

Mot

orsc

ore

SD/V

S/D

(%)

1–3

50

4–

67

Pupi

lsSD

/VS/

D(%

)

Nor

mal

3

Uni

late

ral

fixed

and

dila

ted

35

Bila

tera

lfix

edan

ddi

late

d50

Wes

ter

(40)

83III

All

GC

SA

llsu

rger

yG

OS

at3

mo

Ret

rosp

ectiv

est

udy

ofth

em

anag

emen

tof

pure

EDH

inpa

tient

sof

all

age

grou

psin

Nor

way

.In

form

atio

non

thei

rin

itial

neur

olog

ical

exam

islim

ited.

The

outc

ome

of11

patie

nts

trea

ted

atth

epr

imar

yho

spita

lw

asw

orse

than

the

outc

ome

ofpa

tient

sth

atw

ere

adm

itted

dire

ctly

.Th

eau

thor

sco

nclu

deth

atpa

tient

ssh

ould

betr

ansf

erre

ddi

rect

lyfo

rne

uros

urgi

cal

man

agem

ent.

aG

CS,

Gla

sgow

Com

aSc

ale;

EDH

,ep

idur

alhe

mat

oma;

MLS

,m

idlin

esh

ift;

GO

S,G

lasg

owou

tcom

esc

ore;

TBI,

trau

mat

icbr

ain

inju

ry;

GR

,go

odre

cove

ry;

MD

,m

oder

ate

disa

bilit

y;SD

,se

vere

disa

bilit

y;V

S,ve

geta

tive

stat

e;D

,de

ath;

CT,

com

pute

dto

mog

raph

icsc

an;

n.s.

,no

tsi

gnifi

cant

.



CHAPTER 4

SURGICAL MANAGEMENT OF ACUTE

SUBDURAL HEMATOMASM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia











Indications for Surgery• An acute subdural hematoma (SDH) with a thickness greater than 10 mm or a midline

shift greater than 5 mm on computed tomographic (CT) scan should be surgicallyevacuated, regardless of the patient’s Glasgow Coma Scale (GCS) score.

• All patients with acute SDH in coma (GCS score less than 9) should undergo intracranialpressure (ICP) monitoring.

• A comatose patient (GCS score less than 9) with an SDH less than 10-mm thick and amidline shift less than 5 mm should undergo surgical evacuation of the lesion if the GCSscore decreased between the time of injury and hospital admission by 2 or more pointson the GCS and/or the patient presents with asymmetric or fixed and dilated pupilsand/or the ICP exceeds 20 mm Hg.

Timing• In patients with acute SDH and indications for surgery, surgical evacuation should be

performed as soon as possible.Methods• If surgical evacuation of an acute SDH in a comatose patient (GCS � 9) is indicated, it

should be performed using a craniotomy with or without bone flap removal andduraplasty.

KEY WORDS: Coma, Computed tomographic parameters, Craniotomy, Decompressive craniectomy, Headinjury, Hematoma, Intracranial pressure monitoring, Salvageability, Subdural, Surgical technique, Timing ofsurgery, Traumatic brain injury

Neurosurgery 58:S2-16-S2-24, 2006 DOI: 10.1227/01.NEU.0000210364.29290.C9 www.neurosurgery-online.com

OVERVIEW

SDH represents one type of intracranialmass lesion, and surgical management at-tempts to define the subset of patients whowould benefit from surgical evacuation of anacute SDH. SDH are diagnosed on a CT scanas extracranial, hyperdense, crescentic collec-tions between the dura and the brain paren-chyma. They can be divided into acute andchronic lesions. Herein, “acute SDH” is de-fined as an SDH diagnosed within 14 days oftraumatic brain injury (TBI).

Incidence

Studies conducted after the introduction ofCT scanning report an incidence of acute SDHbetween 12 and 29% in patients admitted with

severe TBI. Combining several publications,acute SDH was diagnosed in 21% of 2870 pa-tients (8, 26, 27, 36). When including mild, mod-erate, and severe head injuries, 11% (360 of 3397patients) present with SDH (21, 28). The meanage for this combined group is between 31 and47 years, with the vast majority of patients beingmen (6, 11, 17, 21).

Mechanism

The mechanism of injury responsible for thedevelopment of an SDH differs between agegroups. Most SDH are caused by motorvehicle-related accidents (MVA), falls, and as-saults. In one study, 56% of SDH in theyounger group (18–40 yr) were caused byMVA and only 12% were caused by falls,whereas, in the older groups (�65 yr), these


mechanisms were responsible for 22% and 56% of SDH, re-spectively (15). Falls have been identified as the main cause oftraumatic SDH in two studies looking specifically at patientsolder than 75 and 80 years (3, 16). Studies with comatosepatients describe MVA as the mechanism of injury in 53 to75% of SDH. This indicates that MVA causes more severeinjury, possibly because of high-velocity accidents and diffuseaxonal injury (18, 26, 36).

Clinical Presentation

Between 37 and 80% of patients with acute SDH presentwith initial GCS scores of 8 or less (4, 6, 21, 28, 33). A lucidinterval has been described in 12 to 38% of patients beforeadmission but there is no conclusive evidence that this corre-lates with outcome (1, 9, 31, 34, 36). The definition of lucidinterval is vague. Authors interpret the lucid interval differ-ently and analysis of its frequency requires documentationduring the prehospital phase. Pupillary abnormalities are ob-served in 30 to 50% of patients on admission or before surgery(6, 10, 28, 33).

Mortality

Studies looking at patients from all age groups with GCSscores between 3 and 15 with SDH requiring surgery quotemortality rates between 40 and 60% (10, 11, 13, 14, 17, 27, 39).Mortality among patients presenting to the hospital in comawith subsequent surgical evacuation is between 57 and 68% (7,12, 18, 20, 26, 36).

Associated Injuries

Only 30 to 40% of SDH requiring surgery are isolatedlesions (21, 28). In the majority of cases, the SDH is associatedwith other intracranial and extracranial injuries. Contusions andintracerebral hematomas are the most frequently associated in-tracranial abnormalities. Associated intracranial and extracraniallesions have been reported in larger series to occur in 47 to 57%of patients presenting with GCS scores between 3 and 15 (4, 11,15, 28) and in 65 to 82% of patients with GCS scores less than 10(18, 26). In patients with SDH, contusions and fractures arefrequent associated injuries (16, 19, 25, 31, 36, 37). An associatedsubarachnoid hemorrhage has been observed in 14 to 25% ofpatients with SDH (4, 28) and epidural hematomas are observedin 6 to 14% of patients (4, 28). Significant extracranial injuries areobserved in 18 to 51% of patients and the majority of these casesinclude facial fractures, extremity fractures, and thoracic andabdominal trauma (5, 6, 11, 28, 31). Because of the frequentassociation of SDH with parenchymal injury, surgical manage-ment decisions should take into consideration the recommenda-tions for both lesion types.

PROCESS

A MEDLINE computer search using the following key-words for the years 1975 to 2001 was performed: “traumaticbrain injury” or “head injury” and “subdural” or “intradural”

and “hematoma” or “hemorrhage.” The search was narrowedby including the keywords “surgical treatment” or “surgery”or “operation” or “craniotomy” or “craniectomy” or “cranios-tomy” or “burr holes” and excluding “chronic” and “spinal.”These searches combined yielded 161 articles. The referencelists of these publications were reviewed and an additional 18articles were selected for analysis. Case reports, publicationsin books, and publications regarding penetrating brain inju-ries, or spinal or chronic SDH were not included. Chronic SDHwas defined as an SDH occurring or diagnosed more than 14days after trauma. Articles were excluded if the diagnosis ofSDH was not based on CT scanning, or if subgroups of pa-tients who did not undergo CT scanning were not clearlyidentified. Publications with fewer than 10 patients or publi-cations that did not include information on outcome wereexcluded. Of these 179 articles, 21 were selected for analysis.


Indications for Surgery

The decision to operate on an SDH is based on the patient’sGCS score, pupillary exam, comorbidities, CT findings, age, and,in delayed decisions, ICP. Neurological deterioration over time isalso an important factor influencing the decision to operate.Trauma patients presenting to the emergency room with alteredmental status, pupillary asymmetry, and abnormal flexion orextension are at high risk for either an SDH and/or an epiduralhematoma compressing the brain and brainstem.

CT Parameters

Many investigators have tried to define a relationship be-tween CT parameters, such as hematoma volume, clot thick-ness, midline shift (MLS), and patency of the basal cisterns,and outcome. Two studies using multivariate analysis to iden-tify factors affecting outcome from SDH found contradictoryresults. Howard et al. (15) reported on 67 patients, with GCSscores between 3 and 15, who were undergoing surgery, andfound a significant correlation between poor outcome and thevolume of the SDH and the MLS. The volume of the SDH, theMLS, and mortality were significantly greater in older pa-tients. van den Brink et al. (33) found no difference in hema-toma volumes, MLS or status of the basal cisterns when com-paring surgical patients, who had a GCS of 3 to 15, andfavorable versus unfavorable outcome. Zumkeller et al. (39)investigated CT scan parameters in 174 patients with SDH anda GCS between 3 and 15 undergoing surgery. The findingsrevealed a 10% mortality rate in patients with a clot thicknessof less than 10 mm, and a 90% mortality for patients with clotsthicker than 30 mm. For an MLS greater than 20 mm, therewas a steep increase in mortality. Both parameters correlatedwell with the Glasgow outcome score (GOS). In a mixed groupof patients treated with or without surgery, Servadei et al. (28)also found a correlation between outcome and clot thickness,MLS, and status of the basilar cisterns. Kotwica and Brzezinski(18) found a significant relationship between MLS and out-

ACUTE SUBDURAL HEMATOMAS


come in 200 patients with GCS scores lower than 10, who wereundergoing surgery for SDH. In summary, there seems to bea relationship between CT parameters and outcome, but it isdifficult to determine specific threshold values.

Surgical Versus Nonoperative Treatment of SDH

The decision for nonoperative versus surgical managementof SDH is influenced by the GCS score; CT parameters, such asMLS, SDH clot thickness and volume, and patency of the basalcisterns; and the salvageability of the patient (i.e., whether theprimary injury is so extensive that evacuation of the SDH willnot make a difference in outcome). On the basis of the re-viewed literature, a clot thickness greater than 10 mm or aMLS greater than 5 mm are suggested as critical parametersfor surgical evacuation of an acute SDH, regardless of theGCS.

Wong (37) tried to identify parameters that would predictthe failure of initial nonoperative management. No treatmentprotocol was defined. Six of 31 patients with GCS scoresbetween 6 and 15 who were initially treated without surgeryrequired a later craniotomy because of neurological deteriora-tion (performed within 3 d). The authors found that an MLSgreater than 5 mm in patients with a GCS score of lower than15 on the initial CT scan was significantly related to the failureof nonoperative treatment. Hematoma volume and thicknessof the hematoma were not predictive. Good outcome wasachieved in all patients.

Matthew et al. (22) reviewed the data on 23 patients withGCS scores between 13 and 15 who were initially treatednonoperatively. No criteria were defined for nonoperativemanagement. All patients had an isolated SDH and all wereobserved in the neurosurgical intensive care unit. Six patientsrequired delayed (mean, 14 d) evacuation of their SDH. Sig-nificant differences in clot thickness and hematoma volumewere found between the operative and the nonoperativegroups. In addition, all patients with an initial hematomathickness greater than 10 mm required surgery. Finally, Ser-vadei et al. (27) developed a protocol to select comatose pa-tients with SDH for nonoperative management. The criteriaused to select comatose patients for nonoperative treatmentwere clinical stability or improvement during the time frominjury to evaluation at the hospital, hematoma thickness lessthan 10 mm and MLS less than 5 mm on the initial CT scan,and ICP monitoring in the neurosurgical intensive care unit.Surgery was performed if the ICP exceeded 20 mm Hg. Fifteenof 65 comatose patients with SDH were treated nonopera-tively. Of these, two patients were identified that requireddelayed surgery based on increasing ICP and the developmentof intracerebral hematomas. Good outcome was achieved in23% of the patients in the surgery group and 67% of thepatients in the nonoperative group. The authors concludedthat nonoperative treatment can be safely used for a definedgroup of comatose patients with SDH.

Age and Salvageability

Increasing age is a strong independent factor in prognosisfrom severe TBI, with a significant increase in poor outcome inpatients older than 60 years of age (2). Among patients withacute SDH, there is also a tendency for older patients to havea poorer outcome, especially those patients presenting withlow GCS scores (3, 16, 18, 19, 36). In comatose patients withGCS scores less than 9 who underwent craniotomy for SDH,Wilberger et al. (36) found that age older than 65 years wasstatistically correlated with poorer outcome. In patients withGCS scores less than 10 undergoing surgery for SDH, Kotwicaand Brzezinski (18) found that there was a statistically signif-icant difference in 3-months outcome between younger pa-tients (18–30 yr of age, 25% mortality) and older patients (�50yr, 75% mortality). Three smaller studies looked specifically atpatients between 70 and 100 years of age with an admissionGCS (one study) or preoperative GCS (two studies) equal to orless than 9. The 49 patients from these three studies all under-went surgery. Forty-eight patients died and one had a pooroutcome (severely disabled or vegetative) (3, 16, 19). No pa-tient older than the age of 75 years who preoperatively wasextensor posturing, flaccid to pain, or had unilateral or bilat-eral fixed and dilated pupils made a good recovery (GOS, 3–5)(16). In 23 comatose patients aged 66 years and older whopresented with an acute SDH, Howard et al. (15) found that 17died and the others survived in a vegetative state or withsevere disabilities.

Functional outcomes in older patients with low GCS scoreshave also been reported. However, these articles did not doc-ument whether patients showed signs of cerebral herniation.Hatashita et al. (14) reported 9 deaths in 12 patients older than65 years who presented with GCS scores between 4 and 6 andunderwent surgery for SDH, as compared with 34% for thoseaged 19 to 40 years. Two older patients survived with a GOSof 4 or 5. In another publication, 1 of 28 comatose patientsolder than 65 years made a functional recovery after craniot-omy for SDH (36). Although some studies that included pa-tients with all GCS scores undergoing surgery for SDH founda relationship between age and outcome (15, 21, 28), otherauthors failed to describe such a relationship (13, 17, 26, 33,39). Three studies using multivariate analysis in patients op-erated on for SDH did not identify age as an independentpredictor of outcome (15, 26, 33). In summary, there is arelationship between poor outcome and age, low GCS, andsigns of herniation, but it is not possible to predict death onthe basis of old age and poor GCS with certainty.

Timing of Surgery

The time from injury to entering the operating room is oneof the few factors that can be affected by intervention. Unfor-tunately, the relationship between time from injury to opera-tion and outcome is difficult to study because patients who areoperated on soon after TBI tend to have more severe injuriesthan those who undergo delayed surgery. Therefore, outcomein patients operated on a short time after injury is frequently



worse when compared with patients undergoing delayed sur-gery. Furthermore, time from TBI to surgery may not be asimportant as time from clinical deterioration or onset of cere-bral herniation to surgery. The literature supports the state-ment that the length of time from clinical deterioration tooperative treatment of an SDH is significantly related to out-come. Haselsberger et al. (13) studied the time interval fromonset of coma to surgery in 111 patients with SDH. Thirty-fourpatients were operated on within 2 hours after onset of coma.Of those patients, 47% died and 32% recovered with goodoutcome or moderate disability. However, 54 patients whounderwent surgery longer than 2 hours after the onset of comahad a mortality of 80% and only 4% had a favorable outcome.These differences were statistically significant.

Seelig et al. (26) studied the delay to surgery in 82 patientswith SDH who were all comatose on admission. They found a30% mortality rate in patients operated on within 4 hours afterinjury and a 90% mortality in patients who had surgery morethan 4 hours after injury. The mean time for evacuation was390 � 39 minutes in patients who died and 170 � 18 minutesin patients who made a functional recovery. Multivariate anal-ysis identified time to surgery as one of the factors determin-ing outcome from SDH. The weaknesses of this study are thata proportion of patients did not undergo CT scanning and thatSDH was diagnosed using air ventriculography. In comatosepatients undergoing surgery for SDH, Wilberger et al. (36)found that the time interval from TBI to surgery was 374 � 31minutes for patients who died and 280 � 26 minutes forpatients who made a functional recovery. Mortality in patientsundergoing surgery within 4 hours of injury was 59% versus69% in patients operated on after 4 hours. A statistically sig-nificant difference could only be found in patients who un-derwent surgery after 12 hours, in which case, mortality roseto 82%.

Sakas et al. (24) looked at outcome from surgery for intra-cerebral hematoma, epidural hematoma, and SDH in 40 pa-tients who developed bilateral pupillary abnormalities duringtheir hospital course. The authors found a significant relation-ship between the time from onset of bilateral pupillary abnor-malities and 6-months outcome. Patients who had surgerymore than 3 hours after herniation had a higher morbidity andmortality than those undergoing surgery earlier (mortality,63% versus 30%).

Most studies focusing on the time between injury and sur-gery did not find a correlation with outcome (15, 17, 18, 21, 28,32). Some investigators even reported that early surgery wasassociated with worse results than delayed surgery (6, 14, 29).As mentioned, this may be related to the fact that most inves-tigators do not control for other variables affecting outcome,such as prehospital hypotension, hypoxia, GCS score, andassociated intracranial lesions. In 82 patients undergoing sur-gery for SDH, Dent et al. (6) found that time to surgery of lessthan 4 hours was associated with a significantly lower rate offunctional outcome when compared with surgery delayed forlonger than 4 hours (24% versus 51%). Mortality was approx-imately 30% in both groups. The authors also found that

patients who underwent surgery within 4 hours were morelikely to have obliterated basal cisterns and showed a ten-dency for lower GCS scores and more associated intracranialinjuries, suggesting a more severe TBI.

The only large study with patients with low GCS scores(GCS � 10) that did not find a relationship between earlysurgery and better outcome was the study by Kotwica andBrzezinski (18). In that study, mortality was approximately60% in all patients, regardless of whether they had surgerywithin 4 hours or between 4 and 16 hours after TBI. A detailedanalysis reveals that although GCS scores were the same,almost 90% of patients undergoing early surgery had associ-ated intracranial lesions. Associated lesions were found in 78%and 64% of patients surviving the first 5 and 12 hours, respec-tively. This indicates that patients with early surgery hadmore severe injuries. In summary, there is evidence that pa-tients who undergo surgery within 2 to 4 hours after clinicaldeterioration have a better outcome than those who undergodelayed surgery.

Surgical Technique

Different surgical techniques have been advocated for theevacuation of an SDH. The most commonly used techniquesare:

• Twist drill trephination/craniostomy procedures.• Burr hole trephination.• Craniotomy with or without dural grafting.• Subtemporal decompressive craniectomy.• Large decompressive hemicraniectomy, with or without du-

ral grafting.

Most investigators do not specify the type of surgical treat-ment used for evacuation of the SDH and, if they do, theyusually do not address the effectiveness of the procedure.Except for two studies (14, 30) no papers were found lookingat the impact of procedure type on outcome. The choice ofoperative technique is influenced by the surgeon’s expertise,training, and evaluation of the particular situation. Some cen-ters treat all SDH with decompressive craniectomies (18, 23),whereas other centers used solely osteoplastic craniotomies(36). Most studies report a mixture of procedures dependingon the clinical and radiographic evaluation (13–15, 17, 38), orcombined approaches in the same patient, i.e., subtemporaldecompression plus subsequent craniotomy (26) or cranioto-mies with contralateral decompressive craniectomies in somechildren (31). One study evaluated decompressive hemicrani-ectomies for the treatment of selected patients with SDH (30).

Only two investigators addressed the effect of the operativetechnique on outcome from SDH. Hatashita et al. (14) lookedat 3-months GOS in 60 patients with GCS scores between 3and 15 admitted for SDH evacuation. All patients underwentsurgery. The authors performed 24 burr holes, 25 cranioto-mies, 8 craniotomies with dural grafting, and 3 decompressivecraniectomies. In patients with GCS scores between 4 and 6,the authors found a statistically significant increased mortality



and reduced functional recovery rate in patients undergoingburr hole trephination versus craniotomy. Koc et al. (17) com-pared craniotomy, craniotomy with dural grafting, and de-compressive craniectomy in 113 patients with GCS scoresbetween 3 and 15 undergoing SDH evacuation. Seventeenpatients underwent decompressive craniectomy and all died.No other significant differences were found between treat-ment groups. The results of all of these studies have to beviewed with caution because groups undergoing differenttypes of surgical treatment were not comparable.

SUMMARY

In patients with an acute SDH, clot thickness or volume andthe MLS on the preoperative CT correlate with outcome. Instudies analyzing CT parameters that may be predictive fordelayed surgery in patients undergoing initial nonoperativemanagement, an MLS greater than 5 mm or a clot thicknessgreater than 10 mm on the initial CT scan emerged as signif-icant prognostic factors (see Appendices for measurement tech-niques). Therefore, patients with SDH presenting with a clotthickness greater than 10 mm or an MLS greater than 5 mmshould undergo surgical evacuation, regardless of their GCS.Patients who present in a coma (GCS � 9) but with an SDHwith a thickness less than 10 mm and an MLS less than 5 mmcan be treated nonoperatively, providing that they undergoICP monitoring, they are neurologically stable since the injury,they have no pupillary abnormalities, and they have no intra-cranial hypertension (ICP � 20 mm Hg). Because of the fre-quent association of SDH with parenchymal injury, surgicalmanagement decisions should take into consideration the rec-ommendations for both lesion types.

KEY ISSUES FOR FUTURE INVESTIGATION• Craniotomy versus decompressive craniectomy and dural

grafting for the initial evacuation of SDH. Effect of differentprehospital ambulance systems on timing of surgery andoutcome from SDH.

• Incidence and impact of prehospital hypotension and hyp-oxia on outcome from SDH.

• Identification of subgroups that do not benefit from surgery:older patients with low GCS scores, pupillary abnormalities,and associated intracerebral lesions.

• Prospective evaluation of the treatment option for comatosepatients (GCS � 9) presented above: does operating on allcomatose patients, regardless of their hematoma thicknessand MLS lead to a better outcome than following the treat-ment option presented above.

REFERENCES

1. Bowers S, Marshall L: Outcome in 200 consecutive cases of severe headinjury treated in San Diego County: A prospective analysis. Neurosurgery6:237–242, 1980.

2. Brain Trauma Foundation: Early indicators of prognosis in severe traumaticbrain injury. J Neurotrauma 17:535–627, 2000.

3. Cagetti B, Cossu M, Pau A, Rivano C, Viale G: The outcome from acutesubdural and epidural intracranial haematomas in very elderly patients. BrJ Neurosurg 6:227–231, 1992.

4. Cordobes F, Lobato R, Rivas J, Munoz M, Chillon D, Portillo J, Lamas E: Observa-tions on 82 patients with extradural hematoma. Comparison of results before andafter the advent of computerized tomography. J Neurosurg 54:179–186, 1981.

5. Croce M, Dent D, Menke P, Robertson J, Hinson M, Young B, Donovan T,Pritchard F, Minard G, Kudsk K, Fabian TC: Acute subdural hematoma:Nonsurgical management of selected patients. J Trauma 36:820–826, 1994.

6. Dent D, Croce M, Menke P, Young B, Hinson M, Kudsk K, Minard G,Pritchard F, Robertson J, Fabian T: Prognostic factors after acute subduralhematoma. J Trauma 39:36–42, 1995.

7. Domenicucci M, Strzelecki J, Delfini R: Acute posttraumatic subdural he-matomas: �Intradural� computed tomographic appearance as a favorableprognostic factor. Neurosurgery 42:51–55, 1998.

8. Ersahin Y, Mutluer S: Posterior fossa extradural hematomas in children.Pediatr Neurosurg 19:31–33, 1993.

9. Espersen J, Petersen O: Computerized tomography (CT) in patients withhead injuries. Relation between CT scans and clinical findings in 96 patients.Acta Neurochir (Wien) 56:201–217, 1981.

10. Fell D, Fitzgerald S, Moiel R, Caram P: Acute subdural hematomas. Reviewof 144 cases. J Neurosurg 42:37–42, 1975.

11. Gabl M, Mohsenipour I, Benedetto K: Acute posttraumatic subdural hema-toma in advanced age [in German]. Unfallchirurgie 15:273–278, 1989.


13. Haselsberger K, Pucher R, Auer L: Prognosis after acute subdural or epi-dural haemorrhage. Acta Neurochir (Wien) 90:111–116, 1988.

14. Hatashita S, Koga N, Hosaka Y, Takagi S: Acute subdural hematoma:Severity of injury, surgical intervention, and mortality. Neurol Med Chir(Tokyo) 33:13–18, 1993.

15. Howard MA 3rd, Gross AS, Dacey RJ Jr, Winn HR: Acute subdural hema-tomas: An age-dependent clinical entity. J Neurosurg 71:858–863, 1989.

16. Jamjoom A: Justification for evacuating acute subdural haematomas inpatients above the age of 75 years. Injury 23:518–520, 1992.

17. Koc R, Akdemir H, Oktem I, Meral M, Menku A: Acute subdural hematoma:Outcome and outcome prediction. Neurosurg Rev 20:239–244, 1997.

18. Kotwica Z, Brzezinski J: Acute subdural haematoma in adults: An analysisof outcome in comatose patients. Acta Neurochir (Wien) 121:95–99, 1993.

19. Kotwica Z, Jakubowski J: Acute head injuries in the elderly. An analysis of136 consecutive patients. Acta Neurochir (Wien) 118:98–102, 1992.

20. Lobato R, Cordobes F, Rivas J, de la Fuente M, Montero, A, Barcena A, Perez C,Cabrera A, Lamas E: Outcome from severe head injury related to the type ofintracranial lesion. A computerized tomography study. J Neurosurg 59:762–774,1983.

21. Massaro F, Lanotte M, Faccani G, Triolo C: One hundred and twenty-sevencases of acute subdural haematoma operated on. Correlation between CTscan findings and outcome. Acta Neurochir (Wien) 138:185–191, 1996.

22. Mathew P, Oluoch-Olunya D, Condon B, Bullock R: Acute subduralhaematoma in the conscious patient: Outcome with initial nonoperativemanagement. Acta Neurochir (Wien) 121:100–108, 1993.

23. Paterniti S, Fiore P, Macri E, Marra G, Cambria M, Falcone F, Cambria S:Extradural haematoma. Report of 37 consecutive cases with survival. ActaNeurochir (Wien) 131:207–210, 1994.


25. Schecter W, Peper E, Tuatoo V: Can general surgery improve the outcome ofthe head-injury victim in rural America? A review of the experience inAmerican Samoa. Arch Surg 120:1163–1166, 1985.

26. Seelig J, Becker D, Miller J, Greenberg R, Ward J, Choi S: Traumatic acutesubdural hematoma: Major mortality reduction in comatose patients treatedwithin four hours. N Engl J Med 304:1511–1518, 1981.

27. Servadei F, Nasi M, Cremonini A, Giuliani G, Cenni P, Nanni A: Importanceof a reliable admission Glasgow Coma Scale score for determining the needfor evacuation of posttraumatic subdural hematomas: A prospective studyof 65 patients. J Trauma 44:868–873, 1998.



28. Servadei F, Nasi M, Giuliani G, Cremonini A, Cenni P, Zappi D, Taylor G:CT prognostic factors in acute subdural haematomas: The value of the’worst’ CT scan. Br J Neurosurg 14:110–116, 2000.

29. Shenkin H: Acute subdural hematoma. Review of 39 consecutive cases withhigh incidence of cortical artery rupture. J Neurosurg 57:254–257, 1982.

30. Shigemori M, Syojima K, Nakayama K, Kojima T, Ogata T, Watanabe M,Kuramoto S: The outcome from acute subdural haematoma following de-compressive hemicraniectomy. Acta Neurochir (Wien) 54:61–69, 1980.

31. Spanu G, Pezzotta S, Silvani V, Leone V: Outcome following acute supratentorialsubdural hematoma in pediatric age. J Neurosurg Sci 29:31–35, 1985.

32. Uzan M, Yentur E, Hanci M, Kaynar MY, Kafadar A, Sarioglu AC, Bahar M,Kuday C: Is it possible to recover from uncal herniation? Analysis of 71 headinjured cases. J Neurosurg Sci 42:89–94, 1998.

33. van den Brink WA, Zwienenberg M, Zandee SM, van der Meer L, Maas AI,Avezaat CJ: The prognostic importance of the volume of traumatic epidural andsubdural haematomas revisited. Acta Neurochir (Wien) 141:509–514, 1999.

34. van den Heever CM, van der Merwe DJ: Management of depressed skullfractures. Selective conservative management of nonmissile injuries.J Neurosurg 71:186–190, 1989.

35. Wilberger JJ, Harris M, Diamond D: Acute subdural hematoma: Morbidityand mortality related to timing of operative intervention. J Trauma 30:733–736, 1990.

36. Wilberger JJ, Harris M, Diamond D: Acute subdural hematoma: Morbidity,mortality, and operative timing. J Neurosurg 74:212–218, 1991.

37. Wong CW: Criteria for conservative treatment of supratentorial acute sub-dural haematomas. Acta Neurochir (Wien) 135:38–43, 1995.

38. Yanaka K, Kamezaki T, Yamada T, Takano S, Meguro K, Nose T: Acutesubdural hematoma—Prediction of outcome with a linear discriminantfunction. Neurol Med Chir (Tokyo) 33:552–558, 1993.

39. Zumkeller M, Behrmann R, Heissler H, Dietz H: Computed tomographiccriteria and survival rate for patients with acute subdural hematoma.Neurosurgery 39:708–712, 1996.



TAB

LE1.

Surg

ical

man

agem

ent

ofac

ute

subd

ural

hem

atom

asa

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Cag

etti

etal

.(3

)26

IIIA

llG

CS

All

surg

ery

Earl

ym

orta

lity

Ret

rosp

ectiv

est

udy

of28

patie

nts

betw

een

80an

d10

0yr

ofag

e.2

patie

nts

had

anED

Han

d26

patie

nts

had

aSD

H.

Patie

nts

olde

rth

an80

yrha

da

88%

mor

talit

y.A

ll19

patie

nts

with

aG

CS

betw

een

3an

d9

died

.

GC

SN

o.of

pati

ents

Die

dSu

rviv

ed

13–1

54

22

10–1

25

41

3–9

1919

0

Den

tet

al.

(6)

211

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

GO

Sat

am

ean

of25

3d

Ret

rosp

ectiv

ean

alys

isof

fact

ors

affe

ctin

gou

tcom

ein

211

patie

nts

with

SDH

.83

patie

nts

unde

rwen

tsu

rger

yan

d12

8di

dno

t.A

trea

tmen

tpr

otoc

olw

asno

tsp

ecifi

ed.

Adm

issi

onG

CS,

inju

ryse

veri

tysc

ore,

and

pupi

llary

reac

tivity

wer

ein

depe

nden

tpr

edic

tors

ofou

tcom

ein

the

entir

epa

tient

popu

latio

n.

<4

hfr

omTB

Ito

surg

ery

(%)

>4

h(%

)P

valu

e

GR

/MD

2451

0.02

SD/V

S48

19

D28

30n.

s.

Ope

nci

ster

ns5

190.

0004

Effa

ced

cist

erns

7653

0.00

2

GC

S7.

0po

ints

8.4

poin

tsn.

s.

Has

elsb

erge

ret

al.

(13)

111

IIIA

llG

CS

All

surg

ery

GO

Sat

9m

oA

retr

ospe

ctiv

ere

view

of17

1pa

tient

sw

hopr

esen

ted

with

eith

era

subd

ural

(111

patie

nts)

oran

epid

ural

(60

patie

nts)

hem

atom

a.Th

ein

fluen

ceof

surg

ical

timin

gw

asan

alyz

edin

rela

tion

tom

orta

lity

and

func

tiona

lre

cove

ry.

Patie

nts

with

anac

ute

subd

ural

orep

idur

alhe

mat

oma

had

alo

wer

mor

talit

yan

dim

prov

edfu

nctio

nal

reco

very

whe

nop

erat

edon

�2

haf

ter

onse

tof

com

a.

Tim

efr

omco

ma

onse

tto

surg

ery

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS

(%)

D(%

)

�2

h34

3221

47

�2

h55

416

80

Hat

ashi

taet

al.

(14)

60III

All

GC

SA

llsu

rger

yG

OS

at3

mo

Ret

rosp

ectiv

ean

alys

isof

60pa

tient

san

dth

ein

fluen

ceof

surg

ical

timin

gon

the

outc

ome

from

SDH

.

Patie

nts

oper

ated

onw

ithin

4h

ofin

jury

had

ahi

gher

mor

talit

yco

mpa

red

with

thos

eop

erat

edon

afte

r4

h.Pa

tient

sw

itha

GC

Sbe

twee

n4

and

6w

houn

derw

ent

acr

anio

tom

yha

da

sign

ifica

ntly

bette

rou

tcom

eco

mpa

red

with

thos

ew

how

ere

trea

ted

with

burr

hole

s.

GO

S<

4h

from

TBI

tosu

rger

y,n

�43

(%)

4–1

0h,

n�

17(%

)

GR

/MD

2641

D63

35

How

ard

etal

.(1

5)67

IIIA

llG

CS

All

surg

ery

GO

Sat

2m

oR

etro

spec

tive

anal

ysis

of2

age

grou

ps(1

8–

40yr

and

�65

yr)

ofpa

tient

sw

ithac

ute

SDH

.A

ge�

65yr

,si

zeof

the

hem

atom

a,an

dM

LSw

ere

rela

ted

toou

tcom

e,bu

tno

tin

depe

nden

tly.

Size

ofth

ehe

mat

oma

and

MLS

wer

egr

eate

rin

the

olde

rpa

tient

grou

p.M

orta

lity

inth

eol

der

patie

ntgr

oup

was

74%

,in

the

youn

ger

grou

p,it

was

18%

(P�

0.00

1).

Jam

joom

(16)

27III

All

GC

SA

llsu

rger

yG

OS

at6

mo

Are

view

of27

patie

nts

aged

75yr

orol

der

who

requ

ired

oper

atio

nfo

ran

acut

eSD

H.

No

patie

ntol

der

than

75yr

who

preo

pera

tivel

yw

asex

tens

orpo

stur

ing,

flacc

idto

pain

,or

had

unila

tera

lor

bila

tera

lfix

edan

ddi

late

dpu

pils

mad

ea

good

reco

very

(GO

S3–

5).

Evac

uatio

nof

acut

esu

bdur

alhe

mat

oma

isno

tre

com

men

ded

inth

issu

bgro

upof

patie

nts.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/V(%

)D

(%)

Preo

pera

tive

dete

rior

atio

nto

GC

S�

8

138

92

Pupi

lsre

activ

e15

2727

46

Uni

/bila

tera

lun

reac

tive

pupi

ls12

100

Adm

issi

onG

CS

�4

1921

1663

Adm

issi

onG

CS

�5

813

87



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Koc

etal

.(1

7)11

3III

All

GC

SA

llsu

rger

yG

OS

at3

mo

Rev

iew

of11

3co

nsec

utiv

epa

tient

sof

all

age

grou

psop

erat

edon

for

acut

eSD

H.

Preo

pera

tive

GC

San

dpu

pilla

ryex

amco

rrel

ated

with

func

tiona

lou

tcom

e.A

ssoc

iate

din

trac

ereb

ral

hem

atom

a,ce

rebr

alco

ntus

ion,

and

SAH

wer

eal

sore

late

dto

poor

erou

tcom

e.Ti

me

from

inju

ryto

surg

ery

did

not

affe

ctm

orta

lity.

No.

ofpa

tien

tsG

R(n

o.)

MD

(no.

)SD

(no.

)D

(no.

)

GC

S3–

437

235

b

GC

S5–

628

42

22b

GC

S7–

825

21

1125

GC

S9

–12

1110

1

GC

S13

–15

1211

1

Bila

tera

lpr

esen

tlig

htre

flex

6337

32

21

Uni

late

ral

abse

ntlig

htre

flex

111

19c

Bila

tera

lab

sent

light

refle

x39

138

c

b P�

0.05

,c P

�0.

01

Kot

wic

aan

dB

rzez

insk

i(1

8)20

0III

GC

S�

10A

llsu

rger

yG

OS

at3

mo

Ret

rosp

ectiv

ean

alys

isof

200

adul

tpa

tient

sw

houn

derw

ent

surg

ical

evac

uatio

nof

anac

ute

SDH

.

AG

CS

betw

een

3an

d6,

age

�50

yr,

and

MLS

�1.

5cm

wer

eas

soci

ated

with

poor

outc

ome.

Ther

ew

asno

diffe

renc

ein

outc

ome

betw

een

earl

y(le

ssth

an4

h)an

dla

tesu

rgic

alev

acua

tion.

No.

ofpa

tien

tsG

R(n

o.)

MD

(no.

)SD

(no.

)V

S(n

o.)

D(n

o.)

GC

S3–

616

629

625

799

GC

S7–

934

56

26

15

MLS

�1.

5cm

4115

26

216

MLS

1.5–

3cm

9618

612

1050

MLS

�3

cm63

14

91

4

Age

18–

40yr

7016

512

829

Age

50–

65yr

703

47

353

Kot

wic

aan

dJa

kubo

wsk

i(1

9)27

IIIA

llG

CS

All

surg

ery

Not

docu

men

ted

Ret

rosp

ectiv

ean

alys

isof

27pa

tient

s�

70yr

with

SDH

unde

rgoi

ngsu

rger

y.O

ldag

eis

asso

ciat

edw

itha

high

mor

talit

yfr

omSD

H.

All

17pa

tient

sw

itha

GC

S�

9di

ed.

Mas

saro

etal

.(2

1)12

7III

All

GC

SA

llsu

rger

yG

OS

at18

mo

Ret

rosp

ectiv

ean

alys

isof

127

patie

nts

unde

rgoi

ngsu

rger

yfo

rSD

H.

GC

Sis

the

mos

tim

port

ant

pred

icto

rof

outc

ome.

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

D(%

)

GC

S�

128

100

GC

S9

–12

3534

2046

GC

S�

984

1121

68

Mat

hew

etal

.(2

2)23

IIIG

CS

13–1

5Su

rger

yan

dno

nsur

gica

lG

OS

at3–

8m

oR

etro

spec

tive

anal

ysis

of23

patie

nts

with

SDH

who

wer

ein

itial

lym

anag

edno

nope

rativ

ely.

6pa

tient

sun

derw

ent

dela

yed

surg

ery.

All

patie

nts

had

ago

odou

tcom

eat

3–8

mo.

All

patie

nts

with

acl

ot�

10m

mth

ick

onin

itial

CT

scan

requ

ired

surg

ery.

Saka

set

al.

(24)

22III

Com

atos

eA

llsu

rger

yG

OS

at1

yrA

naly

sis

of40

seve

reTB

Ipa

tient

sw

houn

derw

ent

cran

ioto

my

afte

rde

velo

ping

bila

tera

lfix

edan

ddi

late

dpu

pils

.

Patie

nts

with

SDH

had

asi

gnifi

cant

incr

ease

inm

orta

lity

(64%

)co

mpa

red

with

thos

ew

ithED

H(1

8%).

Patie

nts

who

unde

rwen

tde

laye

d(�

3h)

surg

ery

had

aw

orse

outc

ome.

Tim

efr

ompu

pilla

ryno

nrea

ctiv

ity

tosu

rger

y

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS/

D(%

)

�3

h20

3030

40

�3

h16

2512

63

Serv

adei

etal

.(2

7)65

IIIG

CS

�9

Surg

ery

and

nons

urgi

cal

GO

Sat

6m

oR

etro

spec

tive

anal

ysis

ofco

mat

ose

patie

nts

with

SDH

who

wer

etr

eate

dno

nope

rativ

ely

acco

rdin

gto

pred

efin

edcr

iteri

a.

15of

65co

mat

ose

patie

nts

with

SDH

wer

etr

eate

dno

nope

rativ

ely

and

2pa

tient

sre

quir

edde

laye

dsu

rger

ybe

caus

eof

incr

easi

ngIC

Pan

din

trac

ereb

ral

hem

atom

as.

The

auth

ors

conc

lude

dth

atno

nope

rativ

etr

eatm

ent

can

besa

fely

used

for

sele

cted

com

atos

epa

tient

sw

ithSD

H.

Serv

adei

etal

.(2

8)20

6III

All

GC

SSu

rger

yan

dno

nsur

gica

lG

OS

at6

mo

Pros

pect

ive

anal

ysis

of20

6pa

tient

sof

all

age

grou

pspr

esen

ting

with

anac

ute

SDH

ofat

leas

t5

mm

thic

knes

s.14

8pa

tient

sun

derw

ent

oper

ativ

etr

eatm

ent.

The

initi

alC

Tsc

anid

entif

ies

patie

nts

atri

skfo

run

favo

rabl

eou

tcom

e.H

emat

oma

thic

knes

s,M

LS,

stat

usof

basa

lci

ster

ns,

and

pres

ence

ofSA

Hin

all

patie

nts

(sur

gica

lan

dno

nsur

gica

l)ar

ere

late

dto

outc

ome.

Shig

emor

iet

al.

(30)

15III

GC

S�

9A

llsu

rger

yG

OS

at6

mo

Ret

rosp

ectiv

ean

alys

isof

15pa

tient

sw

ithSD

Hun

derg

oing

deco

mpr

essi

vehe

mic

rani

ecto

my

beca

use

ofra

pid

neur

olog

ical

dete

rior

atio

n.

10pa

tient

sdi

edan

d2

patie

nts

mad

ea

good

reco

very

.G

ood

reco

very

was

only

seen

inpa

tient

sw

ithlo

wpo

stop

erat

ive

ICP.

The

patie

nts

who

mad

ea

good

reco

very

wer

ebo

th9

yrol

d.



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Uza

net

al.

(32)

18III

Com

atos

eA

llsu

rger

yG

OS

at6

mo

Pros

pect

ive

stud

yof

71pa

tient

sw

ithED

H,

SDH

,an

din

trac

ereb

ral

hem

orrh

age,

ofal

lag

egr

oups

oper

ated

onfo

rsi

gns

ofun

cal

hern

iatio

n.

Tim

ing

ofsu

rger

ydi

dno

taf

fect

outc

ome.

GC

Sco

rrel

ated

with

GO

S.

van

den

Bri

nket

al.

(33)

91III

All

GC

SA

llsu

rger

yG

OS

at6

mo

Ret

rosp

ectiv

ean

alys

isof

the

CT

scan

para

met

ers

in91

patie

nts

with

acut

eSD

H.

Vol

ume

ofth

esu

bdur

albl

ood

did

not

corr

elat

ew

ithou

tcom

e.Su

bara

chno

idbl

ood

and

pupi

llary

dysf

unct

ion

wer

eth

eon

lysi

gnifi

cant

para

met

ers

corr

elat

ing

with

apo

orou

tcom

e.

SD/V

S/D

(%)

Bila

tera

lab

norm

alpu

pils

89

GC

Sm

otor

scor

e�

480

GC

Sm

otor

scor

e�

352

Wilb

erge

ret

al.

(35)

101

IIIG

CS

�9

All

surg

ery

GO

Saf

ter

18m

oR

etro

spec

tive

anal

ysis

of10

1co

mat

ose

patie

nts

ofal

lag

egr

oups

with

acut

eSD

H.

Asi

gnifi

cant

incr

ease

inm

orta

lity

was

asso

ciat

edw

ithG

CS

3or

4,ag

egr

eate

rth

an65

yr,

ICP

grea

ter

than

45m

mH

gan

dev

acua

tion

ofhe

mat

oma

�12

haf

ter

inju

ry.

No.

ofpa

tien

tsG

R/M

DD

Pva

lue

Tim

efr

omin

jury

toop

erat

ion

101

280

�26

min

374

�31

min

n.s.

Age

(yr)

No.

ofpa

tien

tsG

R/M

D(%

)D

(%)

Pva

lue

�35

3531

.454

.3n.

s.

35–5

021

23.8

61.9

n.s.

51–

6517

11.8

70.6

n.s.

6528

3.6

82.1

P�

0.01

GC

SN

o.of

pati

ents

%G

R/M

D%

DP

valu

e

319

590

P�

0.05

432

1076

P�

0.05

527

1862

n.s.

6–7

2344

51n.

s.

Won

g(3

7)31

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

GC

Sat

2w

kto

5yr

Ret

rosp

ectiv

ean

alys

isof

31ad

ult

patie

nts

(1ch

ild)

of30

0pa

tient

sw

ithSD

Hw

how

ere

initi

ally

trea

ted

nono

pera

tivel

y.6

patie

nts

requ

ired

dela

yed

surg

ery

for

neur

olog

ical

dete

rior

atio

n.N

odi

ffere

nce

was

foun

din

outc

ome.

MLS

�5

mm

(onl

yin

patie

nts

with

GC

S�

15)

and

thic

knes

sof

the

hem

atom

a�

10m

mon

the

initi

alC

Tsc

anar

esi

gnifi

cant

lyre

late

dto

the

failu

reof

nono

pera

tive

trea

tmen

t.H

emat

oma

volu

me

was

not

pred

ictiv

e.

Yan

aka

etal

.(3

8)17

0III

All

GC

SSu

rger

yan

dno

nsur

gica

lG

OS

at3

mo

Are

tros

pect

ive

stud

yon

170

patie

nts

ofal

lag

egr

oups

with

SDH

adm

itted

duri

ng7

yr,

trea

ted

eith

ersu

rgic

ally

orno

nsur

gica

lly.

Ast

rict

prot

ocol

was

follo

wed

and

patie

nts

with

MLS

�5

mm

unde

rwen

tsu

rger

y.

77pa

tient

sun

derw

ent

surg

ery.

Prog

nost

icin

dica

tors

ofou

tcom

efo

rth

ew

hole

grou

pof

patie

nts

wer

e:G

CS,

pupi

ls,

age,

hem

atom

asi

ze,

MLS

,cl

otth

ickn

ess,

asso

ciat

edco

ntus

ions

,SA

H,

stat

usof

basa

lci

ster

ns,

and

ICP.

Zum

kelle

ret

al.

(39)

174

IIIA

llG

CS

All

surg

ery

Post

oper

ativ

em

orta

lity

Ret

rosp

ectiv

ean

alys

isof

174

patie

nts

oper

ated

onfo

rac

ute

SDH

.SD

Hth

ickn

ess

ofle

ssth

an1

cmor

mid

line

shift

less

than

12m

mar

ew

ell

tole

rate

d,w

ithhi

ghsu

rviv

alra

tes.

Mor

talit

ycl

imbs

stee

ply

whe

nM

LSsu

rpas

ses

clot

thic

knes

s.

aG

CS,

Gla

sgow

Com

aSc

ale;

EDH

,epi

dura

lhem

atom

a;SD

H,s

ubdu

ralh

emat

oma;

GO

S,G

lasg

owou

tcom

esc

ore;

TBI,

trau

mat

icbr

ain

inju

ry;G

R,g

ood

reco

very

;MD

,mod

erat

edi

sabi

lity;

SD,

seve

redi

sabi

lity;

VS,

vege

tativ

est

ate;

D,

deat

h;n.

s.,

not

sign

ifica

nt;

MLS

,m

idlin

esh

ift;

SAH

,su

bara

chno

idhe

mor

rhag

e;IC

P,in

trac

rani

alpr

essu

re;

CT,

com

pute

dto

mog

raph

ic.



CHAPTER 5

SURGICAL MANAGEMENT OF TRAUMATIC

PARENCHYMAL LESIONSM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia








Jack Wilberger, M.D.Department of Neurological Surgery,Allegheny General Hospital,Pittsburgh, Pennsylvania



Indications• Patients with parenchymal mass lesions and signs of progressive neurological

deterioration referable to the lesion, medically refractory intracranial hypertension,or signs of mass effect on computed tomographic (CT) scan should be treatedoperatively.

• Patients with Glasgow Coma Scale (GCS) scores of 6 to 8 with frontal or temporalcontusions greater than 20 cm3 in volume with midline shift of at least 5 mm and/orcisternal compression on CT scan, and patients with any lesion greater than 50 cm3

in volume should be treated operatively.• Patients with parenchymal mass lesions who do not show evidence for neurological

compromise, have controlled intracranial pressure (ICP), and no significant signs ofmass effect on CT scan may be managed nonoperatively with intensive monitoringand serial imaging.

Timing and Methods• Craniotomy with evacuation of mass lesion is recommended for those patients with

focal lesions and the surgical indications listed above, under Indications.• Bifrontal decompressive craniectomy within 48 hours of injury is a treatment option

for patients with diffuse, medically refractory posttraumatic cerebral edema andresultant intracranial hypertension.

• Decompressive procedures, including subtemporal decompression, temporal lo-bectomy, and hemispheric decompressive craniectomy, are treatment options forpatients with refractory intracranial hypertension and diffuse parenchymal injurywith clinical and radiographic evidence for impending transtentorial herniation.

KEY WORDS: Coma, Computed tomographic parameters, Craniotomy, Decompressive craniectomy, Headinjury, Herniation, Intracranial pressure monitoring, Parenchymal mass lesion, Surgical technique, Timingof surgery, Traumatic brain injury

Neurosurgery 58:S2-25-S2-46, 2006 DOI: 10.1227/01.NEU.0000210365.36914.E3 www.neurosurgery-online.com

OVERVIEW

Traumatic parenchymal mass lesions arecommon sequelae of traumatic brain injury(TBI), occurring in up to 8.2% of all TBI (37)and 13 to 35% of severe TBI, (6, 35, 44, 47, 57,58) and comprising as much as 20% of opera-tive intracranial lesions in representative se-ries (44, 68). Most small parenchymal lesionsdo not require surgical evacuation (18, 19, 57).However, the development of mass effect fromlarger lesions may result in secondary brain in-jury, placing the patient at risk of further neu-

rological deterioration, herniation, and death(6). Because parenchymal lesions tend to evolve,(36, 54, 55, 58) and because timing of surgerywith respect to the occurrence of neurologicaldeterioration clearly affects outcome (41), mucheffort has been directed at defining patients atrisk of progressive neurological compromise asa result of their traumatic injuries (6, 50). Beckeret al. (3) advocated early evacuation of traumaticintracranial hematomas and contusions for thepurpose of avoiding secondary complications.A selective approach was envisioned by Gall-braith and Teasdale (18), who stated, “if those


who are going to deteriorate could be identified soon after thehematoma has been detected they could be operated on imme-diately without incurring the risks of delay, and the remainderwould be spared unnecessary operation.” The literature address-ing these issues in the CT era is reviewed here in an attempt todefine appropriate surgical indications, methods, and timing forthe patient with traumatic parenchymal injuries.

Although the majority of relevant data retrospectively ad-dresses prognostic variables, differential classification, andclinical course of parenchymal lesions, several studies attemptto assess the efficacy of surgical intervention in a retrospective,historically controlled fashion. Only one randomized, con-trolled trial has been published (61). In addition, the topics ofdelayed traumatic intracerebral hematoma (DTICH) and de-compressive operations for diffuse parenchymal injury andpersistent intracranial hypertension deserve specific consider-ation, and are also addressed here.

PROCESS

A MEDLINE computer search using the following keywords: “intracerebral” or “intraparenchymal” or “ICH” or“IPH and “hematoma” or “hematoma” or “hemorrhage” and“surgery” or “craniotomy” or “craniectomy” or “burr hole” or“craniostomy” and “trauma” or “traumatic” or “TBI” or“CHI” between 1975 and 2001 and limited to humans wasperformed. A total of 330 documents were found. An addi-tional search using the following key words: “brain” or “cor-tex” and “laceration” between 1975 and 2001 was performed,yielding 101 additional articles. This search was narrowed toinclude the following key words: “surgery” or “operative” or“craniotomy” or “craniectomy” or “decompressive craniec-tomy” or “repair” and “outcome,” yielding 49 articles. A thirdsearch using the following key words: “contusion” or “hem-orrhagic contusion” and “brain” and “surgery” or “craniot-omy” or “craniectomy” or “burr hole” or “craniostomy” wasperformed, yielding 174 articles. A fourth search, using thekey words “DTICH” or “DTIPH,” yielded 11 articles. A fifthsearch, using the key word “TICH,” yielded eight articles. Allfive searches were combined. Duplicates between searcheswere discarded. A total of 495 references resulted. In addition,the reference lists of selected articles were reviewed, and atotal of 51 articles were selected for critical analysis. Theresults of this analysis were incorporated into the reviewpresented here. Papers primarily addressing the followingtopics were not included: nontraumatic lesions (e.g., sponta-neous intraparenchymal hemorrhage or infarction), patientswith other associated nontraumatic lesions (e.g., tumors orarteriovenous malformations), posttraumatic aneurysms,chronic lesions, penetrating trauma, carotid-cavernous fistu-lae, patients undergoing anticoagulation therapy, patientswith associated illnesses (e.g., acquired immunodeficiencysyndrome; concomitant infection; hemophilia; thromboticthrombocytopenic purpura; or hemolysis, elevated liver en-zymes, and low platelet count), pregnant patients, birthtrauma, traumatic intraventricular hemorrhage, traumatic hy-

drocephalus, external ventricular drainage, sagittal sinus in-jury, pre-CT era reports, and book chapters. Papers with thefollowing characteristics were also excluded: case series withfewer than 10 patients evaluated by CT scan and with incom-plete outcome data (mortality or Glasgow outcome score[GOS]), case reports, operative series with operations occur-ring greater than 14 days from injury. Posterior fossa paren-chymal injuries are addressed in Surgical Management of Pos-terior Fossa Mass Lesions. Selected articles were evaluated fordesign, prognostic significance, therapeutic efficacy, and over-all outcome. In addition, several articles were reviewed for thepurposes of historical perspective.


Traumatic Parenchymal Lesions

Traumatic parenchymal lesions are a heterogeneous group,and are traditionally divided into focal and nonfocal lesions.Focal lesions include intracerebral hematomas (ICH), DTICH,contusions, and infarctions. Nonfocal lesions include cerebraledema, hemispheric swelling, and diffuse injury. Althoughthese lesions often do not occur in isolation, their presence hasbeen shown to adversely affect prognosis (16, 44). Indeed,Fearnside et al. (16) prospectively collected data on 315 se-verely head-injured patients and found that “the model whichprovided the most accurate prediction of poor outcome in-cluded age, hypotension and three different CT characteristics:subarachnoid blood, ICH or intracerebral contusion (accuracy72.5%).” Parenchymal lesions have been further subclassifiedby multiple authors, and outcome has clearly been shown todiffer among lesion types (19, 35, 39–41, 44, 65). Marshall (39)demonstrated that CT-defined injury type was a highly sig-nificant independent predictor of mortality, even when ageand GCS motor score were included in the predictive model.Given the heterogeneity of the pathophysiology and prognos-tic significance of “parenchymal” lesions, the task of definingclear surgical indications and methods becomes difficult.

Prognostic Factors

Despite proven differences among lesion types, outcomewithin the broad category of “parenchymal” lesions correlateswith known prognostic variables of TBI in general (5). Theseinclude age (26, 44, 45, 56, 58, 66, 70), admission or postresus-citation GCS (6, 19, 23, 40, 44, 45, 49, 51, 58), presence of cranialfracture (56), presence of pupillary response/brainstem re-flexes (7, 44), respiratory insufficiency (8), ICP (6, 7, 23, 39, 44,50, 51), and the status of the basal cisterns (6, 41, 62) or thirdventricle (6, 41, 62) on CT scan. Moreover, other variables signif-icantly correlate with outcome. These include location of thelesion (2, 32, 50, 58), ICH volume (8, 63), GCS at time of follow-upCT (63), lowest recorded GCS (7), severity of surrounding edema(6), timing of surgery (41, 51, 53), occurrence of preoperativeneurological deterioration (41), and presence of acute hemi-spheric swelling or concomitant subdural hematoma (7, 35). Al-though their study included nontraumatic lesions, Andrews et al.



(2) showed that patients with a temporal or temporoparietal ICHof 30 cm3 or greater, as defined by product of anteroposterior,mediolateral, and superoinferior diameters on CT scan, weresignificantly more likely to develop signs of brainstem compres-sion or tentorial herniation, implying that these patients shouldundergo early evacuation of the offending mass lesion. However,these prognostic variables alone do not define the patient whoshould undergo operative intervention.

Operative Indications

As stated above, in Overview, several studies have focusedon defining the patient at risk for subsequent neurologicaldeterioration, making the assumption that operative interven-tion for this patient will improve the likelihood of a morefavorable outcome. Predictors of failure of nonoperative man-agement (defined by subsequent neurological deteriorationand need for craniotomy) include lesion location (50), intra-cranial hypertension (6, 18, 50), presence of subarachnoidhemorrhage (41), cisternal effacement (41), lesion volume (41),and hypoxic events (41).

Bullock et al. (6) prospectively studied 85 patients with ICHwhose initial need for craniotomy was uncertain. These pa-tients underwent ICP monitoring in an attempt to better de-fine the need for surgical intervention. The authors then ret-rospectively reviewed the CT scans of those patients for whomICP monitoring failed to predict late deterioration and, thus,the need for ICH evacuation. With multiple linear regressionanalysis, they found the peak ICP to be the strongest predictorof outcome. However, ICP monitoring failed to predict latedeterioration or death secondary to high ICP in 5 of 30 patientswho did not undergo initial surgery. After critical analysis ofCT factors, they concluded that the decision to operate“should be based on a spectrum of clinical, CT scanning andICP findings”. CT and clinical predictors included cisternalstatus, edema severity, and admission GCS. Interestingly, theauthors found that the weight of each predictor depended onthe location of the ICH. For temporoparietal lesions, hema-toma size, degree of edema, GCS, basal cistern status, and ICPdata correlated with outcome. However, for frontal lesions,peak ICP alone was predictive of outcome. These findingsexpand on those reported by Gallbraith and Teasdale (18),who found that all patients with “intradural” lesions (includ-ing subdural hematoma, ICH, and “burst lobes”) and sus-tained ICP greater than 30 mm Hg, versus only one patientwith an ICP less than 20 mm Hg, required operative interven-tion, as defined by clinical deterioration or failure to improvein the setting of increased ICP.

Mathiesen et al. (41) reviewed data collected prospectively forthe Head Injury Trial-2 nimodipine trial on 218 TBI patients notobeying commands within 24 hours of injury. These authorsfound that the initial CT characteristics of presence of subarach-noid hemorrhage, presence of focal lesion with volume greaterthan 40 cm3, and compressed or absent cisterns were associatedwith neurological deterioration, defined as a fall in GCS by 2points or from 4 to 3, or as the development of pupillary dilation.

They also found that the incidence of secondary (greater than 5 dafter injury) deterioration was associated with the occurrence ofhypoxic events. The occurrence of neurological deterioration, inturn, was found to adversely affect outcome from craniotomy,suggesting that patients with factors strongly associated withneurological deterioration should be considered for early surgery(i.e., before the onset of neurological deterioration). The authorsdirectly demonstrated that a subgroup of patients with admis-sion GCS of at least 6 and focal lesion volume of at least 20 cm3

who underwent surgery without previous neurological deterio-ration had significantly better outcomes compared with thosewho either did not undergo surgery or who underwent surgeryafter deterioration. Furthermore, if a radiological sign of masseffect (i.e., compression or obliteration of the cisterns and/or amidline shift � 5 mm) was present, craniotomy significantlyimproved the outcome in this group. Craniotomy was also di-rectly shown to improve outcome in a small subgroup of patientswith admission GCS of at least 10, temporal contusions, and aradiological sign of mass effect (i.e., a midline shift and/or com-pression or obliteration of the basal cisterns. Additionally, pa-tients admitted with a GCS of at least 6 and a lesion volume of atleast 50 cm3 had better outcome with surgery before or immedi-ately after deterioration than without surgery or with delayedoperation.

Although the goal of identifying those patients who arelikely to deteriorate neurologically is paramount, the questionof whether surgery itself is beneficial remains unanswered.Few studies compare surgical outcome with matched, nonsur-gically managed controls. In a retrospective study of 21 pa-tients with frontal lobe contusions and medically intractableintracranial hypertension (�40 mm Hg), mortality was signif-icantly decreased in the surgical group compared with a non-surgical historical group (22% versus 88%, respectively) (28).These patients were matched for age, sex, GCS, and ICP levels,although statistics for these variables are not provided by theauthors. Choksey et al. (8) retrospectively reviewed 202 pa-tients with traumatic ICH and showed, with logistic regres-sion analysis, that craniotomy significantly improved theprobability of good outcome. Factors taken into considerationfor this analysis included low GCS and hematoma volumegreater than 16 cm3, each of which independently predictedpoor outcome in these patients. Several other studies examineoutcome relative to specific decompressive procedures, andare discussed in the surgical treatment section.

For the purposes of CT classification, Marshall (39) defineda �mass lesion� as a lesion of volume greater than 25 cm3. Theyshowed differential outcome between patients with evacuatedand nonevacuated mass lesions (23% versus 11% favorableoutcome, respectively) in a series of 746 severe TBI patients(i.e., after resuscitation, GCS � 8). In contrast, a recent paperfrom the European Brain Injury Consortium (54) evaluating aseries of 724 TBI patients with a GCS of 3 to 12 showed a 45%rate of favorable results in evacuated mass lesions versus 42%in nonevacuated mass lesions using the same classificationsystem. Sample size between these two studies was noticeablydifferent: the former series included 276 patients with evacu-

TRAUMATIC PARENCHYMAL LESIONS


ated mass lesions and 36 with nonevacuated mass lesions,whereas the latter included 195 and 148 patients, respectively.These studies reviewed illustrate that a classification systembased solely on lesion volume is unable to consistently showthe relationship between surgery and outcome. Surgical indi-cations are, in fact, related to many factors, including CTparameters (i.e., volume, midline shift, and basal cistern com-pression), clinical status, and the occurrence of clinical deteri-oration, among others.

One fundamental problem in using initial CT parameters asindependent indications for surgery is that CT pathology hasclearly been shown to be a dynamic process. Using the TraumaticComa Data Bank classification system (39), Lobato et al. (36)showed that 51.2% of 587 severely injured patients (GCS � 8)developed significant changes between initial and “control” CTscans, the latter of which more accurately predicted outcome.Similarly, Servadei et al. (54) showed that 16% of moderately-to-severely injured patients (GCS of 3–12) with diffuse injuryshowed radiological progression across Traumatic Coma DataBank classes. Yamaki et al. (69) showed that only 56% of ICHgreater than 3 cm in diameter developed within 6 hours of injury,and that only 84% of ICH reached maximal size within 12 hours.These studies highlight the dynamic nature of parenchymal in-juries and the dangers inherent in placing too much emphasis ona single, static CT scan for management decisions.

The data reviewed shows that there are subpopulations ofpatients with traumatic intraparenchymal lesions that willbenefit from surgical intervention. However, the precise char-acteristics of these subpopulations are not, as yet, clearly de-fined. The literature supports taking into account an amalgamof clinical and radiographic criteria, including GCS, location,volume, CT appearance, ICP, and the presence of neurologicaldeterioration, to make an informed decision to subject a pa-tient with a parenchymal lesion to a craniotomy. It seems thatall factors must be taken into consideration to best define thepatient population that will benefit from surgery.

DTICH

ICH have been shown to evolve over time (55, 58, 69). Theentity of DTICH was initially described by Bollinger in 1891(4), and is now defined by most authors as occurring in areasof radiographically normal brain in patients with otherwiseabnormal initial CT scans (20, 22, 59, 63). It is defined byGentleman et al. (20) as a �lesion of increased attenuationdeveloping after admission to hospital, in a part of the brainwhich the admission CT scan had suggested was normal”.Other authors, however, have noted DTICH to occur in areasof contusion on initial, high-resolution CT scan (72). The inci-dence of DTICH ranges from 3.3 to 7.4% in patients withmoderate-to-severe TBI (15, 20, 22, 29, 59, 63). EvacuatedDTICH represent approximately 1.6% of all evacuated trau-matic ICH (20), and mortality ranges from 16 to 72% (15, 20,22, 59, 63) Therefore, the importance of careful monitoring andof serial CT scanning cannot be overemphasized (55, 63, 72).

In a retrospective review of 32 patients with DTICH, Tseng(63) found that greater lesion volume, cisternal compression,earlier timing of appearance, occurrence of clinical deteriora-tion, and lower GCS at time of a second, follow-up CT ad-versely affected outcome. Mortality occurred only with clini-cal presentation of DTICH within 48 hours of injury. In thissmall series, no patient with DTICH requiring craniotomypresented after 72 hours of injury. Sprick et al. (59) similarlyfound that approximately 70% of clinically significant DTICHpresented within 48 hours of injury. Additionally, DTICH hasbeen shown to be significantly associated with an increasedincidence of secondary, systemic insults (22), an increasedincidence after decompressive surgery for other mass lesions(22), and an increased incidence of abnormal clotting param-eters (29), suggesting a complex etiology for this lesion beyondthe mechanical disruption of the parenchyma (22, 27).

Although DTICH is most likely a distinct pathophysiologi-cal entity, it is still a parenchymal lesion from which manypatients either fail to recover or clinically deteriorate. Thefindings of Mathiesen et al. (41) indicate that patients withintracerebral lesions undergoing surgery before neurologicaldeterioration have improved outcomes. It follows logicallythat a subset of DTICH patients would benefit from rapidsurgical intervention after discovery of the lesion. However noCT-era studies have critically examined surgical outcome. Be-cause the majority of studies show that all patients who de-velop clinically relevant DTICH have abnormal initial CTscans (20, 63, 72), it is essential that patients with initiallyabnormal scans undergo intensive monitoring and serial im-aging to ensure rapid intervention, if necessary.

Surgical Methods

The standard surgical treatment of focal lesions, such asintracerebral hemorrhages or contusions, is craniotomy withevacuation of the lesion. Location of the lesion and proximityto critical structures are considerations when contemplatingthe choice of surgical options. Evacuation of traumatic masslesions is often effective in amelioration of brain shift andreduction of ICP, and can decrease the requirement for inten-sive medical treatment. Other methods, such as stereotacticevacuation of focal mass lesions, have also been used, al-though much less commonly (12). These procedures, however,become less effective when the patient’s intracranial pathol-ogy is diffuse and involves intracranial hypertension as aresult of posttraumatic edema or hemispheric swelling—fac-tors known to be associated with poor outcome (6, 19, 35, 38,44, 51). Even with focal ICH, a significant proportion of pa-tients have medically intractable intracranial hypertension af-ter standard craniotomy, and these patients fare the worst (44).

Surgical Treatment of Intracranial Hypertension

Rationale

A variety of operations have been developed for, or applied to,decompression of the brain at risk for the sequelae of traumati-



cally elevated ICP. These include Cushing’s subtemporal decom-pression (13), temporal lobectomy (34, 48), manual reduction ofthe temporal lobe (64), circumferential craniotomy (9), and themore widely used hemispheric decompressive craniectomy (52)and bifrontal decompressive craniectomy (30). The rationale forsuch an approach is supported from a physiological perspectiveby both human and animal studies. Particular attention has beendirected at the study of changes in ICP and CT scan features,such as cisternal effacement and midline shift, after decompres-sion, with the goal of correlating significant postoperative reduc-tion of these parameters with improved outcome.

Hatashita and Hoff (25) showed that decompressive fronto-parietal craniectomy in cats led to significant reduction in ICP,reduction in cortical gray and white matter tissue pressure, in-creased pressure-volume index, and increased tissue compliance.The authors found that craniectomy significantly increased thevolumetric compensatory capacity of the intracranial cavity, afinding consistent with those of Hase et al. (24), in which ven-tricular catheter ICP measurements in 47 severe TBI patients, 33of whom underwent external decompression, documented a dra-matic increase in intracranial compliance after decompression.Yoo et al. (71) studied intraoperative ventricular pressures in acohort of 20 patients with refractory intracranial hypertensionafter both traumatic and nontraumatic insults who underwentbilateral frontotemporoparietal decompressive craniectomy withdural expansion and grafting. These authors found a 50.2 �16.6% reduction of initial ICP after craniectomy, and a furtherreduction to 15.7 � 10.7% of initial ICP after dural opening.

Polin et al. (51) showed a significant decrease in ICP afterbifrontal decompressive craniectomy, as well as a significantdifference in postoperative ICP when compared with ICPmeasured 48 to 72 hours after injury in a cohort of historicallymatched controls. Gower et al. (21) found that 7 of 10 patientswho underwent subtemporal decompression for medically re-fractory intracranial hypertension had an average decrease inICP of 34%. Kunze et al. (33) found a reduction in mean ICP from41.7 to 20.6 mm Hg in 28 patients after unilateral or bilateraldecompressive craniectomy for posttraumatic edema refractoryto maximal medical therapy. And Whitfield et al. (67) demon-strated a significant reduction of ICP from 37.5 to 18.1 mm Hg (P� 0.003) in 26 patients who underwent bifrontal decompressivecraniectomy for refractory intracranial hypertension managedusing a standardized ICP/cerebral perfusion pressure (CPP)treatment algorithm. The amplitude of the ICP waveform and ofslow waves were significantly reduced, and compensatory re-serve was significantly increased postoperatively in a subgroupof eight patients in this study. Munch et al. (45), however, failedto show a significant postoperative reduction in ICP or increasein CPP after unilateral hemispheric decompression. From a ra-diological perspective, however, this group found a significantimprovement in the visibility of mesencephalic cisterns and asignificant decrease in midline shift after decompressive craniec-tomy—both known to correlate with improved outcome. Fur-thermore, the change in cistern visibility correlated with thedistance between the lower craniectomy border and cranial base.Additionally, Alexander et al. (1) found an average increase in

intracranial volume of 26 cm3 in an analysis of CT scans inpatients with traumatic and nontraumatic lesions undergoingsubtemporal decompression. In contrast, experimental evidencefrom Cooper et al. (10) supports the notion that decompressiveprocedures may aggravate cerebral edema formation, thus, re-sulting in increased secondary injury. Such studies may helpexplain why, despite laboratory success, actual improvement inpatient outcome has not been consistently demonstrated.

Procedures and Outcome

There are several studies that suggest an important role fordecompressive procedures in the management of parenchy-mal injury. In a retrospective review of 28 patients undergoingunilateral or bilateral decompressive craniectomy for posttrau-matic edema and intracranial hypertension refractory to headelevation, moderate hyperventilation, osmodiuretics, barbitu-rates, tromethamine, and cerebrospinal fluid drainage, 57% ofpatients had good outcome or moderate disability at 1 year.However, the authors excluded patients with “vast” primarylesions, hypoxia, ischemic infarction, brainstem injury, “central”herniation, and primary anisocoria, thus, biasing results towardfavorable recovery (33). Nussbaum et al. (48) showed that com-plete temporal lobectomy performed within 2 hours of the de-velopment of clinical signs of transtentorial herniation in 10patients with unilateral hemispheric swelling and a GCS of lessthan 7 resulted in 40% functional independence. All patientsdemonstrated displacement of the brainstem, compression of thecontralateral peduncle, and progressive obliteration of the para-sellar and interpeduncular cisterns on CT scan, along with fixedpupillary dilation, and therefore, represent a particularly com-promised patient population.

Guerra et al. (23) prospectively performed decompressivecraniectomy following a standardized treatment protocol withstandardized surgical technique for posttraumatic diffusebrain swelling in 57 severe TBI patients (GCS, 4–6). Thirty-nine of these patients underwent primary decompression, and18 others underwent secondary decompression because ofpersistent intracranial hypertension after evacuation of a sur-gical mass lesion. Intracranial hypertension in these patientswas refractory to a standardized medical protocol that in-cluded hemodynamic stabilization, head elevation, sedationwith or without muscle relaxation, controlled hyperventilationto an arterial carbon dioxide pressure of 28 to 32 mm Hg,mannitol, tromethamine for acute rises in ICP, and electroen-cephalogram burst suppression with barbiturates. Fifty-eightpercent of the first group and 65% of the second group expe-rienced good outcome or moderate disability at 1 year. Theseresults compare favorably with published outcomes of alter-native second-tier therapy. However, a direct comparisonwith matched nonsurgical controls was not performed. Whit-field et al. (67) demonstrated a favorable outcome (GOS, 4–5)in 61% of a severe TBI population that underwent bifrontaldecompressive craniectomy for ICP greater than 30 mm Hgwith CPP less than 70 mm Hg, or for ICP greater than 35 mmHg, irrespective of CPP, despite a medical management pro-



tocol that involved head elevation, propofol sedation, manni-tol and/or hypertonic saline, normocarbia, mild hypothermiato 33°C to 35°C, and electroencephalogram burst suppression.Only 55% of eligible patients, however, underwent craniec-tomy, and the reasons for nonsurgical management despiteeligibility were not documented or discussed.

Munch et al. (45) were able to show a significant differencein outcome between patients undergoing rapid decompressivecraniectomy and those undergoing delayed decompressivecraniectomy. However, the nature of the lesions differed be-tween the two groups, and, thus, the outcomes cannot bemeaningfully compared. Tseng (64) noted that the addition ofgentle elevation of the temporal lobe until the tentorium wasvisualized and CSF egress was noted decreased mortality andimproved the incidence of good outcome when comparedwith a standard craniotomy with hematoma evacuation andcontusion resection in a series of 32 severe TBI patients withanisocoria, hemiparesis, and CT evidence for uncal herniation.However, no statistical analysis was performed, and the au-thor notes that preoperative selection was biased and thatoperative timing and intraoperative judgements may haveinfluenced the choice of surgical procedure. Gower et al. (21)retrospectively studied 115 patients with admission GCS of 8or less who had adequate ICP monitoring data and no oper-ative mass lesion on admission. These patients were managedunder a standard treatment protocol. Outcome was comparedbetween 10 patients who underwent subtemporal decompres-sion and 17 patients managed by induction of pentobarbitalcoma. They found that subtemporal decompression affordedsignificantly lower mortality than pentobarbital coma, despitethe fact that the surgical group had a lower (but not statisti-cally significant) average admission GCS. However, 10 of thepatients treated medically had been determined not to beoperative candidates and subsequently died, greatly biasingresults in favor of the operative group. In a retrospectivereview of 29 patients undergoing operation for a combinationof acute subdural hematoma and severe contusion and swell-ing of the temporal lobe with uncal herniation, Lee et al. (34)documented a significant improvement in outcome with theaddition of temporal lobectomy to subtemporal decompres-sion and debridement of contused brain. Mortality decreasedfrom 56% to 8%, with a concomitant increase in average GOSfrom 2.2. to 4.0. These two surgical groups did not differ withrespect to preoperative GCS, age, or sex. However, patientswith intraoperative “overswelling,” defined as herniation ofbrain more than 2 cm above the craniectomy window, wereexcluded and underwent decompressive craniectomy withdural expansion, thus, potentially biasing these results.

Coplin et al. (11) retrospectively reviewed 29 consecutivepatients with GCS of at most 9 and CT scans with a midlineshift greater than the volume of a surgically amenable lesion toevaluate the safety and feasibility of decompressive craniec-tomy with duraplasty versus traditional craniotomy as theinitial surgical procedure. No significant differences in age,gender, admission GCS, time to surgery, or serum ethanolconcentration existed between the two groups. Despite a sig-

nificantly lower GCS at time of surgery and significantlygreater percentage of Diffuse Injury III and IV injuries (39),there was no significant difference in mortality, GOS, acutehospital length of stay, functional independence measurescore on admission to a rehabilitation unit, change in func-tional independence measure score, or length of rehabilitationstay between craniectomy and craniotomy groups. Althoughthis study is subject to the biases inherent in a retrospective,uncontrolled design, it strongly supports the safety of decom-pressive craniectomy as a first-line surgical intervention asopposed to its traditional role as a salvage procedure.

The studies briefly outlined here support the potential use-fulness of decompressive procedures, but are clearly ham-pered by inherent biases.

The study by Polin et al. (51) deserves particular mention,because it offers more concrete evidence that a decompressiveprocedure may result in improved patient outcome. Theseauthors report a retrospective evaluation of outcome in 35patients undergoing bifrontal decompressive craniectomy forrefractory posttraumatic cerebral edema matched for age, ad-mission GCS, sex, and maximal ICP with historical controlsselected from the Traumatic Coma Data Bank. Only patientswith Diffuse Injury III (39) were eligible as controls. Highestpostoperative ICP less than 24 hours after surgery in cases wasmatched with highest ICP 48 to 72 hours after injury in con-trols. Preoperative ICP in cases was matched with highest ICPless than 48 hours after injury in controls. Several findings areparticularly relevant. In the operative group, surgery per-formed less than 48 hours after injury was significantly asso-ciated with favorable outcome when compared with surgeryperformed longer than 48 hours after injury (46% versus 0%,respectively). Medical management alone carried a 3.8 timesrelative risk of unfavorable outcome compared with decom-pressive craniectomy. Maximum benefit was achieved in pa-tients undergoing decompression within 48 hours of injuryand whose ICP elevations had not yet been sustained above 40mm Hg (60% favorable outcome versus 18% in matched con-trols). This study argues strongly in favor of bifrontal decom-pressive craniectomy for patients with medically refractoryposttraumatic cerebral edema and resultant intracranial hy-pertension not yet sustained above 40 mm Hg within 48 hoursof injury, but does not have contemporaneous controls (51).

Overall, the literature suggests, but does not prove, that de-compressive procedures may be the intervention of choice giventhe appropriate clinical context. A recent study by Taylor et al.(61) examined the use of early (median 19.2 h after injury) bitem-poral craniectomy in addition to intensive medical managementversus intensive medical management alone in 27 children withsustained intracranial hypertension in a prospective, random-ized, controlled fashion. Their results showed a trend towardsgreater improvement in ICP, less time required in the intensivecare unit, and improved outcome with surgical decompression.These trends, although promising, did not reach statistical sig-nificance. Additional prospective, controlled studies are, thus,needed to strengthen the argument for the use of surgical de-



compression in the management of intracranial hypertensionand refractory cerebral edema.

SUMMARY

The majority of studies regarding surgical treatment ofparenchymal lesions are case series. Only one prospectiveclinical trial of treatment using surgical versus nonsurgicalmanagement has been published (61). The majority of evi-dence indicates that the development of parenchymal masslesions, which are associated with progressive neurologicaldysfunction, medically refractory intracranial hypertension, orradiological signs of mass effect, are associated with a pooroutcome if treated nonsurgically. Specific surgical criteria,however, have not been firmly established.

Evidence also suggests that decompressive craniectomymay be the procedure of choice in patients with posttraumaticedema, hemispheric swelling, or diffuse injury, given the ap-propriate clinical context. This context has yet to be defined.


The majority of studies on traumatic parenchymal lesions isobservational, and the studies offer no means to meaningfullycompare outcome between surgical and nonsurgical groups.Those studies that attempt this comparison fail to adequatelycontrol for known prognostic variables between surgicallyand nonsurgically managed groups in a prospective fashion.Prospective, controlled trials, such as that of Taylor et al. (61)need to be pursued and supported to define appropriate clin-ical criteria and surgical methods.

Additionally, the data of Coplin et al. (11) and Taylor et al.(61) strongly support the feasibility of performing trials ofdecompressive operations in the first instance, as opposed to asecond- or third-tier therapy for posttraumatic intracranialhypertension.

REFERENCES

1. Alexander E, Ball M, Laster D: Subtemporal decompression: Radiologicalobservations and current surgical experience. Br J Neurosurg 1:427–433,1987.

2. Andrews BT, Chiles BW, Olsen WL, Pitts LH: The effect of intracerebralhematoma location on the risk of brain-stem compression and on clinicaloutcome. J Neurosurg 69:518–522, 1988.

3. Becker D, Miller J, Ward J, Greenberg R, Young H, Sakalas R: The outcomefrom severe head injury with early diagnosis and intensive management.J Neurosurg 47:491–502, 1977.

4. Bollinger O: Uber traumatische Spat-Apoplexie: Ein Beitrag zur Lehre derHirnerschutterung, in Internationale Beitrage zur Wissenschaftlichen Medizin.Berlin: Festschrift, Rudolf Virchow, A Hirsrschwald, 1891, vol 2, pp 457–470.

5. Brain Trauma Foundation: Early indicators of prognosis in severe traumaticbrain injury. J Neurotrauma 17:535–627, 2000.

6. Bullock R, Golek J, Blake G: Traumatic intracerebral hematoma—Whichpatients should undergo surgical evacuation? CT scan features and ICPmonitoring as a basis for decision making. Surg Neurol 32:181–187, 1989.

7. Caroli M, Locatelli M, Campanella R, Balbi S, Martinelli F, Arienta C:Multiple intracranial lesions in head injury: Clinical considerations, prog-nostic factors, management, and results in 95 patients. Surg Neurol 56:82–88, 2001.

8. Choksey M, Crockard HA, Sandilands M: Acute traumatic intracerebralhaematomas: Determinants of outcome in a retrospective series of 202 cases.Br J Neurosurg 7:611–622, 1993.

9. Clark K, Nash TM, Hutchison GC: The failure of circumferential craniotomyin acute traumatic cerebral swelling. J Neurosurg 29:367–371, 1968.

10. Cooper PR, Hagler H, Clark WK, Barnett P: Enhancement of experimentalcerebral edema after decompressive craniectomy: Implications for the man-agement of severe head injuries. Neurosurgery 4:296–300, 1979.

11. Coplin WM, Cullen NK, Policherla PN, Vinas FC, Wilseck JM, Zafonte RD,Rengachary SS: Safety and feasibility of craniectomy with duraplasty as theinitial surgical intervention for severe traumatic brain injury. J Trauma50:1050–1059, 2001.

12. Coraddu M, Floris F, Nurchi G, Meleddu V, Lobina G, Marcucci M: Evac-uation of traumatic intracerebral haematomas using a simplified stereotacticprocedure. Acta Neurochir (Wien) 129:6–10, 1994.

13. Cushing H: Subtemporal decompressive operations for the intracranial com-plications associated with bursting fractures of the skull. Ann Surg 47:641–644, 1908.

14. De Luca GP, Volpin L, Fornezza U, Cervellini P, Zanusso M, Casentini L,Curri D, Piacentino M, Bozzato G, Colombo F: The role of decompressivecraniectomy in the treatment of uncontrollable posttraumatic intracranialhypertension. Acta Neurochir 76 [Suppl]:401–404, 2000.

15. Diaz FG, Yock DH, Larson D, Rockswold GL: Early diagnosis of delayedposttraumatic intracerebral hematomas. J Neurosurg 50:217–223, 1979.

16. Fearnside MR, Cook RJ, McDougall P, McNeil RJ: The Westmead HeadInjury Project outcome in severe head injury. A comparative analysis ofpre-hospital, clinical and CT variables. Br J Neurosurg 7:267–279, 1993.

17. Gaab M, Rittierodt M, Lorenz M, Heissler H: Traumatic brain swelling andoperative decompression: A prospective investigation. Acta Neurochir(Wien) 51 [Suppl]:326–328, 1990.

18. Gallbraith S, Teasdale G: Predicting the need for operation in the patientwith an occult traumatic intracranial hematoma. J Neurosurg 55:75–81, 1981.


20. Gentleman D, Nath F, Macpherson P: Diagnosis and management of de-layed traumatic intracerebral haematomas. Br J Neurosurg 3:367–372, 1989.

21. Gower D, Lee K, McWhorter J: Role of subtemporal decompression in severeclosed head injury. Neurosurgery 23:417–422, 1988.

22. Gudeman S, Kishore P, Miller J, Girevendulis A, Lipper M, Becker D: Thegenesis and significance of delayed traumatic intracerebral hematoma.Neurosurgery 5:309–313, 1979.

23. Guerra W, Gaab M, Dietz H, Mueller J, Piek J, Fritsch M: Surgical decom-pression for traumatic brain swelling: Indications and results. J Neurosurg90:187–196, 1999.

24. Hase U, Reulen HJ, Meinig G, Schurmann K: The influence of the decom-pressive operation on the intracranial pressure and the pressure-volumerelation in patients with severe head injuries. Acta Neurochir (Wien) 45:1–13, 1978.

25. Hatashita S, Hoff JT: The effect of craniectomy on the biomechanics ofnormal brain. J Neurosurg 67:573–578, 1987.

26. Jamjoom A: The influence of concomitant intradural pathology on thepresentation and outcome of patients with acute traumatic extraduralhaematoma. Acta Neurochir (Wien) 115:86–89, 1992.

27. Katayama Y, Tsubokawa T, Miyazaki S: Two types of delayed traumaticintracerebral hematoma: Differential forms of treatment. Neurosurg Rev12:231–236, 1989.

28. Katayama Y, Tsubokawa T, Miyazaki S, Kawamata T, Yoshino A: Oedemafluid formation within contused brain tissue as a cause of medically uncon-trollable elevation of intracranial pressure: The role of surgical therapy. ActaNeurochir (Wien) 51 [Suppl]:308–310, 1990.



29. Kaufman HH, Moake JL, Olson JD, Miner ME, duCret, RP, Pruessner JL,Gildenberg PL: Delayed and recurrent intracranial hematomas related todisseminated intravascular clotting and fibrinolysis in head injury.Neurosurgery 7:445–449, 1980.

30. Kjellberg RN, Prieto A: Bifrontal decompressive craniotomy for massivecerebral edema. J Neurosurg 34:488–493, 1971.

31. Kotwica Z, Jakubowski J: Acute head injuries in the elderly. An analysis of136 consecutive patients. Acta Neurochir (Wien) 118:98–102, 1992.

32. Kumchev Y, Dimitrov Z, Kalnev B, Argirov S: Traumatic intracerebralhematomas—Diagnostic and therapeutic problem. Folia Med (Plovdiv) 40:52–57, 1998.

33. Kunze E, Meixensberger J, Janka M, Sorensen N, Roosen K: Decompressivecraniectomy in patients with uncontrollable intracranial hypertension. ActaNeurochir (Wien) 71 [Suppl]:16–18, 1998.

34. Lee EJ, Chio CC, Chen HH: Aggressive temporal lobectomy for uncalherniation in traumatic subdural hematoma. J Formos Med Assoc 94:341–345, 1995.

35. Lobato R, Cordobes F, Rivas J, de la Fuente M, Montero, A, Barcena A, PerezC, Cabrera A, Lamas E: Outcome from severe head injury related to the typeof intracranial lesion. A computerized tomography study. J Neurosurg59:762–774, 1983.

36. Lobato RD, Gomez PA, Alday R, Rivas JJ, Dominguez J, Cabrera A,Turanzas FS, Benitez A, Rivero B: Sequential computerized tomographychanges and related final outcome in severe head injury patients. ActaNeurochir (Wien) 139:385–391, 1997.

37. Mandera M, Zralek C, Krawczyk I, Zycinski A, Wencel T, Bazowski P:Surgery or conservative treatment in children with traumatic intracerebralhaematoma. Childs Nerv Syst 15:267–269, 1999.

38. Marmarou A, Anderson RL, Ward JD, Choi SC, Young HF, Eisenberg HM,Foulkes MA, Marshall LF, Jane JA: Impact of ICP instability and hypoten-sion on outcome in patients with severe head trauma. J Neurosurg 75[Suppl]:S59–S66, 1991.

39. Marshall L: A new classification of head injury based on computerizedtomography. J Neurosurg 75 [Suppl]:S14–S20, 1991.

40. Marshall L, Gautille T, Klauber M, Eisenberg H, Jane J, Luerssen T,Marmarou A, Foulkes M: The outcome of severe closed head injury.J Neurosurg 75 [Suppl]:S28–36, 1991.

41. Mathiesen T, Kakarieka A, Edner G: Traumatic intracerebral lesions withoutextracerebral haematoma in 218 patients. Acta Neurochir (Wien) 137:155–163, 1995.

42. Meier U, Heinitz A, Kintzel D: Surgical outcome after severe craniocerebraltrauma in childhood and adulthood. A comparative study [in German].Unfallchirurg 97:406–409, 1994.

43. Meier U, Zeilinger FS, Henzka O: The use of decompressive craniectomy forthe management of severe head injuries. Acta Neurochir 76 [Suppl]:475–478,2000.

44. Miller JD, Butterworth JF, Gudeman SK, Faulkner JE, Choi SC, Selhorst JB,Harbison JW, Lutz HA, Young HF, Becker DP: Further experience in themanagement of severe head injury. J Neurosurg 54:289–299, 1981.

45. Munch E, Horn P, Schurer L, Piepgras A, Paul T, Schmiedek P: Managementof severe traumatic brain injury by decompressive craniectomy.Neurosurgery 47:315–322, 2000.

46. Nagabhand A, Sangcham K: The study of traumatic intracerebral hematomaat Buri Ram Hospital. J Med Assoc Thai 76:399–404, 1993.

47. Nordstrom C, Messeter K, Sundbarg G, Wahlander S: Severe traumatic brainlesions in Sweden. Part I: Aspects of management in non-neurosurgicalclinics. Brain Inj 3:247–265, 1989.

48. Nussbaum E, Wolf A, Sebring L, Mirvis S: Complete temporal lobectomy forsurgical resuscitation of patients with transtentorial herniation secondary tounilateral hemispheric swelling. Neurosurgery 29:62–66, 1991.

49. Papo I, Caruselli G, Luongo A, Scarpelli M, Pasquini U: Traumatic cerebralmass lesions: Correlations between clinical, intracranial pressure, and com-puted tomographic data. Neurosurgery 7:337–346, 1980.

50. Patel N, Hoyt D, Nakaji P, Marshall L, Holbrook T, Coimbra R, Winchell R,Mikulaschek A: Traumatic brain injury: Patterns of failure of nonoperativemanagement. J Trauma 48:367–374, 2000.

51. Polin R, Shaffrey M, Bogaev C, Tisdale N, Germanson T, Bocchicchio B, JaneJ: Decompressive bifrontal craniectomy in the treatment of severe refractoryposttraumatic cerebral edema. Neurosurgery 41:84–92, 1997.

52. Ransohoff J, Benjamin MV, Gage EL Jr, Epstein F: Hemicraniectomy in themanagement of acute subdural hematoma. J Neurosurg 34:70–76, 1971.


54. Servadei F, Murray GD, Penny K, Teasdale GM, Dearden M, Iannotti F,Lapierre F, Maas AJ, Karimi A, Ohman J, Persson L, Stocchetti N,Trojanowski T, Unterberg A: The value of the �worst� computed tomo-graphic scan in clinical studies of moderate and severe head injury. Euro-pean Brain Injury Consortium. Neurosurgery 46:70–75, 2000.

55. Servadei F, Nanni A, Nasi M, Zappi D, Vergoni G, Giuliani G, Arista A:Evolving brain lesions in the first 12 hours after head injury: Analysis of 37comatose patients. Neurosurgery 37:899–906, 1995.

56. Servadei F, Piazza GC, Padovani R, Fagioli L, Gaist G: �Pure� traumaticcerebral lacerations. A review of 129 cases with long-term follow-up.Neurochirurgia (Stuttg) 28:170–173, 1985.

57. Singounas E: Severe head injury in a paediatric population. J Neurosurg Sci36:201–206, 1992.

58. Soloniuk D, Pitts LH, Lovely M, Bartkowski H: Traumatic intracerebralhematomas: Timing of appearance and indications for operative removal.J Trauma 26:787–794, 1986.

59. Sprick C, Bettag M, Bock W: Delayed traumatic intracranial hematomas—Clinical study of seven years. Neurosurg Rev 12 [Suppl 1]:228–230, 1989.

60. Statham PF, Johnston RA, Macpherson P: Delayed deterioration in patientswith traumatic frontal contusions. J Neurol Neurosurg Psychiatry 52:351–354, 1989.

61. Taylor A, Butt W, Rosenfeld J, Shann F, Ditchfield M, Lewis E, Klug G,Wallace D, Henning R, Tibballs J: A randomized trial of very early decom-pressive craniectomy in children with traumatic brain injury and sustainedintracranial hypertension. Childs Nerv Syst 17:154–162, 2001.

62. Teasdale E, Cardoso E, Galbraith S, Teasdale G: CT scan in severe diffusehead injury: Physiological and clinical correlations. J Neurol NeurosurgPsychiatry 47:600–603, 1984.

63. Tseng SH: Delayed traumatic intracerebral hemorrhage: A study of prog-nostic factors. J Formos Med Assoc 91:585–589, 1992.

64. Tseng SH: Reduction of herniated temporal lobe in patients with severehead injury and uncal herniation. J Formos Med Assoc 91:24–28, 1992.

65. Uzzell BP, Dolinskas CA, Wiser RF, Langfitt TW: Influence of lesions de-tected by computed tomography on outcome and neuropsychological re-covery after severe head injury. Neurosurgery 20:396–402, 1987.

66. Vollmer D, Torner J, Jane J, Sadovnic B, Charlebois D, Eisenberg H, FoulkesM, Marmarou A, Marshall L: Age and outcome following traumatic coma:Why do older patients fare worse? J Neurosurg 75:S37–49, 1991.

67. Whitfield PC, Patel H, Hutchinson PJ, Czosnyka M, Parry D, Menon D,Pickard JD, Kirkpatrick PJ: Bifrontal decompressive craniectomy in themanagement of posttraumatic intracranial hypertension. Br J Neurosurg15:500–507, 2001.

68. Wu J, Hsu C, Liao S, Wong Y: Surgical outcome of traumatic intracranialhematoma at a regional hospital in Taiwan. J Trauma 47:39–43, 1999.

69. Yamaki T, Hirakawa K, Ueguchi T, Tenjin H, Kuboyama T, Nakagawa Y:Chronological evaluation of acute traumatic intracerebral haematoma. ActaNeurochir (Wien) 103:112–115, 1990.

70. Yamaura A, Uemura K, Makino H: Large decompressive craniectomy inmanagement of severe cerebral contusion. A review of 207 cases. NeurolMed Chir (Tokyo) 19:717–728, 1979.

71. Yoo DS, Kim DS, Cho KS, Huh PW, Park CK, Kang JK: Ventricular pressuremonitoring during bilateral decompression with dural expansion. JNeurosurg 91:953–959, 1999.

72. Young HA, Gleave JR, Schmidek HH, Gregory S: Delayed traumatic intra-cerebral hematoma: Report of 15 cases operatively treated. Neurosurgery14:22–25, 1984.

73. Zumkeller M, Hollerhage HG, Proschl M, Dietz H: The results of surgery forintracerebral hematomas. Neurosurg Rev 15:33–36, 1992.





TAB

LE1.

Surg

ical

Man

agem

ent

ofTr

aum

atic

Pare

nchy

mal

Lesi

onsa

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Bul

lock

etal

.(6

)85

IIIM

ean

GC

S9

Non

surg

ical

GO

S3

mo

Pros

pect

ive

stud

yof

85pa

tient

sw

ithIC

Hw

ithun

cert

aint

yre

gard

ing

need

for

imm

edia

tecr

anio

tom

y.Th

ese

patie

nts

wer

ese

lect

edfo

rIC

Pm

onito

ring

inan

atte

mpt

tobe

tter

defin

ene

edfo

rcr

anio

tom

y.A

retr

ospe

ctiv

ean

alys

isof

CT

scan

char

acte

rist

ics

was

then

perf

orm

edfo

rth

ose

patie

nts

for

who

mIC

Pm

onito

ring

did

not

pred

ict

late

dete

rior

atio

nto

crea

tea

bette

rpr

edic

tive

mod

elof

thos

ein

need

ofIC

Hev

acua

tion.

●M

anag

emen

t�s

houl

dbe

base

don

asp

ectr

umof

clin

ical

,C

Tsc

anni

ng,

and

ICP

findi

ngs.

�●

Peak

ICP

was

the

stro

nges

tpr

edic

tor

ofou

tcom

e,bu

tfa

iled

topr

edic

tla

tede

teri

orat

ion

orde

ath

in16

.7%

.●

For

tem

pora

l/par

ieta

lle

sion

s,he

mat

oma

size

,de

gree

ofed

ema,

GC

S,ba

sal

cist

ern

stat

us,

and

peak

ICP

corr

elat

edw

ithou

tcom

e;fo

rfr

onta

lle

sion

s,pe

akIC

Pal

one

pred

icte

dou

tcom

e.●

Prem

esen

ceph

alic

cist

ern

stat

usfa

lsel

ypr

edic

ted

low

ICP

inon

ly2.

3%of

patie

nts

with

elev

ated

ICP

(fals

ene

gativ

e).

●A

utho

rsco

nclu

de:

1)B

asal

cist

ern

effa

cem

ent

onin

itial

CT

scan

isin

dica

tion

for

surg

ery,

rega

rdle

ssof

GC

S. 2)Ex

tens

ive

surr

ound

ing

edem

ash

ould

prom

ptIC

Pm

onito

ring

orse

rial

CT

scan

ning

.If

cist

erns

are

effa

ced,

evac

uatio

nsh

ould

beco

nsid

ered

.

3)IC

Pm

onito

ring

shou

ldbe

carr

ied

out

inpa

tient

sin

com

a.

No.

ofpa

tien

tsG

R/m

inim

aldi

sabi

lity

(%)

MD

(%)

VS/

D(%

)

Surg

ery

5547

2033

No

initi

alsu

rger

y30

4723

30

Car

oli

etal

.(7

)36

IIIA

vera

geG

CS

9.3

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Ret

rosp

ectiv

ere

view

of95

patie

nts

with

mul

tiple

trau

mat

icin

trac

rani

alle

sion

sto

iden

tify

clin

ical

fact

ors

and

radi

olog

icpr

edic

tors

ofpr

ogno

sis

toes

tabl

ish

man

agem

ent

crite

ria.

Patie

nts

wer

edi

vide

din

to3

grou

psba

sed

onpr

edom

inan

tle

sion

type

;36

patie

nts

with

pure

ICH

asth

epr

edom

inan

tle

sion

wer

ein

clud

edin

this

anal

ysis

.Pa

tient

sw

ithpa

renc

hym

alle

sion

s�

2cm

wer

eex

clud

edby

the

auth

ors.

●C

iste

rns

(P�

0.01

),sh

ift(P

�0.

05),

prol

onge

din

crea

sed

ICP

(P�

0.00

01)

diffe

red

sign

ifica

ntly

betw

een

surg

ical

and

nons

urgi

cal

grou

ps.

●80

.6%

neur

olog

ical

lyw

orse

ned

with

in12

–24

hof

initi

alas

sess

men

t.●

44.4

%in

cide

nce

ofD

TIC

H.

●Ty

peof

lesi

on(S

DH

�IC

H),

low

est

reco

rded

GC

S,pr

esen

ceof

prol

onge

din

crea

sed

ICP,

abse

nce

ofpu

pilla

ryre

flexe

sw

ere

sign

ifica

ntpr

edic

tors

ofba

dou

tcom

e(V

S/D

)ba

sed

onm

ultip

lere

gres

sion

anal

ysis

.●

Dec

ompr

essi

vecr

anie

ctom

yw

asa

“use

ful

mea

nsof

cont

rolli

ngin

trac

rani

alpr

essu

rein

19/2

2pa

tient

sun

derg

oing

proc

edur

e.”

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

ICH

3633

.338

.916

.70

11.1

Ope

rate

d13

.3

Non

oper

ated

219.

5

Cho

ksey

etal

.(8

)20

2III

All

GC

SSu

rger

yan

dno

nsur

gica

lG

OS

6m

oR

etro

spec

tive

stud

yof

202

patie

nts

with

trau

mat

icIC

Hto

asce

rtai

nfa

ctor

sin

fluen

cing

outc

ome.

●H

emat

oma

volu

me

�16

cm3

incr

ease

dpr

obab

ility

ofcl

inic

alde

teri

orat

ion.

●R

espi

rato

ryin

suffi

cien

cy,

low

GC

San

dhe

mat

oma

volu

me

�16

cm3

wer

esi

gnifi

cant

lyas

soci

ated

with

apo

orou

tcom

e.

●W

ithth

ese

fact

ors

take

nin

toco

nsid

erat

ion,

cran

ioto

my

sign

ifica

ntly

impr

oved

prob

abili

tyof

good

outc

ome.

●Lo

catio

nof

peri

pher

alhe

mat

omas

did

not

pred

ict

outc

ome.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

1H

emat

oma

169

5842

2H

emat

omas

1921

79

�3

Hem

atom

as14

010

0

Hyp

oxia

,re

spir

ator

yin

suffi

cien

cy35

2080

No

resp

irat

ory

insu

ffici

ency

167

5743

Evac

uatio

n84

6238

Non

evac

uatio

n11

842

58

Age

0–2

0yr

5629

71

Age

21–

40yr

5467

33

Age

41–

64yr

4468

32

Age

�65

yr48

4258

GC

S3–

711

231

69

GC

S8

–11

4562

38



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

GC

S12

–15

4587

13

Vol

ume

1–5

cm3

2681

19

Vol

ume

6–1

5cm

358

5743

Vol

ume

16–3

5cm

367

4654

Vol

ume

�36

cm3

5133

67

Neu

rolo

gica

lde

teri

orat

ion

135

4753

No

neur

olog

ical

dete

rior

atio

n49

8020

Initi

alG

CS

�4

200

100

Fron

tal

9348

52

Pari

etal

1250

50

Tem

pora

l63

6337

Occ

ipita

l9

6733

Cen

tral

2520

80

Cop

linet

al.

(11)

29III

GC

S3–

9D

ecom

pres

sive

cran

iect

omy

GO

Sat

disc

harg

e;FI

Mat

adm

issi

onan

ddi

scha

rge

from

reha

bilia

tion

Ret

rosp

ectiv

ere

view

of29

cons

ecut

ive

patie

nts

with

GC

S�

9an

dC

Tsc

ans

with

mid

line

shift

grea

ter

than

the

volu

me

ofa

surg

ical

lyam

enab

lele

sion

toev

alua

teth

esa

fety

and

feas

ibili

tyof

deco

mpr

essi

vecr

anie

ctom

yw

ithdu

rapl

asty

vers

ustr

aditi

onal

cran

ioto

my

asth

ein

itial

surg

ical

proc

edur

e.Ex

clus

ion

crite

ria

incl

uded

age

�16

yr,

inju

ry�

24h

befo

read

mis

sion

,D

iffus

eIn

jury

II(T

CD

Bcl

assi

ficat

ion)

,is

olat

edED

H,

pree

xist

ing

illne

sslim

iting

life

expe

ctan

cyto

�1

yrfr

omic

tus,

age

�40

yrw

ithex

tens

orpo

stur

ing,

and

bila

tera

lun

reac

tive

pupi

ls�

4m

min

diam

eter

.

●N

onsi

gnifi

cant

diffe

renc

ebe

twee

ncr

anio

tom

yan

dcr

anie

ctom

ygr

oups

with

rega

rds

toag

e(P

�0.

8),

gend

er(P

�1.

0),

adm

issi

onG

CS

scor

e(P

�0.

07),

time

tosu

rger

y(P

�1.

0),

seru

met

hano

lco

ncen

trat

ion

(P�

0.6)

.●

Cra

niec

tom

ygr

oup

had

sign

ifica

ntly

low

erG

CS

scor

eat

surg

ery

(P�

0.04

).●

Cra

niec

tom

ygr

oup

had

sign

ifica

ntly

mor

epa

tient

sw

ithD

iffus

eIn

jury

IIIan

dIV

(56%

vers

us9%

).●

No

sign

ifica

ntdi

ffere

nce

betw

een

grou

psw

ithre

spec

tto

mor

talit

y,G

OS,

acut

eho

spita

lle

ngth

ofst

ay,

initi

alFI

Msc

ore,

chan

gein

FIM

scor

e,or

leng

thof

reha

bilit

atio

nst

ay.

●D

espi

tesi

gnifi

cant

lylo

wer

GC

Sat

time

ofsu

rger

yan

dw

orse

inju

rycl

assi

ficat

ion

byC

Tsc

an,

deco

mpr

essi

vecr

anie

ctom

yw

ithdu

rapl

asty

affo

rded

equi

vale

ntou

tcom

eto

trad

ition

alcr

anio

tom

y,es

tabl

ishi

ngits

safe

tyan

dfe

asib

ility

asa

first

-lin

esu

rgic

alin

terv

entio

nas

oppo

sed

toits

trad

ition

alro

leas

a“s

alva

ge”

proc

edur

e.●

Aut

hors

emph

asiz

eim

port

ance

ofcr

anie

ctom

yex

tend

ing

dow

nto

the

floor

ofth

em

iddl

ecr

ania

lfo

ssa

atth

ero

otof

the

zygo

ma.

All

Cra

niot

omy

Cra

niec

tom

yP

valu

e

No.

ofpa

tient

s29

1712

Dis

char

geG

OS

44

40.

8

Mor

talit

y(%

)34

.541

.225

0.4

Hos

pita

lle

ngth

ofst

ay(d

)17

1818

0.4

Adm

issi

onFI

M63

6438

0.7

Dis

char

geFI

M90

9271

0.6

Cha

nge

inFI

M27

2723

1.0

Cor

addu

etal

.(1

2)37

IIIG

CS

3–5

to�

7St

ereo

tact

icev

acua

tion

Mor

talit

yat

disc

harg

eR

etro

spec

tive

seri

esof

37pa

tient

sw

ithtr

aum

atic

ICH

trea

ted

byst

ereo

tact

icev

acua

tion

thro

ugh

anen

larg

edbu

rrho

leto

eval

uate

outc

ome

usin

gth

iste

chni

que.

Surg

ical

indi

catio

nsin

clud

edco

ma,

dete

rior

atin

gco

nsci

ousn

ess,

orfo

cal

neur

olog

ical

defic

it.

●N

om

orta

lity

with

adm

issi

onG

CS

�7.

●A

dmis

sion

GC

Sre

late

din

vers

ely

tom

orta

lity

(no

stat

istic

s).

●Te

mpo

ral

loca

tion

had

high

erm

orta

lity

(no

stat

istic

s).

No.

ofpa

tien

tsD

(%)

ICH

,to

tal

3730

GC

S3–

510

80

GC

S6

–712

25

GC

S�

715

0

Fron

tal

185.

5

Tem

pora

l18

55



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

De

Luca

etal

.(1

4)22

IIIG

CS

3–12

Dec

ompr

essi

vecr

anie

ctom

yG

OS

atdi

scha

rge

Ret

rosp

ectiv

ere

view

of22

patie

nts

who

unde

rwen

tde

com

pres

sive

cran

iect

omy

for

cere

bral

edem

aw

ithre

frac

tory

intr

acra

nial

hype

rten

sion

orfo

r�i

mm

inen

the

rnia

tion�

duri

ngth

eev

acua

tion

ofa

spac

e-oc

cupy

ing

trau

mat

icle

sion

.Su

rgic

alm

etho

dsar

ede

scri

bed.

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS

(%)

D(%

)

Dec

ompr

essi

vecr

anie

ctom

y22

4118

2318

Dia

zet

al.

(15)

9III

Unk

now

nSu

rger

yM

orta

lity

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of9

patie

nts

with

DTI

CH

toev

alua

tecl

inic

alch

arac

teri

stic

san

dou

tcom

e.

●55

.6%

mor

talit

y.●

Incl

uded

with

n�

9fo

rw

ell-

pres

ente

dda

taon

DTI

CH

.●

Def

initi

onof

DTI

CH

:1)

hist

ory

ofhe

adin

jury

with

head

inm

otio

npr

oduc

ing

tran

sien

tor

perm

anen

tLO

C,

foca

lne

urol

ogic

alfin

ding

s,or

cran

ial

frac

ture

;2)

inte

rval

betw

een

inju

ryan

dde

velo

pmen

tof

DTI

CH

of�

2w

k.

No.

ofpa

tien

tsD

(%)

DTI

CH

955

.6

Gaa

bet

al.

(17)

37III

Unk

now

nD

ecom

pres

sive

cran

iect

omy

GO

S1

yrPr

ospe

ctiv

ese

ries

of37

patie

nts

who

unde

rwen

tde

com

pres

sive

cran

iect

omy

beca

use

oftr

aum

atic

brai

nsw

ellin

gto

asse

ssef

ficac

yof

trea

tmen

tpr

otoc

olat

1yr

afte

rtr

aum

a.Su

rgic

alin

dica

tions

and

excl

usio

ncr

iteri

ade

taile

d.

●14

.7%

mor

talit

yus

ing

stan

dard

ized

prot

ocol

.●

Initi

alG

CS

�6,

day

1G

CS

�8

asso

ciat

edw

ithbe

tter

outc

ome

(no

stat

istic

s).

●A

llpa

tient

sw

ithG

OS

1–2

had

GC

S3

with

pupi

llary

abno

rmal

ity.

●R

epre

sent

sea

rly

repo

rtof

Gue

rra

etal

.st

udy.

No.

ofpa

tien

tsFu

ll(%

)D

isab

led

(%)

VS

(%)

D(%

)

Dec

ompr

essi

vecr

anie

ctom

y34

41.2

35.3

8.8

14.7

Gen

nare

lliet

al.

(19)

1107

IIIG

CS

3–8

Surg

ery

and

nons

urgi

cal

GO

S3

mo

Mul

ticen

ter

pros

pect

ive

seri

esof

1107

patie

nts

with

seve

rebl

unt

trau

mat

icbr

ain

inju

ry(G

CS

�8)

toex

amin

edi

ffere

nces

inty

peof

inju

ry,

seve

rity

ofin

jury

,an

dG

OS

at3

mo.

Subg

roup

sof

patie

nts

with

�oth

erfo

cal

inju

ry�

(n�

205)

and

�diff

use

inju

ry�

(n�

487)

wer

eex

amin

edfo

rth

isch

apte

r.O

fno

te,

�oth

erfo

cal

inju

ry�

incl

uded

ICH

,co

ntus

ions

,bu

tal

sode

pres

sed

cran

ial

frac

ture

,SD

H,

and

EDH

that

wer

eno

tth

epr

imar

yle

sion

s,an

dco

mbi

ned

EDH

/SD

Hin

whi

chth

em

ost

impo

rtan

tle

sion

was

unkn

own.

●Lo

wer

GC

Sw

asst

rong

lyco

rrel

ated

with

wor

seou

tcom

efo

rea

chle

sion

type

.●

Giv

enth

esa

me

pres

entin

gG

CS,

lesi

onty

pes

diffe

red

mar

kedl

yw

ithre

spec

tto

outc

ome.

●A

utho

rsm

ain

poin

tis

that

outc

ome

diffe

rsm

arke

dly

amon

gle

sion

type

sw

ithin

the

real

mof

seve

rehe

adin

jury

,an

dth

atou

tcom

ede

pend

sin

part

onty

peof

lesi

onas

wel

las

onin

jury

seve

rity

asm

easu

red

byG

CS.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Oth

erfo

cal

lesi

on,

tota

l20

521

1819

339

GC

S3–

511

27

1418

456

GC

S6

–8

9338

2319

218

Ope

rate

d71

2115

204

39

Ope

rate

d,G

CS

3–5

359

920

657

Ope

rate

d,G

CS

6–

836

3322

193

22

Non

oper

ated

134

2119

183

39

Non

oper

ated

,G

CS

3–5

776

1717

456

Diff

use

inju

ry,

tota

l48

734

1711

532

GC

S3–

523

716

1014

851

GC

S6

–8

250

5224

92

13

Com

a6

–24

h92

6315

21

15

Com

a6

–24

h,G

CS

3–5

2752

77

430

Com

a6

–24

h,G

CS

6–

865

6818

50

9

Com

a�

24h

395

2817

136

36



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Com

a�

24h,

GC

S3–

521

011

1015

954

Com

a�

24h,

GC

S6

–8

185

4625

103

15

Com

a�

24h,

not

dece

rebr

ate

219

3821

125

24

Com

a�

24h,

not

dece

rebr

ate,

GC

S3–

577

218

1312

47

Com

a�

24h,

not

dece

rebr

ate,

GC

S6

–8

142

4729

111

11

Com

a�

24h,

dece

rebr

ate

176

1513

147

51

Com

a�

24h,

dece

rebr

ate,

GC

S3–

513

66

1216

957

Com

a�

24h,

dece

rebr

ate,

GC

S6

–8

4048

158

328

Gen

tlem

anet

al.

(20)

23III

Com

ave

rsus

not

com

a

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Ret

rosp

ectiv

ere

view

of27

01he

ad-i

njur

edpa

tient

sdi

still

edto

aco

hort

of23

patie

nts

with

DTI

CH

.V

olum

eof

hem

atom

aan

dtim

ing

ofre

peat

CT

scan

are

rela

ted

tom

anag

emen

t.O

pera

tive

vers

usno

nope

rativ

em

anag

emen

tis

rela

ted

toG

OS

at6

mo.

●O

pera

tive

vers

usno

nope

rativ

ele

sion

sdi

ffere

din

aver

age

volu

me

(24

vers

us7

cm3 )

and

time

tore

peat

CT

base

don

clin

ical

crite

ria

(41

vers

us99

h).

Out

com

ebe

twee

ngr

oups

was

not

stat

istic

ally

sign

ifica

nt.

●39

%of

DTI

CH

requ

ied

evac

uatio

n;cr

iteri

ano

tde

scri

bed.

●A

llpa

tient

sw

ithD

TIC

Hha

dab

norm

alC

Ton

adm

issi

on.

●D

efin

ition

ofD

TIC

H:

�les

ion

ofin

crea

sed

atte

nuat

ion

deve

lopi

ngaf

ter

adm

issi

onto

hosp

ital,

ina

part

ofth

ebr

ain

whi

chth

ead

mis

sion

CT

scan

had

sugg

este

dw

asno

rmal

.�

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

Ope

rate

dD

TIC

H9

5644

b

Non

oper

ated

DTI

CH

1457

43

DTI

CH

,to

tal

2357

43b

P�

n.s.

Gow

eret

al.

(21)

10III

GC

S�

8Su

btem

pora

lde

com

pres

sion

vers

uspe

ntob

arbi

tal

com

a

Mor

talit

y1

yrR

etro

spec

tive

seri

esof

115

patie

nts

with

GC

S�

8w

ithno

oper

ativ

em

ass

lesi

onat

time

ofad

mis

sion

man

aged

ona

stan

dard

trea

tmen

tpr

otoc

ol.

10pa

tient

sun

derw

ent

subt

empo

ral

deco

mpr

essi

onbe

caus

eof

med

ical

lyre

frac

tory

intr

acra

nial

hype

rten

sion

,an

dou

tcom

ew

asco

mpa

red

with

thos

epa

tient

sm

anag

edby

pent

obar

bita

lco

ma

with

out

surg

ery.

Tech

niqu

ede

scri

bed.

●Pa

tient

sun

derg

oing

subt

empo

ral

deco

mpr

essi

onha

dsi

gnifi

cant

lylo

wer

mor

talit

yth

anth

ose

trea

ted

with

pent

obar

bita

lco

ma

desp

iteha

ving

low

er(b

utno

tst

atis

tical

lysi

gnifi

cant

)av

erag

ead

mis

sion

GC

S(4

.9ve

rsus

6.1)

.H

owev

er,

10pa

tient

sun

derg

oing

pent

obar

bita

lco

ma

wer

ede

term

ined

not

tobe

oper

ativ

eca

ndid

ates

and

subs

eque

ntly

died

;bi

ases

resu

ltsin

favo

rof

oper

ated

grou

p.●

70%

ofop

erat

edpa

tient

sha

dav

erag

ede

crea

sein

ICP

of34

%.

No.

ofpa

tien

tsD

(%)

Pent

obar

bita

lco

ma

1782

.4

Subt

empo

ral

deco

mpr

essi

on10

40c

cP

�0.

039.

Gud

eman

etal

.(2

2)III

12G

CS

4–1

0Su

rger

yan

dno

nsur

gica

lG

OS

3m

o,1

yrPr

ospe

ctiv

est

udy

of16

2pa

tient

sw

ithse

vere

head

inju

ry,

12of

who

mde

velo

ped

DTI

CH

byC

Tsc

an.

Thes

epa

tient

sw

ere

anal

yzed

for

clin

ical

asso

ciat

ions

and

outc

omes

tobe

tter

unde

rsta

ndth

isle

sion

.

●92

%D

TIC

Hde

tect

ed�

48h

from

inju

ry.

●50

%de

velo

ped

afte

rde

com

pres

sive

surg

ery.

●D

TIC

Hha

d50

%m

orta

lity;

high

erbu

tno

tst

atis

tical

lysi

gnifi

cant

com

pare

dw

ithal

lse

vere

head

inju

ries

duri

ngth

esa

me

peri

od.

●Si

gnifi

cant

lyhi

gher

inci

denc

eof

seco

ndar

ysy

stem

icin

sults

attim

eof

adm

issi

onin

DTI

CH

patie

nts

vers

usal

lse

vere

CH

Ipa

tient

s(9

2%ve

rsus

44%

).●

No

sign

ifica

ntte

mpo

ral

rela

tions

hip

betw

een

deve

lopm

ent

ofD

TIC

H,

ICP

leve

ls,

neur

olog

ical

chan

ge.

●A

utho

rsbe

lieve

that

deve

lopm

ent

ofD

TIC

His

anep

iphe

nom

enon

ofse

cond

ary

insu

ltto

ada

mag

edbr

ain

asop

pose

dto

aca

usat

ive

fact

orfo

rne

urol

ogic

alde

teri

orat

ion—

thus

,tr

eatm

ent

shou

ldbe

aim

edat

prev

entio

nof

seco

ndar

yhy

poxi

cin

sults

.

●D

efin

ition

ofD

TIC

H:

new

pare

nchy

mal

high

dens

ityle

sion

whe

nno

lesi

onor

ane

glig

ible

abno

rmal

ityw

aspr

esen

ton

initi

alC

T.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

DTI

CH

1225

250

050



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Gue

rra

etal

.(2

3)57

IIIG

CS4

–6

Dec

ompr

essi

vecr

anie

ctom

yG

OS

1yr

Pros

pect

ive

seri

esof

57pa

tient

sw

houn

derw

ent

deco

mpr

essi

vecr

anie

ctom

yun

der

ast

anda

rdiz

edtr

eatm

ent

prot

ocol

and

stan

dard

ized

surg

ical

tech

niqu

efo

rpo

sttr

aum

atic

diffu

sebr

ain

swel

ling.

The

patie

nts

wer

edi

vide

din

toa

grou

pun

derg

oing

prim

ary

deco

mpr

essi

on(n

�39

)an

da

grou

pun

derg

oing

seco

ndar

yde

com

pres

sion

beca

use

ofth

erap

y-re

sist

ant

intr

acra

nial

hype

rten

sion

afte

rev

acua

tion

ofa

surg

ical

mas

sle

sion

(n�

18).

Surg

ical

indi

catio

nsan

dm

etho

dsar

ede

taile

d.Pa

tient

sw

ithG

CS

3an

d/or

bila

tera

lfix

ed/

dila

ted

pupi

lsw

ere

excl

uded

.A

ge�

30yr

,�

40yr

,an

d�

50yr

vari

ably

used

asex

clus

ion

crite

ria

thro

ugho

utst

udy.

●37

%G

OS

5,19

%m

orta

lity

over

all.

●G

CS

onpo

sttr

aum

ada

y1

pred

ictiv

eof

outc

ome.

●Si

gnifi

cant

lybe

tter

outc

ome

with

nolo

ssof

Bw

aves

and

pres

ence

ofpl

atea

uw

aves

onIC

Pm

onito

ring

.●

Incl

usio

nof

SEP,

ICP/

CPP

,an

dA

EPm

onito

ring

tech

niqu

eshe

lped

tode

fine

indi

catio

nsfo

rth

eau

thor

s.●

Aut

hors

conc

lude

that

deco

mpr

essi

vecr

anie

ctom

ysh

ould

be1s

tam

ong

seco

nd-t

ier

ther

apie

sfo

rre

sist

ant

intr

acra

nial

hype

rten

sion

.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

Prim

ary

deco

mpr

essi

on38

5842

Seco

ndar

yde

com

pres

sion

1765

35P

�n.

s.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

All

patie

nts

5537

2111

919

Kat

ayam

aet

al.

(28)

21III

Unk

now

nSu

rger

yan

dno

nsur

gica

lM

orta

lity

Ret

rosp

ectiv

eca

seco

ntro

lst

udy

of21

patie

nts

with

fron

tal

lobe

cont

usio

nsan

dm

edic

ally

intr

acta

ble

intr

acra

nial

hype

rten

sion

�40

mm

Hg,

com

pari

ngm

orta

lity

betw

een

the

grou

ptr

eate

dm

edic

ally

(n�

8)an

dth

egr

oup

trea

ted

with

surg

ical

evac

uatio

nof

cont

used

brai

n(n

�13

).A

utho

rsst

ate

that

age,

sex,

GC

S,an

dIC

Ple

vels

wer

eco

mpa

rabl

ebe

twee

ngr

oups

,bu

tdo

not

prov

ide

stat

istic

s.

●M

orta

lity

was

sign

ifica

ntly

decr

ease

din

the

surg

ical

grou

p(2

2%ve

rsus

88%

).●

Aut

hors

note

that

prog

ress

ive

ICP

incr

ease

and

neur

olog

ical

dete

rior

atio

nof

ten

cont

inue

dpa

stth

etim

ing

ofm

axim

alen

larg

emen

tof

intr

acer

ebra

lle

sion

son

CT.

●A

utho

rsad

voca

tesu

rgic

alex

cisi

onof

cont

used

brai

nin

the

setti

ngof

med

ical

lyun

cont

rolla

ble

ICP

elev

atio

n.

No.

ofpa

tien

tsD

(%)

Non

oper

ativ

e8

88

Surg

ery

1322

d

dP

�0.

01

Kau

fman

etal

.(2

9)8

IIIA

llGC

SSu

rger

yan

dno

nsur

gica

lM

orta

lity

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of12

patie

nts

with

DTI

CH

orre

curr

ent

hem

atom

aaf

ter

head

inju

ryto

exam

ine

the

asso

ciat

ion

ofdi

ssem

inat

edin

trav

ascu

lar

clot

ting

and

fibri

noly

sis

with

occu

rren

ceof

thes

ele

sion

s.A

subs

etof

8pa

tient

sw

ithD

TIC

Hw

asev

alua

ted

for

this

chap

ter.

●7

of8

DTI

CH

patie

nts

had

abno

rmal

clot

ting

para

met

ers.

●In

clud

edw

ithn

�8

beca

use

ofpr

esen

tatio

nof

data

and

eval

uatio

nof

clot

ting

para

met

ers.

No.

ofpa

tien

tsD

(%)

Clo

ttin

gab

norm

alit

ies

(%)

DTI

CH

,to

tal

837

.587

.5

Kot

wic

aan

dJa

kubo

wsk

i(3

1)48

IIIA

llGC

SSu

rger

yan

dno

nsur

gica

lG

OS,

not

spec

ified

Ret

rosp

ectiv

ese

ries

of13

6pa

tient

sag

ed70

–91

yrw

ithTB

Ito

eval

uate

outc

ome

from

TBI

inan

aged

popu

latio

n.48

patie

nts

with

ICH

/con

tusi

onre

view

ed.

●O

vera

llm

orta

lity

from

ICH

/con

tusi

ons

was

50%

;8

of13

inop

erat

edca

ses,

and

16of

35in

nono

pera

ted

case

s.●

Mor

talit

yco

rrel

ated

with

adm

issi

onG

CS

(no

stat

istic

s).

●A

utho

rsfe

elth

atbe

caus

eth

em

orta

lity

rate

for

patie

nts

�70

yrw

ithG

CS

�9

isso

high

(�80

%),

limite

dat

tem

pts

shou

ldbe

mad

eat

resu

scita

tion,

inte

nsiv

eca

re,

and

surg

ery.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

ICH

�co

ntus

ion,

surg

ery

137.

77.

715

.47.

761

.5

ICH

�co

ntus

ion,

nosu

rger

y35

11.4

11.4

22.9

8.6

45.7



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Kum

chev

etal

.(3

2)79

IIIG

CS

3–6

to12

–15

Surg

ery

and

nons

urgi

cal

Mor

talit

yR

etro

spec

tive

revi

ewof

79pa

tient

sw

ithin

trac

ereb

ral

hem

atom

as,

com

pari

nglo

catio

nof

hem

atom

aw

ithm

orta

lity,

and

outc

ome

ofsu

rger

yve

rsus

nons

urgi

cal

ther

apy.

Trea

tmen

tpr

otoc

olw

asno

tst

anda

rdiz

ed;

surg

ical

indi

catio

nsw

ere

not

repo

rted

.

●Th

ose

patie

nts

with

tem

pora

lIC

Hw

ere

mor

elik

ely

todi

e.●

Surg

ery

did

not

alte

rm

orta

lity.

Surg

ical

lym

anag

edpa

tient

sw

ere

mor

elik

ely

tosh

owcl

inic

alim

prov

emen

t.

Loca

tion

No.

ofpa

tien

tsD

(%)

Pva

lues

Fron

tal

166

�0.

001

Tem

pora

l36

57.6

�0.

01

Pari

eto-

occi

pita

l14

18.2

�0.

01

Mul

tiple

site

s13

18.2

�0.

01

Kun

zeet

al.

(33)

28III

All

GC

SD

ecom

pres

sive

cran

iect

omy

GO

S1

yrR

etro

spec

tive

stud

yof

28pa

tient

sun

derg

oing

uni/b

ilate

ral

deco

mpr

essi

vecr

anie

ctom

yfo

rpo

sttr

aum

atic

edem

aaf

ter

�max

imal

�m

edic

alth

erap

y.

●57

%ov

eral

lfa

vora

ble

outc

ome

(GO

S4

–5)

at1

yr.

●M

ean

ICP

decr

ease

dfr

om41

.7to

20.6

mm

Hg

afte

rde

com

pres

sion

.●

Patie

nts

with

“vas

t”pr

imar

yle

sion

s,hy

poxi

a,is

chem

icin

farc

tion,

brai

nste

min

jury

,“c

entr

al”

hern

iatio

n,an

dpr

imar

yan

isoc

oria

excl

uded

.

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS

(%)

D(%

)

Dec

ompr

essi

vecr

anie

ctom

y28

5718

1411

Lee

etal

.(3

4)29

IIIG

CS

�8

Subt

empo

ral

deco

mpr

essi

on�

tem

pora

llo

bect

omy

GO

S1

yrR

etro

spec

tive

seri

esof

29pa

tient

sop

erat

edon

for

acut

eSD

Han

dse

vere

cont

usio

n/sw

ellin

gof

tem

pora

llo

bew

ithun

cal

hern

iatio

n.16

patie

nts

unde

rwen

tsu

btem

pora

lde

com

pres

sion

and

debr

idem

ent

ofco

ntus

edbr

ain

whe

reas

13su

bseq

uent

patie

nts

unde

rwen

tte

mpo

ral

lobe

ctom

yin

addi

tion.

Thes

etw

ogr

oups

wer

eco

mpa

red

with

resp

ect

toou

tcom

eat

1yr

orlo

nger

.

●A

dditi

onof

tem

pora

llo

bect

omy

tosu

btem

pora

lde

com

pres

sion

and

debr

idem

ent

ofco

ntus

edbr

ain

sign

ifica

ntly

decr

ease

dm

orta

lity

from

56%

to8%

and

sign

ifica

ntly

impr

oved

aver

age

GO

Sfr

om2.

2to

4.0.

Preo

pera

tive

GC

S,ag

e,an

dse

xw

ere

not

sign

ifica

ntly

diffe

rent

betw

een

grou

ps.

●In

cide

nce

ofm

otor

reco

very

,co

gniti

vede

ficits

,an

dse

izur

esw

ere

not

sign

ifica

ntly

diffe

rent

betw

een

grou

ps.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)D

(%)

STD

,deb

ridem

ent

1625

7556

STD

,de

brid

emen

t,TL

1377

238e

STD

,de

brid

emen

t,co

mpl

ete

TL10

100

00

STD

,de

brid

emen

t,an

teri

orTL

30

100

33

eP

�0.

01.

Loba

toet

al.

(35)

277

IIIG

CS

3–7

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Pros

pect

ive

seri

esof

277

seve

rely

head

-in

jure

dpa

tient

scl

assi

fied

byC

Tsc

anin

to8

inju

ryty

pes

toex

amin

eG

OS

at6

mo

rela

ted

tole

sion

type

.IC

Pw

asm

onito

red

inal

lpa

tient

s.Pa

tient

sbr

ain

dead

onar

riva

lw

ere

excl

uded

.A

stan

dard

ized

man

agem

ent

prot

ocol

was

used

and

isde

scri

bed.

GC

S,IC

P,an

din

terv

alto

oper

atio

nva

ried

betw

een

CT

grou

ps.

●Pa

tient

sw

ithpu

reex

trac

ereb

ral

hem

atom

a,si

ngle

brai

nco

ntus

ion,

gene

ral

brai

nsw

ellin

g,an

dno

rmal

CT

scan

sha

da

sign

ifica

ntly

bette

rou

tcom

eth

anpa

tient

sw

ithex

trac

ereb

ral

hem

atom

aan

dac

ute

hem

isph

eric

swel

ling,

mul

tiple

brai

nco

ntus

ions

,an

dpa

tient

sw

ithdi

ffuse

axon

alin

jury

.●

Acu

tehe

mis

pher

icsw

ellin

gsi

gnifi

cant

lyin

crea

sed

mor

talit

y.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Sing

leco

ntus

ion

(SC

)25

728

44

12

SC�

EDH

714

860

00

SC�

SDH

1315

4615

023

MU

C20

015

100

75

MU

C�

SDH

120

178

075

Mul

tiple

bila

tera

lco

ntus

ion

4219

265

050

GB

S25

768

40

12

GB

S�

EDH

1688

60

06

DA

I43

59

2112

53

Nor

mal

CT

2839

2918

114



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Loba

toet

al.

(36)

587

IIIG

CS

3–8

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Ret

rosp

ectiv

ese

ries

of58

7pa

tient

sw

ithG

CS�

8an

dat

leas

ton

eco

ntro

lC

Tsc

an(a

vera

ge36

.7h

afte

rin

itial

CT)

toev

alua

tein

cide

nce

ofch

ange

sin

CT

clas

sific

atio

nov

ertim

ean

dto

asse

ssth

ere

lativ

epr

ogno

stic

sign

ifica

nce

ofa

seco

ndC

Tsc

anve

rsus

the

first

.

●51

.2%

deve

lope

dsi

gnifi

cant

CT

chan

ges

requ

irin

gch

ange

indi

agno

stic

cate

gory

.●

Fina

lou

tcom

ew

asm

ore

accu

rate

lypr

edic

ted

byco

ntro

lC

Tdi

agno

sis

than

byin

itial

CT

diag

nosi

s.●

19%

ofdi

ffuse

inju

ries

evol

ved

tom

ass

lesi

ons.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS

(%)

D(%

)

Diff

use

inju

ryI

2475

8.3

16.6

Diff

use

inju

ryII

247

54.6

19.4

25.9

Diff

use

inju

ryIII

123

20.3

10.5

69.1

Diff

use

inju

ryIV

2516

480

Man

dera

etal

.(3

7)31

IIIG

CS

3–8

to13

–15

Surg

ery

and

nons

urgi

cal

GO

S,no

tsp

ecifi

edR

etro

spec

tive

stud

yof

31ch

ildre

nw

ithtr

aum

atic

ICH

toex

amin

efa

ctor

sre

late

dto

surg

ical

deci

sion

mak

ing

and

asse

ssou

tcom

e.

●O

utco

me

did

not

diffe

rbe

twee

nsu

rgic

alan

dco

nser

vativ

egr

oups

desp

iteth

efa

ctth

atm

ore

patie

nts

with

seve

rein

jury

wer

etr

eate

dsu

rgic

ally

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Surg

ery

2035

2030

015

Non

surg

ical

1145

1818

018

Mar

shal

l(3

9)41

4III

GC

S3–

8N

onsu

rgic

alG

OS

11d–

2yr

Pros

pect

ive

seri

esof

746

patie

nts

from

TCD

Bto

iden

tify

patie

nts

atri

skfo

rde

velo

ping

intr

acra

nial

hype

rten

sion

inth

eab

senc

eof

sign

ifica

ntm

ass

lesi

ons

onin

itial

CT.

The

auth

ors

prop

ose

ane

wcl

assi

ficat

ion

ofhe

adin

jury

base

don

CT

and

rela

teth

ene

wcl

assi

ficat

ion

cate

gori

esto

outc

ome.

Non

surg

ical

diffu

sein

jury

cate

gori

esar

ein

clud

ed.

●C

Tdi

agno

sis

was

ahi

ghly

sign

ifica

ntin

depe

nden

tpr

edic

tor

ofm

orta

lity

whe

nag

ean

dm

otor

scor

ew

ere

incl

uded

inth

em

odel

.●

InD

iffus

eIn

jury

III,

high

est

ICP

and

post

resu

scita

tion

pupi

llary

reac

tivity

wer

esi

gnifi

cant

pred

icto

rsof

outc

ome.

Mar

shal

let

al.

(40)

502

IIIG

CS

3–8

Surg

ery

(ICH

)an

dno

nsur

gica

l(D

iffus

eIn

jury

I–IV

)

GO

S11

d–2

yrPr

ospe

ctiv

ese

ries

of74

6pa

tient

sse

lect

edfr

omTC

DB

toas

sess

impo

rtan

ceof

prog

nost

icva

riab

les

inde

term

inin

gou

tcom

efr

omse

vere

head

inju

ry.

Subg

roup

sof

nono

pera

ted

patie

nts

with

Diff

use

Inju

rypa

ttern

sI–

IVan

dpa

tient

sw

ithev

acua

ted

ICH

are

incl

uded

.

●90

%of

deat

hsoc

curr

edw

ithin

2w

k.●

Intr

acra

nial

diag

nosi

sst

rong

lyco

rrel

ated

with

outc

ome;

e.g.

,hi

gher

%of

Diff

use

Inju

ryIII

/IVpa

tient

sw

ere

vege

tativ

e.●

InD

iffus

eIn

jury

IIgr

oup,

age

�40

yrst

rong

lyco

rrel

ated

with

poor

outc

ome.

●G

OS

for

evac

uate

dIC

Hpa

tient

sco

rrel

ated

nega

tivel

ywith

age

�40

yr.

●Fo

ral

lpa

tient

s,po

stre

susc

itatio

nG

CS,

pupi

llary

reac

tivity

,an

dag

eco

rrel

ated

with

outc

ome.

CT

diag

nosi

sf

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Diff

use

Inju

ryI

5227

34.6

19.2

9.6

9.6

Diff

use

Inju

ryII

177

8.5

2640

.711

.313

.5

Diff

use

Inju

ryIII

153

3.3

13.1

26.8

22.9

34

Diff

use

Inju

ryIV

323.

13.

118

.818

.856

.2

Diff

use

Inju

ryII,

age

�40

yr15

310

28.7

41.1

11.1

9.1

Diff

use

Inju

ryII,

age

�40

yr24

08.

337

.512

.541

.7

fP

�0.

001



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Mat

hies

enet

al.

(41)

218

IIIM

ean

GC

S6.

9Su

rger

yan

dno

nsur

gica

lG

OS

6m

oR

etro

spec

tive

revi

ewof

data

accu

mul

ated

inpr

ospe

ctiv

e,ra

ndom

ized

,do

uble

-blin

dH

IT-2

stud

yon

nim

odip

ine

inhe

adin

jury

.21

8pa

tient

sno

tfo

llow

ing

com

man

dsw

ithin

24h

oftr

aum

aw

ithin

trac

ereb

ral

lesi

ons

only

onC

Tw

ere

chos

enfo

ran

alys

is.

Patie

nts

with

extr

acer

ebra

lle

sion

s,no

rmal

CT,

guns

hot

wou

nds,

preg

nanc

y,bi

late

rally

fixed

/dila

ted

pupi

ls,

GC

S3

for

�2

h,an

dpr

evio

usdr

ugad

min

istr

atio

nin

terf

erin

gw

ithne

urol

ogic

alas

sess

men

tw

ere

excl

uded

.C

linic

alm

anag

emen

tw

asno

tst

anda

rdiz

ed.

Patie

nts

wer

edi

vide

din

to4

subg

roup

sba

sed

onC

Tap

pear

ance

with

in24

hof

trau

ma:

1)co

ntus

ions

;2)

frac

ture

cont

usio

ns(d

irec

tlyun

derl

ying

depr

esse

dfx

);3)

DA

I;an

d4)

ICH

(�10

cm3 ),

and

anal

yzed

with

resp

ect

toG

OS

at6

mo.

Asu

bgro

upof

patie

nts

with

GC

S�

6an

dle

sion

�20

cm3

wer

ean

alyz

edfo

rth

eef

fect

ofsu

rger

yan

dtim

ing

ofsu

rger

yon

outc

ome.

Asu

bgro

upof

patie

nts

who

�tal

ked

and

died

�w

ere

anal

yzed

tode

term

ine

the

valu

eof

earl

ysu

rger

y.

●In

itial

CT

char

acte

rist

ics

asso

ciat

edw

ithne

urol

ogic

alde

teri

orat

ion,

defin

edas

fall

inG

CS

by2,

GC

Sfr

om4

to3,

orpu

pilla

rydi

lata

tion,

wer

eSA

H,

foca

lle

sion

with

volu

me

�40

cm3 ,

and

com

pres

sed/

abse

ntci

ster

ns.

●Se

cond

ary

(�5

d)de

teri

orat

ion

corr

elat

edw

ithSA

Hon

initi

alC

T,hy

poxi

cev

ents

.●

For

the

4gr

oups

:1)

Con

tusi

ons:

outc

ome

adve

rsel

yaf

fect

edby

pres

ence

ofIV

H.

2)Fr

actu

re/c

ontu

sion

s:ou

tcom

esi

gnifi

cant

lyw

orse

than

othe

rgr

oups

;ou

tcom

ead

vers

ely

affe

cted

bypr

esen

ceof

SAH

.3)

DA

I:m

ean

adm

issi

onG

CS

sign

ifica

ntly

wor

seth

anot

her

grou

ps.N

opa

tient

sop

erat

ed.

4)IC

H:m

ean

adm

issi

onG

CS

sign

ifica

ntly

bette

rth

anot

her

grou

ps;o

utco

me

adve

rsel

yaf

fect

edby

pres

ence

ofIV

H.

●Pa

tient

sw

ithle

sion

s�

20cm

3an

dad

mis

sion

GC

S�

6:1)

Patie

nts

oper

ated

with

outp

revi

ous

dete

riora

tion

(n�

9)ha

dbe

tter

outc

ome

than

thos

eno

top

erat

edon

orop

erat

edon

afte

rde

terio

ratio

n(n

�66

).2)

Patie

nts

with

radi

olog

ical

sign

sof

mas

sef

fect

(com

pres

sion

/obl

itera

tion

ofci

ster

nan

d/or

mid

line

shift

�5

mm

)had

bette

rou

tcom

ew

ithsu

rger

yve

rsus

nosu

rger

y,an

dth

ose

oper

ated

befo

rede

terio

ratio

nha

dbe

tter

outc

ome

than

thos

eop

erat

edaf

ter

dete

riora

tion.

3)Su

rger

ydi

dno

tinf

luen

ceou

tcom

ein

patie

nts

adm

itted

with

GC

S�

5.●11

patie

nts

with

tem

pora

lcon

tusi

on,r

adio

logi

cals

igns

ofm

ass

effe

ct,a

dmis

sion

GC

S�

10:

outc

ome

sign

ifica

ntly

bette

rin

thos

eop

erat

edbe

fore

orim

med

iate

lyaf

ter

dete

riora

tion

com

pare

dw

ithde

laye

dor

nosu

rger

y.●A

utho

rspo

ints

:1)

Patie

nts

with

front

alor

tem

pora

lcon

tusi

ons

with

mid

line

shift

orci

ster

nalc

ompr

essi

onha

dst

atis

tical

lybe

tter

outc

ome

with

surg

ery.

2)Su

rgic

alou

tcom

ew

asbe

tter

the

earl

ier

itw

aspe

rfor

med

rela

ted

tode

teri

orat

ion.

3)Pa

tient

sw

ithte

mpo

ral

cont

usio

nsan

dra

diol

ogic

alsi

gns

ofm

ass

effe

ctsh

ould

beop

erat

edon

befo

rede

teri

orat

ion.

4)A

utho

rsof

fer

indi

catio

nsfo

rsu

rger

y:al

lpa

tient

sw

ithco

ntus

ions

�20

cm3

with

radi

olog

ical

sign

sof

mas

sef

fect

orw

ithan

yle

sion

�50

cm3 .

No.

ofpa

tien

tsG

R/M

D(%

)D

(%)

Mea

nG

OS

All

patie

nts

218

5627

.53.

2

Lesi

onty

peN

o.of

pati

ents

GR

(%)

MD

(%)

SD(%

)V

S(%

)D

(%)

Con

tusi

ons

122

2430

115

30

Con

tusi

ons/

depr

esse

dcr

ania

lfr

actu

reg

2412

1725

442

DA

I39

2325

305

17

ICH

3231

2816

322

No.

ofpa

tien

tsG

R/M

D/S

D(%

)V

S/D

(%)

GC

S�

6,vo

lum

e�

20cm

3

No

surg

ery

4250

50

Surg

ery

befo

rede

teri

orat

iong

610

00

Surg

ery

onda

yof

dete

rior

atio

n9

6733

Surg

ery

�1

daf

ter

dete

rior

atio

n9

5544

GC

S�

6,vo

lum

e�

20cm

3 ,C

Tm

ass

effe

ct

No

surg

ery

1926

74

Surg

ery

befo

rede

teri

orat

iong

610

00

Surg

ery

onda

yof

dete

rior

atio

n8

62.5

37.5

Surg

ery

�1

daf

ter

dete

rior

atio

n1

010

0

GC

S�

6,vo

lum

e�

50cm

3

No

surg

ery

1436

64

Surg

ery

befo

rede

teri

orat

iong

310

00

Surg

ery

onda

yof

dete

rior

atio

ng10

7030

Surg

ery

�1

daf

ter

dete

rior

atio

n1

010

0

GC

S�

10cm

3 ,te

mpo

ral

cont

usio

n,C

Tm

ass

effe

ct

Del

ayed

orno

surg

ery

80

100

Surg

ery

befo

reor

attim

eof

dete

rior

atio

ng3

100

0g

P�

0.05

.



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Mei

eret

al.

(42)

38III

All

GC

SSu

rger

yG

OS

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of93

6pa

tient

sw

houn

derw

ent

surg

ery

for

TBI,

38of

who

mw

ere

child

ren

with

ICH

/con

tusi

onas

prim

ary

path

olog

y.G

OS

isre

late

dto

adm

issi

onG

CS.

●C

hild

ren

with

adm

issi

onG

CS

6–

8ha

dbe

tter

over

all

outc

ome

than

adul

tsw

ithad

mis

sion

GC

S6

–8.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS

(%)

D(%

)

Chi

ldre

nw

ithco

ntus

ions

3853

1037

Mei

eret

al.

(43)

19III

GC

S3–

12D

ecom

pres

sive

cran

iect

omy

GO

Sat

disc

harg

eR

etro

spec

tive

seri

esof

19pa

tient

sw

ithin

trac

tabl

ein

trac

rani

alhy

pert

ensi

onun

derg

oing

deco

mpr

essi

vecr

anie

ctom

yto

eval

uate

fact

ors

rela

ted

topr

ogno

sis.

Surg

ical

indi

catio

nsin

clud

edfa

ilure

ofco

nser

vativ

ein

terv

entio

nfo

rin

trac

rani

alhy

pert

ensi

on,

age

�50

yrw

ithou

tm

ultip

letr

aum

a,ag

e�

30yr

inpr

esen

ceof

maj

orex

trac

rani

alin

juri

es,

seve

rebr

ain

swel

ling

onC

T(o

rin

trao

pera

tivel

y).

11pa

tient

sun

derw

ent

deco

mpr

essi

vecr

anie

ctom

ybe

caus

ege

nera

lized

brai

ned

ema

was

evid

ent

duri

ngev

acua

tion

ofa

spac

e-oc

cupy

ing

lesi

on;

8pa

tient

sun

derw

ent

deco

mpr

essi

vecr

anie

ctom

yse

cond

ary

tofa

ilure

ofco

nser

vativ

etr

eatm

ent.

●2

patie

nts

�60

yrha

dG

OS

2(V

S).

●A

gean

dG

CS

rela

ted

toG

OS

(no

stat

istic

s).

●M

orta

lity

incr

ease

dw

ithth

epr

esen

ceof

extr

acra

nial

inju

ries

(no

stat

istic

s).

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS

(%)

D(%

)

All

1926

526

43

Ope

rate

d63

Non

oper

ated

27

Mill

eret

al.

(44)

20III

GC

S3–

8Su

rgic

alan

dno

nsur

gica

lG

OS

3m

oPr

ospe

ctiv

ese

ries

of22

5pa

tient

sw

ithse

vere

head

inju

rym

anag

edby

stan

dard

ized

prot

ocol

toid

entif

ypr

ogno

stic

vari

able

san

dto

valid

ate

resu

ltsof

apr

evio

usst

udy.

Subg

roup

sw

ithsu

rgic

alan

dno

nsur

gica

lle

sion

sar

eid

entif

ied

and

outc

ome

issp

ecifi

edfo

rea

chle

sion

.Th

ede

finiti

onof

ICH

incl

uded

�5

mm

mid

line

shift

,an

d,th

us,

all

wer

eta

ken

tosu

rger

y.

●C

orre

latio

nbe

twee

nag

e,ab

norm

alm

otor

resp

onse

,bi

late

ral

loss

oflig

htre

flex,

impa

ired

/abs

ent

ocul

ocep

halic

refle

x,an

dou

tcom

ein

nono

pera

tivel

ym

anag

edpa

tient

s.●

Pare

nchy

mal

lesi

ons

ofin

crea

sed

dens

ityon

CT

wer

eas

soci

ated

with

seve

redi

sabi

lity/

vege

tativ

est

ate

com

pare

dw

ithot

her

lesi

ons.

●31

%in

cide

nce

ofun

cont

rolla

ble

post

oper

ativ

eel

evat

edIC

Pin

ICH

patie

nts.

Surg

ical

and

nons

urgi

cal

grou

psw

ere

not

com

para

ble.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS

(%)

D(%

)

ICH

2026

1658

Con

tusi

on,

nosu

rger

y31

6413

23

Diff

use,

nosu

rger

y10

173

1017

Mun

chet

al.

(45)

49III

GC

S�

8to

�8

Dec

ompr

essi

vecr

anie

ctom

yG

OS

6m

oR

etro

spec

tive

seri

esof

49pa

tient

sun

derg

oing

unila

tera

lde

com

pres

sive

cran

iect

omy

for

TBI

unde

rst

anda

rdiz

edtr

eatm

ent

prot

ocol

toev

alua

tepr

eope

rativ

ean

dpo

stop

erat

ive

CT

char

acte

rist

ics,

ICP/

CPP

chan

ge,

and

toas

sess

outc

ome.

Asu

bdiv

isio

nin

tora

pid

and

dela

yed

grou

psw

asal

soan

alyz

edac

cord

ing

toou

tcom

e.Su

rgic

alin

dica

tions

deta

iled.

Patie

nts

with

bila

tera

lin

trac

rani

alle

sion

s;bi

late

rally

fixed

and

dila

ted

pupi

ls;

and

life-

thre

aten

ing

conc

omita

ntm

edic

aldi

seas

ew

ere

excl

uded

.A

llpa

tient

sun

derg

oing

rapi

dde

com

pres

sion

had

acut

eSD

H.

All

patie

nts

unde

rgoi

ngde

laye

dde

com

pres

sion

had

diffu

sebr

ain

swel

ling.

●41

%go

odou

tcom

eat

6m

o.●

Vis

ibili

tyof

mes

ence

phal

icci

ster

nsan

dm

idlin

esh

iftim

prov

edsi

gnifi

cant

lyaf

ter

cran

iect

omy;

gyra

lpa

ttern

and

visi

bilit

yof

vent

ricl

esdi

dno

t.●

No

chan

gein

mea

nth

erap

eutic

inte

nsity

leve

lbe

fore

vers

usaf

ter

deco

mpr

essi

on.

●A

ge�

50yr

,G

CS

�8

corr

elat

edsi

gnifi

cant

lyw

ithbe

tter

outc

ome.

●G

OS

at6

mo

sign

ifica

ntly

bette

rin

rapi

dve

rsus

dela

yed

deco

mpr

essi

ongr

oups

,bu

tG

OS

atdi

scha

rge

from

ICU

not

sign

ifica

ntly

diffe

rent

.●

No

sign

ifica

ntdi

ffere

nce

betw

een

befo

rean

daf

ter

deco

mpr

essi

onIC

Pan

dC

PP.

●C

hang

ein

mes

ence

phal

icci

ster

nvi

sibi

lity

corr

elat

edw

ithdi

stan

cebe

twee

nlo

wer

cran

iect

omy

bord

eran

dcr

ania

lba

se(m

ore

than

area

ofde

com

pres

sion

).

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

At

disc

harg

efr

omIC

U49

2872

At

6m

o49

4159

GC

S8

GC

S<

8R

apid

Del

ayed

Age

<50

yrA

ge50

yr

Mea

nG

OS

3.9

1.4h

3.1

1.9i

3.0

1.9i

hP

�0.

023,

iP

�0.

046.



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Nag

abha

ndan

dSa

ngch

am(4

6)10

IIIU

nkno

wn

Surg

ery

Mor

talit

yR

etro

spec

tive

revi

ewof

71pa

tient

sun

derg

oing

surg

ery

for

intr

acra

nial

hem

atom

a.M

orta

lity

com

pare

dbe

twee

ndi

ffere

ntin

trac

rani

alle

sion

s,an

dbe

twee

ngr

oups

trea

ted

befo

rean

daf

ter

intr

oduc

tion

ofC

Tto

the

hosp

ital.

Asu

bset

of10

patie

nts

with

ICH

rece

ivin

gpr

eope

rativ

eC

Tis

incl

uded

No.

ofpa

tien

tsD

(%)

ICH

/con

tusi

on10

50

Nor

dstr

omet

al.

(47)

587

IIIG

CS

3–8

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Ret

rosp

ectiv

ese

ries

of58

7pa

tient

sw

ithse

vere

TBI,

prim

arily

toan

alyz

eef

fect

ofin

stitu

tion

ofm

anag

emen

tpr

otoc

olin

are

gion

inSw

eden

onou

tcom

e.Pr

otoc

olin

stitu

ted

in19

83.

For

our

purp

oses

,ep

idem

iolo

gyan

dou

tcom

eof

ICH

and

seve

reTB

Iw

ithno

mas

sle

sion

are

eval

uate

d.

●Pa

tient

sw

ithIC

H(in

clud

ing

asso

ciat

edSD

H/E

DH

)ha

d64

%m

orta

lity

and

15%

good

reco

very

at6

mo.

●%

ofex

plor

ator

ycr

anio

tom

ies

and

%m

orta

lity

inlo

cal

hosp

itals

decr

ease

dfo

llow

ing

inst

itutio

nof

man

agem

ent

prot

ocol

.C

Tw

asno

tun

iver

sally

avai

labl

eat

loca

lho

spita

ls.

No.

ofpa

tien

tsG

R(%

)D

(%)

ICH

205

1564

No

mas

sle

sion

,ex

plor

ator

ycr

anio

tom

y30

73j

No

mas

sle

sion

,no

expl

orat

ion

150

61jP

�n.

s.

Nus

sbau

met

al.

(48)

10III

GC

S3–

6Te

mpo

ral

lobe

ctom

yB

arth

elin

dex,

min

imum

8m

oR

etro

spec

tive

seri

esof

10pa

tient

sw

ithG

CS

3–6

and

clin

ical

sign

sof

tran

sten

tori

alhe

rnia

tion

who

unde

rwen

tst

anda

rdiz

edte

mpo

ral

lobe

ctom

yfo

run

ilate

ral

hem

isph

eric

swel

ling.

All

patie

nts

dem

onst

rate

dfix

edpu

pilla

rydi

latio

n;un

ilate

ral

in7,

bila

tera

lin

3.A

llfa

iled

med

ical

ther

apy

and

unde

rwen

tsu

rger

yw

ithin

2h

ofsi

gns

ofhe

rnia

tion.

●O

vera

ll40

%fu

nctio

nally

inde

pend

ent

surv

ival

with

30%

mor

talit

y.

No.

ofpa

tien

tsFu

ncti

onal

inde

pend

ence

(%)

Min

imal

assi

stan

ce(%

)D

(%)

Tota

l10

4030

30

Papo

etal

.(4

9)29

IIIG

CS

4–5

to�

10Su

rger

yan

dno

nsur

gica

lM

orta

lity

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of29

patie

nts

with

trau

mat

icIC

H/b

rain

lace

ratio

nto

anal

yze

the

corr

elat

ion

betw

een

GC

S,C

T,an

dIC

Pda

taan

dou

tcom

ew

ithor

with

out

oper

ativ

etr

eatm

ent.

ICP

data

reco

rded

via

intr

aven

tric

ular

cath

eter

.

●Pa

tient

sw

ithG

CS

4–5

onad

mis

sion

trea

ted

byop

erat

ion

with

in48

hof

inju

rydi

ed(n

ost

atis

tics)

.

No.

ofpa

tien

tsD

(%)

GC

S4

–55

100

GC

S6

–10

1346

.2

GC

S�

1011

0

Pate

let

al.

(50)

38III

GC

S�

8to

�13

Non

surg

ical

Mor

talit

yR

etro

spec

tive

seri

esof

57pa

tient

sin

itial

lym

anag

edno

nope

rativ

ely

who

subs

eque

ntly

requ

ired

cran

ioto

my

toat

tem

ptto

iden

tify

fact

ors

that

led

tofa

ilure

ofno

nope

rativ

em

anag

emen

t.A

subg

roup

of38

patie

nts

with

ICH

revi

ewed

.

●54

.5%

ofpa

tient

sw

ithIC

P�

20m

mH

gfa

iled

nono

pera

tive

man

agem

ent

in�

24h.

●Fr

onta

lIC

Hw

ere

sign

ifica

ntly

mor

epr

one

toea

rly

vers

usla

tefa

ilure

.●

Mor

talit

yin

faile

dIC

Hpa

tient

sw

as10

.5%

.

No.

ofpa

tien

tsD

(%)

Faile

dIC

H38

10.5



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Polin

etal

.(5

1)35

IIIM

ean

GC

S5.

62B

ifron

tal

deco

mpr

essi

vecr

anie

ctom

y

GO

Sat

disc

harg

eR

etro

spec

tive

case

-con

trol

seri

esof

35pa

tient

sun

derg

oing

bifr

onta

lde

com

pres

sive

cran

iect

omy

for

refr

acto

rypo

sttr

aum

atic

cere

bral

edem

aco

mpa

ring

preo

pera

tive

and

post

oper

ativ

eIC

P,ag

eve

rsus

outc

ome,

and

com

pari

ngth

egr

oup

asa

who

lew

ithth

egr

oup

from

TCD

Bm

atch

edfo

rag

e,ad

mis

sion

GC

S,se

x,an

dm

axim

alIC

Pin

term

sof

outc

ome.

Onl

yTC

DB

patie

nts

with

diffu

sein

jury

,no

sign

ifica

ntm

ass

lesi

on,

nom

idlin

esh

ift(II

I)(M

arsh

all

etal

.[3

9])

wer

eel

igib

lefo

rco

nsid

erat

ion

asco

ntro

ls.

●A

dmis

sion

GC

Sre

late

dto

outc

ome.

●O

pera

tion

�48

has

soci

ated

with

favo

rabl

eou

tcom

e(4

6%)

com

pare

dw

ith�

48h

(0%

).●

Sign

ifica

ntch

ange

inIC

Pbe

fore

vers

usaf

ter

deco

mpr

essi

on.

●Si

gnifi

cant

diffe

renc

ein

post

oper

ativ

eIC

Pve

rsus

cont

rol

ICP

48–7

2h

afte

rin

jury

.●

Coa

gulo

path

ydi

dno

taf

fect

outc

ome

insu

rgic

algr

oup.

●IC

P�

40m

mH

gsu

stai

ned

and

oper

atio

n�

48h

afte

rin

jury

had

sign

ifica

ntly

less

favo

rabl

eou

tcom

es.

●A

ge�

18yr

had

high

erra

teof

favo

rabl

eou

tcom

ein

surg

ical

grou

p.●

Surg

ery

age

�18

yrha

dhi

gher

rate

offa

vora

ble

outc

ome

than

mat

ched

cont

rols

ifop

erat

ed�

48h

and

ICP

�40

mm

Hg.

●M

edic

alm

anag

emen

tal

one

carr

ied

3.86

times

rela

tive

risk

ofun

favo

rabl

eou

tcom

eco

mpa

red

with

deco

mpr

essi

vecr

anie

ctom

y.●

Max

imum

bene

fitis

achi

eved

inpa

tient

sop

erat

edon

�48

haf

ter

inju

ryan

dw

ithIC

Pel

evat

ions

sust

aine

d�

40m

mH

g(6

0%fa

vora

ble

vers

us18

%in

mat

ched

cont

rols

).

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Surg

ery

k35

11.4

25.7

31.4

8.6

22.9

Cas

eco

ntro

l92

7.6

8.8

35.2

18.7

30.8

No.

ofpa

tien

tsG

R/M

D(%

)D

(%)

Coa

gulo

path

y18

33l

28

No

coag

ulop

athy

1741

18

ICP

�40

mm

Hg,

OR

�48

h15

6.7m

ICP

�40

,O

R�

48h

2060

Age

�18

yr18

44

Age

�18

yr17

29

OR

�48

h,IC

P�

4020

60n

ICP

�40

,co

ntro

l58

18

Age

�18

,IC

P�

40,

OR

�48

h10

80o

Age

�18

,IC

P�

40,

cont

rol

2524

kfa

vora

ble

outc

ome,

P�

0.01

4;lP

�n.

s.,

mP

�0.

0004

,n

P�

0.00

01,

oP

�0.

02.

Serv

adei

etal

.(5

4)72

4III

GC

S3–

12Su

rger

yan

dno

nsur

gica

lG

OS

6m

oPr

ospe

ctiv

est

udy

of72

4pa

tient

sw

ithm

oder

ate-

to-s

ever

eTB

Ito

dete

rmin

eth

efr

eque

ncy

ofch

ange

sin

CT

findi

ngs

and

thei

rpr

ogno

stic

impl

icat

ions

.

●16

%of

diffu

sein

juri

eson

initi

alC

Tde

mon

stra

ted

dete

rior

atio

non

subs

eque

ntC

T.●

12%

ofdi

ffuse

inju

ries

evol

ved

tom

ass

lesi

ons.

●Ev

olut

ion

from

diffu

sein

jury

with

out

swel

ling

orsh

ift(T

ype

II)to

am

ass

lesi

onw

asas

soci

ated

with

sign

ifica

ntin

crea

sein

risk

ofun

favo

rabl

eou

tcom

e.

●Tr

end

tow

ards

mor

efa

vora

ble

outc

ome

inpa

tient

sw

ithno

neva

cuat

edve

rsus

evac

uate

dm

ass

lesi

ons.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

Diff

use

inju

ryI,

noch

ange

7587

13

Diff

use

inju

ryI,

evol

utio

n5

8020

Diff

use

Inju

ryII,

noch

ange

185

6238

Diff

use

Inju

ryII,

evol

utio

np34

3862

Diff

use

Inju

ryIII

,no

chan

ge57

3367

Diff

use

Inju

ryIII

,ev

olut

ion

757

43

Diff

use

Inju

ryIV

,no

chan

ge14

2179

Diff

use

Inju

ryIV

,ev

olut

ion

475

25

Mas

sle

sion

343

4357

pP

�0.

01.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD (%

)V

S(%

)D (%

)

Diff

use

inju

ryI

8074

129

05

Diff

use

inju

ryII

219

3622

202

21

Diff

use

inju

ryIII

6420

169

253

Diff

use

inju

ryIV

1811

2211

056



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Sing

ouna

s(5

7)16

IIIG

CS

8–1

0U

nkno

wn

Mor

talit

yR

etro

spec

tive

revi

ewof

48ch

ildre

n�

14yr

with

seve

rehe

adin

jury

(ext

radu

ral

and

subd

ural

hem

atom

asex

clud

ed)

toco

mm

ent

onre

lativ

ely

good

prog

nosi

s.M

orta

lity

for

16pa

tient

sw

ithbr

ain

edem

aon

lyw

asre

port

ed.

No.

ofpa

tien

tsD

(%)

Diff

use

edem

a,no

foca

lle

sion

1612

.5

Solo

niuk

etal

.(5

8)35

IIIG

CS

�8

or�

8Su

rger

yG

OS

6m

oR

etro

spec

tive

revi

ewof

35pa

tient

sun

derg

oing

cran

ioto

my

for

evac

uatio

nof

ICH

toev

alua

tech

arac

teri

stic

sof

timin

gof

appe

aran

ceof

ICH

and

fact

ors

rela

ted

toou

tcom

e.IC

Hde

fined

asho

mog

eneo

usco

ales

cent

area

ofhi

ghat

tenu

atio

nm

easu

ring

2cm

orgr

eate

r.

●O

nly

26%

ofIC

Hre

quir

ing

evac

uatio

nar

epr

esen

ton

CT

with

in6

hof

inju

ry.

●49

%ov

eral

l1-

yrm

orta

lity;

71%

with

GC

S�

8.●

Mor

talit

yva

ries

with

loca

tion,

with

panh

emis

pher

ic,

tem

pora

l,an

dfr

onta

lra

tes

of80

%,

57%

,an

d37

%,

resp

ectiv

ely.

No.

ofpa

tien

tsD

(%)

ICH

,op

erat

ed35

49

Spri

cket

al.

(59)

34III

GC

S3

to�

7Su

rger

yan

dno

nsur

gica

lG

OS,

not

spec

ified

Ret

rosp

ectiv

ese

ries

of34

patie

nts

with

DTI

CH

ofat

leas

t2.

5-cm

diam

eter

inpr

evio

usly

norm

alar

eaon

initi

alC

Tto

eval

uate

outc

ome

and

clin

ical

char

acte

rist

ics.

●53

%D

TIC

Hde

velo

ped

with

in24

haf

ter

adm

issi

onan

d79

%w

ithin

48h

No.

ofpa

tien

tsG

R(%

)D

isab

led

(%)

D(%

)

All

3417

.655

.926

.5O

pera

ted

714

.342

.942

.9

Stat

ham

etal

.(6

0)18

IIIG

CS

5–14

Surg

ery

(n�

1)an

dno

nsur

gica

lG

OS

6m

oR

etro

spec

tive

revi

ewof

18pa

tient

sw

ithpr

imar

yC

Tdi

agno

sis

offr

onta

lco

ntus

ion,

defin

edas

regi

ons

ofm

ixed

high

and

low

orlo

wat

tenu

atio

nin

the

fron

tal

lobe

s,to

atte

mpt

tode

term

ine

fact

ors

that

may

lead

tone

urol

ogic

alde

teri

orat

ion

and

wor

seou

tcom

e.

●G

ood

outc

ome

was

achi

eved

in72

%of

patie

nts

with

fron

tal

cont

usio

n.●

Bot

hca

ses

ofde

teri

orat

ion

wer

eas

soci

ated

with

exte

nsiv

ebi

late

ral

inju

ry(n

ost

atis

tics)

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Uni

late

ral

cont

usio

n10

900

00

10Li

mite

dbi

late

ral

cont

usio

ns5

4040

200

0

Exte

nsiv

ebi

late

ral

cont

usio

ns3

670

00

33

Fron

tal

cont

usio

ns,

tota

l18

72.2

11.1

5.6

011

.1

Tayl

oret

al.

(61)

27II

GC

S4

–11

Dec

ompr

essi

vebi

tem

pora

lcr

anie

ctom

yan

dno

nsur

gica

l

GO

S6

mo;

Hea

lthSt

ate

Util

ityIn

dex

6m

o

Pros

pect

ive,

rand

omiz

ed,

cont

rolle

dtr

ial

of27

child

ren

with

TBI

and

sust

aine

din

trac

rani

alhy

pert

ensi

onto

eval

uate

effe

ctof

earl

yde

com

pres

sive

bite

mpo

ral

cran

iect

omy

onou

tcom

ean

don

cont

rol

ofIC

P.

●Tr

end

tow

ards

grea

ter

decr

ease

inIC

Pin

deco

mpr

essi

ongr

oup

(P�

0.05

7).

●Tr

end

tow

ards

shor

ter

ICU

stay

inde

com

pres

sion

grou

p(P

�0.

12).

●Tr

end

tow

ards

impr

oved

outc

ome

inde

com

pres

sion

grou

p(P

�0.

046

[P�

0.02

21re

quir

edfo

rsi

gnifi

canc

e]).

No.

ofpa

tien

tsFa

vora

ble

(%)

Unf

avor

able

(%)

Dec

ompr

essi

on13

53.8

q46

.1

Con

trol

1414

.385

.7q

P�

0.04

6(P

�0.

0221

requ

ired

for

sign

ifica

nce)

.Fa

vora

ble:

norm

alor

func

tiona

llyno

rmal

but

requ

irin

gm

edic

atio

nor

med

ical

supe

rvis

ion.

Unf

avor

able

:m

oder

ate

disa

bilit

y,de

pend

ent,

seve

rely

disa

bled

/veg

etat

ive,

orde

ad.

Teas

dale

etal

.(6

2)37

IIIG

CS

3–8

Non

surg

ical

GO

S6

mo

Ret

rosp

ectiv

ese

ries

of37

patie

nts

diag

nose

dw

ithse

vere

diffu

sehe

adin

jury

(with

out

mid

line

shift

of�

5m

m)

tode

term

ine

the

rela

tions

hip

betw

een

com

pres

sion

ofba

sal

cist

erns

/3rd

vent

ricl

ean

del

evat

edIC

P.

●H

ighl

ysi

gnifi

cant

rela

tions

hip

betw

een

stat

usof

basa

lci

ster

ns/II

Ive

ntri

cle

and

leve

lof

ICP.

●Pr

esen

ceve

rsus

abse

nce

ofci

ster

ns/II

Ive

ntri

cle

seem

edto

corr

elat

ew

ithou

tcom

ew

ithin

GC

Sgr

oups

(no

stat

istic

s).

No.

ofpa

tien

tsG

R/M

D(%

)SD

(%)

VS/

D(%

)

GC

S3–

5,ab

sent

cist

erns

/3rd

vent

ricl

e11

9.1

090

.9

GC

S3–

5,pr

esen

tci

ster

ns/3

rdve

ntri

cle

520

2060

GC

S6

–8,

abse

ntci

ster

ns/3

rdve

ntri

cle

837

.525

37.5

GC

S6

–8,

pres

ent

cist

erns

/3rd

vent

ricl

e13

61.5

15.4

23.1



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Tsen

g(6

3)32

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

GO

S1

yrR

etro

spec

tive

seri

esof

32pa

tient

sw

ithD

TIC

Hto

eval

uate

clin

ical

and

radi

olog

icfa

ctor

saf

fect

ing

outc

ome.

●H

emat

oma

volu

me,

cist

erna

lef

face

men

t,tim

ing

ofde

tect

ion,

occu

rren

ceof

clin

ical

dete

rior

atio

n,an

dG

CS

attim

eof

follo

w-u

pC

Tw

ere

rela

ted

sign

ifica

ntly

toou

tcom

e.●

16%

over

all

mor

talit

y.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

GC

S3–

7at

follo

w-u

pC

Tr12

3367

GC

S8

–12

atfo

llow

-up

CT

1610

00

GC

S13

–15

atfo

llow

-up

CT

410

00

Cis

tern

effa

cem

ents

50

100

No

cist

ern

effa

cem

ent

2788

.911

.1r

P�

0.00

5,s

P�

0.00

1.

Tsen

g(6

4)32

IIIG

CS

5–7

Cra

niot

omy

�m

anua

lte

mpo

ral

lobe

redu

ctio

n

GO

S3

mo

Ret

rosp

ectiv

est

udy

of32

patie

nts

with

GC

S5–

7,C

Tev

iden

cefo

run

cal

hern

iatio

n,an

isoc

oria

,an

dhe

mip

ares

istr

eate

dei

ther

with

stan

dard

surg

ical

proc

edur

e(i.

e.,

cran

ioto

my

with

hem

atom

aev

acua

tion

and

cont

usio

nre

sect

ion)

orw

ithst

anda

rdpr

oced

ure

plus

man

ual

redu

ctio

nof

hern

iate

dte

mpo

ral

lobe

.G

OS

mea

sure

dat

�3

mo

post

opto

com

pare

outc

ome

betw

een

grou

ps.R

educ

tion

proc

edur

ede

scri

bed

asge

ntle

elev

atio

nof

tem

pora

llo

beun

tilte

ntor

ium

isvi

sual

ized

and

CSF

egre

ssis

note

d.

●A

dditi

onof

tem

pora

llo

bere

duct

ion

toth

est

anda

rdsu

rgic

alpr

oced

ure

seem

edto

resu

ltin

bette

rou

tcom

e(n

ost

atis

tics;

sele

ctio

nbi

as).

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD (%

)V

S(%

)D (%

)

Cra

niec

tom

y�

cont

usio

n�

redu

ctio

n10

2070

00

10

Cra

niec

tom

y�

cont

usio

n,no

redu

ctio

n

220

279

955

Uzz

ell

etal

.(6

5)11

7III

GC

S4

–6

Non

surg

ical

GO

S6

mo;

neur

o-ps

ycho

logi

cal

mea

n4.

8m

o(n

�30

)

Ret

rosp

ectiv

ese

ries

of11

7pa

tient

sw

ithD

AI,

diffu

sesw

ellin

g,or

foca

lin

jury

byC

Tto

char

acte

rize

diffe

rent

ial

outc

ome.

Neu

rops

ycho

logi

cal

test

ing

was

perf

orm

edin

asu

bset

of30

patie

nts

toch

arac

teri

zedi

ffere

ntia

lne

urop

sych

olog

ical

sequ

elae

base

don

lesi

onty

pe.

Patie

nts

with

SDH

,ED

H,

SAH

,or

deve

lopm

ent

ofa

seco

ndm

ajor

lesi

onw

ere

excl

uded

.D

AI

defin

edas

�mid

line

hem

orrh

ages

orm

ultip

lesm

all

hem

orrh

ages

atgr

ay-w

hite

mat

ter

inte

rfac

es�;

diffu

sesw

ellin

gde

fined

as�n

arro

win

gor

oblit

erat

ion

ofve

ntri

cles

orba

sila

rci

ster

ns�;

foca

lin

juri

esde

fined

as�u

nila

tera

lin

trac

ereb

ral

hem

orrh

ages

orco

ntus

ions

.�

●Pa

tient

sw

ithD

AI

had

sign

ifica

ntly

high

erm

orta

lity

and

sign

ifica

ntly

less

“goo

dre

cove

ry”

than

othe

rgr

oups

.●

Mem

ory

and

lear

ning

scor

ebu

tno

tin

telli

genc

eor

visu

omot

orsp

eed

scor

esi

gnifi

cant

lydi

ffere

dam

ong

CT

grou

ps.

●Le

vels

ofm

emor

y,le

arni

ng,

and

visu

omot

orsp

eed

wer

ehi

gher

afte

rdi

ffuse

swel

ling

inju

ries

,bu

tim

prov

emen

tw

asle

ssov

ertim

e.●

Patie

nts

with

DA

Iha

dgr

eate

rim

prov

emen

tov

ertim

eof

mem

ory,

lear

ning

,an

dvi

suom

otor

spee

d.●

Patie

nts

with

foca

lin

juri

esha

dim

prov

emen

tov

ertim

ein

visu

omot

orsp

eed,

but

not

reca

llor

lear

ning

.●

Dia

gnos

ticgr

oups

did

not

diffe

rsi

gnifi

cant

lyin

age,

year

sof

educ

atio

n,G

CS,

timin

gof

neur

opsy

chol

ogic

alte

stin

g,se

x,ha

nded

ness

,or

GO

Sat

6m

o.

CT

grou

ptN

o.of

pati

ents

GR

(%)

MD

(%)

SD (%)

VS

(%)

D (%)

DA

I33

1815

1512

40D

iffus

esw

ellin

g34

6217

60

15

Foca

lin

jury

5048

1514

022

tP

�0.

01.

Vol

lmer

etal

.(6

6)32

1III

All

GC

SSu

rger

yan

dno

nsur

gica

lG

OS

6m

oPr

ospe

ctiv

est

udy

from

1984

–198

7of

patie

nts

with

GC

S3–

15to

eval

uate

effe

ctof

age

onou

tcom

ein

diffe

rent

popu

latio

ns,

incl

udin

gpa

tient

sw

ithla

rge

(�15

cm3 )

ICH

requ

irin

gev

acua

tion

(n�

54)

and

patie

nts

with

�15

cm3

ICH

(n�

267)

.

●Pr

esen

ceof

ICH

and

ofla

rge

ICH

(�15

cm3 )

incr

ease

dsi

gnifi

cant

lyw

ithag

e.●

Incr

easi

ngag

ew

asin

depe

nden

tlypr

edic

tive

ofpo

orou

tcom

ein

larg

ean

dsm

all

ICH

subg

roup

s.

No.

ofpa

tien

tsV

S/D

(%)

ICH

�15

cm3u

267

Age

16–2

5yr

31.8

Age

26–3

5yr

27.7

Age

36–

45yr

42.4

Age

46–5

5yr

62.2

Age

�56

yr82

.2u

P�

0.00

1



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Whi

tfiel

det

al.

(67)

26III

GC

S3–

13B

ifron

tal

deco

mpr

essi

vecr

anie

ctom

y

GO

S6

mo

Ret

rosp

ectiv

ere

view

ofou

tcom

efo

r26

patie

nts

unde

rgoi

ngbi

fron

tal

deco

mpr

essi

vecr

anie

ctom

yfo

rre

frac

tory

intr

acra

nial

hype

rten

sion

(ICP

�30

mm

Hg

with

CPP

�70

mm

Hg;

orIC

P�

35m

mH

gir

resp

ectiv

eof

CPP

)m

anag

edus

ing

ast

anda

rdiz

edIC

P/C

PPtr

eatm

ent

algo

rith

m.

Con

tinuo

usIC

P,A

BP,

and

CPP

data

wer

ean

alyz

edfo

r8

patie

nts

toex

amin

eef

fect

ofbi

fron

tal

deco

mpr

essi

vecr

anie

ctom

yon

thes

eva

riab

les.

Surg

ical

tech

niqu

ede

scri

bed.

●61

%of

seve

reTB

Isu

bgro

upha

dfa

vora

ble

outc

ome

(GO

S4

–5).

●IC

Pw

assi

gnifi

cant

lyre

duce

dfr

om37

.5to

18.1

mm

Hg

(P�

0.00

3).

●C

PPdi

ffere

nces

wer

eno

tsi

gnifi

cant

.●

Am

plitu

deof

ICP

wav

efor

m,

ampl

itude

ofsl

oww

aves

,an

dR

AP

coef

ficie

nt(r

efle

ctin

gin

crea

sed

com

pens

ator

yre

serv

e)w

ere

sign

ifica

ntly

redu

ced

bysu

rger

yin

asu

bgro

upof

eigh

tpa

tient

s.●

Tim

ing

ofsu

rger

yw

asno

tpr

edic

tive

ofou

tcom

e.N

o.of

pati

ents

GR

/MD

(%)

SD(%

)V

S(%

)D

(%)

All

2669

80

23A

ge�

166

83.3

00

16.7

GC

S3–

818

61.1

11.1

027

.8G

CS

9–1

25

100

00

0G

CS

133

66.7

00

33.3

Tim

ing

ofsu

rger

yN

o.of

pati

ents

Favo

rabl

e(%

)U

nfav

orab

le(%

)O

R�

48h

afte

rin

jury

2060

40O

R�

48h

afte

rin

jury

610

00

Wu

etal

.(6

8)35

IIIA

llG

CS

Surg

ery

GO

S6

mo

Ret

rosp

ectiv

ese

ries

of48

9pa

tient

str

eate

dsu

rgic

ally

for

intr

acra

nial

hem

atom

asto

asse

sspr

ogno

stic

fact

ors

rela

ted

toG

OS

at6

mo.

Asu

bgro

upof

35pa

tient

sw

ithIC

His

incl

uded

.

●Po

stre

susc

itatio

nG

CS,

preo

pera

tive

chan

gein

pupi

llary

size

,N

o.of

oper

atio

ns,

No.

ofhe

mat

omas

,an

dty

peof

lesi

onw

ere

sign

ifica

ntly

rela

ted

tosu

rgic

alou

tcom

e.●

Goo

dou

tcom

e:si

ngle

,ev

acua

ted

EDH

�IC

H�

SDH

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Acu

teIC

H,

evac

uate

d35

6022

.98.

62.

85.

7

Yam

aki

etal

.(6

9)48

IIIU

nkno

wn

Surg

ery

and

nons

urgi

cal

GO

S;tim

eof

max

imal

hem

atom

adi

amet

er

Ret

rosp

ectiv

ere

view

of48

patie

nts

with

trau

mat

icIC

H�

3-cm

diam

eter

who

sefir

stC

Toc

curr

edw

ithin

6h

ofin

jury

toev

alua

teth

etim

ing

ofhe

mat

oma

evol

utio

nan

dto

asse

ssou

tcom

e.C

Tsc

ans

of32

nono

pera

ted

patie

nts

wer

eus

edto

asse

ssna

tura

lhi

stor

yof

hem

atom

aev

olut

ion.

●A

llIC

H�

3-cm

diam

eter

deve

lope

dw

ithin

24h

ofin

jury

.●

Onl

y56

%of

ICH

�3-

cmdi

amet

erde

velo

ped

with

in6

hof

inju

ry.

●IC

Hre

ache

dm

axim

alsi

zein

84%

ofpa

tient

sw

ithin

12h.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

ICH

,to

tal

5853

142

328

ICH

,op

erat

ed26

358

47

46N

o.of

pati

ents

GR

(%)

Non

oper

ated

3213

Ope

rate

d26

46A

dmitt

edim

med

iate

lyaf

ter

TBI

1631

Adm

itted

�6

haf

ter

TBI

1070

You

nget

al.

(72)

15III

Gra

dyC

oma

Scal

e1–

5

Surg

ery

Mod

ified

GO

S,no

tsp

ecifi

edR

etro

spec

tive

seri

esaf

ter

intr

oduc

tion

ofC

Tof

15pa

tient

str

eate

dsu

rgic

ally

for

DTI

CH

toex

amin

ecl

inic

alan

dra

diol

ogic

corr

elat

es,

and

tore

late

clin

ical

stat

usto

outc

ome.

●M

orbi

dity

/mor

talit

yas

soci

ated

with

com

asc

ore

onad

mis

sion

(no

stat

istic

s).

●80

%oc

cipi

tal/p

arie

tooc

cipi

tal

impa

ct;

cont

reco

upfr

onto

tem

pora

lle

sion

s.●

Def

initi

onof

DTI

CH

:de

velo

psin

area

ofpr

evio

usco

ntus

ion.

No.

ofpa

tien

tsIn

tact

(%)

Mild

disa

bilit

y(%

)

Sign

ifica

ntdi

sabi

lity

(%)

D(%

)

Gra

dy1–

26

83.3

16.7

00

Gra

dy3–

49

00

22.2

77.8

DTI

CH

,to

tal

1533

.36.

713

.346

.7

Zum

kelle

ret

al.

(73)

53III

GC

S�

5to

11–1

5Su

rger

yan

dno

nsur

gica

lPo

orve

rsus

good

,no

tsp

ecifi

ed

Ret

rosp

ectiv

ese

ries

of10

4pa

tient

sw

ithIC

H,

53of

who

mpr

esen

ted

with

trau

mat

icIC

H,

com

pari

ngop

erat

ive

vers

usno

nope

rativ

etr

eatm

ent

with

outc

ome.

●N

ost

atis

tical

diffe

renc

ein

outc

ome

was

foun

dbe

twee

nop

erat

edan

dno

nope

rate

dgr

oups

.

No.

ofpa

tien

tsG

ood

(%)

Poor

(%)

Ope

rate

d31

7129

Non

oper

ated

2240

.959

.1Si

ngle

lobe

,op

erat

ed17

70.6

29.4

Sing

lelo

be,

non-

oper

ated

875

25P

�n.

s.

aG

CS,

Gla

sgow

Com

aSc

ale;

GO

S,G

lasg

owou

tcom

esc

ore;

ICH

,int

race

rebr

alhe

mor

rhag

e;IC

P,in

trac

rani

alpr

essu

re;C

T,co

mpu

ted

tom

ogra

phic

;GR

,goo

dre

cove

ry;M

D,m

oder

ate

disa

bilit

y;V

S,ve

geta

tive

stat

e;D

,dea

th;S

D,s

ever

edi

sabi

lity;

DTI

CH

,del

ayed

trau

mat

icin

trac

ereb

ralh

emat

oma;

SDH

,sub

dura

lhem

atom

a;n.

s.,n

otsi

gnifi

cant

;FIM

,fun

ctio

nali

ndep

ende

nce

mea

sure

;ED

H,

epid

ural

hem

atom

a;LO

C,l

oss

ofco

nsci

ousn

ess;

TCD

B,T

raum

atic

Com

aD

ata

Ban

k;SE

P,so

mat

osen

sory

evok

edpo

tent

ial;

CPP

,cer

ebra

lper

fusi

onpr

essu

re;A

EP,a

cous

ticev

oked

pote

ntia

l;ST

D,

subt

empo

ral

deco

mpr

essi

on;

TL,

tem

pora

llo

bect

omy;

MU

C,

mul

tiple

unila

tera

lco

ntus

ion;

GB

S,ge

nera

lized

brai

nsw

ellin

g;D

AI,

diffu

seax

onal

inju

ry;

TBI,

trau

mat

icbr

ain

inju

ry;

SAH

,su

bara

chno

idhe

mor

rhag

e;IV

H,

intr

aven

tric

ular

hem

orrh

age;

AB

P,ar

teri

albl

ood

pres

sure

;R

AP,

rena

lar

tery

pres

sure

;IC

U,

inte

nsiv

eca

reun

it;O

R,

oper

atio

n.

CHAPTER 6

SURGICAL MANAGEMENT OF POSTERIOR FOSSA

MASS LESIONSM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia











Indications• Patients with mass effect on computed tomographic (CT) scan or with neurological

dysfunction or deterioration referable to the lesion should undergo operative inter-vention. Mass effect on CT scan is defined as distortion, dislocation, or obliterationof the fourth ventricle; compression or loss of visualization of the basal cisterns, orthe presence of obstructive hydrocephalus.

• Patients with lesions and no significant mass effect on CT scan and without signs ofneurological dysfunction may be managed by close observation and serial imaging.

Timing• In patients with indications for surgical intervention, evacuation should be performed as

soon as possible because these patients can deteriorate rapidly, thus, worsening theirprognosis.

Methods• Suboccipital craniectomy is the predominant method reported for evacuation of

posterior fossa mass lesions, and is therefore recommended.

KEY WORDS: Cerebellum, Coma, Computed tomographic parameters, Contusion, Head injury, Occipital,Posterior fossa, Surgical technique, Timing of surgery, Traumatic brain injury

Neurosurgery 58:S2-47-S2-55, 2006 DOI: 10.1227/01.NEU.0000210366.36914.38 www.neurosurgery-online.com

OVERVIEW

Posterior fossa injury is rare, occurring in lessthan 3% of head injuries in most published series(8, 12, 22). The vast majority of these series dealexclusively with posterior fossa epidural hemato-mas (EDH), representing 1.2 to 12.9% of all EDH(7, 20, 21, 30). A small number of observationalstudies address subdural and intraparenchymalhematomas of the posterior fossa (5, 8, 17, 22, 24),representing 0.5 to 2.5% and 1.7% of all subduralhematomas and intraparenchymal hematomas,respectively (22, 24). Additionally, there is a sepa-rate literature that focuses on parturitional hemor-rhages, up to 48% of which primarily involve theposterior fossa (27). Because of the physiology andanatomy of the neonate, and the unique mecha-nism of these injuries, this subgroup of patientswarrants independent analysis, and will not beaddressed in these Guidelines.

Despite the rarity of these lesions, the impor-tance of timely recognition and surgical evacu-

ation, when indicated, cannot be overstated.Many patients can undergo rapid clinical dete-rioration because of the limited size of the pos-terior fossa and the propensity for these lesionsto produce brainstem compression.

PROCESS

A MEDLINE computer search using the fol-lowing key words: “posterior fossa” or “cerebel-lum” or “cerebellar” or “occipital” and “sub-dural” or “epidural” or “extradural” or“intradural” or “parenchymal” or “intraparen-chymal” or “intracerebellar” or “fracture” be-tween 1975 and 2001 was performed. A total of1828 documents were found. The search wasnarrowed to include the key words: “surgery”or “operative” or “craniotomy” or “craniec-tomy” or “decompressive craniectomy” or “re-pair” and “trauma” or “traumatic” or “TBI” or“CHI.” A total of 430 articles were found. Atertiary search adding the key words “contu-


sion,” “hemorrhagic contusion,” “surgical decompression,” “cra-niostomy,” “TICH,” and “DTICH” was performed, yielding 421articles. The secondary and tertiary searches were combined,yielding a total of 433 articles. In addition, the reference lists ofselected articles were reviewed, and 24 articles were selected forcritical analysis. The results of this analysis were incorporatedinto the review presented here. Papers primarily addressing thefollowing topics were not included: nontraumatic lesions, pa-tients with associated posterior fossa anomalies (e.g., Chiari mal-formation), posttraumatic aneurysms, chronic subdural hemato-mas, vertebral artery dissection, patients undergoinganticoagulation therapy, patients with associated illnesses (e.g.,acquired immunodeficiency syndrome, idiopathic thrombocyto-penia purpura, hemophilia, arteriovenous malformation, aftercraniotomy, or von Willebrand’s disease), pre-CT era reports,and book chapters. In general, papers with the following char-acteristics were also excluded: case series with less than 10 pa-tients evaluated by CT scan and with incomplete outcome data(mortality or Glasgow outcome score [GOS]), case reports, andoperative series with operations occurring longer than 14 daysfrom injury. Several articles with case series of less than 10patients were examined and reviewed because of the limitednumber of patient series evaluating primary traumatic posteriorfossa mass lesions that exist in the literature. Selected articleswere evaluated for design, prognostic significance, therapeuticefficacy, and overall outcome. In addition, several articles werereviewed for the purposes of historical perspective.


Because of the rapid and life-threatening nature of neuro-logical deterioration secondary to expanding mass lesions inthe limited compartment of the posterior fossa, surgery isgenerally viewed as required therapy in symptomatic patientswith progressive dysfunction. Because of the potential adverseconsequences of withholding or delaying surgery for such pa-tients, studies depend on retrospective analyses. As a result,there is no Class I or Class II evidence to support recommenda-tions for the surgical management of these injuries. However, thepredominantly observational studies that were reviewed yieldan important and relatively clear picture of the prognosis for thepatient with a posterior fossa mass lesion as patients are cur-rently managed. Admission Glasgow Coma Scale (GCS) score (5,7, 8, 12, 24, 26) and GCS score at surgery (6, 14, 19, 20, 22)correlate with outcome (GOS and mortality). D’Avella et al. (5)retrospectively reviewed the clinical and radiographic character-istics of 81 patients with traumatic intracerebellar hemorrhages.Subjecting their data to multivariate analysis, they found thatonly GCS and the presence of concomitant supratentorial lesionsindependently predicted outcome at 6 months. Outcome wasfavorable (GOS, 4 or 5) in 95% of patients with admission GCSscore of at least 8, whereas outcome was poor (GOS, 1–3) in 81%of patients with a GCS score less than 8. Class III data suggeststhat a neurologically deteriorating patient should undergo emer-gent evacuation of the mass lesion.

Neurologically intact patients with a posterior fossa lesionand no CT evidence for mass effect (compression of cisterns,distortion of 4th ventricle, hydrocephalus) have been success-fully managed nonoperatively with close observation and se-rial imaging (1, 5, 15, 25, 28).

Wong (28) conducted a retrospective study of 25 patientswith posterior fossa EDH and compared clinical and radiolog-ical characteristics and outcomes between 17 patients under-going early surgery and 8 patients managed nonoperatively.Patients with a posterior fossa EDH of at most 10 cm3 involume, at most 15 mm in thickness, and responsible for atmost a 5-mm midline shift had excellent survival rates witheither surgical or nonsurgical treatment. Patients managednonsurgically with a posterior fossa EDH greater than 10 cm3

in volume, greater than 15 mm in thickness, and responsiblefor greater than 5 mm of midline shift had a significantlygreater mortality than those with similar CT characteristicsundergoing early surgery. The disparity in mortality, how-ever, is confounded by the strong correlation between thepresence of an associated frontal lesion (which was foundmore commonly in the latter group) and death. This reportraises the concept of conservative management for patientswith posterior fossa lesions on the basis of objective CT char-acteristics. This concept is also supported by a two-centerstudy performed by Bozbuga et al. (1). The authors dividedpatients into management groups on the basis of CT charac-teristics. All patients (n � 14) without evidence of mass effecton CT scan, defined by obliteration of the perimesencephaliccisterns, compression and/or displacement of the fourth ven-tricle, or the presence of hydrocephalus, were managed non-operatively and had a good outcome (GOS, 5). According tothe authors, these objective criteria were “earlier, more pre-dictive, and more reliable” than the clinical findings, althoughno statistics regarding this statement were performed. Severalother case series add additional support (9, 16, 25).

There are several prognostic factors that adversely affectoutcome regardless of management. These include the pres-ence of associated intracranial lesions ( 4, 5, 12, 14, 18–22, 24,28), extension of an infratentorial lesion into the supratentorialcompartment (19), the location of the lesion (e.g., intraparen-chymal versus extra-axial, and midline versus hemispheric)(17, 22), the presence or absence of associated hydrocephalus(7, 8, 14), and the acuity of presentation, with subacute pre-sentation portending a better outcome than acute presentation(4, 14, 19, 23–25). There are no controlled studies measuringthe impact of these variables on surgical versus nonsurgicalmanagement of posterior fossa mass lesions.

SUMMARY

There are no controlled, prospective clinical trials of treat-ment using surgical versus nonsurgical management of pos-terior fossa mass lesions. The available data support rapidevacuation of posterior fossa mass lesions that 1) show CTevidence of mass effect, or 2) result in progressive neurologicaldysfunction. Moreover, data support expectant management



with serial imaging for select cases in which there is neuro-logical stability and no radiological evidence for mass effect.


There are several patient groups in which the distinction be-tween surgical and conservative management is blurred. Onesuch group includes patients who present with neurologicaldeficit and a traumatic posterior fossa mass lesion without clin-ical evidence for neurological deterioration or radiological evi-dence for mass effect. Conversely, another group includes theneurologically intact patient with radiological evidence for masseffect from an offending hematoma. These groups have not beenadequately addressed in the current literature, and, when re-ported, are managed at the discretion of the individual neuro-surgeon, thus, precluding an accurate assessment of efficacy oftreatment. The literature contains methodological problems out-lined in this supplement that preclude the establishment of man-agement standards, and even of treatment guidelines, for poste-rior fossa injury. Most series present prognostic data regardingoutcome after either conservative or surgical treatment of poste-rior fossa mass lesions. In those few studies that attempt tocompare outcomes, important prognostic factors known to berelevant to TBI outcome, such as cardiorespiratory instability,other systemic injuries, comorbidities, etc. (2), are not controlledbetween the surgical and nonsurgical cohorts. As a result, wehave important prognostic information regarding operativelyand nonoperatively managed posterior fossa injury, but nomeans for valid, direct comparison between the two. This com-parison is essential if we are attempting to establish a standard ofcare. Thus, attention needs to be directed to controlled studies ofpatients with similar CT and clinical characteristics who aremanaged with operative versus nonoperative intervention.

REFERENCES

1. Bozbuga M, Izgi N, Polat G, Gurel I: Posterior fossa epidural hematomas:Observations on a series of 73 cases. Neurosurg Rev 22:34–40, 1999.

2. Brain Trauma Foundation, American Association of Neurological Surgeons,Joint Section on Neurotrauma and Critical Care: Guidelines for the manage-ment of severe traumatic brain injury. J Neurotrauma 17:457–462, 2000.

3. Brambilla G, Rainoldi F, Gipponi D, Paoletti P: Extradural haematoma of theposterior fossa: A report of eight cases and a review of the literature. ActaNeurochir (Wien) 80:24–29, 1986.

4. Ciurea AV, Nuteanu L, Simionescu N, Georgescu S: Posterior fossa extra-dural hematomas in children: Report of nine cases. Childs Nerv Syst 9:224–228, 1993.

5. d’Avella D, Servadei F, Scerrati M, Tomei G, Brambilla G, Angileri FF,Massaro F, Cristofori L, Tartara F, Pozzati E, Bruni P, Delfini R, TomaselloF: Traumatic intracerebellar hemorrhages: A clinicoradiological analysis of81 cases. Neurosurgery 50:16–25, 2002.

6. Ersahin Y, Mutluer S: Posterior fossa extradural hematomas in children.Pediatr Neurosurg 19:31–33, 1993.

7. Holzschuh M, Schuknecht B: Traumatic epidural haematomas of the poste-rior fossa: 20 new cases and a review of the literature since 1961. BrJ Neurosurg 3:171–180, 1989.

8. Karasawa H, Furuya H, Naito H, Sugiyama K, Ueno J, Kin H: Acutehydrocephalus in posterior fossa injury. J Neurosurg 86:629–632, 1997.

9. Kawakami Y, Tamiya T, Tanimoto T, Shimamura Y, Hattori S, Ueda T,Ishida T: Nonsurgical treatment of posterior fossa epidural hematoma.Pediatr Neurol 6:112–118, 1990.

10. Koc RK, Pasaoglu A, Menku A, Oktem S, Meral M: Extradural hematoma ofthe posterior cranial fossa. Neurosurg Rev 21:52–57, 1998.

11. Lui T, Lee S, Chang C, Cheng W: Epidural hematomas in the posteriorcranial fossa. J Trauma 34:211–215, 1993.

12. Mahajan R, Sharma B, Khosla V, Tewari M, Mathuriya S, Pathak A, Kak V:Posterior fossa extradural haematoma—Experience of nineteen cases. AnnAcad Med Singapore 22:410–413, 1993.

13. Mohanty A, Kolluri V, Subbakrishna D, Satish S, Mouli B, Das B: Prognosisof extradural haematomas in children. Pediatr Neurosurg 23:57–63, 1995.

14. Neubauer UJ: Extradural haematoma of the posterior fossa. Twelve yearsexperiences with CT-scan. Acta Neurochir (Wien) 87:105–111, 1987.

15. Otsuka S, Nakatsu S, Matsumoto S, Sato S, Motozaki T, Ban S, Yamamoto T:Study on cases with posterior fossa epidural hematoma—Clinical featuresand indications for operation. Neurol Med Chir (Tokyo) 30:24–28, 1990.

16. Pang D, Horton J, Herron J, Wilberger JJ, Vries J: Nonsurgical managementof extradural hematomas in children. J Neurosurg 59:958–971, 1983.

17. Pozzati E, Grossi C, Padovani R: Traumatic intracerebellar hematomas. JNeurosurg 56:691–694, 1982.

18. Pozzati E, Tognetti F, Cavallo M, Acciarri N: Extradural hematomas of theposterior cranial fossa. Observations on a series of 32 consecutive casestreated after the introduction of computed tomography scanning. SurgNeurol 32:300–303, 1989.

19. Prusty GK, Mohanty A: Posterior fossa extradural haematoma. J IndianMed Assoc 93:255–258, 1995.

20. Rivano C, Altomonte M, Capuzzo T, Borzone M: Traumatic posterior fossaextradural hematomas. A report of 22 new cases surgically treated and areview of the literature. Zentralbl Neurochir 52:77–82, 1991.

21. Roda J, Gimenez D, Perez-Higueras A, Blazquez M, Perez-Alvarez M:Posterior fossa epidural hematomas: A review and synthesis. Surg Neurol19:419–424, 1983.

22. Sripairojkul B, Saeheng S, Ratanalert S, Pheunpathom N, Sriplung H: Trau-matic hematomas of the posterior cranial fossa. J Med Assoc Thai 81:153–159, 1998.

23. St John JN, French BN: Traumatic hematomas of the posterior fossa. Aclinicopathological spectrum. Surg Neurol 25:457–466, 1986.

24. Tsai FY, Teal JS, Itabashi HH, Huprich JE, Hieshima GB, Segall HD: Com-puted tomography of posterior fossa trauma. J Comput Assist Tomogr4:291–305, 1980.

25. Vrankovic DJ, Splavski B, Hecimovic I, Kristek B: Acute traumatic hemato-mas of the posterior fossa: Experience with eleven cases. Neurol Croat43:21–30, 1994.

26. Wang EC, Lim AY, Yeo TT: Traumatic posterior fossa extraduralhaematomas (PFEDH). Singapore Med J 39:107–111, 1998.

27. Welch K, Strand R: Traumatic parturitional intracranial hemorrhage. DevMed Child Neurol 28:156–164, 1986.

28. Wong CW: The CT criteria for conservative treatment—but under closeclinical observation—of posterior fossa epidural haematomas. ActaNeurochir (Wien) 126:124–127, 1994.

29. Zuccarello M, Andrioli G, Fiore D, Longatti P, Pardatscher K, Zampieri P:Traumatic posterior fossa haemorrhage in children. Acta Neurochir (Wien)62:79–85, 1982.

30. Zuccarello M, Pardatscher K, Andrioli GC, Fiore DL, Iavicoli R, Cervellini P:Epidural hematomas of the posterior cranial fossa. Neurosurgery 8:434–437,1981.

POSTERIOR FOSSA MASS LESIONS




TAB

LE1.

Surg

ical

man

agem

ent

ofpo

ster

ior

foss

am

ass

lesi

onsa

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Boz

buga

etal

.(1

)73

IIIU

nkno

wn

Surg

ery

and

nons

urgi

cal

GO

S,no

tsp

ecifi

edPr

ospe

ctiv

est

udy

of73

patie

nts

with

PFED

Hm

anag

edby

CT

crite

ria

eith

ersu

rgic

ally

orno

nsur

gica

llyba

sed

onpr

esen

ceof

sign

sof

mas

sef

fect

onC

T.

●Pa

tient

sw

ithPF

EDH

and

noas

soci

ated

CT

evid

ence

for

mas

sef

fect

may

bem

anag

edef

fect

ivel

yw

ithou

tsu

rger

y.●

No

patie

ntw

ithm

ass

effe

cton

CT

was

man

aged

nons

urgi

cally

,th

us,

prec

ludi

ngdi

rect

com

pari

son

betw

een

grou

ps.

●N

om

orbi

dity

orm

orta

lity

insu

bacu

teca

ses

(pre

sent

atio

n24

h–7

d).

No.

ofpa

tien

tsG

R(%

)M

D(%

)D

(%)

Non

surg

ical

1410

00

0

Surg

ical

5986

77

Acu

te(�

24h)

5485

77

Suba

cute

(24

h–7

d)19

100

00

Bra

mbi

llaet

al.

(3)

8III

GC

S4

–15

Surg

ery

Dea

thve

rsus

com

plet

ere

cove

ry,

not

spec

ified

Ret

rosp

ectiv

ese

ries

of8

patie

nts

surg

ical

lytr

eate

dfo

rPF

EDH

toev

alua

tecl

inic

alch

arac

teri

stic

san

dou

tcom

e.

●A

utho

rsem

phas

ize

high

mor

talit

yra

teth

ough

tse

cond

ary

toco

ncom

itant

brai

nste

man

dba

sal

gang

liain

jury

seen

atau

tops

y.*N

ote:

pape

rin

clud

edw

ithn

�8

beca

use

oflim

ited

num

ber

ofca

sese

ries

with

�10

patie

nts.

No.

ofpa

tien

tsG

R(%

)D

(%)

PFED

H8

62.5

37.5

Ciu

rea

etal

.(4

)9

IIIG

CS

13–1

4Su

rger

yG

OS,

not

spec

ified

Ret

rosp

ectiv

ese

ries

of9

patie

nts

aged

2–12

ysu

rgic

ally

trea

ted

for

PFED

Hto

eval

uate

clin

ical

char

acte

rist

ics

and

outc

ome.

●Su

bacu

tepr

esen

tatio

npr

edom

inat

ed(8

8.8%

)an

d,th

us,

may

have

bias

edou

tcom

eda

ta.

●A

ssoc

iate

din

trac

rani

alin

juri

esse

emed

asso

ciat

edw

ithw

orse

outc

ome

(no

stat

istic

s).

Not

e:pa

per

incl

uded

with

n�

9be

caus

eof

limite

dnu

mbe

rof

case

seri

esw

ith�

10pa

tient

s.

No.

ofpa

tien

tsG

R(%

)M

D(%

)

PFED

H9

8911

PFED

H�

asso

ciat

edle

sion

475

25

d’A

vella

etal

.(5

)81

IIIG

CS

3–5

to14

–15

Surg

ery

and

nons

urgi

cal

GO

S6

mo

Ret

rosp

ectiv

ese

ries

of81

patie

nts

with

trau

mat

icin

trac

ereb

ella

rhe

mor

rhag

e/co

ntus

ion

toev

alua

tepr

ogno

stic

fact

ors

and

exam

ine

clin

icor

adio

logi

cal

pres

enta

tion.

●44

.4%

poor

resu

lt(G

OS

1–3)

;55

.6%

favo

rabl

ere

sult

(GO

S4

–5).

●G

CS,

pres

ence

ofsu

prat

ento

rial

lesi

on,

stat

usof

basa

lci

ster

nsan

dfo

urth

vent

ricl

e,m

echa

nism

ofin

jury

,an

dcl

otsi

zeco

rrel

ated

with

outc

ome.

●G

CS

and

pres

ence

ofco

ncom

itant

supr

aten

tori

alle

sion

wer

ein

depe

nden

tpr

ogno

stic

fact

ors.

●9.

4tim

esre

lativ

eri

skof

poor

outc

ome

with

GC

S�

8.

●2.

3tim

esre

lativ

eri

skof

poor

outc

ome

with

asso

ciat

edsu

prat

ento

rial

lesi

on.

●Tr

aum

atic

intr

acer

ebel

lar

clot

sob

serv

edto

incr

ease

insi

zeup

to4

daf

ter

inju

ry—

auth

ors

advo

cate

repe

atC

Tsc

ans

until

lesi

onst

abili

zes.

●A

utho

rsre

com

men

dsu

rger

yfo

rpa

tient

sw

ithla

rger

clot

sca

usin

gpo

ster

ior

foss

am

ass

effe

ct.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

Intr

acer

ebel

lar

hem

orrh

age/

cont

usio

n,to

tal

8155

.644

.4

GC

S�

839

946

GC

S�

8b42

1981

OR

c27

4852

bP

�0.

000,

cP

�n.

s.

Hol

zsch

uhan

dSc

hukn

echt

(7)

20III

GC

S3–

15Su

rger

yG

OS

6m

oR

etro

spec

tive

seri

esof

20pa

tient

ssu

rgic

ally

trea

ted

for

PFED

Hto

eval

uate

clin

ical

and

radi

olog

icch

arac

teri

stic

san

dou

tcom

e.

●G

CS

corr

elat

edw

ithG

OS

(no

stat

istic

s)●

Mor

talit

yof

acut

e(�

24h)

and

suba

cute

(24

h–7

d)ca

ses

was

50%

and

20%

,re

spec

tivel

y.●

Aut

hors

emph

asiz

eim

port

ance

ofC

Tfo

rpa

tient

sw

ithsi

gns

ofoc

cipi

tal

trau

ma

todi

agno

sePF

EDH

befo

recl

inic

alde

teri

orat

ion.

Not

e:C

Tin

18/2

0ca

ses.

No.

ofpa

tien

tsG

R(%

)M

D(%

)D

(%)

PFED

H,

tota

l20

4520

35

GC

S15

–14

210

00

0

GC

S13

–8

862

.512

.525

GC

S7–

310

2030

50

Acu

te(�

24h)

1428

.621

.450

Suba

cute

(24

h–7

d)5

6020

20

Chr

onic

(�7

d)1

100

00



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Kar

asaw

aet

al.

(8)

53III

GC

S3–

8to

13–1

5Su

rger

yan

dno

nsur

gica

lG

OS

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of53

patie

nts

with

post

erio

rfo

ssa

inju

ryco

mpa

ring

thos

ew

ithan

dw

ithou

tas

soci

ated

HC

Pto

eval

uate

CT

fact

ors

asso

ciat

edw

ithth

ede

velo

pmen

tof

HC

Pan

dth

edi

ffere

nces

inou

tcom

ebe

twee

nsu

bgro

ups

with

and

with

out

HC

P.

●Su

prat

ento

rial

exte

nsio

n,he

mat

oma

thic

knes

s�

15m

m,

and

abno

rmal

mes

ence

phal

icci

ster

nsw

ere

sign

ifica

ntly

asso

ciat

edw

ithH

CP

(P�

0.05

).●

Bila

tera

lle

sion

san

din

abili

tyto

visu

aliz

eth

efo

urth

vent

ricl

ew

ere

sign

ifica

ntly

asso

ciat

edw

ithH

CP

inin

trac

ereb

ella

rhe

mat

oma/

cont

usio

n.●

No

sign

ifica

ntdi

ffere

nce

inm

orta

lity

betw

een

patie

nts

with

PFED

Hw

ithor

with

out

HC

P.●

HC

Pin

crea

sed

mor

talit

yin

patie

nts

with

ICH

/con

tusi

on(1

00%

vers

us15

.4%

;P

�0.

05).

●Su

rger

ydi

dno

tal

ter

mor

talit

yin

patie

nts

with

asso

ciat

edH

CP,

thou

ghnu

mbe

rof

patie

nts

was

too

smal

lfo

rco

mpa

riso

n.

●A

dmis

sion

GC

Sw

asin

vers

ely

rela

ted

tom

orta

lity

(P�

n.s.

).

●A

utho

rsad

voca

teex

tern

alve

ntri

cula

rdr

aina

gein

patie

nts

with

ICH

/con

tusi

onan

dH

CP,

not

with

PFED

H.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Post

erio

rfo

ssa

inju

ry,

tota

l53

747

22

15

EDH

2588

80

04

ICH

/con

tusi

on13

46.2

07.

77.

738

.5

SAH

710

00

00

0

SDH

580

00

020

Bra

inst

emhe

mor

rhag

e3

066

.70

033

.3

EDH

�su

rger

y�

HC

P7

8614

00

0

EDH

�su

rger

y,no

HC

P7

7114

00

14

ICH

�su

rger

y�

HC

P2

00

00

100

ICH

�su

rger

y,no

HC

P1

100

00

00

HC

P�

surg

ery

966

.711

.10

022

.2

HC

P,no

surg

ery

366

.70

00

33.3

Koc

etal

.(1

0)11

IIIG

CS

7–15

Surg

ery

GC

Sat

disc

harg

eR

etro

spec

tive

seri

esof

14pa

tient

s(1

1w

ithC

Tsc

ans)

trea

ted

surg

ical

lyfo

rPF

EDH

toex

amin

ecl

inic

alan

dra

diol

ogic

feat

ures

and

toas

sess

outc

ome.

●A

utho

rsre

port

that

lesi

onsi

ze,

pres

ence

ofco

exis

ting

intr

acra

nial

lesi

ons,

and

incr

easi

ngag

ear

eno

tas

soci

ated

with

prog

nosi

s,ho

wev

er,

all

mor

bidi

tyan

dm

orta

lity

occu

rred

inpa

tient

sw

ithou

tC

Tsc

ans.

No.

ofpa

tien

tsG

R(%

)D

(%)

PFED

Hw

ithC

T11

100

0

Lui

etal

.(1

1)89

IIIA

llG

CS

Surg

ery

GO

S,1

yr;

mor

talit

y,1

wk

Ret

rosp

ectiv

ere

view

of89

patie

nts

trea

ted

surg

ical

lyfo

rPF

EDH

toev

alua

teou

tcom

ere

late

dto

CT

char

acte

rist

ics,

age,

GC

S.

●M

orta

lity

17.9

%.

●M

orta

lity

inve

rsel

yre

late

dto

GC

Sbe

fore

surg

ery.

●A

ge�

16yr

asso

ciat

edw

ithde

crea

sed

mor

talit

y.

●Pr

esen

ceof

mix

edle

sion

(i.e.

,ex

tens

ion

into

occi

pita

lre

gion

)an

dco

ncom

itant

,no

ncon

tiguo

ussu

prat

ento

rial

lesi

onas

soci

ated

with

incr

ease

dm

orta

lity.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H,

tota

l89

74.2

5.6

1.1

1.1

17.9

GC

S(s

urge

ry)

3–5

1764

.7

GC

S6

–8

195.

2

GC

S9

–12

2913

.8

GC

S13

–15

240

Age

�16

yr29

10.3

Age

�16

yr60

21.7



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Mah

ajan

etal

.(1

2)19

IIIG

CS

�7

to13

–15

Surg

ery

GO

S,no

tsp

ecifi

edR

etro

spec

tive

seri

esof

19pa

tient

str

eate

dsu

rgic

ally

for

PFED

Hto

exam

ine

clin

ical

char

acte

rist

ics

and

outc

ome.

●M

orta

lity

15.8

%.

●A

dmis

sion

GC

Sco

rrel

ated

with

outc

ome

(no

stat

istic

s).

●B

oth

patie

nts

with

conc

omita

ntsu

prat

ento

rial

intr

acra

nial

lesi

ons

died

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H,

tota

l19

57.9

26.3

00

15.8

GC

S(a

dmis

sion

)15

–13

710

00

00

0

GC

S12

–8

757

.128

.60

014

.3

GC

S7–

35

060

00

40

Moh

anty

etal

.(1

3)24

IIIG

CS

3–5

to13

–15

Surg

ery

GO

Sat

disc

harg

eR

etro

spec

tive

seri

esof

489

patie

nts

with

surg

ical

lyev

acua

ted

EDH

com

pari

ngpr

ogno

sis

betw

een

child

ren

and

adul

tsan

dev

alua

ting

diffe

renc

ein

asso

ciat

edcl

inic

alpa

ram

eter

sbe

twee

nch

ildre

nan

dad

ults

.A

subg

roup

of24

patie

nts

with

surg

ical

lytr

eate

dPF

EDH

isin

clud

ed.

●In

cide

nce

ofPF

EDH

/ED

Hhi

gher

inch

ildre

nve

rsus

adul

ts(1

1%ve

rsus

3%).

●PF

EDH

had

bette

rou

tcom

eve

rsus

supr

aten

tori

alED

Hin

both

age

grou

ps(n

ost

atis

tics)

.

No.

ofpa

tien

tsG

R/M

D(%

)SD

/VS/

D(%

)

PFED

H,

tota

l24

87.5

12.5

age

�15

yr11

919

age

�16

yr13

8515

Neu

baue

r(1

4)18

IIIG

CS

3–14

Surg

ery

GO

Sno

tsp

ecifi

edR

etro

spec

tive

seri

esof

18pa

tient

ssu

rgic

ally

trea

ted

for

PFED

Hto

eval

uate

outc

ome

rela

tive

toac

uity

ofpr

esen

tatio

n,pr

esen

ceof

hydr

ocep

halu

s,an

dpr

esen

ceof

asso

ciat

edle

sion

s.

●M

orta

lity

22%

—al

lpr

esen

ted

with

acut

eco

urse

.●

GC

Sim

med

iate

lybe

fore

surg

ery,

acut

epr

esen

tatio

n,an

dpr

esen

ceof

hydr

ocep

halu

sse

emed

toco

rrel

ate

with

mor

talit

y(n

ost

atis

tics)

.●

Con

com

itant

lesi

ons

did

not

alte

rm

orta

lity

but

corr

elat

edw

ithin

crea

sed

disa

bilit

y(n

ost

atis

tics)

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H,

tota

l18

5511

5.5

5.5

22

With

asso

ciat

edle

sion

825

2512

.512

.525

No

asso

ciat

edle

sion

1080

00

020

Acu

te(�

24h)

1040

Suba

cute

(2–7

d)8

0

With

HC

P6

33

No

HC

P12

16

Ots

uka

etal

.(1

5)11

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

GO

Sat

disc

harg

eR

etro

spec

tive

seri

esof

11pa

tient

sw

ithPF

EDH

toev

alua

tecl

inic

alan

dra

diol

ogic

feat

ures

,an

dto

com

pare

outc

ome

betw

een

oper

ated

and

nono

pera

ted

grou

ps.

●M

orta

lity

18.2

%.

●O

utco

me

was

sim

ilar

betw

een

grou

psw

hen

surg

ical

indi

catio

nsw

ere

dete

rior

atio

n/no

nim

prov

emen

tof

neur

olog

ical

sign

san

dsy

mpt

oms

ofhe

adac

he,

naus

ea,

and

vom

iting

inpr

esen

ceof

CT

para

met

ers

ofm

axim

alhe

mat

oma

thic

knes

s�

15m

m,

poor

visu

aliz

atio

nof

post

erio

rfo

ssa

cist

erns

,m

arke

dde

form

ity/d

ispl

acem

ent

ofIV

vent

ricl

e,an

dex

tens

ion

ofPF

EDH

into

supr

aten

tori

alre

gion

(no

stat

istic

s).

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Surg

ery

666

.716

.70

016

.7

Non

surg

ical

580

00

020

Pozz

ati

etal

.(1

7)32

IIIG

CS

�7

to14

(14

poin

tsc

ale)

Surg

ery

(n�

30)

and

nons

urgi

cal

(n�

2)

Mod

ified

GO

S,no

tsp

ecifi

ed

Ret

rosp

ectiv

ese

ries

of32

patie

nts

with

PFED

H,

30of

who

mun

derw

ent

surg

ery,

toev

alua

tecl

inic

alch

arac

teri

stic

san

dou

tcom

e.

●M

orta

lity,

15.6

%.

●Pr

esen

ceof

asso

ciat

edin

trac

rani

alle

sion

corr

elat

esw

ithpo

orou

tcom

e(i.

e.,

disa

bilit

y,m

orta

lity)

com

pare

dw

ith“p

ure”

PFED

H(n

ost

atis

tics)

.

●A

llpa

tient

sw

ithG

OS

�5

had

asso

ciat

edle

sion

s.

●N

onsu

rgic

alm

anag

emen

tw

assu

cces

sful

in2

patie

nts

with

smal

lcl

ots.

No.

ofpa

tien

tsG

R(%

)D

isab

led

(%)

D(%

)

Pure

PFED

H,

oper

ated

1410

00

0

Ass

ocia

ted

lesi

on,

oper

ated

1650

18.8

31.3

Smal

l,no

nsur

gica

l2

100

00



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Pozz

ati

etal

.(1

8)7

IIIG

CS

5–11

Surg

ery

and

nons

urgi

cal

GO

Sat

disc

harg

eR

etro

spec

tive

seri

esof

7pa

tient

sto

exam

ine

the

role

ofC

Tin

cons

erva

tive

vers

ussu

rgic

altr

eatm

ent

ofis

olat

edin

trac

ereb

ella

rhe

mat

oma.

●Po

orer

outc

ome

was

asso

ciat

edw

ithm

idlin

eve

rsus

hem

isph

eric

clot

s(n

ost

atis

tics)

.

●C

Tal

low

edsu

cces

sful

cons

erva

tive

man

agem

ent

ofse

lect

edpa

tient

s.

●N

ogo

odou

tcom

ew

asac

hiev

edw

ithcl

ot�

3cm

.N

ote:

pape

rin

clud

edw

ithn

�7

beca

use

oflim

ited

num

ber

ofca

sese

ries

with

�10

patie

nts.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Mid

line

clot

30

00

010

0

Hem

isph

eric

clot

450

500

00

Surg

ery

30

330

067

Non

surg

ical

450

250

025

�3

cm3

067

00

33

�3

cm4

500

00

50

GC

S�

74

5025

00

25

GC

S�

73

033

00

67

Prus

tyan

dM

ohan

ty(1

9)17

IIIG

CS

4–1

5Su

rger

y(n

�16

)an

dno

nsur

gica

l(n

�1)

GO

Sat

disc

harg

ean

d6

mo

Ret

rosp

ectiv

ese

ries

of17

patie

nts

with

PFED

H,

16of

who

mun

derw

ent

surg

ery,

toex

amin

ecl

inic

alch

arac

teri

stic

san

dou

tcom

e.

●M

orta

lity

17.6

%—

all

patie

nts

who

died

pres

ente

dac

utel

y(�

24h

from

inju

ry)

and

had

conc

omita

ntin

trac

rani

alle

sion

s.

●G

OS

5at

6m

ofo

ral

lpa

tient

spr

esen

ting

with

suba

cute

orch

roni

cco

urse

.N

ote:

CT

perf

orm

edin

14/1

7pa

tient

s.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H,

tota

l17

5923

00

18

Riv

ano

etal

.(2

0)22

IIIG

CS

4–1

5Su

rger

yG

OS

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of22

patie

nts

unde

rgoi

ngsu

rger

yfo

rPF

EDH

toev

alua

tecl

inic

alan

dra

diol

ogic

feat

ures

with

resp

ect

toou

tcom

e.

●M

orta

lity

13.6

%—

all

patie

nts

who

died

pres

ente

dw

ithac

ute

clin

ical

cour

se.

●G

CS

atsu

rger

yre

late

dto

outc

ome

(no

stat

istic

s).

●Pr

esen

ceof

asso

ciat

edin

trac

rani

alle

sion

sde

crea

sed

good

outc

ome,

incr

ease

dm

orbi

dity

,an

din

crea

sed

mor

talit

yco

mpa

red

with

patie

nts

with

isol

ated

PFED

H(n

ost

atis

tics)

.N

ote:

2/22

patie

nts

eval

uate

dby

angi

ogra

phy.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Acu

te(O

R�

24h)

1464

77

022

Suba

cute

(OR

24h–

10d)

683

170

00

GC

S(s

urge

ry)

15–1

45

100

00

00

GC

S13

–8

1385

150

00

GC

S7–

34

00

250

75

Ass

ocia

ted

lesi

on7

4314

140

29

No

asso

ciat

edle

sion

1384

.67.

70

07.

7

Srip

airo

jkul

etal

.(2

2)22

IIIG

CS

3T–1

5Su

rger

yG

OS

6m

oR

etro

spec

tive

seri

esof

22pa

tient

sw

ithsu

rgic

ally

trea

ted

post

erio

rfo

ssa

hem

atom

asev

alua

ting

outc

ome

byag

e,ty

peof

hem

atom

a,an

din

terv

albe

twee

nin

jury

and

surg

ery.

●M

orta

lity

38%

.●

GC

S�

9at

time

ofsu

rger

yco

rrel

ated

with

poor

outc

ome.

●In

terv

albe

twee

nin

jury

and

surg

ery

�1

dco

rrel

ated

with

bette

rou

tcom

eth

ansu

rger

y�

6h—

this

may

beca

used

bya

sele

ctio

nof

patie

nts

with

suba

cute

pres

enta

tion.

●C

onco

mita

ntsu

prat

ento

rial

and

infr

aten

tori

alle

sion

sin

crea

sed

mor

bidi

tyan

dm

orta

lity

(mor

talit

y20

%ve

rsus

0%).



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

Age

�40

yr16

756.

26.

20

12.5

Age

�40

yr6

16.7

16.7

16.7

050

GC

S(s

urge

ry)

3–8

d9

33.3

11.1

11.1

044

.4

GC

S9

–12

666

.70

16.7

016

.7

GC

S13

–15

785

.714

.30

00

EDH

710

00

00

0

EDH

�as

soci

ated

lesi

on

955

.622

.20

022

.2

SDH

110

00

00

0

SDH

�IC

H1

00

100

00

ICH

20

050

050

ICH

�su

prat

ento

rial

lesi

on

20

00

010

0

Inju

ryto

OR

�6

h9

55.6

011

.10

33.3

Inju

ryto

OR

6h–

1d

742

.914

.314

.30

28.6

Inju

ryto

OR

�1

d6

83.3

16.7

00

0

dO

dds

ratio

,0.

057

(sta

tistic

ally

sign

ifica

nt).

St.

John

and

Fren

ch(2

3)8

IIIG

CS

6–1

4Su

rger

yG

OS,

not

spec

ified

Ret

rosp

ectiv

ese

ries

of8

patie

nts

surg

ical

lytr

eate

dfo

rtr

aum

atic

post

erio

rfo

ssa

hem

atom

ato

eval

uate

clin

ical

char

acte

rist

ics

and

outc

ome.

●M

orta

lity

25%

.●

Suba

cute

pres

enta

tion

seem

edto

corr

elat

ew

ithbe

tter

outc

ome,

but

num

ber

ofpa

tient

sto

osm

all

toev

alua

te.

Not

e:pa

per

incl

uded

with

n�

8be

caus

eof

limite

dnu

mbe

rof

case

seri

esw

ith�

10pa

tient

s.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

All

patie

nts

850

12.5

12.5

025

Acu

te(�

24h)

633

.316

.716

.70

33.3

Suba

cute

(24

h–7

d)2

100

00

00

Tsai

etal

.(2

4)57

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

GO

Sva

riab

leR

etro

spec

tive

seri

esof

57pa

tient

sw

ithpr

imar

ypo

ster

ior

foss

ain

jury

toev

alua

teC

Tch

arac

teri

stic

san

dco

rrel

ate

type

ofin

jury

with

prog

nosi

s.

●PF

EDH

patie

nts

who

wer

eco

mat

ose

onad

mis

sion

died

with

orw

ithou

tsu

rger

y.

●2/

2PF

SDH

patie

nts

with

asso

ciat

edin

jury

wer

eco

mat

ose

onad

mis

sion

,di

dno

tun

derg

osu

rger

y,an

ddi

ed.

●Ev

iden

cefo

rbr

ains

tem

inju

ry,

post

erio

rfo

ssa

swel

ling/

cist

ern

oblit

erat

ion

corr

elat

edw

ithin

crea

sed

mor

talit

y(n

ost

atis

tics)

.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H14

500

00

50

PFSD

H8

62.5

12.5

00

25

Cer

ebel

lar

inju

ry14

00

28.6

14.3

57.1

Bra

inst

emin

jury

210

00

23.8

76.2

Vra

nkov

icet

al.

(25)

11III

Unk

now

nSu

rger

yan

dno

nsur

gica

lG

OS

atdi

scha

rge

Ret

rosp

ectiv

ese

ries

of11

patie

nts

with

ATP

FHtr

eate

dei

ther

surg

ical

ly(n

�8)

orno

nope

rativ

ely

(n�

3)to

eval

uate

clin

ical

and

radi

olog

icch

arac

teri

stic

san

dou

tcom

e.

●9

of11

patie

nts

had

occi

pita

lfr

actu

res

and

8.1%

ofpa

tient

sw

ithoc

cipi

tal

frac

ture

sha

dA

TPFH

,em

phas

izin

gim

port

ance

ofC

Tfo

rpa

tient

sw

ithsi

gns

ofoc

cipi

tal

trau

ma.

●3

patie

nts

with

intr

acer

ebel

lar

hem

atom

asan

dno

sign

sof

mas

sef

fect

onC

Tw

ere

succ

essf

ully

man

aged

with

out

surg

ery.

●A

utho

rsad

voca

teim

med

iate

surg

ery

for

all

case

sof

PFED

Hbe

caus

eof

unpr

edic

tabi

lity

ofde

teri

orat

ion.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

All

patie

nts

1154

.536

.40

09.

1



TAB

LE1.

Con

tinu

ed

Aut

hors

(ref

.no

.)N

o.of

pati

ents

Cla

ssIn

clus

ion

GC

STr

eatm

ent

Out

com

eD

escr

ipti

onC

oncl

usio

n

Wan

get

al.

(26)

13III

All

GC

SSu

rger

yG

OS

3–6

mo

Ret

rosp

ectiv

ese

ries

of13

patie

nts

who

unde

rwen

tsu

rger

yfo

rPF

EDH

toid

entif

ycl

inic

alfe

atur

esth

atm

ayle

adto

earl

ydi

agno

sis

and

toev

alua

teou

tcom

ew

ithre

spec

tto

initi

alG

CS.

●Ea

rly

CT

shou

ldbe

prom

pted

byth

epr

esen

ceof

occi

pita

lso

fttis

sue

inju

ry,

drow

sine

ss,

occi

pita

lfr

actu

re,

and

dias

tasi

sof

lam

bdoi

dsu

ture

.

●G

CS

inve

rsel

yre

late

dto

outc

ome

(no

stat

istic

s).

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

GC

S15

–14

710

00

00

0

GC

S13

–8

250

500

00

GC

S7–

34

500

250

25

Won

g(2

8)25

IIIA

llG

CS

Surg

ery

and

nons

urgi

cal

Mor

talit

yat

disc

harg

eR

etro

spec

tive

seri

esof

25pa

tient

sw

ithPF

EDH

com

pari

ngou

tcom

eof

surg

ical

vers

usno

nsur

gica

lm

anag

emen

tw

ithre

spec

tto

initi

alC

Tch

arac

teri

stic

s,in

clud

ing

volu

me,

thic

knes

s,de

gree

ofsh

ift,

and

pres

ence

ofas

soci

ated

fron

tal

cont

usio

n.

●Pa

tient

sm

anag

edex

pect

antly

with

PFED

Hvo

lum

e�

10cm

3 ,th

ickn

ess

�15

mm

,an

dsh

ift�

5m

mha

dsi

gnifi

cant

lyhi

gher

mor

talit

yth

anpa

tient

sm

anag

edsu

rgic

ally

with

equi

vale

ntC

Tch

arac

teri

stic

s.

●10

0%su

rviv

alin

all

patie

nts

with

PFED

Hvo

lum

e�

10cm

3 ,th

ickn

ess

�15

mm

,an

dsh

ift�

5m

m,

rega

rdle

ssof

trea

tmen

t.

●A

ssoc

iate

dfr

onta

lIC

Hsi

gnifi

cant

lyin

crea

sed

mor

talit

yin

both

grou

ps.

No.

ofpa

tien

tsD

(%)

Vol

�10

cm3 /e

arly

OR

e9

11.1

Vol

�10

cm3 /e

xpec

tant

475

Vol

�10

cm3 /e

arly

OR

f8

0

Vol

�10

cm3 /e

xpec

tant

40

Thic

knes

s�

15m

m/e

arly

OR

g13

7.7

Thic

knes

s�

15m

m/e

xpec

tant

475

Thic

knes

s�

15m

m/e

arly

OR

f4

0

Thic

knes

s�

15m

m/e

xpec

tant

40

Shift

�5

mm

/ear

lyO

Re

50

Shift

�5

mm

/exp

ecta

nt1

100

Shift

�5

mm

/ear

lyO

Rf

128.

3

Shift

�5

mm

/exp

ecta

nt7

28.6

Ass

ocia

ted

fron

tal

cont

usio

nh6

66.7

No

fron

tal

cont

usio

n19

0e

P�

0.05

,fP

�n.

s,g

P�

0.01

,h

P�

0.00

1.

Zuc

care

lloet

al.

(29)

10III

Unk

now

nSu

rger

yan

dno

nsur

gica

lM

orta

lity

Ret

rosp

ectiv

ese

ries

of10

patie

nts

with

trau

mat

icpo

ster

ior

foss

am

ass

lesi

ons

tode

scri

befa

ctor

sre

late

dto

man

agem

ent

and

prog

nosi

s.

●M

orta

lity

40%

(incl

udes

all

lesi

onty

pes)

.●

No

com

pari

son

betw

een

grou

psab

leto

bem

ade.

No.

ofpa

tien

tsD

(%)

EDH

20

Cer

ebel

lar

hem

orrh

age

�su

rger

y2

50

Cer

ebel

lar

hem

orrh

age,

nons

urgi

cal

30

Bra

inst

emin

jury

310

0

Zuc

care

lloet

al.

(30)

8III

GC

Sm

otor

4–5

Surg

ery

GO

Sat

disc

harg

eR

etro

spec

tive

clin

ical

seri

esof

8pa

tient

sw

ithsu

rgic

ally

trea

ted

PFED

Hto

eval

uate

outc

ome.

●Im

port

ance

ofC

Tis

emph

asiz

ed.

Not

e:pa

per

incl

uded

with

n�

8be

caus

eof

limite

dnu

mbe

rof

case

seri

esw

ith�

10pa

tient

s.

No.

ofpa

tien

tsG

R(%

)M

D(%

)SD

(%)

VS

(%)

D(%

)

PFED

H,

tota

l8

5037

.50

012

.5

aG

CS,

Gla

sgow

Com

aSc

ale;

GO

S,G

lasg

owou

tcom

esc

ore;

PFED

H,p

oste

rior

foss

aep

idur

alhe

mat

oma;

CT,

com

pute

dto

mog

raph

icsc

an;G

R,g

ood

reco

very

;MD

,mod

erat

edi

sabi

lity;

D,d

eath

;SD

,sev

ere

disa

bilit

y;V

S,ve

geta

tive

stat

e;n.

s.,n

otsi

gnifi

cant

;HC

P,hy

droc

epha

lus;

ATP

FH,a

cute

trau

mat

icpo

ster

ior

foss

ahe

mat

oma;

ICH

,int

race

rebr

alhe

mor

rhag

e;ED

H,e

pidu

ralh

emat

oma;

SAH

,su

bara

chno

idhe

mor

rhag

e;SD

H,

subd

ural

hem

atom

a;O

R,

oper

atio

n.

CHAPTER 7

SURGICAL MANAGEMENT OF DEPRESSED

CRANIAL FRACTURESM. Ross Bullock, M.D., Ph.D.Department of Neurological Surgery,Virginia Commonwealth UniversityMedical Center,Richmond, Virginia











Indications• Patients with open (compound) cranial fractures depressed greater than the thick-

ness of the cranium should undergo operative intervention to prevent infection.• Patients with open (compound) depressed cranial fractures may be treated nonop-

eratively if there is no clinical or radiographic evidence of dural penetration,significant intracranial hematoma, depression greater than 1 cm, frontal sinusinvolvement, gross cosmetic deformity, wound infection, pneumocephalus, or grosswound contamination.

• Nonoperative management of closed (simple) depressed cranial fractures is a treat-ment option.

Timing• Early operation is recommended to reduce the incidence of infection.Methods• Elevation and debridement is recommended as the surgical method of choice.• Primary bone fragment replacement is a surgical option in the absence of wound

infection at the time of surgery.• All management strategies for open (compound) depressed fractures should include

antibiotics.

KEY WORDS: Antibiotic prophylaxis, Burr hole, Cranial fracture, Craniotomy, Depressed cranial fracture,Depressed skull fracture, Head injury, Skull fracture, Surgical technique, Traumatic brain injury

Neurosurgery 58:S2-56-S2-60, 2006 DOI: 10.1227/01.NEU.0000210367.14043.0E www.neurosurgery-online.com

OVERVIEW

The presence of a cranial fracture has con-sistently been shown to be associated with ahigher incidence of intracranial lesions, neu-rological deficit, and poorer outcome (4, 8, 12,14). Indeed, Chan et al. (4) found cranial frac-ture to be the only independent significantrisk factor in predicting intracranial hemato-mas in a cohort of 1178 adolescents. Macpher-son et al. (12) found that 71% of 850 patientswith a cranial fracture had an intracranial le-sion (i.e., contusion or hematoma), comparedwith only 46% of 533 patients without a cra-nial fracture. Hung et al. (8) determined thatpatients with both loss of consciousness andcranial fracture were at significantly greaterrisk of developing a “surgically significant in-tracranial hematoma” than those with one or

neither condition. Servadei et al. (14) showedthe importance of cranial fracture in predict-ing the presence of intracranial lesions, evenin minor head injuries (Glasgow Coma Scalescore 14 or 15). These studies underscore theimportance of cranial fractures as indicators ofclinically significant injuries, as well as theimportance of computed tomographic (CT)scans in evaluation of all patients with knownor clinically suspected cranial fractures.

Depressed cranial fractures may complicateup to 6% of head injuries in some series (7),and account for significant morbidity andmortality. Compound fractures account for upto 90% of these injuries (3, 6, 17), and areassociated with an infection rate of 1.9 to10.6% (9, 13, 16, 17), an average neurologicalmorbidity of approximately 11% (6), an inci-dence of late epilepsy of up to 15% (10), and a


mortality rate of 1.4 to 19% (3, 5–7, 17). By convention, com-pound depressed cranial fractures are treated surgically, withdebridement and elevation, primarily to attempt to decreasethe incidence of infection. Closed (“simple”) depressed cranialfractures undergo operative repair if the extent of depressionis greater than the full thickness of the adjacent calvarium,with the theoretical benefits of better cosmesis, a diminution inlate-onset posttraumatic epilepsy, and a reduction in the inci-dence of persistent neurological deficit. There is, however,very little literature to support these management strategies,despite their widespread, and theoretically sound, practice.There is Class III literature that addresses the efficacy ofsurgical management of these injuries, and it argues againstautomatic surgical treatment of all compound fractures (7).

Most of the literature reviewed focuses predominantly oninfectious complications, seizures, surgical technique (e.g., bonefragment replacement versus removal), or the predictive powerof cranial fracture for the presence of other intracranial pathol-ogy. Several large studies of patients with cranial fracture shedlight on the breadth of issues associated with such lesions and arediscussed below, under Scientific Foundation. However, some ofthese studies were conducted before the CT-scan era, and thus,although important for our understanding of the injury itself, arenot included for critical analysis.

PROCESS

A MEDLINE computer search using the following key words:“skull” and “fracture” and “depressed” between 1975 and 2001was performed. A total of 224 documents were found. The searchwas narrowed to include the key words: “surgery” or “opera-tion” or “elevation”. A total of 122 articles were found, 5 of whichmet the criteria for critical analysis. In addition, the reference listsof all articles were reviewed, and additional articles were se-lected for background information. The results of this analysiswere incorporated into the review presented here. Papers pri-marily addressing the following topics were not included: pa-tients with associated medical illnesses, sinus fractures, cranialbase fractures, isolated orbital or facial fractures, and pre-CT erareports. In general, papers with the following characteristicswere also excluded: case series with less than 10 patients evalu-ated by CT scan and with incomplete outcome data (mortality orGlasgow outcome score), case reports, operative series with op-erations occurring longer than 14 days from injury. Several arti-cles with case series of less than 10 patients were examined andreviewed because of the limited number of patient series evalu-ating the acute surgical management of depressed cranial frac-tures in the CT era. Selected articles were evaluated for design,prognostic significance, therapeutic efficacy, and overall out-come. In addition, several articles were reviewed for the pur-poses of historical perspective.


Closed, linear cranial fractures are considered nonoperativelesions unless associated with surgical intracranial masses.

Controversy surrounds appropriate management of de-pressed cranial fractures. Compound depressed cranial frac-tures are depressed fractures with an overlying scalp lacera-tion in continuity with the fracture site and with galealdisruption, and have conventionally been treated with de-bridement and surgical elevation (3, 6, 9, 14). Simple de-pressed cranial fractures have no galeal disruption and aretraditionally managed with surgical elevation only if the ex-tent of depression equals or exceeds the thickness of adjacent,intact bone, or if there is an associated intracranial hematomawith mass effect that requires evacuation.

The rationale for aggressive treatment of depressed cranialfractures stems from their association with infection and lateepilepsy. Cosmetic deformity also plays a role in surgicaldecision making. Such complications, and their potential se-quelae, are well documented. In a series of 359 patients withcompound cranial fractures, Jennett and Miller (9) docu-mented a 10.6% incidence of infection, which was associatedwith a significantly higher incidence of persistent neurologicaldeficit, late epilepsy (defined as seizures longer than 1 wkfrom injury), and death. Operative debridement reduced theincidence of infection to 4.6% in their series. Operative delaygreater than 48 hours from injury dramatically increased theincidence to 36.5%. There was no difference in infection ratebetween surgical cohorts who had bone fragments replacedversus removed—results supported by a series of 225 patientswith depressed cranial fracture reported by Braakman (3), anda treatment strategy reported as early as Macewan in 1888 (9).In a separate report of 1000 patients with nonmissile de-pressed cranial fractures, Jennett et al. (10) documented a 15%incidence of late epilepsy, which was significantly associatedwith posttraumatic amnesia longer than 24 hours, torn dura,the presence of focal neurological signs, and the presence ofearly epilepsy (i.e., within 1 wk of injury). In the closed-fracture patients in this series, there was no difference inincidence of epilepsy between the elevated and nonelevatedcases. Additionally, there was a higher incidence of late epi-lepsy in patients with elevated compound fractures. The au-thors explain this finding by documenting a higher incidenceof those factors independently associated with late epilepsy,such as dural tearing and long posttraumatic amnesia, in theelevated-fracture patient cohort. These series were reportedbefore the CT era, however, they offer us a clear picture ofboth the range of complications associated with nonmissiledepressed cranial fractures and the controversies surroundingmanagement strategies.

The primary question facing the neurosurgeon regardingdepressed cranial fracture is whether to operate. Heary et al.(7) reported a group of patients with compound depressedcranial fractures in which nonsurgical therapy was used for asubgroup of 26 patients without clinical or radiographic evi-dence of dural violation or significant underlying brain injury.They concluded that patients with open (compound) de-pressed cranial fractures may be treated nonoperatively ifthere is no clinical or radiographic evidence of dural penetra-tion, significant intracranial hematoma, depression greater

DEPRESSED CRANIAL FRACTURES


than 1 cm, frontal sinus involvement, gross deformity, woundinfection, pneumocephalus, or gross wound contamination.No infectious complications occurred. Similarly, van denHeever and van der Merwe (16) reported an equally lowincidence of infection in a group of nonoperatively treatedpatients that included 139 compound depressed fractures.Surgical indications in their series included clinical character-istics of the wound. CT scans were not routinely used unlessa neurological deficit was present on admission.

Although these studies are retrospective and nonrandom-ized, and, thus, subject to inherent biases, they clearly dem-onstrate that at least a select group of patients with compounddepressed cranial fractures will do well without surgery.

Another challenge to traditional thinking that has surfaced inthe literature involves the proper surgical management of com-pound depressed cranial fractures with respect to the bone frag-ments. Conventional treatment involves operative debridement,elevation of the fracture, removal of bone fragments, and delayedcranioplasty. However, this subjects the patient to a second op-eration (i.e., cranioplasty), with its attendant risks and complica-tions. Kriss et al. (11), Jennett and Miller (9), and Braakman (3)showed that infectious complications are not increased by pri-mary bone fragment replacement. Wylen et al. (17) retrospec-tively reviewed a series of 32 patients who underwent elevationand repair of a compound depressed cranial fracture with pri-mary replacement of bone fragments within 72 hours of injury.Patients treated longer than 72 hours after injury and patientswho presented with existing infection were excluded from thestudy. There were no infectious complications. Blankenship et al.(2) also demonstrated a 0% infection rate in 31 children withcompound depressed cranial fractures treated with primarybone fragment replacement, regardless of the degree of contam-ination of the wound at the time of surgery. Thirty patients in thisseries were treated within 16 hours of injury. Likewise, Adeloyeand Shokunbi (1) report the success of immediate bone replace-ment, without infectious sequelae, in 12 patients with compounddepressed fractures, 11 of whom were treated within 10 hours ofinjury. Four patients in their series were treated with free-fragment removal secondary to the greater severity of parenchy-mal injury, suggesting benefit from the decompression that boneremoval would provide. Despite the retrospective, uncontrolled,nonrandomized design of these observational studies, theyclearly demonstrate the feasibility of immediate bone fragmentreplacement without a corresponding increase in infectious se-quelae, thus, obviating the need for a second surgical procedure.

SUMMARY

The majority of studies are case series. No controlled, prospec-tive clinical trials of treatment using surgical versus nonsurgicalmanagement have been published. The majority of data supportdebridement and elevation of grossly contaminated compounddepressed cranial fractures as soon as possible after injury. How-ever, several retrospective studies demonstrate successful non-operative management of some patients with less-severe com-pound depressed cranial fractures on the basis of CT and clinical

criteria. In the absence of gross wound infection at the time ofpresentation, immediate replacement of bone fragments seemsnot to increase the incidence of infection if surgery is performedexpeditiously, and this replacement eliminates the need for sub-sequent cranioplasty and its attendant risks and complications.No controlled data exist to support the timing of surgery or theuse of one technique over another.


To improve the strength of recommendations above theoption level, well-controlled trials of surgical technique arewarranted, and should examine issues of bone fragment re-placement versus removal, dural laceration repair, etc., andtheir respective relationship to outcome variables, such asincidence of infection, incidence of epilepsy, need for reopera-tion, surgical complications, and, most importantly, neurolog-ical and neuropsychological outcomes.

REFERENCES

1. Adeloye A, Shokunbi MT: Immediate bone replacement in compound de-pressed skull fractures. Cent Afr J Med 39:70–73, 1993.

2. Blankenship JB, Chadduck WM, Boop FA: Repair of compound-depressedskull fractures in children with replacement of bone fragments. PediatrNeurosurg 16:297–300, 1990.

3. Braakman R: Depressed skull fracture: Data, treatment, and follow-up in 225consecutive cases. J Neurol Neurosurg Psychiatry 35:395–402, 1972.

4. Chan KH, Mann KS, Yue CP, Fan YW, Cheung M: The significance of skullfracture in acute traumatic intracranial hematomas in adolescents: A pro-spective study. J Neurosurg 72:189–194, 1990.

5. Colak A, Berker M, Ozcan OE: Occipital depression fractures in childhood.A report of 14 cases. Childs Nerv Syst 7:103–105, 1991.

6. Cooper PR: Skull fracture and traumatic cerebrospinal fluid fistulas, in Cooper PR(ed): Head Injury. Baltimore, Williams and Wilkins, 1993, pp 115–136 ed 3.

7. Heary RF, Hunt CD, Krieger AJ, Schulder M, Vaid C: Nonsurgical treatmentof compound depressed skull fractures. J Trauma 35:441–447, 1993.

8. Hung CC, Chiu WT, Lee LS, Lin LS, Shih CJ: Risk factors predictingsurgically significant intracranial hematomas in patients with head injuries.J Formos Med Assoc 95:294–297, 1996.

9. Jennett B, Miller J: Infection after depressed fracture of skull. Implicationsfor management of nonmissile injuries. J Neurosurg 36:333–339, 1972.

10. Jennett B, Miller J, Braakman R: Epilepsy after monmissile depressed skullfracture. J Neurosurg 41:208–216, 1974.

11. Kriss F, Taren J, Kahn E: Primary repair of compound skull fractures byreplacement of bone fragments. J Neurosurg 30:698–702, 1969.

12. Macpherson BC, MacPherson P, Jennett B: CT evidence of intracranialcontusion and haematoma in relation to the presence, site and type of skullfracture. Clin Radiol 42:321–326, 1990.

13. Mendelow AD, Campbell D, Tsementzis SA, Cowie RA, Harris P, Durie TB,Gillingham FJ: Prophylactic antimicrobial management of compound de-pressed skull fracture. J R Coll Surg Edinb 28:80–83, 1983.

14. Servadei F, Ciucci G, Pagano F, Rebucci GG, Ariano M, Piazza G, Gaist G:Skull fracture as a risk factor of intracranial complications in minor headinjuries: A prospective CT study in a series of 98 adult patients. J NeurolNeurosurg Psychiatry 51:526–528, 1988.

15. Steinbok P, Flodmark O, Martens D, Germann ET: Management of simpledepressed skull fractures in children. J Neurosurg 66:506–510, 1987.

16. van den Heever CM, van der Merwe DJ: Management of depressed skullfractures. Selective conservative management of nonmissile injuries.J Neurosurg 71:186–190, 1989.

17. Wylen EL, Willis BK, Nanda A: Infection rate with replacement of bone fragmentin compound depressed skull fractures. Surg Neurol 51:452–457, 1999.



DEPRESSED CRANIAL FRACTURES


TAB

LE1.

Surg

ical

man

agem

ent

ofde

pres

sed

cran

ial

frac

ture

sa

Aut

hors

No.

ofpa

tien

tsC

lass

Incl

usio

nG

CS

Trea

tmen

tO

utco

me

Des

crip

tion

Con

clus

ion

Ade

loye

and

Shok

unbi

(1)

16III

GC

S15

Bon

efr

agm

ent

repl

acem

ent

vers

usre

mov

al

Infe

ctio

nra

te;

GO

S;tim

ing

not

spec

ified

Ret

rosp

ectiv

ese

ries

of16

patie

nts

trea

ted

eith

erw

ithim

med

iate

bone

frag

men

tre

plac

emen

t(n

�12

)or

free

frag

men

tre

mov

al(n

�4)

toco

mpa

rein

fect

ion

rate

and

outc

ome

betw

een

grou

ps.

●N

oin

cide

nce

ofin

fect

ious

com

plic

atio

nin

eith

ergr

oup,

supp

ortin

gth

esa

fety

ofbo

nefr

agm

ent

repl

acem

ent.

●Se

lect

ion

bias

and

smal

lnu

mbe

rof

patie

nts

prec

lude

com

pari

son

ofou

tcom

e.

No.

ofpa

tien

tsG

R(%

)In

fect

ion

(%)

Rep

lace

men

t12

100

0

No

repl

acem

ent

410

00

Bla

nken

ship

etal

.(2

)31

IIIU

nkno

wn

Bon

efr

agm

ent

repl

acem

ent

Infe

ctio

nor no

nuni

on,

aver

age

26.5

mo

Ret

rosp

ectiv

ere

view

of31

child

ren

aged

20m

oto

17yr

surg

ical

lytr

eate

dw

ithpr

imar

ybo

nefr

agm

ent

repl

acem

ent

for

com

poun

dde

pres

sed

cran

ial

frac

ture

toex

amin

eth

ein

cide

nce

ofin

fect

ious

com

plic

atio

ns.

Thir

tyof

31pa

tient

sw

ere

trea

ted

with

in16

hof

inju

ry.

●N

oin

stan

ces

ofin

fect

ious

com

plic

atio

nsan

dno

need

for

subs

eque

ntcr

anio

plas

ty,

rega

rdle

ssof

degr

eeof

wou

ndco

ntam

inat

ion.

No.

ofpa

tien

tsIn

fect

ion

(%)

Rep

lace

men

t31

0

Bra

akm

an(3

)22

5N

/AN

/AB

one

frag

men

tre

plac

emen

tve

rsus

rem

oval

Infe

ctio

nra

te;

inci

denc

eof

epile

psy

Ret

rosp

ectiv

ere

view

of22

5pa

tient

sw

ithde

pres

sed

cran

ial

frac

ture

toex

amin

eep

idem

iolo

gy,

asse

ssou

tcom

ew

ithre

spec

tto

inci

denc

eof

epile

psy

and

pers

iste

ntne

urol

ogic

alde

ficit,

and

toas

sess

the

effe

ctof

prim

ary

bone

frag

men

tre

plac

emen

ton

infe

ctio

nra

te.

●4.

4%in

cide

nce

ofea

rly

epile

psy.

●7.

1%in

cide

nce

ofla

teep

ileps

y.●Pr

imar

yre

plac

emen

tof

bone

frag

men

tsdi

dno

tal

ter

infe

ctio

nra

te.

No.

ofpa

tien

tsIn

fect

ion

(%)

Rep

lace

men

t82

2.4

Part

ial

repl

acem

ent

2711

.1

Rem

oval

5610

.7

Hea

ryet

al.

(7)

54III

Exce

llent

,go

od,

fair

,po

or

Surg

ery

and

nons

urgi

cal

Mea

nG

OS,

9.5

mo

Ret

rosp

ectiv

ere

view

of54

pros

pect

ivel

ytr

eate

dpa

tient

sw

ithco

mpo

und

depr

esse

dcr

ania

lfr

actu

res

com

pari

ngpa

tient

str

eate

dsu

rgic

ally

vers

usno

nsur

gica

llyba

sed

ona

stan

dard

ized

trea

tmen

tpr

otoc

olin

volv

ing

clin

ical

and

CT

crite

ria.

●N

odi

ffere

nce

inou

tcom

ebe

twee

npa

tient

str

eate

dsu

rgic

ally

vers

usno

nsur

gica

llyw

ithco

mpo

und

depr

esse

dcr

ania

lfr

actu

res

that

have

nocl

inic

alor

radi

ogra

phic

evid

ence

ofdu

ral

pene

trat

ion,

sign

ifica

ntin

trac

rani

alhe

mat

oma,

depr

essi

on�

1cm

,fr

onta

lsi

nus

invo

lvem

ent,

gros

sde

form

ity,

wou

ndin

fect

ion,

orgr

oss

wou

ndco

ntam

inat

ion.

●N

oin

fect

ious

com

plic

atio

nsoc

curr

edin

eith

ergr

oup

with

stan

dard

ized

trea

tmen

tpr

otoc

olin

clud

ing

prop

hyla

ctic

antib

iosi

s.●Pa

tient

sw

ithsu

rgic

alle

sion

sha

dsi

gnifi

cant

lyin

crea

sed

inci

denc

eof

SAH

and

cere

bral

cont

usio

n(P

�0.

01an

dP

�0.

05,

resp

ectiv

ely)

.●M

orta

lity

5.6%

ofth

ose

anal

yzed

,19

%to

tal.

No.

ofpa

tien

tsEx

celle

nt(%

)G

ood

(%)

Fair

(%)

Poor

(%)

D(%

)

Surg

ery

2864

.321

.410

.70

3.6

Non

surg

ical

2680

.83.

87.

70

7.7

Jenn

ett

and

Mill

er(9

)35

9N

/AN

/AEa

rly

vers

usde

laye

d;bo

nefr

agm

ent

repl

acem

ent

vers

usre

mov

al;

dura

clos

edve

rsus

open

Infe

ctio

nra

te,

CN

Ssi

gns,

late

epile

psy,

mor

talit

y

Ret

rosp

ectiv

ere

view

of35

9pa

tient

sw

ithco

mpo

und,

depr

esse

dcr

ania

lfr

actu

res

toex

amin

eth

eca

uses

and

cons

eque

nces

ofin

fect

ion

and

toas

sess

the

effe

ctof

surg

ical

timin

gan

dm

etho

dson

infe

ctio

nra

te.

●10

.6%

over

all

infe

ctio

nra

te.

●In

fect

ion

was

sign

ifica

ntly

asso

ciat

edw

ithin

crea

sed

mor

talit

y,pr

olon

ged

CN

Ssi

gns,

late

epile

psy.

●In

cide

nce

ofin

fect

ion

was

sign

ifica

ntly

grea

ter

inpa

tient

sw

ith�

48h

dela

ybe

twee

nin

jury

and

oper

atio

n.●N

odi

ffere

nce

inin

fect

ion

rate

betw

een

the

grou

pw

hose

dura

was

clos

edan

dth

egr

oup

who

sedu

raw

asle

ftop

en.

●Pr

imar

ybo

nefr

agm

ent

repl

acem

ent

did

not

sign

ifica

ntly

affe

ctpo

stop

erat

ive

infe

ctio

nra

teor

inci

denc

eof

late

epile

psy.

Jenn

ett

etal

.(1

0)10

00N

/AN

/ASu

rger

yve

rsus

nosu

rger

yIn

cide

nce

ofea

rly

and

late

epile

psy

Ret

rosp

ectiv

ere

view

of10

00pa

tient

sw

ithno

nmis

sile

depr

esse

dcr

ania

lfr

actu

res

tost

udy

risk

fact

ors

asso

ciat

edw

ithpo

sttr

aum

atic

epile

psy.

●10

%in

cide

nce

ofea

rly

epile

psy

(1w

k).

●15

%in

cide

nce

ofla

teep

ileps

y(�

1w

k).

●In

crea

sed

risk

ofla

teep

ileps

yw

ithpo

sttr

aum

atic

amne

sia

�24

h,to

rndu

ra,

foca

lne

urol

ogic

alsi

gns,

occu

rren

ceof

earl

yep

ileps

y.●Lo

wer

risk

ofea

rly/

late

epile

psy

with

occi

pita

lfr

actu

res.



TAB

LE1.

Con

tinu

ed

Aut

hors

No.

ofpa

tien

tsC

lass

Incl

usio

nG

CS

Trea

tmen

tO

utco

me

Des

crip

tion

Con

clus

ion

No.

ofpa

tien

tsIn

fect

ion

(%)

Initi

ally

clea

n,in

jury

toO

R�

36h

283

3.91

Initi

ally

clea

n,in

jury

toO

R�

48h

214.

8

Inju

ryto

OR

�48

hor

neve

r76

36.5

2

Bon

efr

agm

ent

repl

acem

ent

166

3.6

Bon

efr

agm

ent

rem

oval

135

4.4

1P

�n.

s,2

P�

0.00

1

Men

delo

wet

al.

(13)

176

N/A

N/A

Prop

hyla

ctic

antib

iotic

sve

rsus

noan

tibio

tics

Infe

ctio

nra

teR

etro

spec

tive

seri

esof

176

patie

nts

with

com

poun

d,de

pres

sed

cran

ial

frac

ture

sto

dete

rmin

efa

ctor

sth

atpr

edis

pose

toin

fect

ion

and

the

effe

ctof

prop

hyla

ctic

antib

iotic

son

infe

ctio

nra

te.

●6.

3%ov

eral

lin

fect

ion

rate

.●Pr

ophy

lact

icus

eof

ampi

cilli

nan

da

sulp

hona

mid

ew

asas

soci

ated

with

asi

gnifi

cant

lylo

wer

infe

ctio

nra

te(1

.9%

)th

anw

ithan

yot

her

com

bina

tion

(15.

6%)

orno

antib

iotic

sat

all

(10.

5%).

●D

ural

pene

trat

ion,

prim

ary

vers

usse

cond

ary

clos

ure,

sinu

sin

volv

emen

t,su

rger

yve

rsus

nosu

rger

y,bo

nere

plac

emen

tve

rsus

rem

oval

did

not

sign

ifica

ntly

affe

ctin

fect

ion

rate

.

No.

ofpa

tien

tsIn

fect

ion

(%)

Am

pici

llin

�su

lpho

nam

ide

107

1.9

No

antib

iotic

s19

10.5

1

All

othe

ran

tibio

tics

4515

.62

1P

�0.

05,

2P

�0.

01,

com

pare

dw

itham

pici

llin

�su

lpho

nam

ide

Stei

nbok

etal

.(1

5)11

1III

Dur

atio

nof

com

aSu

rger

yan

dno

nsur

gica

lIn

cide

nce

ofse

izur

es,

neur

olog

ical

dysf

unct

ion,

cosm

esis

;tim

ing

not

spec

ified

Ret

rosp

ectiv

ese

ries

of11

1ch

ildre

n�

16yr

with

sim

ple

depr

esse

dcr

ania

lfr

actu

res

toex

amin

eou

tcom

ew

ithre

spec

tto

the

occu

rren

ceof

seiz

ures

,ne

urol

ogic

aldy

sfun

ctio

n,an

dco

smes

isbe

twee

nsu

rgic

ally

and

nons

urgi

cally

man

aged

patie

nts.

●N

odi

ffere

nce

inse

izur

es,

neur

olog

ical

dysf

unct

ion,

cosm

etic

defo

rmity

betw

een

surg

ical

and

nons

urgi

cal

trea

tmen

t.G

roup

sdi

ffere

dw

ithre

spec

tto

max

imum

depr

essi

onof

frac

ture

(P�

0.00

0),

but

not

todu

ratio

nof

com

a,ag

e.●D

ural

lace

ratio

nw

assi

gnifi

cant

lyas

soci

ated

with

neur

olog

ical

defic

it.

No.

ofpa

tien

tsSe

izur

es(%

)C

osm

etic

defo

rmit

y(%

)R

esid

ual

defic

it(%

)

Surg

ery

1910

.521

0

Non

surg

ical

234.

317

.40

van

den

Hee

ver

and

van

der

Mer

we

(16)

284

IIIU

nkno

wn

Surg

ery

vers

usno

nsur

gica

l

Sept

icco

mpl

icat

ion;

outc

ome

offo

cal

neur

olog

ical

abno

rmal

ities

;m

orta

lity

Ret

rosp

ectiv

ere

view

of28

4pa

tient

sw

hose

prim

ary

inju

ryw

asno

nmis

sile

depr

esse

dcr

ania

lfr

actu

reto

com

pare

outc

ome

betw

een

surg

ical

and

nons

urgi

cal

grou

ps.

Surg

ical

indi

catio

nsde

taile

d.

●5.

3%in

fect

ion

rate

—2.

8%co

nser

vativ

eve

rsus

8%su

rgic

al.

●5%

inci

denc

eof

dete

rior

atio

nof

neur

olog

ical

defic

it—4%

cons

erva

tive

vers

us7%

surg

ical

.●1.

4%m

orta

lity,

all

insu

rgic

algr

oup—

sele

ctio

nbi

asno

ted.

Not

e:ca

nnot

sepa

rate

patie

nts

with

CT.

Pape

rin

clud

edto

show

com

para

ble

outc

ome

betw

een

surg

ical

and

nons

urgi

cal

man

agem

ent

ofde

pres

sed

cran

ial

frac

ture

.

No.

ofpa

tien

tsIn

fect

ion

(%)

Mor

talit

y(%

)

Non

surg

ical

139

2.8

0

Surg

ery

124

83.

2

No.

ofpa

tien

tsD

eter

iora

tion

(%)

Non

surg

ical

624

Wyl

enet

al.

(17)

32III

GC

S6

–15

Bon

efr

agm

ent

repl

acem

ent

Infe

ctio

n,av

erag

e22

mo

Ret

rosp

ectiv

ere

view

of32

patie

nts

surg

ical

lytr

eate

dw

ithpr

imar

ybo

nefr

agm

ent

repl

acem

ent

for

com

poun

dde

pres

sed

cran

ial

frac

ture

with

in72

hof

inju

ryto

exam

ine

infe

ctio

nra

te.

●N

oin

stan

ces

ofin

fect

ious

com

plic

atio

nsor

new

/wor

sene

dne

urol

ogic

alde

ficit

follo

win

gpr

imar

ybo

nefr

agm

ent

repl

acem

ent

with

in72

hof

inju

ry.

●25

%in

cide

nce

ofpe

rsis

tent

neur

olog

ical

defic

itat

6m

o.

No.

ofpa

tien

tsIn

fect

ion

(%)

Rep

lace

men

t32

0

aG

CS,

Gla

sgow

Com

aSc

ale;

GO

S,G

lasg

owou

tcom

esc

ore;

GR

,go

odre

cove

ry;

N/A

,no

tap

plic

able

;C

T,co

mpu

ted

tom

ogra

phic

scan

;SA

H,

suba

rach

noid

hem

orrh

age;

D,

deat

h;C

NS,

cent

ral

nerv

ous

syst

em;

OR

,op

erat

ion;

n.s.

,no

tsi

gnifi

cant

.

APPENDICES

APPENDIX I: POST-TRAUMATIC MASS VOLUME MEASUREMENT IN

TRAUMATIC BRAIN INJURY PATIENTS

1. Direct volumetric measurement with imaging softwareusing a modern computer tomographic CT scanner is thegold standard. This has been applied only on rare occa-sions.

2. The “ellipsoid method” was developed to calculate thevolume of arteriovenous malformations (3). It is based onthe concept that the volume of an ellipsoid is approxi-mately one-half of the volume of the parallelepiped (asix-faced polyhedron, all of whose faces are parallelo-grams lying in pairs of parallel planes) into which it isplaced. By measuring three diameters of a given lesion inthe arterial phase of an angiogram, a parallelepiped isconstructed, and its volume, divided in half, is close tothe actual volume of the malformation. By extending thisconcept from angiography to CT scanning, calculation ofspace-occupying lesions becomes possible (4). The“ABC” method has been described by Kothari et al. (2)for the measurement of intracerebral hemorrhages, andis also based on the concept of measuring the volume ofan ellipsoid. The formula for an ellipsoid is:

Ve � 4/3 � (A/2) (B/2) (C/2)where A, B, and C are the three diameters.For � � 3, the formula becomes Ve � ABC/2

The volume of an intracerebral hemorrhage can be approx-imated by following the steps listed below:

• Identify the CT slice with the largest area of hemorrhage(Slice 1)

• A: measure the largest diameter, A.• B: measure the largest diameter 90° to A on the same slice,

B.

• C: count the number of 10-mm slices.• Compare each slice with slice 1.• If the hemorrhage is greater than 75% compared with slice 1,

count the slice as 1.• If the hemorrhage is 25 to 75%, count the slice as 0.5.• If the hemorrhage less than 25%, do not count the slice.• Add up the total C.

3. More recently, the “Cavalieri direct estimator” methodhas been introduced (1). It breaks down the lesion on theCT scan into a corresponding number of points. Thevolume of a lesion is the product of the sum of the pointsthat fall into the lesion, the area associated with eachpoint, and the distance between the scan slices. A gridthat is used to determine the number of points can beobtained by photocopying a template provided in theoriginal article or by preparing a uniformly spaced pointgrid by computer (4).

REFERENCES

1. Clatterbuck R, Sipos E: The efficient calculation of neurosurgically relevantvolumes from computed tomographic scans using Cavalieri’s Direct Estima-tor. Neurosurgery 40:339–342, 1997.

2. Kothari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR, Zuccarello M,Khoury J: The ABCs of measuring intracerebral hemorrhage volumes. Stroke27:1304–1305, 1996.

3. Pasqualin A, Barone G, Cioffi F, Rosta L, Scienza R, Da Pian R: The relevanceof anatomic and hemodynamic factors to a classification of cerebral arterio-venous malformations. Neurosurgery 28:370–379, 1991.

4. Stocchetti N, Croci M, Spagnoli D, Gilardoni F, Resta F, Colombo A: Massvolume measurement in severe head injury: Accuracy and feasibility of twopragmatic methods. J Neurol Neurosurg Psychiatry 68:14–17, 2000.


APPENDICES

APPENDIX II: EVALUATION OF RELEVANT COMPUTED

TOMOGRAPHIC SCAN FINDINGS

Computed tomographic (CT) scanning is the imagingmodality of choice for traumatic brain injury because ofits widespread availability, the rapid imaging time, the

low associated costs, and its safety. CT scanning measures thedensity of tissues using x-rays. To standardize the imagingprocedure, 5-mm slices should be obtained from the foramenmagnum to the sella and 10-mm slices should be obtainedabove the sella, parallel to the orbitomeatal line. The followingearly CT scan findings correlate with outcome (1):

• Status of the basal cisterns.• Midline shift.• Subarachnoid hemorrhage in the basal cisterns.

Basal Cisterns at the Midbrain Level

Compressed or absent basal cisterns indicate a threefoldrisk of raised intracranial pressure and the status of the basalcisterns is related to outcome. The degree of mass effect isevaluated at the level of the midbrain. Cerebrospinal fluidcisterns around the midbrain are divided into three limbs, oneposterior and two laterally (Fig. 1). Each limb can be assessed

separately as to whether ornot it is open or com-pressed. Basal cisterns canbe:

• Open (all limbs open).• Partially closed (one or

two limbs obliterated).• Completely closed (all

limbs obliterated).

Midline Shift at theForamen of Monro

The presence of midlineshift is inversely related toprognosis. However, inter-action exists with the pres-ence of intracranial lesions and other CT parameters (1). Mid-line shift at the level of the foramen of Monro should bedetermined by first measuring the width of the intracranialspace to determine the midline (“A”). Next, the distance fromthe bone to the septum pellucidum is measured (“B”) (Fig. 2).The midline shift can be determined by calculating:

Midline shift � (A/2) � B

Traumatic Subarachnoid Hemorrhage

Traumatic subarachnoid hemorrhage occurs in between 26and 53% of all patients with severe traumatic brain injury.Mortality is increased twofold in the presence of traumaticsubarachnoid hemorrhage. The presence of subarachnoidhemorrhage in the basal cisterns carries a positive predictivevalue of unfavorable outcome of approximately 70%.

REFERENCE

1. Chesnut R, Ghajar J, Maas A, Marion D, Servadei F, Teasdale G, Unterberg A,von Holst, Walters B: Early indicators of prognosis in severe traumatic braininjury. J Neurotrauma 17:535–627, 2000.FIGURE 1. Evaluation of the basal cisterns on computed tomographic scan.

FIGURE 2. Assessment of midlineshift on computed tomographic scan.


Surgical guidelines article_2

Documents

Transcript of Surgical guidelines article_2