Ai

A

DC

dhta

C

0d

The American Journal of Surgery (2010) 200, 810–813

ppendicitis outcomes are better at resident teachingnstitutions: a multi-institutional analysis

rezou Yaghoubian, M.D., Christian de Virgilio, M.D., Steven L. Lee, M.D.*

epartment of Surgery, Harbor-UCLA Medical Center, Torrance, CA, USA; Kaiser Permanente Los Angeles Medical

enter, Los Angeles, CA

AbstractBACKGROUND: This study compared the outcomes of appendicitis between teaching and nonteach-

ing institutions.METHODS: A retrospective review was performed of all appendicitis patients aged �18 years from

1998 to 2007. The outcomes from 2 teaching institutions (each with its own general surgery residencyprogram) were compared with those from 11 nonteaching institutions. Study outcomes includedpostoperative morbidity and length of hospitalization.

RESULTS: A total of 3,242 patients were treated at the teaching institutions (mean age, 41 years;61% men) and 14,483 at the nonteaching institutions (mean age, 38 years; 54% men). The perforatedappendicitis rate was 29% at the teaching institution and 28% at the nonteaching institutions (P � .20).For nonperforated appendicitis, there was no difference in the incidence of wound infection between theteaching and nonteaching institutions (2.7% vs 2.3%, P � .30). There was a lower rate of abscessdrainage (.4% vs 1%, P � .02), a lower readmission rate (1.7% vs 3.5%, P � .0001), and shorterlengths of stay (1.7 � 1.5 vs 1.8 � 1.6 days, P � .002) at teaching institutions. For perforatedappendicitis, there were also lower rates of wound infection (4.8% vs 7%, P � .03), abscess drainage(4.9% vs 10%, P � .0001), and need for readmission (4.2% vs 8.4%, P � .0001) at the teachinghospitals. The lengths of stay were similar (5.0 � 4.2 vs 5.2 � 3.1 days, P � .30). Use of laparoscopywas lower and nonoperative management of perforated appendicitis higher at the teaching hospitals.

CONCLUSIONS: Teaching institutions were more likely to perform appendectomy using an opentechnique and to manage perforated appendicitis nonoperatively. Infectious complications and read-mission rates for both perforated and nonperforated appendicitis were lower at teaching institutions.© 2010 Elsevier Inc. All rights reserved.

KEYWORDS:Resident education;Surgery residency;Appendicitis;Surgical outcomes

beNp

mtgrs

There is an ongoing debate over the quality of careelivered by teaching hospitals. Patients frequently expressesitation at the prospect of being treated by physicians inraining. Patients’ concerns are further magnified when theyre undergoing surgical procedures at teaching hospitals,

Presented at the 2010 annual meeting of the Southwestern Surgicalongress, Tucson, AZ.

* Corresponding author. Tel.: 310-222-2706; fax: 310-782-1562.E-mail address: slleemd@yahoo.com

tManuscript received March 6, 2010; revised manuscript July 29, 2010

002-9610/$ - see front matter © 2010 Elsevier Inc. All rights reserved.oi:10.1016/j.amjsurg.2010.07.028

ecause they assume that the participation of less experi-nced surgical residents will adversely affect outcomes.evertheless, practical training is essential to produce com-etent physicians.

Surgical residents gain extensive experience in the manage-ent of appendicitis early in their training. In fact, appendec-

omy is the most common emergency procedure performed byeneral surgeons. Open and laparoscopic appendectomies areelatively straightforward procedures to teach and to learn. Asuch, they are ideal procedures for senior residents to serve as

he teaching assistants to junior residents.

tpwdotTci

M

ryfiCpattpgapia

pmmdpi

(ntvnmvt

R

tsiSi

ps

pwid3ttc

oa8tript

C

ohcf

811A. Yaghoubian et al. Appendicitis outcomes better at teaching hospitals

Some surgeons and patients believe that educating andraining future surgeons conflicts with providing the bestossible care to patients. Yet few data are available as tohether resident involvement in common surgical proce-ures adversely affects patient outcomes. In the current eraf a heightened emphasis on patient outcomes, it is impor-ant to document the effect of residents as primary surgeons.hus, the purpose of this study was to determine the out-omes of appendicitis between teaching and nonteachingnstitutions.

ethods

After obtaining institutional review board approval, aetrospective review of all appendicitis patients aged � 18ears between 1998 and 2007 at 12 hospitals was per-ormed. Two were teaching hospitals (Harbor-UCLA Med-cal Center and Kaiser Permanente Los Angeles Medicalenter), and 11 were nonteaching Kaiser Permanente hos-itals. At the teaching institutions, the surgical residentsre actively involved in all aspects of patient care. They arehe first to see consultations in the emergency room, serve ashe primary surgeons, direct the postoperative care, and seeatients back in clinic after discharge. Senior resident sur-eons at the teaching institutions typically serve as teachingssistants for the appendectomy cases under attending su-ervision. In contrast, there are no residents at the nonteach-ng institutions, and the attending surgeons perform allspects of patient care.

Patient factors collected included age, gender, and theresence of perforation. Outcome variables were 30-dayorbidity and length of hospitalization (LOH). Thirty-dayorbidity included wound infection, postoperative abscess

rainage, and readmission. Outcomes of patients with ap-endicitis were compared between teaching and nonteach-ng hospitals.

All patient data were collected in an Excel spreadsheetMicrosoft Corporation, Redmond, WA) and translated intoative SAS format using DBMS/Copy (Dataflux Corpora-ion, Cary, NC). Descriptive statistics were calculated for allariables. Numerical variables were compared using theonparametric Wilcoxon rank-sum test and are reported asedians with interquartile ranges. Categorical or nominal

ariables were compared using the �2 test or Fisher’s exactest, as appropriate.

esults

Overall, 3,242 patients with appendicitis were treated athe teaching institutions and 14,483 at the nonteaching in-titutions. The mean ages were 41 years at the teachingnstitutions and 38 years at the nonteaching institutions.ixty-one percent of patients were male at the teaching

nstitutions and 54% at the nonteaching institutions. The q

erforated appendicitis rates were 29% at the teaching in-titutions and 28% at the nonteaching institutions (P � .20).

Outcomes data are summarized in the Table 1. For non-erforated appendicitis, there was no difference in theound infection rate between teaching and nonteaching

nstitutions. However, there were lower rates of abscessrainage (.4% vs 1.0%, P � .02) and readmission (1.7% vs.5%, P � .0001) at teaching than at nonteaching institu-ions, respectively. LOH was shorter at the teaching insti-utions. The use of laparoscopy for nonperforated appendi-itis was lower at the teaching institutions.

For perforated appendicitis, there were also lower ratesf wound infection (4.8% vs 7.0%, P � .03), abscess drain-ge (4.9% vs 10.0%, P � .02), and readmission (4.2% vs.4%, P � .0001) at teaching than at nonteaching institu-ions, respectively. The LOH was similar. The use of lapa-oscopy for perforated appendicitis was lower at the teach-ng institutions. However, nonoperative management oferforated appendicitis was higher at the teaching institu-ions.

omments

There has always been controversy regarding the qualityf care delivered at teaching hospitals. Academic surgeonsave the dual responsibility of ensuring the best possibleare to patients while simultaneously training and educatinguture generations of surgeons. Thus, the ability to ade-

Table 1 Morbidity and LOH outcomes in patients withappendicitis managed at teaching and nonteachinginstitutions

Outcome Teaching Nonteaching P

Nonperforated appendicitis 2,289 10,385Wound infection 2.7% 2.3% .30Abscess drainage .4% 1% .02Readmission within 30

days 1.7% 3.5% �.0001LOH (days) 1.7 � 1.5 1.8 � 1.6 .002Laparoscopic

appendectomy 48% 68% �.0001Perforated appendicitis:

operative 953 4098Wound infection 4.8% 7% .03Abscess drainage 4.9% 10% �.0001Readmission within 30

days 4.2% 8.4% �.0001LOH (days) 5.0 � 4.2 5.2 � 3.1 .30Laparoscopic

appendectomy 23% 42% �.0001Perforated appendicitis:

nonoperative 10% 6% �.0001

Data are expressed as numbers, percentages, or mean � SD.

uately train surgical residents without compromising the

qi

ttwfasmprMctfrlpadtocgprdvtgrcttt

itoclplsawLcibgisca

aTtwsrtbbc

pnewosptrtsmctns

aWucdmcmbatto

powawtsp

R

812 The American Journal of Surgery, Vol 200, No 6, December 2010

uality of care is a concern for all faculty members involvedn surgical education.

Previous studies examining other procedures have shownhat the quality of care at teaching and nonteaching institu-ions is similar.1–5 A prior pilot study by our institution,hich compared 1 teaching and 1 nonteaching hospital,

ound comparable quality of care when outcomes for pedi-tric appendicitis were analyzed.6 Unlike our previoustudy, this multi-institutional study demonstrated lowerorbidity for both adult perforated and nonperforated ap-

endicitis at teaching hospitals with respect to readmissionate, wound infection, and postoperative abscess drainage.

eguid et al,7 in a retrospective study of outcomes of lungancer resections, demonstrated a lower mortality rate ateaching hospitals. Similarly, a study by the same groupound a lower mortality rate for abdominal aortic aneurysmepair at teaching hospitals although the length of stay wasonger.8 They attributed their improved outcomes to theresence of specialty training and increased volume. Theuthors hypothesized that the increase in length of stay wasue to sicker patients at the teaching hospitals. The limita-ion of the prior mentioned studies is that they evaluated theutcomes of complex procedures in which residents typi-ally play the role of assistants rather than primary sur-eons. Looking at more routine operations, Hwang et al1

erformed an analysis of outcomes of bowel resection, lapa-oscopic cholecystectomy, hernia, mastectomy, and appen-ectomy of 4 attending surgeons who worked with residentsersus 4 who did not. Comparing all procedures together,here were no differences in complications between theroups, although there was greater mortality, a greater du-ation of stay, and higher costs in the resident group. Whenomparing the 5 most common procedures individually,here was no difference in complications or mortality, al-hough a greater length of stay and higher costs were seen inhe teaching group.

In general, similar or improved outcomes at teachingnstitutions have come at the expense of increase LOH andhus higher cost. A study by Hutter et al2 in the evaluationf pancreatic resections also demonstrated improved out-omes at the teaching hospitals, but they found a longerength of stay and attributed this to their care of uninsuredatients. In a subset analysis, the uninsured patients had theongest length of stay (27 vs 20 days). Unlike previoustudies, in our pilot study comparing outcomes of pediatricppendicitis, we found a decrease in the LOH for childrenith nonperforated appendicitis.6 Because the decrease inOH was only .4 days, one can argue that this may not belinically significant. One possible explanation for this find-ng is that more patients at the teaching institution may haveeen discharged in the evening rather than the next morningiven the availability of the resident team. If this werendeed the case, then this would have led to substantial costavings in this group of patients, making this finding clini-ally significant. In this current multi-institutional study, we

lso found a slight decrease in LOH for nonperforated

ppendicitis and a similar LOH for perforated appendicitis.he LOH at teaching hospitals may have been even shorter

han stated because the rate of laparoscopic appendectomyas also lower at teaching hospitals. We have recently

hown that LOH was shorter in patients undergoing lapa-oscopic appendectomy versus open appendectomy.9 Al-hough we did not perform a formal cost analysis, weelieve that cost at the teaching institutions was lowerecause both morbidity was lower and LOH was shorterompared with nonteaching institutions.

In the present study, we also found that patients witherforated appendicitis were more likely to be managedonoperatively at the teaching institutions. A potentialxplanation for this finding is that teaching institutionsere more willing to implement a fairly novel techniquef nonoperative management on the basis of researchtudies.10 –12 By contrast, the nonteaching institutionserformed a much higher rate of laparoscopic appendec-omies compared with the nonteaching institutions. Theeason behind this observation was because one of theeaching institutions transitioned into performing laparo-copic appendectomy at a much later date. This was dueainly to resistance from the attending surgeons to

hange in techniques, and it was the residents’ influencehat ultimately led to the change in practice. Currently,early all patients with appendicitis will undergo laparo-copic appendectomy at both institutions.

There were several limitations of this study. First, this isretrospective analysis and by nature had its limitations.e did not evaluate the length of the operation, which may

nderstandably take longer at the academic institutions andontribute to increased costs of the operating room, asemonstrated by other studies. Furthermore, the overall loworbidity and mortality of appendicitis may offset the in-

reased risk for resident-related errors in patient manage-ent. We also did not review the number of cases that

egan laparoscopically and required conversion to openppendectomy between the 2 types of institutions. Last,eaching hospitals may be inherently different from non-eaching hospitals independent of the presence of residentsr students.

Our data demonstrate that teaching hospital care ofatients with appendicitis is associated with improvedutcomes. In addition, patients at teaching institutionsith perforated appendicitis were more likely to be man-

ged nonoperatively at teaching institutions comparedith nonteaching institutions. In general, our data refute

he notion that there is a conflict between training futureurgeons and simultaneously providing the best possibleatient care.

eferences

1. Hwang C, Pagano CR, Wichterman KA. Resident versus no resident:a single institutional study on operative complications, mortality, and

cost. Surgery 2008;144:339–44.

1

1

1

D

q2oo

tidsItddpmqw

tntreotdtwtiCsnaanga

813A. Yaghoubian et al. Appendicitis outcomes better at teaching hospitals

2. Hutter MM, Glasgow RE, Mulvihill SJ. Does the participation of asurgical trainee adversely impact patient outcomes? A study of majorpancreatic resections in California. Surgery 2000;128:286–92.

3. Coates KW, Kuehl TJ, Bachofen C, et al. Analysis of surgical com-plications and patient outcomes in a residency training program. Am JObstet Gynecol 2001;184:1380–5.

4. Paisley AM, Madhavan KK, Paterson-Brown S, et al. Role of thesurgical trainee in upper gastrointestinal resectional surgery. Ann RColl Surg Engl 1999;81:40–59.

5. Doerr RJ, Yildiz I, Flint LM. Pancreaticoduodenectomy: universityexperience and resident education. Arch Surg 1990;125:463–5.

6. Lee SL, Shekherdimian S, Chiu VY. Comparison of pediatric appen-dicitis outcomes between teaching and nonteaching hospitals. J PediatrSurg 2010;45:894–7.

7. Meguid RA, Brooke BS, Chang DC, et al. Are surgical outcomes forlung cancer resections improved at teaching hospitals? Ann ThoracSurg 2008;85:1015–24.

8. Meguid RA, Brooke BS, Perler BA, et al. Impact of hospital teachingstatus on survival from ruptured abdominal aortic aneurysm repair. JVasc Surg 2009;50:243–50.

9. Lee SL, Yaghoubian A, Aboulian A, et al. Laparoscopic versus openappendectomy in children: outcomes comparison based on age. In:International Pediatric Endoscopy Group Proceedings. In press.

0. Nadler EP, Reblock KK, Vaughan KG, et al. Predictors of outcome forchildren with perforated appendicitis initially treated with non-opera-tive management. Surg Infect 2004;5:349–56.

1. Yamini D, Vargas H, Bongard F, et al. Perforated appendicitis: is ittruly a surgical urgency? Am Surg 1998;64:970–5.

2. Oliak D, Yamini D, Udani VM, et al. Nonoperative management ofperforated appendicitis without periappendiceal mass. Am J Surg2000;179:177–81.

iscussion

Gary Vercruysse, M.D. (Atlanta, GA): I have a fewuestions and observations. Why in the period of 1998 to007 well after published work on CT for the diagnosisf appendicitis did both institutions have such a high rate

f perforation, 29 vs 28%? Also I would be interested in h

he rates of negative appendectomies as well if that datas available. Were there significant delays in time toiagnosis, emergency department delays, delays in con-ultation, or response to consultation, etc. to explain this?n the 1990s appendectomy changed from largely an openo laparoscopic procedure. Your institution publishedata that suggested it leads to a shorter length of stay andecreased cost. Why the delay in adapting to newerrocedure at your institution? Is there still a role in yourind for open appendectomy? This is sort of a dated

uestion I guess, because I think there is for sure. Andho is the ideal patient for an open appendectomy?Arezou Yaghoubian, M.D. (Torrance, CA): As far as

he high rate of perforation seen at both the teaching andonteaching institutions, I think that is more of a reflec-ion of the prehospital course of the patients and just mayeflect the patient population that the institutions werexposed to. We didn’t collect information on the numberf negative appendectomies unfortunately and whetherhere was a delay in diagnosis or management. We alsoid not evaluate the time from evaluation in the ED toime of CT scan or arrival to the OR. As far as why thereas such a lower rate of laparoscopic appendectomy at

he two teaching facilities, one of the two teaching facil-ties adopted laparoscopic appendectomy at a later date.urrently, for both institutions, all appendectomies are

tarted laparoscopically and only converted to open ifecessary. As far as who is the ideal patient for openppendectomy, typically at our institution again we beginll cases laparoscopically and only convert to open whenecessary. Anecdotally I have noticed there is a certainroup of patients, the very young and old that tend to bepproached from the open method from the beginning. I

ope I have addressed all your questions.