O

maoe

i

otc

J Oral Maxillofac Surg70:1040-1044, 2012

Logistic Regression Analysis of RiskFactors for the Development of

Alveolar OsteitisDiego Halabí, BDS,* José Escobar, BDS,† Carlos Muñoz, DDS,‡

and Sergio Uribe, DDS, MSc§

Purpose: To assess risk factors for alveolar osteitis.

Materials and Methods: A prospective nested case-control study was conducted in an urban com-munity dental clinic in Valdivia, Chile. A cohort of 1,355 patients who underwent dental extractions wasincluded. Eight predictor variables (risk factors), namely patient gender, hygiene, tooth location, previ-ous surgical site infection, traumatic extraction, systemic diseases, alcohol consumption, and tobaccouse, were considered in a risk factor model. A binary regression logistic analysis was performed todetermine significant associations.

Results: In total 1,302 participants completed the follow-up. Eighty incident case patients with alveolarosteitis and 80 matched control patients were included. A statistically significant association was foundbetween traumatic extraction (odds ratio [OR], 13.1; 95% confidence interval [CI], 5.4 to 31.7), tobaccosmoking after extraction (OR, 3.5; 95% CI, 1.3 to 9.0), previous surgical site infection (OR, 3.3; 95% CI,1.4 to 7.7), and the development of alveolar osteitis.

Conclusions: Previous surgical site infection, traumatic extraction, and tobacco smoking are associatedwith an increased risk of alveolar osteitis.© 2012 American Association of Oral and Maxillofacial Surgeons

J Oral Maxillofac Surg 70:1040-1044, 2012

0

f all postextraction complications,1 alveolar oste-itis occurs most commonly, with an incidence of4.8% for the total number of tooth extractions.2 The

ain symptom is severe pain within the socketfter extraction, although some patients reportther symptoms such as headaches, fever, or par-sthesia.3

Risk factors reported to be associated with alve-olar osteitis include patient gender,2,4 herpes infec-tion,5 mandibular location of the tooth,2,3 level ofmpaction, previous surgical site infection,6 trau-

matic exodontia,5 poor oral hygiene,7 handednessf the operator (right vs left), lingual flap retrac-ion,8 systemic diseases,3,6 smoking, and alcoholonsumption.3

Received from the School of Dentistry, Faculty of Medicine, Uni-

versidad Austral de Chile, Valdivia, Chile.

*Clinical Instructor, School of Dentistry, Faculty of Medicine,

Universidad Austral de Chile, Valdivia, Chile.

†Clinical Instructor, School of Dentistry, Faculty of Medicine,

Universidad Austral de Chile, Valdivia, Chile.

‡Adjunct Instructor, School of Dentistry, Faculty of Medicine,

Universidad Austral de Chile, Valdivia, Chile.

§Associate Professor, School of Dentistry, Faculty of Medicine,

Universidad Austral de Chile, Valdivia, Chile.d

1040

Numerous studies have attempted to explain theetiology of this complication, which seems to bemultifactorial. However, the available evidence iscompounded by statistical and methodologic issues.9

Inconsistent and often conflicting results are found inthe literature9 because they come from individualopinions, or from case studies that lack analysis, or areperformed with methods that are poorly designed, ornot reproducible.

A better knowledge of how risk factors are corre-lated with alveolar osteitis would help to anticipatethe development of this complication in patients atrisk. This could decrease the incidence of alveolarosteitis, in turn decreasing the successive surgeryrequired to treat it and thus health care costs.10

This article is based on a thesis submitted in partial fulfillment of

the requirements for the degree of Bachelor in Dental Surgery in

the School of Dentistry, Universidad Austral de Chile, 2011.

Address correspondence and reprint requests to Dr Uribe:

School of Dentistry, Faculty of Medicine, Universidad Austral de

Chile, Rudloff 1640, Valdivia, Chile; e-mail: sergiouribe@uach.cl

© 2012 American Association of Oral and Maxillofacial Surgeons

278-2391/12/7005-0$36.00/0

oi:10.1016/j.joms.2011.11.024

scha

gwpptt

metop1brsmmtase((gwd

ltaae

pfpatwa

ttt

HALABÍ ET AL 1041

Several therapies have been proposed for the treat-ment of alveolar osteitis,11 but these are limited toymptomatic management and are of uncertain clini-al efficiency11; hence, current literature reviewsave reinforced preventative methods as the key tovoiding this complication.10 Therefore, it is crucial

to correctly identify probable risk factors to identifythose patients having a high risk of developing alve-olar osteitis.

Using a logistic regression model, the aim of thisstudy was to assess the relation between 8 risk factors(patient gender, oral hygiene, tooth location, previ-ous surgical site infection, traumatic extraction, sys-temic diseases, alcohol consumption, and tobaccouse) and the development of alveolar osteitis.

Materials and Methods

A prospective nested case-control study was con-ducted. The baseline was a cohort of 1,423 adultsolder than 18 years, recruited from March throughJune 2011, who underwent dental extractions in apublic community dental clinic in Valdivia, Chile(population, 127,750). The study protocol was ap-proved by the research ethics committee of the Pub-lic Health Service of Valdivia, and all participantssigned an informed consent according to recommen-dations of the Declaration of Helsinki.

Exclusion criteria were narrowed to individuals re-quiring extraction in the operating theater, residentsof rural areas who manifested difficulty in returningfor follow-up, and those who were undergoing anti-microbial therapy.

Dental extractions were performed by 6 dentalsurgeons from the emergency department of theclinic in accordance with standard procedures as de-fined by the National Health Service (Ministry ofHealth). Before surgery, details of patient gender,tooth location, plaque index, previous surgical siteinfection, trauma, and systemic diseases were regis-tered using a clinical record designed for this study.Tobacco use, alcohol consumption, and a re-evalua-tion of any underlying systemic diseases were re-corded for each participant in a self-report consistingof a simple 3-question patient history questionnaire.

Clinical variables were recorded in accordancewith previous research3 as follows: patient gender(male or female), location of the tooth (maxilla ormandible), and oral hygiene (good: Simplified OralHygiene Index12 � 0 or 1; poor: Simplified Oral Hy-iene Index � �2). Previous surgical site infectionas evaluated as present (clinical diagnosis of chroniceriodontitis, acute periodontal conditions, apicaleriodontitis, pericoronitis, fungal infections, or den-al pulp gangrene) or absent. Trauma during extrac-

ion was evaluated as simple extraction (nontrau-

atic) or surgical extraction (lifting a flap, use oflevators for �4 min, and/or rotary instruments). Sys-emic disease was evaluated from the medical historyf the patient and re-evaluated from the self-reportedatient history questionnaire that considered at leastof the following conditions: high blood pressure,

lood disorders, diabetes, rheumatic disorders, respi-atory disorders, kidney disorders, heart disease,troke, liver disease, specific nutritional require-ents, or any other condition requiring medical treat-ent. Alcohol consumption was self-reported by par-

icipants as alcohol consumption (�2 measures oflcohol 48 hrs after extraction) or no alcohol con-umption (1 or 0 measures of alcohol 48 hrs afterxtraction). Tobacco use was evaluated as smoker�5 cigarettes 24 hrs after extraction) or nonsmoker�5 cigarettes 24 hrs after extraction). After the sur-ical intervention, all participants received verbal andritten instructions and information on the proce-ure and postoperative care.Four days later, completed questionnaires were col-

ected and participants were clinically examined. Pa-ients who did not return for follow-up were locatednd examined in their homes. The home clinical ex-mination was performed with a sterile oral mirror,xamination gloves, and artificial light.Incident cases were identified as patients with a

ositive clinical diagnosis of alveolar osteitis and theollowing characteristics: 1) increasing postoperativeain intensity for 4 days within and around the socketnd 2) total or partial breakdown of the blood clot inhe socket with or without bone exposure.3 Patientsho did not meet 1 of these 2 criteria were identified

s controls.Patients who developed alveolar osteitis, and pa-

ients who developed other complications, werereated in accordance with the available clinical pro-ocols of the Chilean Health Ministry.13

Registration of the 80th incident case completedthe recruitment required for the present study. Eightymatched controls were obtained at random.

The sample size was estimated to comply with anincidence of 10 events per variable.14 Data were tab-ulated in Calc 3.2 (Apache Found., Los Angeles, CA,USA) and exported to SPSS 19 (SPSS, Inc, Chicago, IL)to conduct a binary logistic regression analysis andestablish a risk model to correlate the explanatoryvariables with the development of alveolar osteitis.Participants lost to follow-up were excluded from thestudy. For all tests, statistical relations were deter-mined to be significant at P � .05.

Results

During a period of 93 days, 1,422 patients sought

care in the emergency department of the clinic. Of

Surg 2

O

A

1042 RISK FACTORS FOR ALVEOLAR OSTEITIS

these, 1,355 fulfilled the inclusion criteria and gavewritten informed consent to take part in the study.Fifty-three participants were lost during follow-up, sothe complete data were from 1,302 participants.

Alveolar osteitis was detected in 80 participants,representing an overall incidence of 6.14%. This wascontrasted with 80 controls randomized from all con-trol participants completing the follow-up (1,222).

The mean age of the participants was 39.7 years(standard deviation, 16 years). Further details arelisted in Table 1.

A binary logistic regression analysis was conductedto assess whether the risk model, with 8 predictors,varied significantly by whether patients developedalveolar osteitis. The predictor variables were patientgender, hygiene, tooth location, previous surgical siteinfection, traumatic extraction, systemic diseases, al-cohol consumption, and tobacco use. When all 8predictor variables were considered together, theysignificantly predicted whether a patient would de-velop alveolar osteitis (n � 160, �2

8 � 63.8, P �.001). Logistic regression analysis after adjustmentshowed a statistically significant association between3 variables and the development of alveolar osteitis(Table 2). Those 3 factors were previous surgical siteinfection (odds ratio [OR], 3.3; 95% confidence inter-val [CI], 1.4 to 7.7), traumatic extraction (OR, 13.1;95% CI, 5.4 to 31.7), and smoking of tobacco afterextraction (OR, 3.5; 95% CI, 1.3 to 9.0).

Discussion

The aim of the present study was to analyze somerisk factors for alveolar osteitis using a nested case-control study. In this study, the authors identifiedcases of alveolar osteitis that occurred within a de-fined cohort, and a specified number of matchedcontrols was selected from among those in the cohort

Table 1. DISTRIBUTION OF MEAN AGE IN YEARS, FREQPREVIOUS INFECTION IN SURGICAL SITE, TRAUMATIC ECONSUMPTION, AND TOBACCO SMOKING BETWEEN C

Alveolar Ostei

Age (yrs), mean � SD 39.1 � 17.1Mandibular exodontia 43Systemic disease 22Female gender 43Alcohol consumption 20Poor oral hygiene 62Previous surgical site infection 63Tobacco smoke 33Traumatic extraction 52

Abbreviation: SD, standard deviation.

Halabí et al. Risk Factors for Alveolar Osteitis. J Oral Maxillofac

who did not develop alveolar osteitis at the time of

disease occurrence in each case. There was an in-creased risk of alveolar osteitis in patients with previ-ous surgical site infection, those who had undergonea traumatic tooth extraction, and those who hadsmoked tobacco within 24 hours after the extraction.

The incidence of alveolar osteitis over 3 monthswas 6.14%, an incidence higher than reported byNussair and Younis,2 Bortoluzzi et al,15 and Oginni etal.16 This difference may be explained by the fact thatthe present study design included a follow-up checkfor all participants, whereas the cited studies did notmention the number of follow-ups. Another possibleexplanation may be the fact that a different unit ofanalysis was used. The present study considered thepatient the unit of analysis, whereas other reportsconsidered the number of extractions the unit ofanalysis, without considering the patients.2,15,16 The

Y OF FEMALE GENDER, MANDIBULAR EXODONTIA,CTION, UNDERLYING SYSTEMIC DISEASE, ALCOHOLAND CONTROLS

Control Total P Value

40.2 � 14.8 39.7 � 16 .6641 84 .7522 44 .9047 90 .5214 34 .2546 108 �.0540 103 �.0124 57 .1412 64 �.01

012.

Table 2. EFFECT OF EXPLANATORY VARIABLES ONDEVELOPMENT OF ALVEOLAR OSTEITIS OBTAINED BYLOGISTIC REGRESSION ANALYSIS

RegressionCoefficient

(�) OR (95% CI) P Value

Mandibular exodontia �0.49 1.0 (0.4-2.1) .903Systemic disease �0.38 0.7 (0.3-1.7) .410Female gender �0.09 0.9 (0.4-2.1) .816Alcohol 0.06 1.1 (0.4-2.9) .911Poor oral hygiene 0.45 1.6 (0.7-3.7) .310Previous infection in

surgical site 1.18 3.3 (1.4-7.7) .008Tobacco smoke 1.25 3.5 (1.3-9.0) .002Traumatic extraction 2.56 13.1 (5.4-31.7) .001Constant �2.36 .001Adjusted R2 0.44

bservations (n) 160

bbreviations: CI, confidence interval; OR, odds ratio.

Halabí et al. Risk Factors for Alveolar Osteitis. J Oral Maxillofac

UENCXTRAASES

tis

Surg 2012.

ls

t

ttteoctafia

bitav

atwtsteig

F

HALABÍ ET AL 1043

authors believe this approach violates the principle ofstatistical independence, ie, each observation shouldnot be affected by any other observation.17 This mayead to an underestimation of the incidence,18 ashown by Nusari and Younis2 who reported an inci-

dence of alveolar osteitis of 4.8% per number of ex-tractions, although the reported incidence per subjectwas 6.4%.

Previous studies have lacked sufficient statisticalpower to determine the association between signifi-cant risk factors and the development of alveolarosteitis, or they have used univariate analysis for mul-tiple variables.9 Some differences might be explainedby the different methods used for data analysis. Forexample, using univariate analysis, Al-Belasy19 re-ported an association only in third molar extractions,whereas the authors did not. The authors adjusted the“third molar” variable with trauma and found no as-sociation. This might mean that third molar extrac-tions are more often associated with concomitanttrauma of tissues and, hence, with alveolar osteitis butare not associated per se with an increased incidenceof alveolar osteitis. This finding may also explain thelower incidence of alveolar osteitis when extraction isperformed by an experienced surgeon comparedwith a less experienced surgeon.20,21 It is very clearthat careful surgical techniques are of prime impor-tance in the quest to decrease the incidence of alve-olar osteitis after third molar extraction.

After searching the current literature to determinewhich risk factors to assess in the present study, theauthors agreed that tobacco use should be included.However, published evidence on this risk factor isscarce. Available evidence measures tobacco use asthe number of cigarettes smoked per day and findingshave indicated that a high incidence of alveolar oste-itis is evident in those who smoke at least 5 cigarettesdaily.19 Other studies, however, have not mentionedhis.6,8,22

Thus, the authors decided to score tobacco use assmoking at least 5 cigarettes per day. It is important tomeasure tobacco use on the day after extraction.When these data were analyzed by univariate analysisin the present study, no statistical association wasfound between tobacco smoking and alveolar osteitis,as stated by Heng et al.23 However, when analyzedogether with all variables in a multivariate model,here was a statistically significant increased risk. Fur-her research is needed to analyze the cumulativeffect of tobacco use on the development of alveolarsteitis. Moreover, it is especially difficult to verify theompliance of patients with postextraction instruc-ions (on personal care after surgery) on this issue,nd there is no certainty that this might influence thendings. A multivariate analysis by Parthasarathi et

l22 showed no significant association between to- d

acco smoking and the development of alveolar oste-tis, although an incidence of 12 events is not enougho validate these results, because logistic regressionnalysis requires an analysis of at least 10 events perariable.14 A possible relation between tobacco use

and the development of alveolar osteitis may be ex-plained by the formation of granulation tissue, whichis a key step in postextraction alveolar wound heal-ing, involving mainly certain cell groups such as neu-trophils, monocytes, and macrophages, whose activ-ity is inhibited directly by tobacco products.24 It hasalso been reported that abnormal phagocytosis ofpolymorphonuclear cells is associated with increasedcigarette consumption, producing a decreased localimmune and inflammatory response.25 Smokers havealso been shown to develop statistically less oralbleeding than nonsmoking patients,26 a finding thatmay be significant in clot formation after dental ex-traction. The self-report has proved to be a validinstrument for the detection of tobacco use (Spear-man correlation coefficient, 0.74) compared with theend-expired carbon monoxide (Coexp) analysis as thegold standard.27

Traumatic extraction was found to be the greatestrisk factor associated with the development of alveo-lar osteitis, which was consistent with the results ofother investigations.6,20,21 It has been reported thatremoval of impacted third molars carries a greater riskof developing alveolar osteitis compared with fullyerupted third molars,6 because the osteotomy andflap techniques used to extract impacted third molarsare highly aggressive to tissues. Other studies havereported an increased risk of alveolar osteitis in pa-tients who underwent third molar extractions withinexperienced surgeons, compared with those whounderwent surgery performed by experienced maxil-lofacial surgeons.20,21 As mentioned earlier, thelesser trauma from a more carefully performed sur-gical technique clearly influences the incidence ofalveolar osteitis.

The effect of infection also was examined. Previoussurgical site infection may arise from inoculation withmicroorganisms from the external environment intothe newly exposed socket after extraction.8 In thisdverse environment, tissue trauma would increasehe effect, because more damage to tissues means thatound healing is more likely to be delayed. Smoking

obacco, which directly affects the inflammatory re-ponse to this repair process, would further increasehe risk of alveolar osteitis. Thus, a detailed knowl-dge of microbial flora is important. Further researchs necessary to identify the specific putative patho-ens involved in the development of alveolar osteitis.Other questions emerged from the present study.

or example, a subgroup analysis showed an unequal

istribution of alveolar osteitis incidence in women.

1044 RISK FACTORS FOR ALVEOLAR OSTEITIS

One plausible hypothesis for this is the use of oralcontraceptives.28 Further investigations with all-fe-male cohorts should examine this hypothesis in moredetail by adjusting for oral contraceptive type anddose.

In conclusion, in this study, traumatic extraction,tobacco use, and previous surgical site infection wereidentified as risk factors for the development of alve-olar osteitis. Apart from traumatic extraction, the riskfactors identified cannot be manipulated by clinicians,but according to the present model, the authors couldpredict 8 of 10 patients who were at risk of develop-ing alveolar osteitis. Hence, it is very important thatthe surgeon is aware that a patient with a previousinfection, who is a smoker, or who undergoes a trau-matic extraction has a greater chance of developingalveolar osteitis and thus be able to take preventativemeasures. Reported measures for alleviating alveolarosteitis in high-risk patients include local treatmentwith tetracycline or preoperative and 7-day postoper-ative rinsing with 0.12% chlorhexidine.10

Acknowledgments

The authors acknowledge the collaboration of the patients andstaff of the Department of Dental Emergency “Jorge Sabat Gozalo”in Valdivia, Chile.

References1. Jaafar N, Nor GM: The prevalence of post-extraction complica-

tions in an outpatient dental clinic in Kuala Lumpur Malay-sia—A retrospective survey. Singapore Dent J 23:24, 2000

2. Nussair YM, Younis MH: Prevalence, clinical picture and riskfactors of dry socket in a Jordanian dental teaching centre. JContemp Dent Prac 8:53, 2007

3. Oginni FO: Dry socket: A prospective study of prevalent riskfactors in a Nigerian population. J Oral Maxillofac Surg 66:2290, 2007

4. Amaratunga NA, Senaratne CM: A clinical study of dry socketsin Sri Lanka. Br J Oral Maxillofac Surg 26:410, 1998

5. Hedner E, Vahlne A, Kahnberg KE, et al: Reactivated herpessimplex virus infection as a possible cause of dry socket aftertooth extraction. J Oral Maxillofac Surg 51:370, 1993

6. Chuang SK, Perrott DH, Susarla SM, et al: Risk factors forinflammatory complications following third molar surgery inadults. J Oral Maxillofac Surg 66:2213, 2008

7. Tjernberg A: Influence of oral hygiene measures on the devel-opment of alveolitis sicca dolorosa after surgical removal ofmandibular third molars. Int J Oral Surg 8:430, 1979

8. Baqain ZH, Karaky AA, Sawair F, et al: Frequency estimates andrisk factors for postoperative morbidity after third molar re-moval: A prospective cohort study. J Oral Maxillofac Surg66:2276, 2008

9. Kolokythas A, Olech E, Miloro M: Alveolar osteitis: A compre-

hensive review of concepts and controversies. Int J Dent 2010:249073, 2010

10. Hedström L, Sjögren P: Effect estimates and methodologicalquality of randomized controlled trials about prevention ofalveolar osteitis following tooth extraction: A systematic re-view. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 103:8,2007

11. Noroozi AR, Philbert RF: Modern concepts in understandingand management of the “dry socket” syndrome: Comprehen-sive review of the literature. Oral Surg Oral Med Oral PatholOral Radiol Endod 107:30, 2009

12. Greene JC, Vermillion JR: The simplified oral hygiene index.J Am Dent Assoc 68:7, 1964

13. Ministerio de Salud de Chile (MINSAL): Santiago, Región Met-ropolitana, Chile, Guía Clínica de Urgencia Odontológica Am-bulatoria. 2007

14. Courvoisier DS, Combescure C, Agoritsas T, et al: Performanceof logistic regression modeling: Beyond the number of eventsper variable, the role of data structure. J Clin Epidemiol, 64:993, 2011

15. Bortoluzzi MC, Manfro R, De Déa BE, et al: Incidence of drysocket, alveolar infection, and postoperative pain following theextraction of erupted teeth. J Contemp Dent Pract 11:E033,2010

16. Oginni FO, Fatusi OA, Alagbe AO: A clinical evaluation of drysocket in a Nigerian teaching hospital. J Oral Maxillofac Surg61:6, 2003

17. Sauerland S, Lefering R, Bayer-Sandow T, et al: Fingers, handsor patients? The concept of independent observations. J HandSurg [Br] 28:102, 2009

18. Park MS, Kim SJ, Chung CY, et al: Statistical consideration forbilateral cases in orthopaedic research. J Bone Joint Surg Am92:1732, 2010

19. Al-Belasy FA: The relationship of “shisha” (water pipe) smokingto postextraction dry socket. J Oral Maxillofac Surg 62:10, 2004

20. Jerjes W, El-Maaytah M, Swinson B, et al: Experience versuscomplication rate in third molar surgery. Head Face Med 2:14,2006

21. Jerjes W, Upile T, Nhembe F, et al: Experience in third molarsurgery: An update. Br Dent J 209:E1, 2010

22. Parthasarathi K, Smith A, Chandu A: Factors affecting incidenceof dry socket: A prospective community-based study. J OralMaxillofac Surg 69:1880, 2011

23. Heng CK, Badner VM, Clemens DL, et al: The relationship ofcigarette smoking to postoperative complications from dentalextractions among female inmates. Oral Surg Oral Med OralPathol Oral Radiol Endod 104:757, 2007

24. Pabst MJ, Pabst KM, Collier JA, et al: Inhibition of neutrophiland monocyte defensive functions by nicotine. J Periodontol66:1047, 1995

25. MacFarlane GD, Herzberg MC, Wolff LF, et al: Refractory peri-odontitis associated with abnormal polymorphonuclear leuko-cyte phagocytosis and cigarette smoking. J Periodontol 63:908,1992

26. Calsina G, Ramón JM, Echeverría JJ: Effects of smoking onperiodontal tissues. J Clin Periodontol 29:771, 2002

27. Myles PS, Iacono GA, Hunt JO, et al: Risk of respiratory com-plications and wound infection in patients undergoing ambu-latory surgery: Smokers versus nonsmokers. Anesthesiology97:842, 2002

28. Garcia AG, Grana PM, Sampedro FG, et al: Does oral contra-

ceptive use affect the incidence of complications after extrac-tion of a mandibular third molar? Br Dent J 194:453, 2003