Int. J. Oral Maxillofac. Surg. 2002; 31: 309–317doi:10.1054/ijom.2002.0263, available online at http://www.idealibrary.com on

Evidence-Based MedicineOral Surgery

Contemporary views on drysocket (alveolar osteitis):a clinical appraisal ofstandardization,aetiopathogenesis andmanagement: a critical reviewI. R. Blum: Contemporary views on dry socket (alveolar osteitis): a clinical appraisalof standardization, aetiopathogenesis and management: a critical review. Int. J. OralMaxillofac. Surg. 2002; 31: 309–317. � 2002 International Association of Oral andMaxillofacial Surgeons. Published by Elsevier Science Ltd. All rights reserved.

Abstract. The objective of this article is to harmonize descriptive definitions for thecondition known as alveolar osteitis and to critically review and discuss theaetiology and pathogenesis of alveolar osteitis. In addition, the need for theidentification and elimination of risk factors as well as the preventive andsymptomatic management of the condition are discussed. The aim of this criticalreview is to provide a better basis for clinical management of the condition. Ameta-analysis of data was not done.

I. R. BlumDivision of Oral and Maxillofacial Surgery,Department of Oral and MaxillofacialSciences, University Dental Hospital ofManchester, Higher Cambridge Street,Manchester M15 6FH, UK

Key words: alveolar osteitis; fibrinolysis;prophylactic management; symptomaticmanagement.

Accepted for publication 12 February 2002

Introduction

One of the most common postoperativecomplications following the extractionof permanent teeth is a condition knownas dry socket. This term has been usedin the literature since 1896, when it wasfirst described by C22. Sincethen, several other terms have beenused in referring to this condition, suchas alveolar osteitis (AO), localized ostei-tis, postoperative alveolitis, alveolalgia,alveolitis sicca dolorosa, septic socket,necrotic socket, localized osteomyelitis,and fibrinolytic alveolitis.

So far, authors do not agree on termi-nology for this condition. In his seminalarticles, Birn labelled the complication

0901-5027/02/030309+09 $35.00/0 � 2002 Internatio

‘fibrinolytic alveolitis’5,7,11 which isprobably the most accurate of all theterms, but is also the least used in theliterature. In most cases, the moregeneric lay term ‘dry socket’ tends to beused. In this article, the condition willbe referred to as alveolar osteitis, AO.

Search strategy and literatureselection criteria

A computerized literature search usingMEDLINE was conducted searching forarticles published from 1968–2001. Meshphrases used in the search were: drysocket, alveolar osteitis, localized ostei-tis, fibrinolytic alveolitis, prevention and

nal Association of Oral and Maxillofacial Surgeons. Pu

dry socket, management and dry socket.The search was completed by manualsearches of selected internationallyreviewed journals. Only papers inEnglish and those which stated the diag-nostic criteria were reviewed. For themanagement section randomized con-trolled and controlled trials wereidentified to provide the most powerfulevidence followed in decreasing order ofstrength by cohort studies, case-controlstudies, surveys and case-series.

Standardization of AO

Unfortunately, the literature is repletewith varying descriptive definitions for

blished by Elsevier Science Ltd. All rights reserved.

310 Blum

AO, usually owing to an inconsistency indiagnostic criteria (Table 1). There is alack of absolute and objective clinicalcriteria and varying study designs as wellas efficacy variables between studies,conflicting data (including intermingleddata from ‘cases’, ‘teeth’, and ‘surgi-cal sites’), anecdotal reports, poorlydesigned studies, statistical biases or lackof analysis, and individual opinionscamouflaged as scientific evidence makea scientifically sound comparison verydifficult. Hence, a systematic review wasnot done as there is insufficient evidenceavailable. The variety of subjective diag-nostic definitions for AO appearedto be so great that this author wastempted to come up with a descriptivedefinition that could be used universally

as a standardized definition for AO:postoperative pain in and around theextraction site, which increases in severityat any time between 1 and 3 days after theextraction accompanied by a partially ortotally disintegrated blood clot within thealveolar socket with or without halitosis.It is necessary to exclude any othercause of pain on the same side of theface. Occasionally, patients may alsocomplain of a very unpleasant taste.

The denuded alveolar bone may bepainful and tender. Some patients mayalso complain of intense continuous painirradiating from the empty socket, nor-mally to the ipsilateral ear, temporalregion or the eye. Regional lympha-denopathy is also noted occasionally.Trismus is a rare occurrence, but in cases

of mandibular third molar extractions itis sometimes seen and is probably due tolengthy and traumatic surgery. True AO,in which premature partial or total lossof a formed extraction socket coagulumoccurs, must be distinguished from con-ditions in which pre-existing alveolarbone hypovascularity, such as general-ized vascular or haematological dis-orders, radiotherapy-induced osteo-necrosis, osteopetrosis, Paget’s diseaseand cemento-osseous dysplasia73 pre-vent initial formation of a coagulum.

AO remains a common postoperativeproblem resulting in pain, lost days atwork, loss of productivity, and returnsurgical practice/hospital visits. This isalso costly to the surgeon, as 45% ofpatients who develop AO typicallyrequire at least four additional postop-erative visits in the process of managingthis condition47.

Incidence

The incidence of AO has been reportedas 3–4% following routine dental extrac-tions51 and ranges from 1% to 45%after the removal of mandibular thirdmolars3,26,54,58. This great variability inthe reported incidence of AO is largelydue to differences in diagnostic criteriaand in the methods of assessment; inintermingled and conflicting data fromnon-impacted, partially impacted andfully erupted mandibular third molarsextractions, in intraoperative and post-operative management of extractionsites; in patient populations with respectto age or to surgical techniques or surgi-cal skill. Also, there is a large variationof pain thresholds within the population.Studies claiming 1% incidence lackclinical credibility, whereas those withunusually high incidence rates (>30%)suggest that other, unaddressed vari-ables were introduced or the sample sizewas insufficient.

The better controlled studies havereported the incidence as 25–30% afterthe removal of impacted mandibularthird molars26,37 and this review con-cludes that AO occurs approximately 10times more frequently following theremoval of these teeth than from allother locations.

Onset and duration

Early and recent studies have reportedthat AO onsets 1–3 days after toothextraction26,45,51 and within a weekbetween 95% and 100% of all cases ofAO have been registered24. It is highly

Table 1. The variety of definitions used in the literature for the clinical assessment of alveolarosteitis

Author(s) and Year Definition

A et al.1 (1998) The presence of a disintegrated blood clot, and/orincreased pain in the socket region, and/or foul odour,and/or exudate or pus in the socket

B & L4 (1990) Evidence of a denuded socket with or without necroticdebris or foetid breath

B11 (1972) Partial or complete loss of the blood clot, exaggerated painradiating to the ear and temporal region, and a putridodour

B12 (2000) Complain of pain in the extraction site and the presence ofexposed bone or necrotic debris

C22 (1896) Severe, neuralgiform, irradiating pain and partial or totaldisintegration of the blood clot in the socket have to bepresent simultaneously

D et al.23 (1981) Loss of an adequate clot and development of delayed pain,2 to 5 days after surgery, that was suffice to require activemedical intervention

F & O26 (1990) Absence of a demonstrable clot and symptomatic pain inor around the surgical site 36 h after surgery that wassuffice to require active medical intervention

H et al.46 (1998) Loss of blood clot and/or necrosis of blood clot andpersistent or increasing postoperative pain after thesurgery, with throbbing pain at the surgical site that is notrelieved with mild analgesics

L et al.34 (1972) Evidence of breakdown of clot together with thecharacteristic foul odour

L35 (1991) Persistent or increasing postoperative pain beginning afterthe second day, which is associated with necrotic tissue inthe socket, exposed bone, or loss of the clot on clinicalexamination

M et al.41 (1987) Pain from the extraction site and empty or necroticmaterial containing socket

R et al.50 (1992) The simultaneous presence of a severe irradiating painoriginating from the empty socket and the disintegration(partial or total) of the socket coagulum

R & M51 (1979) A painful socket which is increasing in severity 24 h afterthe extraction

S & P61 (1987) Return of patient 2 or more days postoperativelycomplaining of pain in the extraction area and thepresence of a denuded socket on clinical examination

S & B64 (1977) Severe pain, foul, greyish exudate, and necrotic odour anddebris at the extraction site

T67 (1979) Disintegrated blood clot in combination with pain that isnot adequately relieved by analgesics

V et al.71 (1974) Complete or partial loss of the blood clot with denudedbone in the alveolus and severe irradiating pain

Contemporary views on dry socket (alveolar osteitis) 311

unlikely for AO to occur before the firstpostoperative day, because the bloodclot contains anti-plasmin that must beconsumed by plasmin before clot disin-tegration can take place. The duration ofAO varies to some degree, depending onthe severity of the disease, but it usuallyranges from 5–10 days.

Aetiology

Myriad aetiological and precipitatingfactors for AO have been suggested inthe literature. Although AO is generallybelieved to be of multifactorial origin,the following have been implicated mostcommonly as aetiological, aggravatingand precipitating factors:

1. Oral micro-organisms and AO

The role of bacteria in AO has long beenpostulated39,51. This concept was sup-ported by various reports of theincreased frequency of AO in patientswith poor oral hygiene48, pre-existinglocal infection such as pericoronitis andadvanced periodontal disease55. Acausative relationship with bacteria hasfurther been strengthened by the reducedincidence of AO in conjunction withantibacterial measures19,42,51,62.

There have been numerous attemptsto isolate a specific causative organism.The possible association of Actinomycesviscosus and Streptococcus mutans in AOwas highlighted by R et al.54

where they demonstrated delayed heal-ing of the extraction socket following theinoculation of these organisms in animalmodels.

N et al.44 showed a possiblesignificance of anaerobic organisms(which are also the predominant organ-isms in pericoronitis) in relation to theaetiology of AO.

N et al.44 observed highplasmin-like fibrinolytic activities fromcultures of the anaerobe Treponemadenticola, which is also known to be aputative micro-organism in the develop-ment of periodontal disease. In addition,AO virtually never occurs during child-hood, a period when this organism hasnot yet colonized the mouth.

As bacteria increase in number in AO,and because certain species constantlysecrete pyrogens at the basal level, ithas been postulated that bacterialpyrogens are indirect activators offibrinolysis in vivo17. C17

studied the efficacy of bacterial pyrogensfor treating thromboembolic diseasewhere pyrogens injected intravenously

produced a sustained increase infibrinolysis.

2. Difficulty and trauma during surgeryand AO

Most authors agree that trauma anddifficulty of surgery play an importantrole in the development of AO2,11,13,21.Surgical extractions that involve thereflection of a flap and sectioning of thetooth with some degree of bone removalhave also been reported to be more likelyto cause AO37. One interesting studyindicates that less experienced surgeonscaused a significantly higher incidenceof complications after the removal ofimpacted third molars; the mostcommon complication being AO60.

Excessive trauma has been known toresult in delayed wound healing11. Thishas been attributed to the compressionof the bone lining the socket, whichimpairs its vascular penetration.Alternatively, excessive trauma mayresult in thrombosis in the underly-ing vessels. Several authors have associ-ated trauma with a reduction in tissueresistance and consequent woundinfection36,70.

B11 proposed that trauma duringextraction damages the alveolar bonecells, causing inflammation of thealveolar bone marrow and the subse-quent release of direct tissue activatorsinto the alveolus, where they mayprecipitate fibrinolytic activity, thusplaying a major role in the pathogenesisof AO.

3. Roots or bone fragments remaining inthe wound and AO

B11 suggested this complication as apossible cause of AO. S59 hasshown that such fragments are com-monly present after normal extraction orsurgical removal of teeth, and that smallbone and tooth remnants do not neces-sarily cause complications during heal-ing as they are often externalized by theoral epithelium. The results were derivedafter histological examination of healingextraction wounds in monkeys.

Despite a lack of scientific evidencefor these remnants to be the causativefactor for AO, it seems logical that frag-ment and debris remnants could lead todisturbed wound healing, and therebypossibly contribute to the developmentof AO.

4. Excessive irrigation or curettage ofthe alveolus after extraction and AO

It has been postulated that energeticrepeated irrigation of the alveolus might

interfere with clot formation and giverise to infection, and that violent curet-tage might injure the alveolar bone11.However, scientifically sound investi-gations confirming these contributionsin the development of AO are lacking.Furthermore, since energetic excessiveirrigation is not easily measurable, it isdifficult for it to be assessed.

5. Physical dislodgement of the clotand AO

It has not been substantiated that thephysical dislodgement of the blood cloteither by manipulation or negative press-ure, such as sucking on a straw, wouldbe a major contributory factor to AO.

6. Local blood perfusion, anaesthesiaand AO

Three aspects of blood supply have beenconfused in the literature; the vasculararchitecture, the circulation, and theintegrity of the blood clot. K33

associated poor local blood supply withan increased incidence of AO in man-dibular molar extractions. The presenceof thick cortical bone, it was suggested,resulted in the poor perfusion of bloodand it was suggested that minor perfora-tions into the alveolar marrow cavitywould allow blood vessels to grow inmore easily. This was disputed by B11

who demonstrated that the mandibularmolar region is one of the most richlyvascularized regions of the mandible, itsblood supply being far better than thatof the incisal region.

The vasoconstrictors in local anaes-thetic solutions have been suggested asalternative factors in the pathogenesisof AO41. On the other hand, AO alsofollows tooth extractions carried outunder general anaesthesia where novasoconstrictor was used.

In addition, patients who requirerepeated injections of local anaestheticsolution may have a reduced painthreshold, which may account for com-plaints of pain originating from theextraction socket.

Some investigators claimed anincrease in the incidence of AO whenperiodontal intraligamental (PDL) injec-tions were used rather than block orinfiltration injections41. These findingshave been attributed to the spread ofbacteria, especially with multiple injec-tions to the affected site. This was how-ever, disputed by T et al.69 whohave shown that PDL anaesthesia did

312 Blum

not result in a higher frequency of AOthan when block anaesthesia was used.

7. Oral contraceptives and AO

Contrary to studies conducted prior to1960, studies from the 1970s and latershowed a significantly higher incidenceof AO occurring in females20,58,64. Thiswas attributed to the increased use oforal contraceptives from the 1960sonwards as these were shown to havea definite positive correlation to theincidence of AO65.

It has been proposed that oestrogens,like pyrogens and certain drugs,will activate the fibrinolytic systemindirectly, and thus are believed to con-tribute to the occurrence of AO byincreasing lysis of the blood clot74.C et al.18 concluded that theprobability of AO increases withincreased oestrogen dose in the oral con-traceptive and that fibrinolytic activityappears to be lowest on days 23 through28 of the menstrual cycle. Others studiedthe effect of oral contraceptives on thecoagulation and fibrinolytic systemand demonstrated an increase in thenumber of many factors such as factorII, VII, VIII, X, and in particularplasminogen74.

Interestingly, a recent prospective ran-domized controlled study reported thatfemales have a higher incidence of AOcompared to males regardless of whetherthey are on oral contraceptives19.However, the conclusions drawn fromthis study, owing to the small sample sizeof male participants, are tentative.

In the case of females not using oralcontraceptives, there is little publishedevidence on the effects of the variouspoints in the menstrual cycle on theincidence of AO.

8. Smoking and AO

S & B66 have reported thatamong patients with a total of 400surgically removed mandibular thirdmolars, those who smoked a half-packof cigarettes per day had a four- tofive-fold increase in AO (12% vs 2.6%)compared to non-smoking patients. Theincidence of AO increased to more than20% among patients smoking more thana pack per day, and to 40% amongpatients who smoked on the day ofsurgery, or on the first postoperativeday.

This phenomenon could be due tothe introduction of a foreign substance

that could act as a contaminant in the

surgical site, and/or the suction appliedto the cigarette which might dislodge theclot from the socket and interrupt heal-ing. No references exist in the literaturecorrelating the effects of heat from burn-ing tobacco, contaminants in the smoke,or the systemic effects of the ingredientsin cigarettes with AO.

Pathogenesis

Clinical and laboratory studies haveshown the significance of locallyincreased fibrinolytic activity in thepathogenesis of AO5,7,8,11.

B11 claimed that partial or com-plete lysis and destruction of the bloodclot was caused by tissue kinases liber-ated during inflammation by a direct orindirect activation of plasminogen in theblood. When direct tissue activators arereleased after trauma to the alveolarbone cells, plasminogen (which is laiddown in the fibrin network as it isformed) is converted to plasmin, result-ing in the break up of the clot by disin-tegrating the fibrin. This conversion isaccomplished in the presence of tissue orplasma pro-activators and activators.These activators have been recently clas-sified as direct (physiologic) and indirect(nonphysiologic) and further subclassi-fied according to their origin as intrinsicand extrinsic activators73. Intrinsicactivators originate from plasma compo-nents whereas extrinsic activators origi-nate outside of the plasma/blood per se.Direct intrinsic activators include FactorXII (Hageman factor)-dependent acti-vator and urokinase, which are mediatedby leukocytes. Direct extrinsic activatorsinclude tissue plasminogen activatorsand endothelial plasminogen activators.Tissue plasminogen activators are foundin most tissue types, including alveolarbone6. Indirect activators includesubstances such as streptokinase andstaphylokinase, which are produced bybacteria and bind to plasminogen toform an activator complex that thencleaves other plasminogen molecules toplasmin. This strengthens the theory ofthe involvement of micro-organisms inthe development of AO.

B9 attributed the cause of pain tothe presence and formation of kininlocally in the socket. It has been shownthat kinins activate the primary afferentnerves, which may have already beenpresensitized by other inflammatorymediators and algogenic substances, andin concentrations as low as 1 ng/ml they

9,46

are able to produce intense pain .

Plasmin is also involved in the conver-sion of kallikreins to kinins in thealveolar bone marrow9. Thus, the pres-ence of plasmin may give a possibleexplanation for the two most character-istic features of AO, namely neuralgicpain and disintegrated blood clot. Thisis in accordance with B’s observa-tions11 who noted an increased fibrino-lytic activity in extraction sockets withAO when compared to normal healingsockets. He stated that: ‘fibrinolyticalveolitis resulted when fibrinolysis oranother proteolytic activity in andaround the alveolus was capable ofdestroying the blood clot’.

B & M-J8 have investi-gated the role of alveolar bone inincreasing the local fibrinolytic activity.They concluded that the surroundingbone of the alveolus contains, amongother components, stable tissue activa-tors that may explain the local fibrino-lytic activity in AO, and these stableactivators are linked up with the osteo-blasts of the endosteum. These resultsagreed with other findings that demon-strated a high fibrinolytic activity in theendosteal layer in rats40.

Prophylactic management

In an era of evidence-based care, fewareas of clinical controversy pose assubstantial a dilemma to clinicians, asthe topic of the alleged factors that aretargets for the various preventive regi-ments, and the topic of what prophy-lactic medicaments and materials, if any,should be placed in an alveolar socketfollowing exodontia.

References in the literature correlat-ing to the prevention of AO can bedivided into non-pharmacological andpharmacological preventive measures.Effective non-pharmacological preven-tive measures include a comprehensivehistory of the patient with identification,and if possible, elimination of risk fac-tors associated with an increased riskto develop AO. These risk factors aresummarized in Table 2. Moreover,besides the possible elimination of riskfactors, it is imperative for active non-pharmacological preventive measuresto be implemented. These preventivemeasures are summarized in Table 3.

Because AO is probably the mostcommon local post-extraction compli-cation that exists, a successful methodof pharmacological prevention has longbeen sought. The literature reports avariety of materials and techniques that

have been and still are being investigated

Contemporary views on dry socket (alveolar osteitis) 313

for their success. These pharmacologicalprophylactic interventions are related toone or more of the following groups:

1. Antibacterial agents2. Antiseptic agents and lavage3. Antifibrinolytic agents4. Steroid anti-inflammatory agents5. Obtundent dressings6. Clot support agents

1. Antibacterial agents

Prophylactic antibacterials, either givensystemically or used locally, are consid-ered to reduce the incidence of AO.Systemic antibacterials reported to beeffective in the prevention of AO includethe penicillins32,34, clindamycin15,34,erythromycin15, and metronidazole3,51.

In addition, it has been reported that the

preoperative administration of antibac-terial agents is more effective in reducingthe incidence of AO than when givenpostoperatively34,53. However, of allsystemic antibacterials referred to in theliterature for the prevention of AO, theonly that stood trial successfully inrandomized double-blind studies wasmetronidazole3,51. Owing to its narrowantibacterial spectrum (anaerobicidal),metronidazole is associated with fewerand more infrequent side-effects than thehigh resistance developing penicillinsand erythromycin, and the pseudomem-branous colitis inducing clindamycin.Caution should however be taken withmetronidazole in patients taking war-farin, disulfiram, phenytoin and possiblyantihypertensives because of possibledrug interactions. Concurrent alcoholshould be avoided.

Nowadays, the routine use of systemicpre- and/or postoperative antibacterialsgiven prophylactically is highly disputedand by many considered to be contro-versial because of the development ofresistant bacterial strains and possiblesystemic side effects, such as hypersensi-tivity and unnecessary destruction ofhost commensals.

Myriad studies have been carried outto evaluate the effectiveness of topical(intra-alveolar) medicaments in prevent-ing AO including various types of anti-bacterials used alone or in combinationin varying formulations and dosages.However, very few studies are in agree-ment. The cited incidence in somestudies is higher with antibiotics than inother studies without antibiotic use. Insome cases, the antibacterial or basematerial used to carry the antibiotic hascaused more significant complicationsthan the AO2.

C & D19 investigatedin a double-blind study the effectivenessof topical clindamycin and they reporteda significantly reduced incidence ofAO in mandibular third molar socketsfollowing light socket irrigation withBetadine� and the topical application ofclindamycin in Gelfoam. They attributedtheir findings to the effectiveness of clin-damycin but the irrigant used by themprior to wound closure is an iodophorewith its own antibacterial properties.Furthermore, the subjects of their studyalso received multiple oral doses of sys-temic antibiotics postoperatively; thusmaking it impossible to attribute theirfindings to either of the antibacterialsalone used in their study.

T & S68 examined in a

double-blind crossover study involving

86 patients with 172 bony impactedmandibular third molars, the effect ofthe topical placement of clindamycinsaturated gel sponge inserted into ran-domized unilateral extraction sitesimmediately following surgery. Eachpatient received a placebo in the contra-lateral site and served as his or her owncontrol. The authors reported 7 cases ofAO in the control group and none in thetest group. Based on the significant dif-ference in AO rates at a significancelevel of P<0.5 they concluded that theaetiology of AO is related to an infec-tion with anaerobic bacteria and thatclindamycin applied topically can beeffective in the prevention of AO.

Many studies with topical tetracyclinepowder, aqueous suspensions of tetra-cycline, tetracycline on gauze drain ortetracycline-soaked Gelfoam spongeshave been reported to be effectivein significantly reducing the incidenceof AO1,23,61,62. The latter mixture isthought to provide a firm clot inaddition to preventing infection. How-ever, side-effects including foreign bodygiant-cell reactions have been reportedin association with topically appliedtetracycline43,75.

The topical application of apetroleum-based combination of tetra-cycline and hydrocortisone has also beenreported in several studies to signifi-cantly reduce the incidence of AO26,56.L et al.38 reported however, theoccurrence of the chronic problem ofmyospherulosis in extraction sites thatreceived this combination and they sug-gested that this may arise as a resultof the action of the lipid substances ofthe petrolatum carrier vehicle on theextravasated erythrocytes.

Nevertheless, no adverse reactions tothe topical application of aqueous tetra-cycline suspensions or to impregnatedtetracycline gauze drains in the sockethave been described, and besides theirclaimed effectiveness in decreasing theincidence of AO, they are also consid-ered to be an economical preventivemodality73.

2. Antiseptic agents and lavage

Chlorhexidine (CHX) is a bisdiguanideantiseptic with antimicrobial properties.The use of CHX as both a mouthrinseand as a preoperative irrigant of thegingival crevice has been shown to sig-nificantly reduce the quantity of oralmicrobial populations72. Several studieshave reported that the pre- or periopera-

Table 2. Risk factors associated with true AO

� Previous experience of AO� Deeply impacted mandibular third

molar (risk factor is directlyproportional to increasing severity ofimpaction)

� Poor oral hygiene of patient� Active or recent history of acute

ulcerative gingivitis or pericoronitisassociated with the tooth to be extracted

� Smoking (especially >20 cigarettes perday)

� Use of oral contraceptives� Immunocompromised individuals

Table 3. A summary of non-pharmacologicalmeasures to prevent AO

� Use of good quality current preoperativeradiographs

� Careful planning of the surgery� Use of good surgical principles� Extractions should be performed with

minimum amount of trauma andmaximum amount of care

� Confirm presence of blood clotsubsequent to extraction (if absent,scrape alveolar walls gently)

� Wherever possible preoperative oralhygiene measures to reduce plaque levelsto a minimum should be instituted

� Encourage the patient (again) to stop orlimit smoking in the immediatepostoperative period

� Advise patient to avoid vigorous mouthrinsing for the first 24 h post extractionand to use gentle toothbrushing in theimmediate postoperative period

� For patients taking oral contraceptivesextractions should ideally be performedduring days 23 through 28 of themenstrual cycle

� Comprehensive pre- and postoperativeverbal instructions should besupplemented with written advice toensure maximum compliance

tive use of CHX mouthrinse significantly

314 Blum

reduces the incidence of AO after thesurgical removal of mandibular thirdmolars4,29,35,67. R & S49

noted nearly a 50% reduction in theincidence of AO in patients who pre-rinsed for 30 s with a 0.12% CHXsolution.

F et al.25 examined in a random-ized double-blind placebo-controlledstudy involving 70 patients with 140uncomplicated non-infected thirdmolars, the effect of the topical insertionof an intra-alveolar chlorhexidine gluco-nate solution-soaked Gelfoam into anextraction site and compared it to anintra-alveolar saline-soaked Gelfoaminserted on the contralateral side. Theyreported that the use of the former sig-nificantly reduced postoperative discom-fort, but the incidence of AO was notspecifically documented. They alsoreported that the 0.1% chlorhexidinesolution did not significantly reducepostoperative discomfort whereas theuse of the higher 0.2% concentration wassignificantly efficacious in reducing thesesymptoms. The authors acknowledgedthat Gelfoam exhibits a degree ofhydrophobicity that precludes efficientabsorption of chlorhexidine before intra-alveolar placement. Also, pre-shapedGelfoam morphology does not allow itsplacement to the full depth of the socket.No reference was found in the literaturecorrelating the local applications of thebiodegradable chlorhexidine Periochip�

nor that of chlorhexidine Corsodyl� gelwith AO.

In a crossover study31 the antisepticagent, 9-aminoacridine, saturated inGelfoam was placed in mandibular thirdmolar extraction sites, and was com-pared with the use of Gelfoam aloneplaced in the contralateral mandibularthird molar extraction sites. The authorsconcluded that 9-aminoacridine wasineffective in reducing the incidenceof AO.

H & N28 studied theprophylactic effectiveness of antisepticdressings by suturing a gauze spongesaturated with Whitehead’s varnish (acombination of iodoform, balsamtolutan, and Styrax liquid in a baseliquid) over mandibular third molarextraction sites. The authors claimed torecord a significant decrease in the inci-dence of postoperative pain, haemor-rhage and swelling when compared to acontrol group, but the incidence ofspecifically diagnosed AO was notaddressed.

Following a lavage study63 it was

reported that the incidence of AO after

the removal of mandibular third molarswas significantly reduced from 10.9%using 25 ml normal saline solution forlavage to 5.9% with the use of 175 mllavage. In another lavage study14, nosignificant differences were found in theincidence of AO following the removalof mandibular third molars betweenvolumes of 175 ml and 350 ml of normalsaline solution, but both these volumeswere more effective than a volume of25 ml. The reason for this may be thatsufficient lavage mechanically removesmore of the root remnants and/orbone fragments (and other debris)possibly still left in the extraction socketand which might contribute to thedevelopment of AO.

3. Antifibrinolytic agents

Earlier investigations10,57 into the fibri-nolytic nature of AO indicated that thetopical use of para-hydroxybenzoic acid(PHBA), in extraction wounds signifi-cantly decreased the incidence of AO ina dose-dependent fashion. However, asPHBA is available on the market as acomponent of Apernyl� (Bayer AG,Germany)—an alveolar cone with a for-mulation of 32 mg acetylsalicylic acid,3 mg propyl ester of PHBA and 20 mgunknown tablet mass, it is not possibleto attribute the reported findings toPHBA alone or perhaps to the anti-inflammatory properties of acetyl-salicylic acid. In addition, PHBA hasalso been reported to have some antibac-terial properties which may also havecontributed to the reported findings73.Subsequent histological studies16 how-ever, showed that acetylsalicylic acid incontact with bone causes a local irritat-ing effect accompanied by seriousinflammation of the extraction socket,possibly resulting in AO.

The antifibrinolytic agent Tranexamicacid (TEA) has been reputed to preventAO when applied topically in the extrac-tion socket following exodontia, butcontrolled investigations with specialreference to impacted mandibular thirdmolar extraction wounds have notshown a significant reduction in theincidence of AO when compared to aplacebo group27.

Given the lack of a scientifically con-firmed advantage, and many possibleproblems, there seems to be no rationalefor the use of these agents.

4. Steroid anti-inflammatory agents

Only one reference was found in the

literature regarding the individual use of

topical corticosteroids in the preventionof AO. Even though the corticosteroidhas been reported to decrease immediatepost-operative complications, it failedto reduce the occurrence of AO afterextraction36. The topical application of ahydrocortisone and oxytetracycline mix-ture however, has been shown to signifi-cantly decrease the incidence of AO afterthe removal of impacted mandibularthird molars26,56. Unfortunately, thecontribution of the antibiotic cannotbe separated from that caused by thesteroid.

Given the lack of scientific evidencesubstantiating any benefit to this regi-men its use as a preventive measure forAO is inappropriate.

5. Obtundent dressings

A recent crossover study12 on the pre-vention of AO following the bilateralremoval of 200 mandibular molarsclaimed a significant decrease in the inci-dence of AO, following the immediateplacement of an eugenol containingdressing into randomly selected unilat-eral extraction sockets. The contralateralsockets were not packed and served asthe patients own controls. However, theirritant local effect of eugenol and thedelay in wound healing due to electiveprophylactic packing is well documentedin the literature2,19,57.

6. Clot support agents

In the 1980s, a biodegradeable esterpolymer, polylactic acid (PLA) waswidely promoted as the ultimate solutionfor preventing AO, and it is still avail-able today under the brand name ofDriLac (Osmed, Inc, Costa Mesa, CA,USA). It was suggested that PLA wouldprovide a biological stable support forthe blood clot and for the future granu-lation and osteoid tissue. A study byB & 13 in 1986 reportedan incidence of AO of 2.2% with PLA,placed in mandibular third molar extrac-tion sites, as compared with 18.1% inci-dence without the use of PLA. Thiswas however followed in 1990 by anarticle by M & B43 that high-lighted 18 cases of complications withtetracycline-treated PLA, and in 1995,H & G30 reported a higherincidence of AO when PLA was used inthe control group (23.6% with PLA,13.6% without). The latter prospectivestudy suggests that the use of PLA might

actually increase the incidence of AO.

Contemporary views on dry socket (alveolar osteitis) 315

Given the lack of scientific evidence,there seems to be no benefit in its use.

Symptomatic management

Although numerous authors21,42,52 oftenrefer to the ‘treatment’ of AO, thisappears to be rather misleading, as thecondition cannot be treated as long asthe underlying aetiology has not beenfirmly established. Meanwhile, AO canonly be managed and as management isdirected primarily towards the promptrelief of the patient’s pain during thehealing stages, it takes place primarily bypalliative means.

References in the literature relating tothe management of AO can be dividedinto non-dressing and dressing interven-tions. The non-dressing interventions aresummarized in Table 4.

Several authors42,62,68 advocate, withor without prior non-dressing interven-tions, the use of intra-alveolar medicateddressings. The active components of thedressings reported in the literature formanaging the condition can be broadlyclassified as follows:

1. Antibacterial dressings2. Obtundent dressings3. Topical anaesthetic dressings, and4. Combinations of 1–3.

The use of dressings is empirical andtheir reported effectiveness in reducingpatient discomfort is largely based oncircumstantial personal clinical exper-ience and ample anecdotal reports.Although the placement of dressings iscontroversial in the literature and con-crete evidence regarding their placementis lacking, it has been suggested73 that in

cases of diagnosed AO, dressings might

be an effective malady that could beused as an adjunct to non-dressing inter-ventions. Arguments in the literaturesupporting the use of intra-alveolardressings include the achievement ofgreater local concentration of the sub-stance(s) than can normally be expectedfrom systemic administration, the mini-mizing of possible side effects and sensi-tization that may accompany systemicadministration, the localized obtundenteffect, and the closing of the socket sothat food debris can be kept out. Theexact incidence of complications second-ary to dressings placed in extractionsockets is unknown.

To date, no scientific studies havebeen carried out that specifically investi-gate the incidence of potential sideeffects and tissue damage arising fromthe placement of intra-alveolar dress-ings. Although a theoretical potential forthe development of resistant bacterialstrains with intra-alveolar antibiotic usehas been reported2, there have not beenany reliable data in the literature tosubstantiate this theoretical complica-tion. However, case reports regardingthe occurrence of other local complica-tions have been described in the litera-ture38,43,56,75 and it is generallyacknowledged that dressings delay thehealing of the extraction socket.

Discussion

All the clinical and histological evidencesupports that AO results from disturbedhealing of the extraction wound. Theintroduction of a standardized descrip-tive criterion for AO may providea sound basis for more objectiveand reliable comparisons in futureinvestigations related to AO.

Although the full aetiology of AO hasyet to be firmly established and despitethe plethora of theories available for itsaetiopathogenesis, substantial evidencesuggests that it is most particularlyrelated to a complex interaction betweenexcessive localized trauma, bacterialinvasion and their association to plasminand subsequently, the fibrinolytic sys-tem. Another factor that should meritattention and has not been investigatedearlier is the possibility of a geneticfactor. If there is a genetic predispositionto AO, it is likely to arise from poly-morphisms of one or more genes. Thispossibility could be investigated byundertaking large association studies ortransdisequilibrium testing. The possi-bility of a genetic association with AO is

worthy of further investigation.

Prevention of AO entails reducing thenumber of possible risk factors, meticu-lous attention to procedural details andsurgical skills. Despite a plethora of pub-lished articles, relatively little reliabledata are available for formulating ascientifically sound philosophy regardingthe pharmacological prevention andsymptomatic management of AO. Todate, no single method has gained uni-versal success or acceptance, although alarge number of practitioners continueto use ‘their method’, probably becauseit was passed from one generation toanother, often without controlled studiesto support its use.

Dressings should not be placed intoextraction sockets for merely prophylac-tic reasons as the possible side effectsand unnecessary additional costs to thepatient contraindicate this. Based on thefirst dictum of medicine as stated byHippocrates (421 B.C.): ‘At first do noharm’, it seems prudent to limit thepharmacological preventive interven-tions to measures which are supportedby sufficient evidence to be effective, andequally, show a minimum of side-effects. Besides ample surgical lavage,the reported prophylactic effectiveness,economy and lack of adverse side effectsof chlorhexidine solution justify its useas a preoperative irrigant or mouthrinsein the prevention of AO29,35,49.

The review of the literature with refer-ence to dressings provides a cautionarynote, that even though severe reactionsfrom the use of antibacterials, gelfoam,or other preparations placed in socketsare uncommon, all are accompanied byrisks for reactions, complications, anddelayed healing. Should adverse reac-tions develop in a patient, the prac-titioner may find medicolegal defenceof the use of the material difficult, basedon the documented problems reported inthe literature, rare as they might be. Todate, insufficient scientific evidence existsfor the amelioration of pain followingthe application of dressings.

Despite many years of research therehas been little progress over the years toaddress this very painful condition forpatients. However, further investigationsand well controlled studies are necessaryto draw firm conclusions which can leadto increased clarity regarding the mostbeneficial management of the patientpresenting with AO.

Acknowledgment. The author is mostgrateful to Professor P. Sloan forhis helpful comments during thepreparation of the article.

Table 4. Summary of non-dressing interven-tions to manage AO

� Remove any sutures to allow adequateexposure of the extraction site. As thesocket may be exquisitely tender localanaesthesia may be required

� Irrigate the socket gently with warmsterile isotonic saline or local anaestheticsolution, which is followed by carefulsuctioning of all excess irrigationsolution

� Do not attempt to curette the socket, asthis will increase the level of pain

� Prescription of potent oral analgesics� The patient is given a plastic syringe

with a curved tip for home irrigationwith chlorhexidine solution or saline andinstructed to keep the socket clean. Oncethe socket no longer collects any debris,home irrigation can be discontinued.

316 Blum

References

1. A I, A B, B T.The effect of locally applied gauze drainimpregnated with chlortetracycline oint-ment in mandibular third molar surgery.Acta Odontolol Scand 1998: 56: 25–29.

2. A RE. Dental extractionwound management: A case againstmedicating postextraction sockets. J OralMaxillofac Surg 2000: 58: 538–551.

3. B JK. Metronidazole and drysocket: prophylactic use in mandibularthird molar removal complicated by non-acute pericoronitis. New Zealand Dent J1987: 7: 71–75.

4. B MJE, L ME. Effects of achlorhexidine gluconate oral rinse inmandibular third molar surgery. J OralMaxillofac Surg 1990: 48: 440–448.

5. B H. Bacterial and fibrinolytic activityin ‘dry socket’. Acta Odontolol Scand1970: 28: 773–783.

6. B H. Fibrinolytic activity of normalalveolar bone. Acta Odontolol Scand1971: 29: 141–153.

7. B H. Fibrinolytic activity of alveolarbone in ‘dry socket’. Acta OdontololScand 1972: 30: 23–32.

8. B H, M-J, G. Cellularfibrinolytic activity of human alveolarbone. Int J Oral Surg 1972: 1: 121–125.

9. B H. Kinins and pain in dry socket.Int J Oral Surg 1972a: 1: 34–42.

10. B H. Antifibrinolytic effect of Apernylin dry socket. Int J Oral Surg 1972b: 1:190–194.

11. B H. Etiology and pathogenesis offibrinolytic alveolitis (‘dry socket’). Int JOral Surg 1973: 2: 215–263.

12. B CR. Alveolar osteitis preven-tion by immediate placement of medi-cated packing. Oral Surg Oral Med OralPathol Oral Radiol Endod 2000: 90: 282–284.

13. B JH, B M, R MJ.Effect of surgical trauma and polylactatecubes and granules on the incidence ofalveolar osteitis in mandibular thirdmolar extraction wounds. J Can DentAssoc 1986: 52: 315–320.

14. B DP, S JB. Effect of lavageon the incidence of localized osteitis inmandibular third molar extraction sites.Oral Surg Oral Med Oral Pathol 1977:44: 14–20.

15. B H, N CE, N A.Effect of azidocillin, erythromycin, clin-damycin and doxycyclie on postoperativecomplications after surgical removal ofimpacted mandibular third molars. Int JOral Surg 1980: 9: 157–165.

16. C PB, M RC. The histologiceffect of topically applied acetylsalicylicacid on bone healing in rats. Oral SurgOral Med Oral Pathol 1982: 33: 728–735.

17. C JE. Review of factors contrib-uting to dry socket through enhancedfibrinolysis. J Oral Surg 1979: 37: 42–46.

18. C JE, H S, E SH,C D. Effect of oral contraceptive

cycle on dry socket (localized alveolarosteitis). J Am Dent Assoc 1980: 101:777–780.

19. C P, D L. A review ofdry socket: A double-blind study on theeffectiveness of clindamycin in reducingthe incidence of dry socket. J Can DentAssoc 1992: 58: 43–52.

20. C ME, S JW. Effects ofgender-related factors on the incidence oflocalized alveolar osteitis. Oral Surg OralMed Oral Pathol Oral Radiol Endod1995: 79: 16–22.

21. C RC. The general practitioner’sperspective of the etiology, prevention,and treatment of dry socket. Gen Dent1997: 9: 461–467.

22. C JY. Dry socket. Dent Cosmos1896: 38: 929.

23. D W J, B AU, D W. Theuse of granular gelatin-tetracycline com-pound after third molar removal. J OralSurg 1981: 39: 466–467.

24. F EA, S JA, R E,S J. Dry socket incidence com-pared after a 12 year interval. Br J OralMaxillofac Surg 1988: 23: 419–427.

25. F PG, K GF, S DS,K RJ. Evaluation of intra-alveolarchlorhexidine dressing after removal ofimpacted mandibular third molars. OralSurg Oral Med Oral Pathol 1992: 73:383–388.

26. F KL, O RAJ. Alveolarosteitis following removal of mandibularthird molars. Anaesth Prog 1990: 37:32–41.

27. G-P N. Tranexamic acid inalveolar sockets in the prevention ofalveolitis sicca dolorosa. Int J Oral Surg1979: 8: 421–429.

28. H S, N A. Prevention ofpostoperative symptoms by general anti-biotic treatment and local bandage inremoval of mandibular third molars. Int JOral Surg 1973: 2: 273–278.

29. H CB, H TJ, BAR, B RA, G RW,MC SF. Perioperative use of0.12% chlorhexidine gluconate for theprevention of alveolar osteitis. Oral SurgOral Med Oral Pathol Oral RadiolEndod 1998: 85: 381–387.

30. H JR, G DP. The effect ofpolylactic acid granules on the incidenceof alveolar osteitis after mandibular thirdmolar surgery. A prospective randomizedstudy. Oral Surg Oral Med Oral PatholOral Radiol Endod 1995: 80: 279–283.

31. J WS, B EE. An eval-uation of 9-aminoacridine/Gelfoam toreduce dry socket formation. Oral SurgOral Med Oral Pathol 1988: 66: 167–170.

32. K L, N A. Post-operative complications after surgicalremoval of mandibular third molars. Int JOral Maxillofac Surg 1986: 15: 25–29.

33. K GO. Textbook of Oral andMaxillofacial Surgery. St Louis: Mosby1973: 226.

34. L WRE, S D,MF TW. Control of post-operative infection. Br Dent J 1972: 133:106–109.

35. L PE. The effect of a chlorhexidinerinse on the incidence of alveolar osteitisfollowing the surgical removal ofimpacted mandibular third molars. J OralMaxillofac Surg 1991: 49: 932–937.

36. L MV. Alveolar osteitis. J IndianDent Assoc 1969: 41: 69–72.

37. L GE, O DB, R EM.Alveolar osteitis associated with man-dibular third molar extractions. J AmDent Assoc 1974: 88: 802–806.

38. L DP, N JR, MCJ. Myospherulosis of oral hard and softtissues. J Oral Maxillofac Surg 1984: 42:349–355.

39. MG AJ. Aetiology of drysocket: A clinical investigation. Br J OralSurg 1968: 6: 49–58.

40. M B, G GT. Localis-ation of fibrinolytic activity during tootheruption in rats. Scand J Dent Res 1973:81: 315–318.

41. M JG, V GR, RSN. Local anesthesia and dry socket: Aclinical investigation of single extractionsin male patients. Int J Oral MaxillofacSurg 1987: 16: 279–284.

42. M L. Topical metronidazole inthe treatment of ‘dry socket’. Br Dent J1984: 156: 132–134.

43. M JW, B JH. Foreign bodygiant cell reaction related to placementof tetracycline-treated polylactic acid:Report of 18 cases. J Oral MaxillofacSurg 1990: 48: 808–812.

44. N D, S JF, W D.Fibrinolytic activity of oral anaerobicbacteria. Arch Oral Biol 1978: 23: 465–470.

45. N DW. On the genesis of ‘drysocket’. J Oral Maxillofac Surg 1983: 41:706–710.

46. Nø SE. Treatment of acute painfollowing removal of mandibular thirdmolars. PhD thesis. Int J Oral MaxillfacSurg 1998: 27: 1–41.

47. O TP, F G, S I,T TS. A prospective study ofcomplications related to mandibular thirdmolar surgery. J Oral Maxillofac Surg1985: 43: 767–769.

48. P M, S JM, S, U,B JV. Oral hygiene and post-operative pain after mandibular thirdmolar surgery. Oral Surg Oral Med OralPathol Oral Radiol Endod 2001: 92: 260–264.

49. R JR, S AJ. Evaluation of0.12% chlorhexidine rinse on the pre-vention of alveolar osteitis. Oral SurgOral Med Oral Pathol 1991: 72: 524–526.

50. R M, H S, B PE,E BK. Does metronidazole pre-vent alveolitis sicca dolorsa? Int J OralMaxillofac Surg 1992: 21: 299–302.

Contemporary views on dry socket (alveolar osteitis) 317

51. R JP, M J. Metronida-zole in the prevention of ‘dry socket’. Br JOral Surg 1979: 17: 62–70.

52. R JP, D M. Metronidazole inthe treatment of ‘dry socket’. Int J OralSurg 1981: 10: 345–347.

53. R JP, C J. Metronidazolelevels in alveolar bone. Royal Society ofMedicine International Congress andSymposium Series. London: AcademicPress 1979: 18: 45–47.

54. R J, S IDF, W BA.Is dry socket preventable? J Can DentAssoc 1977: 43: 233–236.

55. R J. Removal of impacted lower thirdmolars with acute pericoronitis andnecrotising gingivitis. Br J Oral Surg1970: 7: 153–160.

56. R JL, M RM. Terra-Cotril for the reduction of the incidenceof localized osteitis following mandibularthird molar removal. J Oral Med 1984:39: 51–53.

57. S JP, F-D G, HG. Fibrinolytic alveolitis and its preven-tion. Int J Oral Maxillofac Surg 1987: 16:175–183.

58. S SR. Evaluation of postoperativelocalized osteitis in mandibular thirdmolar surgery. Oral Surg Oral Med OralPathol 1974: 38: 352–358.

59. S HE. The healing of extractionwounds. Br Dent J 1969: 126: 550–557.

60. S AL, H WB, S DW,J ED. Complications followingremoval of impacted third molars: the

role of the experience of the surgeon. JOral Maxillofac Surg 1986: 44: 855–859.

61. S DC, P JW. The effectof tetracycline on the incidence ofpostextraction alveolar osteitis. J OralMaxillofac Surg 1987: 45: 1029–1033.

62. S AE. A double-blind study inthe effectiveness of tetracycline in reduc-ing the incidence of fibrinolytic alveolitis.J Oral Maxillofac Surg 1989: 47: 165–167.

63. S JB, B DP, D JL.Effects of lavage techniques with thirdmolar surgery. Oral Surg Oral Med OralPathol 1976: 41: 152–168.

64. S JB, B DP. Increased inci-dence of postoperative localized osteitisin mandibular third molar surgery associ-ated with patients using oral contracep-tives. Am J Obstet Gynecol 1977: 127:518–519.

65. S JB, B DP. Predisposing andoperative factors: Effect on the incidenceof localized osteitis in mandibular thirdmolar surgery. Oral Surg Oral Med OralPathol 1978: 46: 206–215.

66. S JB, B DP. The relationshipof smoking to localised osteitis. J OralSurg 1979: 37: 732–735.

67. T A. Influence of oral hygienemeasures on the development of alveolitissicca dolorosa after surgical removal ofmandibular third molars. Int J OralMaxillofac Surg 1979: 8: 430–434.

68. T N, S GD. Preventingdry socket: A simple procedure thatworks. J Am Dent Assoc 1991: 2: 67–68.

69. T AT, I DP, P NA.Dry socket: frequency of occurrence afterintraligamentary anesthesia. Quint Int1992: 23: 575–577.

70. T PS. A clinical study of drysocket. Int J Oral Surg 1982: 11: 226–231.

71. V P, R M, D O.Use of propylic ester of p-hydroxy-benzoic acid after removal of impactedmandibular third molars. Int J OralMaxillofac Surg 1974: 3: 394–399.

72. V AE, K GA, NMG. Reduction of salivary bacteria bypre-procedural rinses with chlorhexidine0.12%. J Periodontolol 1991: 62: 649–651.

73. V PJ. Dental extraction woundmanagement: Medicating PostextractionSockets. J Oral Maxillofac Journal 2000:58: 531–537.

74. Y Y, B S, K-B K,N L. Changes in blood coagulationand fibrinolysis in women receiving oralcontraceptives. Am J Obstet Gynaecol1969: 104: 87–98.

75. Z JR, L JC. Topicaltetracycline-induced neuritis: A casereport. J Oral Maxillofac Surg 1995: 53:196–199.

Address:Dr Igor R. Blum11 Austin DriveManchester M20 6EBUKTel: +44 (0) 161 4343489E-mail: igor@blum.ssword.co.uk