In this particular review, the authors have undertaken the analysis of a very difficult topic, the prevention and treatment of alveolar osteitis (AO) using local measures. `While most clinicians know AO when they see one, for purposes of analysis, the working definition can be quite variable, making comparisons across studies difficult. The reported frequencies of AO range from 0 – 37.5%. The wide range of estimates may be due to differences in the studies' working definition of AO. In this report, diagnostic criteria for AO were not listed. Each study included had the diagnostic criteria agreed upon prospectively. It is not clear, however, that the definitions of AO were the same across studies, and combining data may not be appropriate. Given the unknown frequency of AO, a strength of this review was estimating the treatment effect across several possible frequencies of AO, ie 1%, 5%, and 30%.

The results of this study suggest that the use of chlorhexidine mouthrise or intrasocket application have a minimal to marked effect on decreasing the frequency of AO, ie the number needed to treat ranged from 232 (mouthrinse at a 1% prevalence of AO) to six (intrasocket application at a 30% prevalence of AO). Despite the authors' cautions to the contrary, which seems overestimated in this reviewer's opinion, the risk of adverse effects from brief exposures to chlorhexadine is minimal.

The authors reviewed the use of chlorhexidine in the prevention of AO, but failed to offer any opinions regarding other local treatments to prevent or treat AO other than to say that the data were “insufficient” to determine their effects. Failing to offer any recommendations based on these other single study treatments that were all randomised clinical trials is an abrogation of the reviewers' responsibilities. Evidence-based clinical decision-making is not making decisions with the ‘best' evidence. It is making clinical decisions with the ‘best’ evidence that is available. As such, the reviewers leave it to the reader to make some sense of the other 10 intrasocket interventions to prevent AO and the five interventions for treating AO. Fortunately, the reviewers outlined these other studies with enough detail to make some tentative observations. Based on the material provided (including additional tables), many studies were plagued by inadequate power to detect a difference between local treatments to prevent AO. In two out of three studies, intrasocket application of antibiotics (metronidazole or clindamycin) were statistically effective in reducing AO frequency and the direction of the treatment effect was all in favour of the antibiotic groups, suggesting a promising area of future research efforts.

Treating AO continues to be the bane of both patients and clinicians. The reviewers cited five studies using topical anesthetic, Alvogyl (a propriatery AO paste), Salicept (a proprietary mouthrinse or oral patch used for the relief of pain and management of inflammatory oral lesions or injuries), Alvogyl versus Salicept, or metronidazole. This report failed to offer any additional new information to improve patient outcomes for those with AO.

Based on the results of this report, this reviewer has changed perioperative treatment to include pre- and post-operative rinses of 0.12% chlorhexidine.