Effect modifiers influencing the periodontal healing of the adjacent second molar following mandibular third molar removal

Data sources: Studies were selected, using appropriate key words, from the following databases: Cochrane Central Register of Controlled Trials; Medline (via Pubmed); Scopus/Elsevier; and Embase. A manual search was also conducted of five periodontology and oral and maxillofacial surgery journals. It was not clarified what proportion of studies included were derived from which source. Study selection: Inclusion criteria included prospective studies and randomised controlled trials published in English, with a minimum 6-month follow-up reporting on parameters of periodontal healing distal to the mandibular second molar following removal of M3M in human subjects. These parameters included pocket probing depth (PPD) reduction and final depth (FD), clinical attachment loss (CAL) reduction and FD, alveolar bone defect (ABD) change and FD. Given prognostic indictors and interventions were investigated, the studies were screened using PICO and PECO (Population, Intervention, Exposure, Comparison, Outcome). Cohen's kappa statistic measured the level of agreement between 2 selecting authors (0.96 stage 1 screening, 1.00 stage 2 screening). Disagreements were resolved with a tie-breaker 3rd author. Ultimately, from 918 studies, 17 met the inclusion criteria and 14 were included in the meta-analysis. Studies were excluded on the basis of same patient pools, non-representative outcomes of interest, insufficient follow-up period, and unclear results. Data extraction and synthesis: The 17 studies meeting the inclusion criteria underwent validity assessment and data extraction, including risk of bias analysis. Meta-analysis was performed to calculate mean difference and standard error for each outcome measure. If these were unavailable, a correlation coefficient was calculated. Meta-regression was used on different subgroups to determine factors affecting periodontal healing. For all analyses, statistical significance was determined as p < 0.05. Statistical variability in outcomes beyond the expected was estimated using I-2 analyses, with a value >50% indicating significant heterogeneity. Results: The periodontal parameters investigated yielded the following results after meta-analysis: overall PPD reduction of 1.06 mm at 6 months and 1.67 mm at 12 months; final PPD of 3.81 mm at 6 months; CAL change of 0.69 mm at 6 months; final CAL of 4.28 mm at 6 months and 4.37 mm at 12 months; ABD reduction of 2.62 mm at 6 months; ABD of 3.2 mm at 6 months. The authors found no statistically significant effect on periodontal healing from the following confounding factors: age; M3M angulation (specifically, mesioangular impaction); optimisation of periodontal health prior to surgery; scaling and root planing of the distal second molar at the time of surgery; post-operative antibiotics or chlorhexidine prophylaxis. There were statistically significant correlations between baseline PPD and final PPD. There was improved PPD reduction at 6 months with a three-sided flap compared to others, and regenerative materials and bone grafts improved all periodontal parameters. Conclusions: Although M3M removal results in modest improvement in periodontal health distally of the second mandibular molar, periodontal defects remain onwards of 6 months. There is limited evidence suggesting a three-sided flap is more beneficial than an envelope flap in PPD reduction at 6 months. Regenerative materials and bone grafts result in significant improvements across all periodontal health parameters. The most important predictive factor in final PPD of the distal second mandibular molar is baseline PPD.