Answers are grounded in this paper and cite their sources.
Background Delayed-onset infection (DOI) after mandibular third-molar surgery is inconsistently defined, and its associated factors, microbiology, and management remain uncertain. We characterized DOI timing and synthesized evidence on study-specific frequency, associated factors, microbiology, management, and outcomes. Methods PubMed, Embase, and Scopus were searched through 30 September 2026, with citation searching. The protocol was registered in PROSPERO (CRD420261527462). Overlapping reports were consolidated into study families. Risk of bias was assessed using QUIPS and design-appropriate JBI tools, and certainty using an adapted GRADE approach for prognostic-factor evidence. Because of substantial heterogeneity, evidence was synthesized without meta-analysis using SWiM principles. Results Nineteen reports representing 13 independent study families were included. DOI definitions ranged from onset after 7 days or suture removal to windows extending to 2–3 months, with many dedicated studies reporting onset around 3–6 weeks. Study-specific frequencies in broader extraction cohorts were approximately 0.7%–4.0%, whereas selected Bologna cohorts reported 5.5%–9.2% per procedure, with the higher estimate arising in the germectomy cohort; denominators and ascertainment differed. Distal-space/Class III associations recurred across non-equivalent measures and analyses, but adjusted evidence was limited and certainty very low. Greater impaction depth showed recurrent but less uniform associations. Microbiological evidence came from one small single-centre series, and management evidence was observational and noncomparative. Conclusions DOI is a recognizable late postoperative presentation defined heterogeneously across studies. Anatomical restriction may be associated with DOI, but all formally graded prognostic domains had very low certainty. Standardized definitions and future robust well-designed prospective multicentre studies are needed.
No takes yet. Share an insight, caveat, or question.
Kahraman et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: