RESUMO
BACKGROUND: Dry socket (DS) or fibrinolytic osteitis is a relatively common complication that can occur following tooth extraction. This study aimed to determine the prevalence of DS and identify its associated predictive and mediating variables. MATERIAL AND METHODS: This study is classified as prospective observational, cross-sectional, and multicenter. Patients were consecutively selected in accordance with established criteria for tooth extraction. Data on patient demographics, surgical procedures and postoperative outcomes were collected. Nominal variables were analyzed using the Chi-Square Test, while associations involving ordinal values or considering counts or layers were examined using the Kendall's Tau-B Test or Mantel-Haenszel Test for trend. The GLM Mediation Model was employed to investigate potential mediation or indirect effects or potential underlying mechanisms of predictive variables on the development of DS. Two-tailed significance level of p ≤0.05 was considered statistically significant. RESULTS: A total of 1,357 patients undergoing routine dental extractions were included. DS was observed in 13 patients (prevalence of 1%). DS was associated with younger patients (under 50 years old), longer procedures, and the presence of surgical accidents, but only when mediated by surgical complexity. Smoking, particularly in combination with complex surgeries and surgical accidents, was associated with DS. Postoperative pain for more than two days and reported at moderate to high levels, emerged as a potential warning sign for DS. The use of antibiotics was found to significantly reduce the risk of DS (RR reduction of 36% and absolute risk reduction of 0.63%). CONCLUSIONS: Routine dental extractions revealed a 1% prevalence of dry socket. The obtained results suggests that DS is a multifactorial condition influenced by various factors, including gender, age, smoking, antibiotic prescription and surgical factors such as length, technique and accidents, nevertheless, those associations were observed mainly considering the influence of one variable on another.
