RESUMO
AIM: This retrospective study investigated the prevalence of defective restorations in a public dental service and factors associated with re-intervention in primary teeth. DESIGN: The sample consisted of all clinical records (census) of children presenting restorations in primary teeth, who had undergone dental treatment in a public set during 1-year period. For analysis, only restorations presenting defects related to esthetic, functional, or biological reasons at first clinical examination were included. The outcome 'Success' was set when the restoration received no treatment (monitored), refurbishing, sealing of margins, or was repaired. Otherwise, 'Failure' was set whenever the restoration was either replaced or if another treatment affecting the restoration was necessary (endodontic treatment or tooth extraction). Poisson regression model was used to assess the prevalence of patient- and tooth-related factors that may influence the re-intervention decision (repair or replacement). RESULTS: From a total of 302 restorations placed in 114 children, 37.7% presented some type of defect. Restorations with recurrent caries were more frequently present in caries-active patients (P = 0.03) and were frequently replaced (95% CI, 1.05-3.22, RR = 1.84, P = 0.03). CONCLUSION: Presence of recurrent caries influences the re-intervention decision, leading to restoration replacement in most cases.