RESUMEN
OBJECTIVE: To evaluate the postoperative sensitivity (POS), as well as the clinical performance of posterior restorations using a new chemically-cured bulk-fill composite (Stela Automix and Stela Capsule, SDI) comparing with a light-cured bulk-fill composite after 6 months. METHODS: Fifty-five participants with at least three posterior teeth needing restoration were recruited. A total of 165 restorations were performed on Class I or Class II cavities. After the application of Stela primer, the chemically-cured composite (Stela Automix or Stela Capsule) was inserted. For the light-cured composite group, a universal adhesive (Scotchbond Universal) was applied with a bulk-fill composite (Filtek One). Participants were evaluated for spontaneous and stimulated POS in the baseline, after 48 h, 7 days, and 6 months. Additionally, each restoration was assessed using the updated version of FDI criteria after 6 months. The differences in the proportions of the groups were compared by Cochran test statistics (α = 0.05). RESULTS: Both chemically-cured composites showed a lower risk of POS compared to the light-cured composite at baseline and up to 48 h (p < 0.04). A significantly lower surface luster and texture was observed for the Stela Capsule composite compared to the light-cured bulk-fill composite (baseline and 6 months; p = 0.03). A significant color mismatch was observed for the light-cured bulk-fill composite compared to the chemically-cured composites (baseline and 6 months; p = 0.03). No significant differences were observed in any other item evaluations (p > 0.05). CONCLUSION: Chemically-cured composites exhibit lower postoperative sensitivity and less color mismatch compared to a light-cured bulk-fill composite after 6 months of clinical service. CLINICAL SIGNIFICANCE: The chemically-cured composites appear to be an appealing option for restoring posterior teeth, as they exhibit lower postoperative sensitivity compared to a light-cured bulk-fill composite, both at baseline and up to 48 h, and less color mismatch.