Method: A total of 262 primary molars with approximal caries lesions were restored in 5-8 years old children in Itatiba city, Brazil. The patients were randomly allocated in three groups: G1 – Fuji IX (GC, BE); G2 – HI Dense (Shofu, GE) and G3 – Maxxion R (FGM, BR). After caries removal with hand instruments, the cavities were restored with one of the GICs by two trained operators. Restorations were evaluated after 1, 6, 12, 18, 24, 30 and 36 months according ART criteria. Failed restorations were, if possible, repaired or replaced by another ART restoration using the same GIC. Chi square test (5% significance) was used to evaluate the influence of GIC-brand, tooth-surface and operators in the survival rate in each evaluation. The repaired restorations were also evaluated in relation to the tooth survival and the GIC brand.
Result: There was no difference in success rate among the different brands of GIC. Significant operator and tooth-surface influence were found at all evaluations, with the distal surface presenting the higher number of failures. A drop-out of 18% was observed and overall survival of the restorations was 12% after 36 months. Of the repaired restoration, the tooth survival rate was 74% and there was no difference among the different GIC-brands.
Conclusion: There is no influence of the brand of the GIC in the survival rate of approximal ART restorations. Repairing and replacing failed ART restorations is indicated to increase the survival of the tooth.