Aim: To assess the influence of different variables on the survival of class II glass ionomer restorations a clinical study was carried out among 6-7 year old schoolchildren in and around Dar es Salaam, Tanzania.
Methods: 217 children were randomly divided into three groups. In the conventional group, Class-II cavities were excavated with burs, in the ART group they were excavated with hand instruments and in the Carisolv group excavation took place with a chemical solution and special blunt hand instruments. The preparations in all groups were restored with hand mixed glass ionomer cement (Fuji IX). Four operators were involved in the treatment phase. After restoration, residual caries and cervical gaps were assessed on bite-wings. The quality of the restorations was established and the survival rate determined in two subsequent evaluations.
Results: 195 children (90%) were present at the first evaluation which took place after 7 months and 194 (89%) were present at the second evaluation, after one year. At the first evaluation the survival rate of the ART-restorations was 38%, of the conventional restorations 50%, and of the Carisolv restorations 35%. Survival rates at the time of the second evaluation were 30%, 42% and 28% respectively. The differences between the three treatment groups were, however, not significant (p = 0.200 at t = 1 and p = 0.247 at t = 2). In the first evaluation 161 bitewings were available of the 195 restorations, while for the second evaluation 130 cases were available. There was no significant relation between residual caries and the success/failure rate (p = 0.140 at t = 1 and p = 0,201 at t = 2). Also cervical gaps appeared to have no relation with the failure rate (p = 0.057 at t = 1 and p = 0.833 at t = 2) However, together those variables have a significant influence on the survival rate (p = 0.025).
Conclusion: The combination of two variables (residual caries and cervical gaps) has considerable influence on the survival rate of class II glass ionomer restorations compared to each variable alone.