Tunnel restorations in general practice. Influence of some clinical variables on the success rate

Using bitewing radiographs and clinical inspection, the success rate for tunnel restorations was assessed in a population with low caries activity. The material consisted of 242 tunnel restorations in permanent premolars and molars in 142 individuals (mean age = 18.8 years). The median DFSappr value (decayed and filled approximal surfaces) at the time of restoration was 4.0. The mean follow-up time was 25 months. Bivariate associations between the outcome variable (success/failure of the tunnel restoration) and conceivable explanatory variables were investigated. In a multivariate logistic regression analysis, the independent variables tooth type (premolars vs molars), surface site (mesial vs distal), radiographic stage of approximal carious progression and age of patient at the time of restoration (9-15 years vs > 15 years) were used to estimate the effect on the dependent variable success/failure. Using the life table method, the estimated cumulative proportion of successful restorations was 81% after 2 years and 64% after 3.5 years. The success rate was not related to caries activity and did not differ between the two types of tunnel preparation techniques nor between different follow-up periods. In the multivariate regression analysis, tooth type (molars vs premolars) was the only factor significantly associated with failure. Thus, a failure occurred about 5 times as often in molars as in premolars. Of the failures, half were due to caries; either radiographically observed adjacent to the restoration or progressing enamel caries on the outer proximal surface. Marginal ridge fractures constituted 26% of the failures. From the present results it can be concluded that in a population with low caries activity, the tunnel restoration technique can be recommended for premolars.