Predicting which adult patients will need treatment over the next year

This prospective study was conducted to determine factors important in predicting which regularly attending adult patients would receive first, restorations or extractions for any reason (receiving treatment) and, second, restorations or extractions undertaken specifically for caries (receiving treatment related to caries). Baseline and incremental clinical data were obtained from 24 general dental practitioners on a group of their regularly attending, dentate adult patients over a 12-month period. The patients completed a postal questionnaire with questions relating to dental health behaviour, attitudes, knowledge, and social factors. Complete data were obtained from 2553 patients. Thirty-one variables were identified as potential predictors for the two dependent variables receiving treatment and receiving treatment related to caries, and logistic regression models were fitted. Receiving treatment was associated with having fewer sound teeth and more anterior fillings, posterior fillings and crowns (P < 0.001). The dentist's prediction of the need for treatment related to caries and the patient's own prediction of the need for a filling were also important in the model (P < 0.001). Some of these variables, together with having received recent medical treatment and taking sugar in tea or coffee were also found to predict treatment related to caries. The model for receiving treatment related to caries was more successful at predicting the patient's individual risk but the model for receiving treatment was slightly better at classifying patients into whether or not they received treatment. It is reassuring that the common assumptions made by the dental practitioners of their patient's risk have received statistical validation.