Prognosis of conventional root canal treatment reconsidered

In endodontic literature, the so-called success rate of conventional root canal treatment is reported to range between 70% and 95%. This has been calculated as the percentage of successfully treated teeth of all teeth followed up or included in the clinical trial. This approach, however, does not allow for valid assertions on the prognosis of root canal treatment as the individual observation times are not considered. This article discusses some methodological and statistical aspects of how to design a prognostic study which focuses on the outcome of endodontic therapy and of how to analyse the data appropriately. Methodologically, the response variable should preferably be the individual time required for the occurrence of an event, e.g., success or failure of endodontic therapy which should clearly be defined on the basis of widely accepted criteria in endodontology. Event times can appropriately be analysed by the Kaplan-Meier method, which estimates the probability that the event will not occur within a fixed time. This probability, together with the approximate 95% confidence interval (CI), permits an evaluation of the prognosis of a particular treatment. Two data sets were re-analysed to clarify the rationale behind the analysis of event times. Accordingly, the probability that an endodontically induced lesion will completely heal, e.g., within the first 3 years after root canal therapy ranges between 0.87 (CI: 0.74-1.00) and 0.89 (CI: 0.80-0.98). In this situation, the simple calculation of success rates would overestimate the chance of complete periapical healing within the first years after therapy but underestimate it over longer observation periods. Another example was used to analyse the time to occurrence of periapical pathosis associated with root canal treated teeth not diseased periapically. In this case the chance of recording a successful endodontic treatment is initially underestimated by the percentage of successful cases of all teeth integrated in the study but is markedly overestimated for longer observation periods. Potential risk factors affecting the outcome of the endodontic therapy and thereby the event times can adequately be determined by applying the Cox's or Aalen's regression model.