Endodontic retreatment. Aspects of decision making and clinical outcome.

Epidemiological surveys have reported that 25%-35% of root filled teeth are associated with periapical radiolucencies. Descriptive studies have demonstrated that clinicians' decision making regarding such teeth are subject to substantial variation. A coherent model to explain the observed variation has not been produced. In the present thesis a "Praxis Concept theory" was proposed. The theory suggests that dentists perceive periapical lesions of varying sizes as different stages on a continuous health scale. Interindividual variations can then be regarded as the result of the choice of different cut-off points on the continuum for prescribing retreatment. In the present study experiments among novice and expert decision makers gave evidence in favour of the theory. Data also suggested that the choice of retreatment criterion is affected by values, costs of retreatment and technical quality of original treatment. From a prescriptive point of view, the presence of a persistent periapical radiolucency has often been used as a criterion of endodontic "failure" and as an indication for endodontic retreatment. As an alternative decision strategy, the use of decision analysis has been proposed. Logical display of decision alternatives, values of probabilities, utility values (U-values) of the different outcomes and calculation of optimal decision strategy are features of this theory. The implementation of this approach is impeded by the uncertainty of outcome probabilities and lack of investigations concerning U-values. U-values of two periapical health states in root filled teeth (with and without a periapical lesion respectively) were investigated in a group of 82 dental students and among 16 Swedish endodontists. Two methods were used to elicit U-values: Standard gamble and Visual Analogue Scale. Large interindividual variation for both health states were recorded. The difference in U-values between the two health states was found to be statistically significant regardless of assessment method. Compared with Standard gamble Visual Analogue Scale systematically produced lower ratings. U-values were found to change considerably in both the short and long-term. Any significant correlation between endodontists' U-values and retreatment prescriptions could not be demonstrated. Surgical and nonsurgical retreatment were randomly assigned to 95 "failed" root filled teeth in 92 patients. Cases were followed clinically and radiographically for four years postoperatively. At the 12-month recall a statistically significant higher healing rate was observed for teeth retreated surgically. At the final 48-month recall no systematic difference was detected. Patients were found to be more subject to postoperative discomfort when teeth were retreated surgically compared with nonsurgically. Consequently, surgical retreatment tended to be associated with higher indirect costs than a nonsurgically approach. In the final part of the thesis it is argued that retreatment decision making in everyday clinical practice normally should be based on simple principles. It is suggested that in order to achieve the best overall consequence a periapical lesion in a root filled tooth that is not expected to heal should be retreated. Arguments to withhold retreatment should be based on (i) respect for patient autonomy, (ii) retreatment risks or (iii) retreatment costs.