A statistical model for predicting the retrieval rate of separated instruments and clinical decision-making

Abstract

Background/purpose: There are controversial opinions about the prognosis of the retrieval of a separated instrument from the root canal. The aim of the study was to establish a statistical model to predict the success rate of the retrieval of separated instruments and to aid clinicians with decision-making. Materials and methods: In retrospective studies, information on the tooth position, the root canal curvature, and the depth and length of separated instruments were collected in 210 clinical cases with separated instruments in the lower segments of curved root canals. The correlations of these factors and the retrieval rate of separated instruments were analyzed. Two factors with significant correlations were chosen and a regression equation was established using stepwise multivariate logistic regression analysis. In the verification study, the efficiency of the statistical model was verified by 63 new cases. Results: The root canal curvature and depth of the separated instruments are major factors affecting the retrieval rate of broken instruments. The retrieval rate of separated instruments decreased gradually with the increase of root canal curvature or the depth of the instrument. A regression equation was established correlating these two factors. The predicted accuracy rate of the regression equation was 94.3% for successful retrieval, and 80.0% for failed retrieval. Conclusion: A statistical model relating to root canal curvature and depth of separated instruments was established to evaluate the retrieval rate of separated instruments, and the result of this formulation may provide clues for clinical decision-making.

Keywords

• instrument fracture
• logistic regression analysis
• root canal treatment
• separated instruments
• statistical model