RESUMEN
Teeth containing large defective restorations are frequently indicated to receive crowns to prevent catastrophic fracture. There is a considerable lack of consensus as to when the size of an existing restoration needing retreatment is sufficiently large to warrant a crown. In addition, the treating dentist's gender, age, and location also have been found to influence this decision. An in vitro method was developed to estimate a restoration's volume proportion (VP). This study validates and investigates correlations between the restoration's VP, the decision to indicate a crown, and the responding dentist's demographic information. The coronal portions of occlusal view and bitewing radiographic images of restored posterior teeth were traced to yield four surface areas used in estimating the restoration's VP. An online survey of 15 patients, including intraoral and radiographic images of defective restorations, was created. Dentists were invited to select treatment for each patient's tooth, with options including replacing the restoration only or indicating a crown. Analysis was accomplished using repeated-measures logistic regression. Of the 300 respondents, 17% were female and 14% were from outside the U.S. The relationship between a responding dentist indicating a crown and the restoration's VP, the tooth type (molar or premolar), and nationality (U.S. or foreign dentist) were all deemed statistically significant. Crowns were indicated for molars significantly more often than for premolars, with a significant VP interaction. The final model of tooth, VP, nationality, and the interaction of tooth and VP was used to generate mean predicted proportions for indicating a crown.