Prediction of efficacy of bitewing radiographs for caries detection.

Although bitewing (BW) radiographs play an indispensable role in the detection of interproximal caries, it is not always clear when they should be made. The present study was designed to test the hypothesis that a person's caries and restorative history may be predictive of his or her caries future. Full-mouth radiographic examinations (FMX) and subsequent BW radiographs from 349 adults (age 20 or more) selected at random were examined. Possible risk factors for caries recorded for each patient included the number and type of carious lesions and the restorative experience of the patient. Linear logistic regression models were evaluated to estimate the efficacy of the possible risk factors as predictors of the presence of caries on the BW radiographs. Of the 349 adults examined, 91 (26%) had one or more new carious lesions at the time of their BW examination. The probability of caries in this population at the time of the BW examination increases both with lengthening interval between the FMX and the BW radiographs and with the number of restorations at the time of FMX. A regression model was adopted that estimates the probability of caries being detected on BW radiographs. We found that to provide the most good for the most patients (minimize cost), the threshold level for ordering follow-up radiographs should be when the probability of detecting caries is 5%. It may be seen that even when a person has no caries or restorations at the time of the FMX there will be a 5% chance that BW radiographs will reveal the presence of caries 1.5 years later. The 5% threshold is reached sooner in the presence of caries or restorations.