AN EVALUATION OF THE DIAGNOSTIC YIELD FROM BITEWING RADIOGRAPHS OF SMALL APPROXIMAL AND OCCLUSAL CARIOUS LESIONS IN A LOW-PREVALENCE SAMPLEIN-VITRO USING DIFFERENT FILM TYPES AND SPEEDS

Aim To compare diagnostic yield in caries diagnosis from D- and E-spee d films.Design A laboratory study. Setting A UK dental school between 1992 and 1994. Materials and methods 96 extracted teeth containing app roximal and occlusal lesions, but representing a low caries prevalence sample, were set in occluding dental arches. Bitewing radiographs wer e taken and interpreted by 5 examiners for the presence or absence of caries. Each examiner was also asked which film image he or she subjec tively liked best. Main outcome measures The teeth were subsequently s ectioned and histologically examined to validate diagnostic decisions. Results For all film types the percentage of lesions with caries hist ologically in dentine correctly identified radiologically (sensitivity ) was low (approximal caries 8-22%; occlusal caries 0-30%). The number of sound dentine sites correctly identified (specificity) was high (a pproximal caries 98-100%; occlusal caries 79-100%). There were no sign ificant differences between D- and E-speed films. Sensitivity was unaf fected by each examiner's subjective preference for a particular film. The variation in sensitivity of diagnosis was due to differences betw een examiners. Conclusions The reluctance of many GDPs to use E-speed film because they 'do not like the image' cannot be endorsed or suppor ted. Both E-speed film types examined can be recommended for use in ge neral practice.