Occlusal caries detection: a comparison of DIAGNOdent and two conventional diagnostic methods
The aims of this study were: 1) to compare the DIAGNOdent readings for the detection of caries on three different macroscopically sound and intact occlusal surfaces for the baseline measurement; and 2) to compare this laser device with visual and radiographic inspection. Forty-five extracted premola...
Saved in:
Published in | The Journal of clinical dentistry Vol. 15; no. 3; p. 76 |
---|---|
Main Authors | , , , , |
Format | Journal Article |
Language | English |
Published |
United States
2004
|
Subjects | |
Online Access | Get more information |
Cover
Loading…
Summary: | The aims of this study were: 1) to compare the DIAGNOdent readings for the detection of caries on three different macroscopically sound and intact occlusal surfaces for the baseline measurement; and 2) to compare this laser device with visual and radiographic inspection.
Forty-five extracted premolars and molar human teeth were cleaned with a pumice slurry. Then 74 sites were defined and radiographs were taken from each one. Three examiners performed the examinations and took the DIAGNOdent measurements. Sensitivity and specificity were calculated for each diagnostic system, as well as intra- and inter-examiner reproducibility using Kappa tests. The cut-off point for visual and radiographic inspection was 1-2, and for the DIAGNOdent the manufacturer's instructions were followed.
No statistical difference was found for the three predefined sound sites used for baseline measurements with DIAGNOdent. The visual inspection showed high sensitivity; radiographic inspection and DIAGNOdent measurements showed low sensitivity and high specificity. The highest intra- and inter-reproducibility was found with the DIAGNOdent.
The thickness of the enamel did not have any effect on the readings with DIAGNOdent. It was concluded that in a low prevalence sample, the visual inspection provided the highest proportion of true disease identified correctly, and DIAGNOdent provided the highest proportion of non-disease identified correctly. |
---|---|
ISSN: | 0895-8831 |