An in vitro comparison between two methods of electrical resistance measurement for occlusal caries detection

Because of different measurement techniques and the easier design
of the CRM prototype, this in vitro study aimed to compare the
diagnostic performance and reproducibility of two electrical
methods (Electronic Caries Monitor III, ECM and Cariometer 800,
CRM) for occlusal caries detection, and to evaluate the effect of
staining/ discoloration of fissures on diagnostic performance.
Hundred and seventeen third molars with no apparent occlusal
cavitation were selected. Six examiners inspected all specimens
independently, using the CRM, and a subgroup of 4 using the ECM.
Histological validation using a stereomicroscope was performed
after hemisectioning. Intra- and interexaminer reproducibility was
assessed by Lin's concordance correlation coefficient (CCC) and
Bland and Altman analysis. Diagnostic performance parameters
included sensitivity (SE), specificity (SP) and area under the ROC
curve (A(z)). The CCC yielded an intra- and interexaminer
reproducibility of 0.69/0.62 (ECM) and of 0.79/0.74 (CRM). The mean
intra- and interexaminer 95% range of measurements (range between
Bland and Altman limits of agreement) given in percentages of the
instrument reading were 67%/65% for the ECM and 28%/33% for the
CRM. A(z) at the D3-4 level was 0.74 (ECM) and 0.78 (CRM). The CRM
showed at least equivalent diagnostic performance to the ECM.
However, improvement is still desirable. Diagnostic performance
appeared to be enhanced in discolored lesions; however, this may be
related to sample lesion distribution characteristics. Copyright
(C) 2006 S. Karger AG, Basel.