Statistical The data were encoded into standard formulas
Statistical comparison of the ELISA results obtained with infected and control rabbit groups was performed with the Mann_/Whitney U-test 17. Completed 2_/2 tables and the receiver operating characteristic (ROC) curve were used for diagnostic evaluation of the ELISA 13,18. Sensitivity was calculated as the proportion of samples positive for the disease correctly identified by the ELISA test as positive and specificity as the proportion of true negative specimens correctly identified as negative. The cut-off values for absorbance (A) at 405/490 nm were calculated as the mean value for the negative group of samples plus 0, 1, 2, 3 and 4 standard deviations (SD), respectively. The relationship between sensitivity and specificity of the test was displayed with ROC curves. The ROC curve plots the true positive rate of sensitivity against the one minus the specificity as a false positive rate of different cut-off values. A diagonal line in a plot corresponds to a test that is positive or negative just by chance. The optimal cut-off value was set where ELISA exhibited a high true positive rate and a low false positive rate; this value, therefore, lies close to the top left-hand corner of the ROC curve. The data were encoded into standard formulas for measurement of test reliability. Samples used for the detection of specific antibodies correlating with T. mentagrophytes infection were divided into positive (D_/) and negative (D_/) samples according to whether or not the rabbit source was infected with the disease. ELISA results were divided into positive (T_/) and negative (T_/) on the basis of the cut-off value calculated as a mean of test results for the negative group of samples (D_/), plus the multiplier of numbers of SD as outlined above. ELISA results were analysed as falling into four categories: TP: true positive (D_/, T_/); TN: true negative (D_/, T_/); FN: false negative (D_/, T_/); FP: false positive (D_/, T_/). Sensitivity and specificity were calculated by using 2_/2 table at the five different SD cut-off values mentioned. Sensitivity was calculated as the proportion of true positives identified as positive in ELISA test (TP/TP_/FN) and specificity as the proportion of true negatives identified as negative in ELISA (TN/TN_/FP). The predictive value of a positive result was calculated as the proportion of animals with a positive ELISA result that really were positive (TP/TP_/FP) and the predictive value of a negative result as the proportion of animals showing a negative ELISA result that really were negative (TN/TN_/FN).