Technical and financial evaluation of assays for progesterone in canine practice in the UK.

The concentration of progesterone was measured in 60 plasma samples from bitches at various stages of the oestrous cycle, using commercially available quantitative and semi-quantitative ELISA test kits, as well as by two commercial laboratories undertaking radioimmunoassay (RIA). The RIA, which was assumed to be the 'gold standard' in terms of reliability and accuracy, was the most expensive method when analysing more than one sample per week, and had the longest delay in obtaining results, but had minimal requirements for practice staff time. When compared with the RIA, the quantitative ELISA had a strong positive correlation (r=0.97, P<0.05) and a sensitivity and specificity of 70.6 per cent and 100.0 per cent, respectively, and positive and negative predictive values of 100.0 per cent and 71.0 per cent, respectively, with an overall accuracy of 90.0 per cent. This method was the least expensive when analysing five or more samples per week, but had longer turnaround times than that of the semi-quantitative ELISA and required more staff time. When compared with the RIA, the semi-quantitative ELISA had a sensitivity and specificity of 100.0 per cent and 95.5 per cent, respectively, and positive and negative predictive values of 73.9 per cent and 77.8 per cent, respectively, with an overall accuracy of 89.2 per cent. This method was more expensive than the quantitative ELISA when analysing five or more samples per week, but had the shortest turnaround time and low requirements in terms of staff time.