Examination of the correlation between the manual and automated serological testing methods for syphilis

Abstract

We evaluated the correlation between the conventional manual serological testing method for syphilis and a novel automated serological testing method and between six different reagents used in the automated method. Twenty-six serum samples, which were positive on non-treponemal manual serological testing, were obtained from 19 patients with early syphilis. The samples were manually analyzed using the non-treponemal serological test for syphilis kit and automatically analyzed using six different reagents approved by the Ministry of Health, Labor and Welfare in Japan. Statistically significant correlations were observed between most of the reagents used in the automated testing (r = 0.652-0.996, P < 0.001), except for one combination of the reagents. In the simple regression analysis, the slope of the simple regression line (range, 0.014-3.040) and some of the regression coefficients were not equal to 1.0. Therefore, it is recommended that when the automated serological testing method is used to test for syphilis, the same reagent should be consistently selected to evaluate the changes in antibody titers. Statistically significant correlations were also observed between the manual method and all the reagents used in the automated method (r = 0.682-0.811, P < 0.001). In this case, the regression coefficients ranged 0.375-6.270, and the simple regression line intercept ranged -71.926 to 4.184. The regression coefficient and the intercept between the manual method and some of the reagents used in the automated method were not similar to the values described in the documentation attached to the reagents used in this study.