ABSTRACT
OBJECTIVES: We propose to analyse the positive and false-positive results of treponemal and nontreponemal tests in blood donors from Brazil and to evaluate possible factors associated with the results of treponemal tests. BACKGROUND: Treponemal tests have been used widely for syphilis screening in blood banks. The introduction of these tests in donor screening has caused an impact and a loss of donors who need to be assessed. METHODS: This was a retrospective cross-sectional study of syphilis screening and confirmatory test results of blood donors that were obtained before and after adopting a chemiluminescent immunoassay (CLIA). A comparative analysis was performed using a second sample drawn from positive donors. The possible factors associated with CLIA-positive or CLIA-false-positive results were investigated in a subgroup. Statistical tests were used to compare the proportions and adjusted estimates of association. RESULTS: The reactivity rate increased from 1·01% (N = 28 158) to 2·66% (N = 25 577) after introducing the new test. Among Venereal Disease Research Laboratory (VDRL)- and CLIA-confirmed results, the false-positive rates were 40·5% (N = 180) and 37·4% (N = 359), respectively (P = 0·5266). Older donors (OR = 1·04; P = 0·0010) and donors with lower education levels (OR = 6·59; P = 0·0029) were associated with a higher risk of positivity for syphilis. CONCLUSIONS: CLIA represents an improvement in blood bank serological screening. However, its use in a healthy population appears to result in high rates of false positives. Identifying which characteristics can predict false positives, however, remains a challenge.