Résumé
Background Serological screening tests play a crucial role to diagnose gambiense human African trypanosomiasis (gHAT). Presently, they preselect individuals for microscopic confirmation, but in future “screen and treat” strategies they will identify individuals for treatment. Variability in reported specificities, the development of new rapid diagnostic tests (RDT) and the hypothesis that malaria infection may decrease RDT specificity led us to evaluate the specificity of 5 gHAT screening tests. Methods During active screening, venous blood samples from 1095 individuals from Côte d’Ivoire and Guinea were tested consecutively with commercial (CATT, HAT Sero- K -SeT, Abbott Bioline HAT 2.0) and prototype (DCN HAT RDT, HAT Sero- K -SeT 2.0) gHAT screening tests and with a malaria RDT. Individuals with ≥ 1 positive gHAT screening test underwent microscopy and further immunological (trypanolysis with T.b. gambiense LiTat 1.3, 1.5 and 1.6; indirect ELISA/ T.b. gambiense ; T.b. gambiense inhibition ELISA with T.b. gambiense LiTat 1.3 and 1.5 VSG) and molecular reference laboratory tests (PCR TBRN3, 18S and TgsGP; SHERLOCK 18S Tids, 7SL Zoon , and TgsGP; Trypanozoon S 2 -RT-qPCR 18S2, 177T, GPI-PLC and TgsGP in multiplex ; RT-qPCR DT8, DT9 and TgsGP in multiplex). Microscopic trypanosome detection confirmed gHAT, while other individuals were considered gHAT free. Differences in fractions between groups were assessed by Chi square and differences in specificity between 2 tests on the same individuals by McNemar. Results One gHAT case was diagnosed. Overall test specificities ( n = 1094) were: CATT 98.9% (95% CI : 98.1–99.4%); HAT Sero- K -SeT 86.7% (95% CI : 84.5–88.5%); Bioline HAT 2.0 82.1% (95% CI : 79.7–84.2%); DCN HAT RDT 78.2% (95% CI : 75.7–80.6%); and HAT Sero- K -SeT 2.0 78.4% (95% CI : 75.9–80.8%). In malaria positives, gHAT screening tests appeared less specific, but the difference was significant only in Guinea for Abbott Bioline HAT 2.0 ( P = 0.03) and HAT Sero- K -Set 2.0 ( P = 0.0006). The specificities of immunological and molecular laboratory tests in gHAT seropositives were 98.7–100% ( n = 399) and 93.0–100% ( n = 302), respectively. Among 44 reference laboratory test positives, only the confirmed gHAT patient and one screening test seropositive combined immunological and molecular reference laboratory test positivity. Conclusions Although a minor effect of malaria cannot be excluded, gHAT RDT specificities are far below the 95% minimal specificity stipulated by the WHO target product profile for a simple diagnostic tool to identify individuals eligible for treatment. Unless specificity is improved, an RDT-based “screen and treat” strategy would result in massive overtreatment. In view of their inconsistent results, additional comparative evaluations of the diagnostic performance of reference laboratory tests are indicated for better identifying, among screening test positives, those at increased suspicion for gHAT.