Molecular, parasitological and clinical diagnosis of female genital schistosomiasis: A diagnostic yield and comparative accuracy study using latent class analysis.

BACKGROUND: Female genital schistosomiasis (FGS) is a neglected manifestation of Schistosoma haematobium infection whose diagnosis is restricted by the absence of a flawless "gold standard" diagnostic tool. This study compared the yield, accuracy, performance of polymerase chain reaction (PCR), urine egg microscopy, and colposcopy. METHODS: A cross-sectional study was conducted among 143 female participants from S. haematobium-endemic communities in the West Region of Cameroon between July and August 2024. Participants underwent urine microscopy, genital swab PCR, and colposcopy, and a two-class latent class analysis model was developed using the R package poLCA. RESULTS: PCR yielded the highest positivity rate (73.4%), followed by urine egg microscopy (29.4%) and colposcopy/Visual FGS (23.08%). Among women with visual FGS, 68.75% were PCR-positive and had a lower median Ct value (29.6) than PCR-negative women (32.8) by SYBR Green PCR. Within the LCA framework, PCR provided a baseline sensitivity of 100.00% (95% CI: 94.85%-100.00%) and a specificity of 95.20% (95% CI: 91.88%-98.52%). urine egg microscopy and colposcopy/visual FGS served as highly specific rule-out tools with specificities of 83.32% (95% CI: 62.75%-100.00%) and 83.33% (95% CI: 62.76%-100.00%), despite their lower sensitivities of 29.06% (95% CI: 17.5%-40.62%) and 38.27% (95% CI: 29.92%-46.62%) respectively. CONCLUSION: The LCA model confirmed that a negative result from traditional tools cannot rule out infection necessitating the use of complementary diagnostic tools. However, clinicians should interpret positive PCR outcomes with care, as molecular detection cannot distinguish between active infection, recent parasite exposure, or residual DNA from past infections.