Prevalence and Diagnostic Performance of Molecular and Serological Tests for Strongyloides stercoralis Infection in Immunosuppressed Patients from North India

The infection burden of Strongyloides stercoralis in India remains poorly defined. We aimed to estimate the burden of S. stercoralis infection among immunocompromised patients at our tertiary care hospital in north India and evaluated the performance of multiple diagnostic tests. Paired stool and blood samples were collected from 498 immunosuppressed individuals. Screening included stool microscopy without concentration, Baermann's funnel method, Agar plate and Harada Mori cultures, 18S ribosomal RNA gene real-time polymerase chain reaction (RT-PCR), and anti-Strongyloides IgG antibody ELISA. Bayesian latent class analysis (BLCA) was performed to estimate diagnostic accuracy without assuming a gold standard. Overall, 16.9% (n = 84/498) of immunosuppressed individuals were positive by a composite reference standard, most commonly detected by a combination of ELISA and RT-PCR. Among single tests, BLCA estimated ELISA sensitivity of 74.5% (95% credible interval [CrI]: 62.7-84.5) and specificity of 88.3% (95% CrI: 85.4-90.9). Culture and RT-PCR showed sensitivities of 62.9% (95% CrI: 50.0-75.1) and 62.6% (95% CrI: 50.2-74.6) and specificities of 99.8% (95% CrI: 99.4-100.0) and 95.5% (95% CrI: 93.6-97.2), respectively. ELISA plus RT-PCR demonstrated the best performance with 87.5% (95% CrI: 76.9-95.0) sensitivity and 89.0% (95% CrI: 83.4-94.7) specificity. Strongyloides stercoralis infection was most frequent among patients with hypereosinophilia syndromes (33.3%), and increasing absolute eosinophilia count independently predicted infection. This study highlights a high burden of S. stercoralis infection in immunocompromised patients in north India and underscores the utility of combining serological and molecular assays for optimal detection in high-risk populations.