Performance of a rapid diagnostic test for the detection of Cryptosporidium spp. in African children admitted to hospital with diarrhea.

<h4>Background</h4>Cryptosporidium is a protozoan parasite that causes mild to severe diarrhoeal disease in humans. To date, several commercial companies have developed rapid immunoassays for the detection of Cryptosporidium infection. However, the challenge is to identify an accurate, simple and rapid diagnostic tool for the estimation of cryptosporidiosis burden. This study aims at evaluating the accuracy of CerTest Crypto, a commercialized rapid diagnostic test (RDT) for the detection of Cryptosporidium antigens in the stool of children presenting with diarrhoea.<h4>Methods</h4>A cross-sectional study was conducted in four study sites in Sub-Saharan Africa (Gabon, Ghana, Madagascar, and Tanzania), from May 2017 to April 2018. Stool samples were collected from children under 5 years with diarrhoea or a history of diarrhoea within the last 24 hours. All specimens were processed and analyzed using CerTest Crypto RDT against a composite diagnostic panel involving two polymerase chain reaction (PCR) tests (qPCR and RFLP-PCR,) as the gold standard.<h4>Results</h4>A total of 596 stool samples were collected. Evaluation of the RDT yielded a very low overall sensitivity of 49.6% (confidence interval (CI) 40.1-59.0), a specificity of 92.5% (CI 89.8-94.7), positive predictive value of 61.3% (CI 50.6-71.2), and negative predictive value of 88.5% (85.3-91.1) when compared to the composite reference standard of qPCR and RFLP-PCR for the detection of Cryptosporidium species. Moreover, the performance of this test varied across different sites.<h4>Conclusion</h4>The weak performance of the studied RDT suggests the need to carefully evaluate available commercial RDTs before their use as standard tools in clinical trials and community survey of Cryptosporidium infections in pediatric cohorts.

Abstract licence: CC BY

28 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.