Abstract Background We assessed diagnostic agreement of BioFire® FilmArray® multiplex PCR against Seegene™ Allplex™ PCR for testing fecal samples collected during a diarrhea outbreak in resource-limited settings. Methods Fecal samples from consented British military personnel training in Kenya were collected without preservative and tested onsite by FilmArray®. Anonymized corresponding samples frozen near the point-of-care were tested 16-18 months later in the UK using Seegene™ (reference standard). We compared test sensitivity, specificity, and assessed agreement using Cohen’s kappa (k) coefficient. Results Samples from 60 individuals (80% male; median IQR age 24 22-28 years) were analyzed. Overall pathogen detection rates by the FilmArray® and Seegene™ were not significantly different (55/60 91.7% vs 53/59 89.8% p 0.9). Campylobacter spp. detection was significantly higher by Seegene™ (17/59 28.8% vs 6/60 10%; p = .032). FilmArray® sensitivity was moderate for Cryptosporidium spp. (65%; 95% CI, 45.37 - 80.77%), and low for Campylobacter spp. (35.3%, 95% CI 14.21 - 61.67%) and norovirus (7.14%, 95% CI 0.18 - 33.87%). Specificity was good to excellent for detection of Campylobacter spp., Cryptosporidium spp., Enteroaggregative E. coli and sapovirus. Conclusions The study shows moderate FilmArray® concordance with the Seegene™ in the detection of five enteropathogens, and poor to fair concordance for seven others, but high-quality case-control studies are needed to assess agreement between these platforms. However, based on performance characteristics including platform versatility and ease of use, and in the absence of a gold standard test, the FilmArray® remains a suitable NP-FE platform for diarrhea diagnostics in resource limited settings.
Toriro et al. (Fri,) studied this question.