INTRODUCTION: Rapid, highly sensitive and specific multiplex polymerase chain reaction-based stool assays for gastrointestinal pathogens (GI PCR) are increasingly being used alternatively to conventional stool culture. We investigated the concordance between simultaneous GI PCR and stool culture with an ova and parasite (O&P) exam in outpatients presenting with symptoms of infectious gastroenteritis. METHODS: We performed a cross-sectional study of outpatients who received a FilmArray GI PCR test for acute diarrhea at an academic medical center from September 2015 to February 2019 to identify patients who had a concomitant stool culture with an ova and parasite exam (conventional testing) at the same time, on the same stool sample. The primary outcome was detection of an infection on GI PCR or conventional stool testing. Correlation was evaluated using McNemar's test for pathogens detected on both tests. Other categorical variables were compared with Chi-square analysis. RESULTS: We identified 150 outpatients who received GI PCR and stool culture with an ova and parasite exam for an episode of acute gastroenteritis. 106 (71%) patients had a pathogen isolated on GI PCR for 144 total pathogens including 128 (88%) bacteria, 13 (9%) viruses, and 3 (2%) parasites; 21 (14%) patients had a pathogen isolated on conventional testing for 18 total pathogens including 9 (50%) bacteria and 9 (50%) parasites (Table 1). Multiple pathogens were found in 38 (26%) GI PCR tests. PCR testing most commonly identified Enteropathogenic Escherichia coli (EPEC), representing 42 (33%) positive PCR tests. Conventional testing most commonly identified Campylobacter jejuni with 13 (54%) positive tests. Of 28 total C. jejuni infections, 15 (54%) were positive only on PCR, 3 (10%) only on conventional testing, and 10 (36%) on both modalities, showing that conventional testing missed 54% of all infections (P = 0.008). Conventional testing missed 4/6 (67%, P = 0.125) Salmonella infections and 9/14 (64%, P = 0.0215) Yersinia infections, nor did it detect any viral or diarrheagenic E. coli infections. Overall, PCR detected 144 of 191 (75%) of possible pathogens whereas conventional testing detected 47 of 179 possible pathogens (26%). CONCLUSION: GI PCR testing identified multiple pathogens unidentified by conventional testing, such as enteric viruses and pathogenic strains of E. coli. Conventional testing missed 88% of enteric bacteria showing poor concordance between simultaneous GI PCR testing and stool culture with an ova and parasite exam.
Read full abstract