PTH-169 Clinical Utility of a Rapid PCR Assay and Traditional CCNA to Detect Clostridium Difficile Infection (CDI) – Comparison to Clinical Diagnosis

• Published in: Gut Vol. 62; no. Suppl 1; p. A280
• Main Authors: Berry, N, Sewell, B, Jafri, S, Vagia, S, Puli, C, Rees, E, Lewis, A, Isaac, M, Ch’ng, C L
• Format: Journal Article
• Language: English
• Published: London BMJ Publishing Group Ltd and British Society of Gastroenterology 01.06.2013; BMJ Publishing Group LTD
• Subjects: Cell culture; Clostridium difficile; Colitis; Cytotoxicity; Diagnosis; Diarrhea; Hospitals; Infections; Laboratories; Medical diagnosis; Parathyroid hormone; Pseudomembranous colitis; Toxin A
• ISSN: 0017-5749; 1468-3288
• DOI: 10.1136/gutjnl-2013-304907.656

Introduction Clostridium difficile causes nosocomial antibiotic associated diarrhoea, with a range of mild to severe disease, pseudomembranous colitis, toxic megacolon and potentially fatal outcome. The diagnosis of this disease in clinical laboratories has traditionally been performed by cell culture cytotoxin neutralisation assay (CCNA) or by toxin A/B detection using EIA. The routine test in our laboratory is CCNA, which takes 24–48 (–72) hours; it is labour intensive, requires specialist facilities, expertise and is not done out of hours. There are other rapid tests now available including GDH as well as molecular (real-time PCR) assays. Methods This study was designed to assess the clinical relevance of a fully automated, random access PCR assay, Xpert C. difficile, for rapid identification of C. difficile infection (CDI) in comparison to clinical diagnosis as the reference method. During March to September 2011, 1040 samples from inpatients in 2 hospitals, with suspected CDI, were prospectively tested by routine cell culture cytotoxin neutralisation assay (CCNA), PCR (GeneXpert, Cepheid), and a GDH/Toxin EIA (Premier, Launch Diagnostics). Cytotoxicity was assessed after 24 and 48 hours. All PCR positive patients (and controls) were reviewed by a multidisciplinary team (Gastroenterologist, Microbiologist, infection control nurse, requesting staff). Results C. difficile detection rates were 10.8% (PCR), 6% (CCNA) and 13.8% (GDH). 974/1035 (94.1%) samples showed concordant CCNA and PCR results, 89% (886/985) were concordant for CCNA, PCR and GDH and 94.4% (930/985) showed concordance between GDH and PCR. With clinical diagnosis as a reference, PCR was 99.1% sensitive, 98.9% specific, with PPV 91.9% and NPV 99.9%. Surprisingly, CCNA on a single sample was only 51% sensitive, 99.4% specific, PPV was 91.9%, NPV 94.3%. GDH sensitivity was 83.8%, specificity 94.5%, PPV 64.7% and NPV 97.9%. 59 more samples were positive by PCR than CCNA (62); 54/59 were clinically CDI. Conclusion We found PCR to be a more sensitive method than CCNA and GDH (sensitivity 83.8%) for the detection of C.difficile infection (CDI). In contrast to using CCNA or an algorithm that includes GDH, the use of Xpert C.difficile PCR allows us to provide accurate and rapid (mostly same day) results to the clinicians. Disclosure of Interest N. Berry Grant/Research Support from: Non promotional Educational research grant from Cepheid, B. Sewell Grant/Research Support from: Non promotional Educational research grant from Cepheid, S. Jafri Grant/Research Support from: Non promotional Educational research grant from Cepheid, S. Vagia Grant/Research Support from: Non promotional Educational research grant from Cepheid, C. Puli Grant/Research Support from: Non promotional Educational research grant from Cepheid, E. Rees Grant/Research Support from: Non promotional Educational research grant from Cepheid, A. Lewis Grant/Research Support from: Non promotional Educational research grant from Cepheid, M. Isaac Grant/Research Support from: Non promotional Educational research grant from Cepheid, C. L. Ch’ng Grant/Research Support from: Non promotional Educational research grant from Cepheid