The impact of PCR on Clostridium difficile detection and clinical outcomes.

PCR has increasingly replaced toxin A and B enzyme immunoassay (EIA) for the testing of Clostridium difficile infection (CDI). This study evaluated the clinical outcomes of CDI and disease epidemiology since the introduction of PCR. Clinical data and outcomes for patients admitted to a tertiary care centre during 2003 to 2012 were extracted using electronic medical records. Outcomes and incidence of disease were compared between types of CDI testing. In total, 15.6% of 108,092 patients admitted were tested for CDI. Among patients tested, 6.1% had positive results. The mean number of tests performed per 1000 admissions by EIA and PCR was 257.4 and 162.6, respectively. A total of 8.2% of PCR tests were positive compared to 5.0% of EIA tests (P < 0.001). The number of tests performed has decreased and the proportion of positive tests increased since PCR introduction. CDI incidence has remained constant. Only albumin (3.09 vs 3.24 g dl(-1), P 0.002) and inflammatory bowel disease (2.6 vs 7.0%, P < 0.001) status differed between the EIA and PCR groups. While hospital mortality did not differ, patients diagnosed by PCR had a shorter median length of stay (10 vs 8 days, P 0.004). Since PCR testing began, less CDI tests have been performed, but the proportion of positive results has increased. The incidence of CDI has remained constant, suggesting no change in disease epidemiology. The length of stay was shorter in the PCR group, reflective of either earlier detection and quicker onset of therapy or detection of less severe disease. Mortality did not change since the introduction of PCR.