634. Longitudinal European Clostridium difficile Infection Diagnosis Surveillance Study (LuCID) Shows Effects of Place, Patient Age and Testing Method on CDI Reporting
Session: Oral Abstract Session: HAI Reporting: The Devil is in the Details
Thursday, October 8, 2015: 2:30 PM
Room: 5--AB
Background:

To use a systematic, prospective large scale sampling approach to investigate variability in C. difficile testing and positivity according to place and time and describe the profile of healthcare associated C. difficile infection (CDI) patients.

Methods:  

Sixty hospitals in 3 countries (France, Italy & UK, each n=20) provided data on C. difficile testing methodology, number of tests performed/month and demographics of CDI patients during 2014-2015. CDI case incidence and patient profiles were compared between hospitals using optimal testing algorithms (GDH/toxin or NAAT/toxin) versus sub-optimal methods (including but not limited to; NOT detecting toxin (e.g. NAAT alone) or ONLY detecting toxin (EIA alone)).

Results:

There were 5817 CDI cases with median age 75 years (IQR 60-84, range 0-104); however 30% of all cases were aged <65. Median age of cases diagnosed in hospitals using sub-optimal methods either NOT detecting toxin (77 yrs) or ONLY detecting toxin (81 yrs) were significantly older than those tested in hospitals using optimal methods (75 yrs, p <0.001).

The average number of CDI cases/10,000 patient bed days (pbds)/month (m)/hospital (h) was significantly higher for those hospitals using sub-optimal vs optimal methods (3.6 vs 2.0/10,000 pbd, p<0.001). For hospitals using sub-optimal methods, those NOT detecting toxin and those ONLY detecting toxin had the highest average CDI rates (5.1 and 4.0/10,000 pbds respectively, p <0.001).  The average number of recurrent CDI samples/10,000 pbds/m/h was higher for those hospitals using sub-optimal vs optimal methods (4.7 vs 3.2/10,000 pbds). For hospitals using sub-optimal methods, those NOT detecting toxin and those ONLY detecting toxin had the highest average rates of recurrence (8.0 and 4.8/10,000 pbds respectively).  The CDI positivity rate was 2.5x higher when methods NOT detecting toxin vs optimal methods were compared (14.1% vs 5.4%,p < 0.001), and 3.5x higher for toxin ONLY detection (18.6 vs 5.4%,p  <0.001).

Conclusion:

Low testing in younger patients and use of sub-optimal methods produces misleading CDI rate data. Hospitals with inclusive testing of all ages, combined with optimal laboratory diagnosis, have lower CDI rates, presumably via more informed infection prevention measures.

Kerrie Davies, MSc1, Georgina Davis, BSc1, Frédéric Barbut, PharmD2, Catherine Eckert, PharmD2, Nicola Petrosillo, MD3 and Mark Wilcox, MD1, (1)University of Leeds, Leeds, United Kingdom, (2)National Reference Laboratory for C. difficile, Paris, France, (3)National Institute for Infectious Diseases, Rome, Italy

Disclosures:

K. Davies, Sanofi-Pasteur: Independent Contractor , Research grant

G. Davis, Sanofi Pasteur: Independent Contractor , Research grant

F. Barbut, Sanofi-Pasteur: Independent Contractor , Consulting fee and Research grant

C. Eckert, Sanofi-Pasteur: Independent Contractor , Research grant

N. Petrosillo, Sanofi-Pasteur: Independent Contractor , Research grant

M. Wilcox, Sanofi-Pasteur: Independent Contractor , Consulting fee and Research grant

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.