1064. Accuracy of Administrative Code Data for Clostridium difficile Infection: A Meta-Analysis
Session: Poster Abstract Session: Surveillance of HAIs: Evaluating National Strategy
Friday, October 4, 2013
Room: The Moscone Center: Poster Hall C
  • CDI poster upload.png (309.3 kB)
  • Background: Administrative code data (ACD), such as ICD-9CM, are widely used for research and quality-assessment purposes.  Currently the Centers for Diseases Control and Prevention National Healthcare Safety Network (NHSN) is utilizing microbiology results as a cornerstone for the surveillance for Clostridium difficile infection (CDI), but using ACD may be more efficient.  Our aim is to summarize the evidence for accuracy of ACD for the diagnosis of CDI, compared to current surveillance method.

    Methods: Data Sources: Electronic literature search of PubMed, EMBASE, Web of Science, and CINAHL (from 1/1/1979 to 12/31/2012). Study Selection: Observational studies that met following inclusion criteria: (1) Reporting accuracy of ACD to detect included CDI; (2) Using ICD-9CM (008.45: intestinal infection due to C. difficile) or ICD-10 (A04.7: enterocolitis due to C. difficile); (3) Using microbiologic test results as reference standards. Data Synthesis: After screening 1548 published studies, we included 7 studies in our final analysis.  Two reviewers independently extracted information on study design, populations, ACD, reference standard method, and accuracy of diagnosis.  Statistical calculations were performed by R version 3.0 with mada package.

    Results: There was uniformly high specificity (range: 0.99-1.00), but sensitivity varied widely from 0.36 to 0.98 with pooled sensitivity 0.760 (95%CI: 0.562-0.887).  The positive predictive values ranged from 0.679 to 0.943.  Three studies also evaluated accuracy of ACD for nosocomial CDI (pooled sensitivity 0.861, 95%CI: 0.609-0.961).  Four studies used ELISA for toxin A/B as the reference, and other three studies had mixed microbiologic methodologies.  There was wide heterogeneity among studies in settings and microbiologic methods.  Also, there was no study which utilized Clostridium difficile PCR as a reference method.

    Conclusion: We estimate ACD has moderate sensitivity and very high specificity for CDI, compared to microbiologic methods.  However, positive predictive value can still potentially be low considering current prevalence of CDI, and may not be appropriate as a screening tool.  ACD may have better sensitivity for nosocomial CDI.    Further studies are needed to compare ACD to new diagnostic technologies, such as Clostridium difficile PCR.

    Michihiko Goto, MD1,2, Michael Ohl, MD, MSPH1,3, Marin Schweizer, PhD3,4 and Eli Perencevich, MD, MS, FIDSA, FSHEA3,5, (1)Division of Infectious Diseases, Department of Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA, (2)VA National Quality Scholars Program, Iowa City VA Health Care System, Iowa City, IA, (3)Center for Comprehensive Access and Delivery Research and Evaluation (CADRE), Iowa City VA Health Care System, Iowa City, IA, (4)Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA, (5)Department of Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA


    M. Goto, None

    M. Ohl, None

    M. Schweizer, None

    E. Perencevich, None

    Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.