421. The Impact of ICD-9-CM Code Rank-Order on the Estimated Prevalence of Clostridium difficile Infection
Session: Poster Session: Hospital-acquired and Transplant Infections
Friday, October 30, 2009: 12:00 AM
Room: Poster Hall A
Background: National estimates of the Clostridium difficile infection (CDI) burden in the U.S. have utilized International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) codes assigned at hospital discharge. Given varying methods of ICD-9-CM code ascertainment by the National Hospital Discharge Survey (NHDS) and the National Inpatient Sample (NIS), it is important to understand whether the ICD-9-CM code rank-order impacts CDI prevalence estimates.
Methods: ICD-9-CM codes for CDI (00845) and dates of admission and discharge were collected from electronic hospital databases for adult patients admitted to 3 hospitals in the U.S. from 07/00-06/06. CDI rates per 1,000 discharges were calculated and compared for the NHDS and NIS definitions. For each discharged patient, the NHDS collects ICD-9-CM code data on < 7 diagnoses and the NIS collects data on 9-15 diagnoses. The number of ICD-9-CM diagnoses collected for the NIS varies by state and year (e.g., 15 in MO, 15 in OH, 9 in UT from 2000-2002, and 10 in UT from 2003-2006). CDI prevalence estimates were compared using the chi-square test. The median rank-order of CDI was calculated.
Results: 7522 CDI cases were included in the analysis, of which 1019 (14%) CDI cases received a primary ICD-9-CM code. Relative to the total CDI cases captured by ICD-9-CM codes in any position, the NHDS and NIS definitions captured 88% (n = 6624) and 98% (n = 7398), respectively. The CDI prevalence measured by the NIS definition was significantly higher than the CDI prevalence measured by the NHDS definition (12.5 vs 11.2 cases per 1,000 discharges; p < .001). Compared to the NHDS estimates, the annual NIS estimates were significantly higher in 4 out of 6 years of the study period. The median rank-order of the ICD-9-CM code for CDI was 3.
Conclusions: Compared to the NHDS definition, the NIS definition captured 12% more CDI cases and reported significantly higher CDI rates.
Anne Butler, MS, Washington University in St. Louis, St. Louis, MO, Erik Dubberke, MD, MSPh, Infectious Disease, Washington University School of Medicine, St. Louis, MO, Victoria Fraser, MD, Washington University School of Medicine, St. Louis, MO, Yosef Khan, MBBS, MPH, The Ohio State University Medical Center, Julie Mangino, MD, Ohio State Univ, Columbus, OH, Jeanmarie Mayer, MD, University of Utah, L. Clifford McDonald, MD, CDC, Atlanta, GA and  E. R. Dubberke,
Viropharma Role(s): Research Relationship, Received: Research Grant.
Merck Role(s): Consultant, Received: Consulting Fee.
Becton Dickinson Role(s): Consultant, Received: Consulting Fee.
A. M. Butler, None..
J. Mayer, None..
J. E. Mangino, None..
Y. M. Khan, None..
L. McDonald, None. 
V. J. Fraser,
Johnson & Johnson Role(s): Speaker's Bureau, Received: Consulting Fee.
Hawaii Chitopure Role(s): Consultant, Received: Consulting Fee.
Ancora Role(s): Consultant, Received: Consulting Fee.