940. Hospital Underutilization of Present on Admission Indicators for Hospital-Acquired Clostridium difficile Infection
Session: Poster Abstract Session: Clostridium difficile Infections: Epidemiology and Diagnostics
Friday, October 9, 2015
Room: Poster Hall
  • POA_PosterIDweek2.pdf (347.7 kB)
  • Background:

    Present on admission (POA) indicators are used by hospitals to record patient diagnoses that are hospital acquired (HA), and are used to identify HA Clostridium difficileinfections (CDIs) from administrative data sources. We analyze hospital utilization of POA indicators by comparing the proportion of CDI cases reported as not POA across time in California and New York.


    HCUP state inpatient databases from 2005-2011, for the states of California and New York, were used to identify inpatient visits containing a CDI diagnosis. CDI diagnoses were characterized based on the diagnosis being a principal diagnosis, secondary diagnosis, diagnosis marked POA, and/or diagnosis marked not POA. Incidence rates of these diagnoses were compared across hospitals and time in the states of California and New York.


    Across hospitals and states, we find substantial variation in the proportion of CDI cases identified as not POA. Hospital’s in New York reported a significantly greater (P<.001) proportion of CDI cases to be not POA. Before 2008, New York hospitals recorded roughly 23% of CDI cases as not POA, while only around 10% were recorded as not POA in California. Beginning in 2008, around 29% of CDI cases were recorded as not POA in New York, while around 25% were recorded as not POA in California. 

    The proportion of CDI cases recorded as a secondary diagnosis was not statistically different between the two states. In each year from 2005-2011, the proportion of secondary CDI cases differed by less than 0.5% between New York and California. In both states, there was a slight downward trend in the proportion of secondary CDI cases from 2005-2008. After 2007 roughly 73% of all CDI diagnoses were recorded as a secondary diagnoses in both states. 


    Because HA CDI cases should also be recorded as a secondary diagnosis, changes in HA-CDI incidence should be reflected by similar changes in secondary-CDI incidence. We find substantial variations in the reported incidence of CDI not POA across hospitals, states and time; however, we find no evidence of such variation within secondary-CDI incidence. These results suggest hospitals may be systematically underutilizing POA indicators when reporting HA CDI. Thus, POA indicators may be a poor marker for incidence of HA CDI across hospitals in different states and time.

    Aaron Miller, PhD, Department of Economics & Business, Cornell College, Mount Vernon, IA, Linnea Polgreen, PhD, Pharmacy Practice & Science, University of Iowa, College of Pharmacy, Iowa City, IA, Joseph Cavanaugh, PhD, Biostatistics, University of Iowa, College of Public Health, Iowa City, IA and Philip M. Polgreen, MD, Division of Infectious Diseases, Department of Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA


    A. Miller, None

    L. Polgreen, None

    J. Cavanaugh, None

    P. M. Polgreen, None

    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.