419. Using Automated Data for Objective and Consistent Definition and Case Identification of Nosocomial Clostridium difficile Infections (NCDI)
Session: Poster Session: Hospital-acquired and Transplant Infections
Friday, October 30, 2009: 12:00 AM
Room: Poster Hall A
Background: There is heightened interest in reporting and limiting reimbursement for NCDI. Little is known on the variations of the definition and case identification of NCDI across hospitals.
Methods: We analyzed electronically captured laboratory and census data from February 27, 2009 to March 23, 2009 across 55 US hospitals in the MedMinedTM Research Database. A NCDI case was defined as positive Clostridium difficile toxin assay specimen collected either on hospitalization day 3 or later (hospital onset CDI) or within 3 days of admission and with previous same hospital discharge within 30 days (community onset hospital associated CDI). Hospital infection control personnels in each of the 55 hospitals independently annotated the electronically identified NCDI cases.
Results: During the study period, a total of 141 NCDI cases were identified electronically. Among them, 103 (73%) were confirmed as NCDI cases and 38 (27%) were rejected by the hospital infection control personnels. The most frequent reason for rejections was that the case was community acquired (89%). Nevertheless, among all rejected cases, 57.9% had specimen collected 3 or more days after admission (hospital onset CDIs); and 42.1% were within 3 days of admission and with previous same hospital discharge between 5 to 27 days (community onset hospital associated CDIs).
Conclusion: Clostridium difficile toxin positive is widely accepted for CDI case definition. However, hospital infection control personnels varied when designating the nosocomial versus community CDI. Majority of rejected NCDI cases had specimen collected 3 or more days after admission, which were valid NCDI cases based on objective definition. Electronic surveillance incorporating real time laboratory and census data offers objective and uniform case definition and might be a more consistent and cost efficient approach when comparing NCDIs across hospitals.
Stephen Brossette, MD, PhD1, Christine Huizinga, MPH1, Xiaowu Sun, PhD2, Ying Tabak, PhD2, Daisy Wong, PhD1 and  S. E. Brossette,
Cardinal Health Role(s): Employee, Received: Salary.
C. Huizinga,
Cardinal Health Role(s): Employee, Received: Salary.
D. Wong,
Cardinal Health Role(s): Employee, Received: Salary.
X. Sun,
Cardinal Health Role(s): Employee, Received: Salary.
Y. P. Tabak,
Cardinal Health Role(s): Employee, Received: Salary., (1)Cardinal Health, Birmingham, AL, (2)Cardinal Health, Marlborough, MA