Program Schedule

888
Performance Characteristics and Associated Outcomes for an Automated Surveillance Tool for Blood Stream Infection

Session: Poster Abstract Session: HAI Surveillance and Public Reporting
Friday, October 10, 2014
Room: The Pennsylvania Convention Center: IDExpo Hall BC
Posters
  • NorthShore Blood NIM IDweek 2014 Poster Final v2.pdf (431.2 kB)
  • Background: In the era of widespread electronic health records, a semi-automated surveillance tool could reduce the time and manual effort required to screen patients for central line-associated blood stream infection (CLABSI).

    Methods: Carefusion's MedMined is an automated surveillance tool that flags potential healthcare associated infections via the Nosocomial Infection Marker (NIM). We sought to determine the positive predictive value (PPV) of the blood NIM and to determine its association with outcomes of mortality, length of stay (LOS), and cost.  We reviewed records of 237 patients with blood NIMs using a gold standard of the National Healthcare Safety Network (NHSN) definition for primary BSI. We developed a propensity model to predict the probability of blood NIM during hospitalization. We matched cases with up to 5 non-cases by propensity score and exposure time. We estimated the attributable mortality, LOS, and cost impacts of NHSN-reportable CLABSI and non-NHSN-reportable BSI (i.e. not reportable because no central line or secondary to another infection) compared to non-cases.

    Results: Among patients with central lines present, the PPV of the blood NIM for CLABSI was 74.2%. For all patients (with or without central lines), the PPV for BSI was 53.6%.  77% of the 'false positive' NIMs met criteria for NHSN-defined infection other than BSI. Table 1 shows outcomes associated with blood NIMs vs. matched non-cases.

    Conclusion:  Although many of the cases detected by the tool were not NSHN-reportable, they were associated with adverse outcomes and higher costs, suggesting that detecting them may be an important component of a quality and cost containment program.

    Table 1: Outcomes for NHSN Reportable and Non-Reportable Cases Compared to Matched Non-Cases

    BSI

    Propensity and Exposure Matched Non-Cases

    P-value

    NHSN-Reportable CLABSI

    n

    57

    276

    Mortality, % (n)

    17.5% (10)

    9.4% (26)

    0.0976

    LOS, Median (Q1, Q3), days

    21 (11, 30)

    16 (11, 24)

    0.0298

    Total Charge, Median (Q1, Q3), $

    143,935 (89,794, 257,447)

    115,267 (74,937, 173,053)

    0.0098

    Days to NIM, Mean (SD), days

    11.5 (8.4)

    Non- Reportable BSI

    n

    89

    445

    Mortality, % (n)

    23.6% (21)

    6.7% (30)

    <0.0001

    LOS, Median (Q1, Q3), days

    14 (9, 20)

    10 (6, 17)

    0.0001

    Total Charge, Median (Q1, Q3), $

    86,927 (54,728, 156,669)

    62,929 (36,743, 115,693)

    <0.0001

    Jessica P. Ridgway, MD1, Xiaowu Sun, PhD2, Ying P. Tabak, PhD2, RS Johannes, MD, MS2 and Ari Robicsek, MD3, (1)Infectious Diseases & Global Health, University of Chicago, Chicago, IL, (2)Clinical Research, CareFusion, San Diego, CA, (3)NorthShore University HealthSystem, Evanston, IL

    Disclosures:

    J. P. Ridgway, CareFusion: Research Contractor, Research support

    X. Sun, CareFusion: Employee, Salary

    Y. P. Tabak, CareFusion: Employee and Shareholder, Salary

    R. Johannes, CareFusion: Employee and Shareholder, Salary

    A. Robicsek, CareFusion: Research Contractor, Research support

    Findings in the abstracts are embargoed until 12:01 a.m. EDT, Oct. 8th with the exception of research findings presented at the IDWeek press conferences.

    Sponsoring Societies:

    © 2014, idweek.org. All Rights Reserved.

    Follow IDWeek