425. Real-Time Antibiotic (ABX) De-escalation (DSCL) Decision Making Among Internal Medicine Post Graduate Physicians (PGPs): A Prospective Study in an Inpatient Geriatric Population
Session: Poster Abstract Session: Innovations in Medical Education
Thursday, October 8, 2015
Room: Poster Hall
Posters
  • Dobrzynski IDSA Poster.pdf (313.6 kB)
  • Background: Knowledge about DSCL practice and decision dynamics in PGPs can enhance ABX stewardship (AS) education initiatives. The aim of this study was to identify decision factors in Internal Medicine PGPs that influence ABX DSCL for geriatric patients (pts).

    Methods: Setting: 885 bed academic medical center. Subjects: Internal medicine PGPs (years 1-3) after informed consent. Design: Prospective, pilot study. Subjects were given a 5 question “pop-up” survey in the Computerized Physician Order Entry (CPOE) system when they prescribed an ABX to a patient ≥ 65 yrs. Questions included the suspected infection and Likert rating of ABX decision making factors. A 2nd survey was given at CPOE login 96 hours later to capture whether DSCL occurred and decision rationale.

    Results: 32 PGPs completed the 1st and 2nd survey on 224 and 217 (97%) pts, respectively. ABX were de-escalated in 124 (55%) pts (Table). When PGPs identified a high concern for bad outcome (Likert 4-5, N=128) or they were highly certain ABX were needed (Likert 4-5, N=144), 62 (48%) and 76 (53%) pts were de-escalated, respectively. DSCL occurred in only 61 (66%) pts when the PGP indicated that antibiotic DSCL would be likely at 72 hrs (N=93). Pts with urinary tract (UTI) and skin/soft tissue (SSSI) infections had the highest proportion of DSCL events. For 124 pts who had ABX DSCL, the presence of clinical improvement was cited by PGPs as the primary reason in 68 (55%) pts. PGPs indicated that supervising physicians were directly involved in ABX decision making for 154 (69%) pts. Of these, DSCL occurred in 94 (61%) pts.  

    Conclusion: Clinical improvement was the most cited reason for ABX DSCL. PGP concerns for bad outcome or certainty about need for ABX led to lower DSCL. Decision dynamics can be studied using CPOE embedded surveys and results may be used to improve PGP AS education. 

    Table

    Infection

    No. Treated

    Likely to de-escalate within 72 hrs*, N=224 (%)

    DSCL, N=224 (%)

    DSCL based on clinical improvement, N=124 (%)

    DSCL based on culture results, N=124 (%)

    Total

    224

    93 (42)

    124 (55)

    68 (55)

    34 (27)

    Pneumonia

    80

    39 (17)

    43 (19)

    26 (21)

    9 (7)

    UTI

    58

    32 (14)

    42 (19)

    18 (15)

    19 (15)

    Bacteremia

    22

    5 (2)

    9 (4)

    5 (4)

    0

    SSSI

    20

    8 (4)

    13 (6)

    10 (8)

    1 (1)

    Septic Shock

    18

    6 (3)

    9 (4)

    2 (2)

    6 (5)

    Other

    18

    3 (1)

    7 (3)

    4 (3)

    1 (1)

    *Initial Survey Likert 4-5/5

    David Dobrzynski Jr., MD1, Christopher Ohl, MD, FIDSA2, John Williamson, PharmD3, James Johnson, PharmD4, James Beardsley, PharmD4, Kevin High, MD, MS, FIDSA5, Jeff Williamson, MD6, Robert Hamm, PhD7 and Vera Luther, MD2, (1)Wake Forest Baptist Medical Center, Winston Salem, NC, (2)Department of Internal Medicine, Section on Infectious Diseases, Wake Forest School of Medicine, Winston-Salem, NC, (3)Department of Pharmacy, Wake Forest Baptist Medical Center, Winston-Salem, NC, (4)Department of Pharmacy, Wake Forest Baptist Health, Winston-Salem, NC, (5)IM-Section On Infectious Diseases, Wake Forest School of Medicine, Winston-Salem, NC, (6)IM-Gerontology and Geriatrics, Wake Forest School of Medicine, Winston-Salem, NC, (7)University of Oklahoma Health Sciences Center, Oklahoma City, OK

    Disclosures:

    D. Dobrzynski Jr., None

    C. Ohl, None

    J. Williamson, None

    J. Johnson, None

    J. Beardsley, None

    K. High, None

    J. Williamson, None

    R. Hamm, None

    V. Luther, 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.