1899. Decreasing Numbers of Blood Cultures: Does it Matter?
Session: Poster Abstract Session: Clinical: Bacteremia and Endocarditis
Saturday, October 7, 2017
Room: Poster Hall CD
  • IDWeek poster ver 4 final .pdf (582.6 kB)
  • Background: Between Jan 2014 and Dec 2016, there was a 22.4% decline in the number of blood cultures (BC) s received in the UF Health Shands Hospital microbiology laboratory, and a 39.4% decrease from the Emergency Department (ED) over the same period.

    Methods: We studied 34,618 admissions from the ED between 7/1/11-12/31/16 who had a blood culture done within 48 hours of ED entry. Based on hospital discharge coding, 8,617 had sepsis present on admission (POA) and 26,001 were never septic. The time from ED admission to obtain a BC and the time to 1st intravenous (IV) antibiotic dose were analyzed by 6 or 12 h intervals up to 48 h. Outcome measures were mean hospital and ICU length of stay days (HLOS) and ICU LOS respectively.

    Results: For non-septic (but not for septic patients) both HLOS and ICULOS were strongly correlated with the time of BC draw. HLOS rose from 5.1d if the culture was obtained within 0-12h to 7.46d if delayed to 36-48h (r=0.982, p<0.005) and ICU LOS rose from 0.67d to 2.54d over the same time (r=0.987, p<0.005). Since the time to antibiotics could explain this effect, we looked at 12h intervals of antibiotic administration. For septic patients, the HLOS rose from 10.9d to18.6d (p < 0.001) and ICULOS rose from 3.5d to 9.5d if antibiotics were given at 0-12h vs 36-48h, (p < 0.001). For non-septic patients, the HLOS rose from 6.0d to 9.7d (p < 0.001) and ICULOS rose from 0.94d to 3.3d, (p < 0.001). To assess the role of getting a BC, we studied the effect of getting a blood culture over 48h within each 12h interval of antibiotic administration. Obtaining a BC in the 1st 12h was significantly associated both shorter HLOS and ICULOS for non-septic patients, but not for septic patients. For antibiotics given in 0-12h, HLOS was 5.46 d vs 6.77 if BC was done in 0 -12h vs 36 – 48h, ICULOS was 0.68d vs 1.44d for 0-12h and 36-48h respectively (t test p< 0.001 for both) and in-hospital mortality was 1.99% if BC was obtained within 12h vs 2.97% after that time, p = 0.0091.

    Conclusion: Delay in administration of IV antibiotics for ED admissions strongly increases HLOS and ICULOS for both septic and non-septic patients. Delay in drawing a BC has an independent effect for non-septic patients that correlated with longer HLOS, ICULOS and higher mortality.

    Kenneth Rand, MD1, Stacy Beal, MD2, Thomas Payton, MD3, Brandon Allen, MD4 and Kimberly Rivera, n/a4, (1)Pathology, University of Florida College of Medicine, Gainesville, FL, (2)Pathology, University of Florida, Gainesville, FL, (3)Emergency Medicine, University of Florida, Gainesville, FL, (4)University of Florida, Gainesville, FL


    K. Rand, None

    S. Beal, None

    T. Payton, None

    B. Allen, None

    K. Rivera, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 4th with the exception of research findings presented at the IDWeek press conferences.