Background: Blood cultures are part of the evaluation of hospital patients with fever. Patients with central lines in place, frequently have blood samples for culture drawn through lines. We sought to assess blood culturing practices at our institution. Methods: Retrospective review of BCs performed in hospitalized patients over a 12-month period (August 2018–July 2019) at an academic, tertiary-care center with 1,297 licensed beds and >62,000 admissions a year. A specialized phlebotomy team is involved in all peripherally drawn blood samples; however, the patient’s nurse obtains a blood sample through a central line. Results: Overall, 35,121 blood cultures were performed for an incidence rate of 106 BC per 1,000 patient days or 566 blood cultures per 1,000 admissions. Most blood samples (67%) were collected via peripheral venipuncture. We detected significant variation in culturing rates and the proportion of blood samples obtained through central lines among collecting units (Table 1). Overall, the blood culture contamination rate was 1.6%. Blood samples obtained through a central line had a higher contamination rate (2.2%) compared to samples obtained through peripheral venipuncture (1.3%; P < .0001). Blood culture rates were highest in intensive care units (ICUs) compared with other types of patient care units (Table 1). The blood culture positivity rate was significantly lower in ICUs (8.8%) compared with hematology-oncology (10%; HR, 0.88; CI, 0.80–0.96; P = .006), general medicine (10%; HR, 0.88; CI, 0.80–0.97; P = .013), and pediatrics (12%; HR, 0.74; CI, 0.59–0.92; P = .008). The ICUs had the lowest rate of BC contamination at 1.3%. Conclusions: Blood samples obtained through central lines for culture are more likely to be contaminated than peripherally drawn blood samples. Despite a relatively high rate of line-drawn blood samples for culture, ICUs had the lowest BC contamination rate, possibly reflecting high familiarity of ICU nurses with line draws. Blood samples collected through lines were most frequently performed in pediatrics and hematology-oncology, and these units had correspondingly higher rates of contamination. This information will be used to inform institutional guidelines on blood culturing and to identify ways to minimize blood culture contamination, which often results in additional testing and/or unnecessary antimicrobial use.

Funding: None

Disclosures: Consulting fee- Merck (Priya Sampathkumar)