By Stan Deresinski, MD, FACP, FIDSA
Clinical Professor of Medicine, Stanford University
SYNOPSIS: In addition to identifying several patient risk factors for contamination of blood culture specimens, the authors also highlighted various adverse clinical and financial adverse effects.
SOURCE: Klucher JM, Davis K, Lakkad M, et al. Risk factors and clinical outcomes associated with blood culture contamination. Infect Control Hosp Epidemiol 2021; Apr 26:1-7. doi: 10.1017/ice.2021.111. [Online ahead of print].
Klucher and colleagues examined risk factors for blood culture contamination as well as clinical outcomes associated with such contamination at a single center in Little Rock, AR, in 2014-2018. Only 2% of specimens for culture were drawn through central venous access devices. During that time, 1,504 (10.9%) of 13,782 blood cultures were true positives. Of the remaining, 1,012 (7.3%) were considered contaminated and 11,266 (81.7%) were negative — these served as cases and controls, respectively.
Multivariate analysis identified the following independent risk factors for contamination: increasing age, Black race, increasing body mass index (BMI), chronic obstructive pulmonary disease (COPD), paralysis, and sepsis with septic shock on presentation. The presence of metastatic cancer was protective. “Code sepsis” cases were associated with a numerically greater risk of contamination.
Blood culture contamination was associated on multivariate analysis with a one day longer length of hospital stay (7.9 days vs. 6.6 days), greater duration of antibiotic administration (6.2 days vs. 5.2 days), greater hospital charges ($35,008 vs. $28,875), higher rate of acute kidney injury (26.7% vs. 26.3%), higher frequency of ordered transthoracic echocardiograms (27.4% vs. 19.2%), and increased in-hospital mortality (8.0% vs. 4.6%). Individual antibiotics were not included in the multivariate model, but, on univariate analysis, blood culture contamination was associated with more frequent and prolonged vancomycin use.
It is no surprise that blood culture contamination leads to unnecessary antibiotic administration, but, as reported here, it leads to an additional array of negative effects. These include the ordering of additional laboratory tests, such as further blood cultures and transthoracic echocardiograms, the occurrence of acute kidney injury, prolongation of hospital stay, and increased costs. Another effect in some cases is unnecessary hospitalizations in patients discharged from the emergency department and then recalled when the laboratory reports the blood culture result.
The authors pointed out that identification of patient characteristics associated with an increased risk of blood culture contamination may allow focusing on them in attempts to reduce the rate. Among the methods employed to avoid contamination is the use of phlebotomy teams. However, in this study, and as is true at many hospitals, blood cultures were obtained by nurses in the emergency department. Another method is avoidance of central line draws. In this study, only 2% were obtained from this site. Enforcement of the use of sterile collection techniques is critical, and feedback of contamination rates to individual phlebotomists may be useful.
Another approach for which there is evidence of benefit is the use of diversion devices, which are a response to evidence that it is the initial 1 mL to 2 mL that often is the source of culture contamination. Such a device actively diverts and sequesters a small amount of the initial recovered blood, with blood for culture collected via a separate flow path.