When is the Best Time to Obtain Blood Cultures from My Potentially Septic Patient?

Abstract & Commentary

Ellen Jo Baron, PhD, D(ABBM), Professor of Pathology and Medicine, Stanford University. Medical School Director; Clinical Microbiology Laboratory, Stanford University Medical Center Dr. Baron reports no financial relationships relevant to this field of study. This article originally appeared in the November 2008 issue of Infectious Disease Alert. It was edited by Stan Deresinski, MD, FACP, and peer reviewed by Connie Price, MD.

Many physicians have followed the historical practice of ordering blood cultures to be drawn as close as possible to the time of the peak of the febrile episode (fever spike).1-5 In the absence of prescient knowledge of this moment, physicians order blood cultures to be drawn at intervals ranging from 30 minutes to 2 hours. A paper by Jaimes et al suggested that many factors, other than fever, such as shaking chills, WBC counts, hypotension, and more were needed to better predict whether a patient was experiencing bacteremia.3

For many years, the only data comparing the yield of blood cultures in relationship to the patient's fever was from a study that was presented as an abstract but never officially published. Dr. Richard Thomson performed the study when he was a new microbiology laboratory director in Akron, Ohio, soon after leaving his post-doctoral fellowship at the Mayo Clinic.6 The results were presented in the abstracts of the 1989 American Society for Microbiology Annual Meeting and included in an American Society for Clinical Pathology Check Sample exercise distributed in 1991.6 Only a few contemporary microbiologists ever even had a copy of the report. Thomson et al looked at numbers of clinically relevant positive blood cultures obtained during four different time periods relative to a patient's fever spike.

Although there was a trend toward more true positive blood cultures being obtained in the period directly before the fever spike, there were no statistically significant differences among the four time periods.

These data served as the basis for most microbiologists' recommendation to obtain all of the blood cultures as soon as a patient becomes febrile, without any time period between draws. A 1994 publication by Li et al basically corroborated the Thomson results.4 The 1994 study showed that the yield of clinically significant blood cultures performed during a 24-hour period did not vary whether the blood was obtained all at once or over a period of several time intervals. Without stronger data, physicians have continued to pursue their idiosyncratic blood culture ordering practices. By asking phlebotomists to obtain blood cultures at intervals spanning several hours, unnecessary additional time is spent in the process and the overall cost and inefficiency of procuring blood cultures is increased. And if antibiotic therapy is withheld while blood cultures are being obtained, patient care also suffers.

Dr. Gary Doern set out to perform the definitive study to answer the question without nuance. He enlisted the aid of six additional medical centers in addition to his own, University of Iowa. Workers collaborating from Geisinger Medical Center (Danville, PA), VA Boston Healthcare System, Johns Hopkins University School of Medicine, Barnes-Jewish Hospital Washington University School of Medicine (St. Louis), University of Texas Health Science Center in San Antonio, TX, and the VA Medical Center (Portland) enrolled 1,436 adult patients with clinically significant episodes of bacteremia and fungemia during 2006.5 For each patient enrolled, the workers noted the time at which the highest temperatures were recorded in both the 24 hours preceding and those following the time that the first positive blood culture was obtained, as well as the temperature of the patient recorded closest to the time of that blood culture. Clinical relevance was determined by criteria in place at each medical center. The patients were two-thirds male, average age 59 years, and their blood cultures grew a variety of microorganisms, including 54% gram-positive bacteria such as staphylococci (38%, 42% of which were coagulase negative) and Enterococcus (10%); 38% gram-negative bacteria such as Enterobacteriaceae (> 23%) and Pseudomonas aeruginosa (4%); 3% anaerobic bacteria; and 5% yeast.

The highest recorded fevers, determined as the one of the three temperatures that was 0.5° C higher than the other two, occurred during the time of the blood culture draw in 44% of episodes. It was noted that 10%-31% of maximum fevers occurred before or after the blood draw in the remaining patients. In general, none of the results were statistically significantly different from each other.

In addition, no significant associations were found between temperatures of patients and their genders, white blood counts, or even when organism types were evaluated. Unfortunately, not enough cultures yielded fungi to allow reliable statistical analysis. One caveat was that for patients 18- to 30-years-old, the maximum temperature was significantly more likely to occur one to < 24 hours after the first positive blood culture. For other age groups (majority of patients enrolled), there were no differences.

Riedel et al concluded that the best practices for col-lecting blood cultures are to obtain enough blood volume (recent studies summarized in the ASM Cumitech and the CLSI guideline on blood cultures have suggested from 40-60 mL), to obtain suitable numbers of separate blood cultures (at least two), and to use stringentaseptic technique to avoid contamination.

References Cited:

1. Baron EJ, et al. 2005.Cumitech 1C. Blood cultures IV. EJ Baron, Coord. editor. In: Cumitech Series, ASM Press, Washington, DC.

2. Clinical and Laboratory Standards Institute (CLSI). Principles and Procedures for Blood Cultures: Approved Guideline. 2007. CLSI document M47-A. Clinical and Laboratory Standards Institute, Wayne, PA.

3. Jaimes F, et al. 2004. Predicting bacteremia at the bedside. Clin Infect Dis. 2004;38:357-362.

4. Li J, et al. 1994. Effects of volume and periodicity on blood cultures. J Clin Microbiol. 32:2829-2831.

5. Riedel S, et al. Timing of specimen collection for blood cultures from febrile patients with bacteremia. J Clin Microbiol. 2008;46:1381-1385.

6. Thomson, RB, et al. 1989. Timing of blood culture collection from febrile patients. Abstr. C-227. Abstr. 89th Annual Meeting Amer. Soc. Microbiology, Washington, DC.