|Values are valid only on day of printing.|
Click CC to turn on closed captioning.
Published: February 2010Print Record of Viewing
Dr. Brad Karon discusses optimum timing of blood culture collections.
Presenter: Dr. Brad Karon
This Hot Topic presentation will be a little different from others that we have posted in the past. Each year Mayo Medical Laboratories sponsors a Phlebotomy Conference, and for the last few years I have given a case-based presentation at the Phlebotomy Conference called Phlebotomy Top Gun. During this session I present difficult cases in phlebotomy to the audience, and ask them to indicate via audience response system what their action or conclusion would be. I then review the data and evidence underlying the question, and at the end allow the audience to respond to the same question again. Cases are selected based on questions submitted by conference participants during the registration process. Today's presentation is one example of a case presented during a previous Phlebotomy Conference.
Before I begin the case I would like to let everyone know that this year's Phlebotomy Conference: "Phlebotomy: A new decade of progress and collaboration", was held from March 11-12, 2010 at the Siebens Medical Education Building on the Mayo Campus in Rochester, MN. Please visit the mayomedicallaboratories.com website for more information on available programs.
This is one case presented a few years ago during the Phlebotomy Top gun session. During the conference the audience had a chance to select the answer they felt was correct using an audience response system, and the percent of attendees indicating each answer was shown before we proceeded with the case.
This was the actual question submitted by a conference attendee during the registration process, so you can see that the cases are created directly from questions submitted by conference participants.
To answer the question I reviewed relevant data and evidence linking timing of blood cultures to yield or success. First of all some definitions. One culture equals one venipuncture for collection of blood. Blood cultures are usually collected in sets or pairs, meaning two or more separate venipunctures to collect one set of cultures.
Why do we generally perform two sticks to obtain blood cultures? In terms of yield, that is the chance that a septic or bactermic patient will have a positive blood culture, the number of sticks is not important. Overall volume cultured is the major factor that influences yield. However occasionally blood cultures can be contaminated with skin flora, making interpretation and treatment difficult. Comparing results between blood culture collections performed from separate venipunctures allows clinicians to determine whether contamination with skin flora may be responsible for a single positive blood culture.
To get to the issue of whether timing of blood cultures matters, I'm going to briefly review two studies that directly addressed this question. Li et al. compared a process whereby 60 cc of blood was cultured at one time to a process where 20 cc cultures were collected 2 hours or more apart.
They found that culturing more blood was better: between 20 and 40 cc of blood cultured the yield increased by 1% per extra cc cultured, and between 40 and 60 cc of blood cultured the yield increased by 0.5% per extra cc cultured. This reinforces the concept that blood culture yield is a function of volume of blood cultured. However timing did not matter, in other words culturing 60 cc at one time was just as good as drawing 3 20 cc blood cultures two or more hours apart.
A second more recent and very large retrospective study was done in 2008. This study specifically asked the question "Are blood cultures more likely to be positive if collected at or near the time that a patient's fever spikes?" The study was powerful because it included a large number of patients, large number of positive blood cultures, and wide variety of organisms causing bacteremia and sepsis. The study was designed to look at the yield of blood cultures collected within one hour of a fever spiking, and compare that to the yield of blood cultures collected either 2-24 hours before or after a fever peaked.
The study found that there was no relationship between culture yield, that is the chance that a bactermic or septic patient would have a positive blood culture, and the timing of the draw relative to fever spiking. Their results were consistent with another smaller study published earlier, and on the whole they recommended that blood cultures be collected when it is convenient to do so. Rather than focusing resources on the timing of blood cultures, the authors recommend focusing on aseptic technique and culturing an adequate volume of blood.
Since there is little good evidence that timing of blood cultures matters, and for many patients it is imperative to get an adequate volume of blood collected for culture before antibiotics are started, our current process in the emergency department is to collect two 30 cc blood cultures when sepsis is suspected, without any artificial delay or wait between collections. This allows us to collect an optimal volume of blood for pathogen detection, 60 cc, in a short period of time to allow for antibiotic treatment of high risk patients to begin after appropriate cultures are obtained.
During the conference each case ends by going back to the original question. The audience is polled again to see how they respond after hearing the data and/or evidence on the question. Evidence of learning is documented in real time, by seeing whether the distribution of answers changes after the information is presented. Finally each case ends with an indication of what I believe to be the correct answer.
In this case the best answer would be collect two 30 cc blood cultures nearly simultaneously at admission.