Predictive Value of Blood Culture Time to Positivity

Nov. 16, 2018

Presence of bacteria in the blood (bacteremia) can be due to one of many factors. Some bacteremias are innocuous, such as transient bacteremia resulting from dental procedures or gastrointestinal biopsies. Other bacteremias are more concerning, such as organisms shedding from an abscess, catheter, or infected heart valve. In some cases, bacteremia can lead to sepsis, an immunological response to pathogens which is associated with significant mortality. As such, detection of bloodstream infection is one of the most important functions of the clinical microbiology laboratory.

Detection of Bloodstream Infection

Despite the potential severity of bloodstream infection, the overall organism burden is often exceptionally low. In many cases, organisms are present at less than one per milliliter of blood. Therefore, direct plating of blood to culture media is uninformative as there simply are not enough organisms to detect with this method. To get adequate numbers of organisms for diagnosis by standard methods, we need to grow them in broth to amplify their numbers.

A modern blood culture instrument with one drawer open to show loaded bottles.
A modern blood culture instrument with one drawer open to show loaded bottles.
Source: K.P. Smith

Modern blood culture instruments (Figure 1) still monitor gas production but perform automatic monitoring of bottles (Figure 2) without breaking sterility. Different manufacturers utilize slightly different mechanisms: BACTEC and BacT/ALERT monitor CO2 production using fluorescent and colorimetric substrates, respectively while VersaTrek measures change in gas pressure in the bottle. All instruments use proprietary algorithms to flag a bottle as positive at bacterial densities ranging from 1 x 108  to 1 x 109  (100 million to 1 billion) colony forming units/mL.

Aerobic (left) and anaerobic (right) blood culture bottles.
Aerobic (left) and anaerobic (right) blood culture bottles.
Source: K.P. Smith

This was originally done by directly inoculating blood into nutrient-rich media, then blindly subculturing at defined time points. However, this is labor intensive and involves sampling a bottle multiple times, potentially leading to increased contamination. So, in the 1980’s the first automated system was introduced which used radiolabeled carbon to detect CO2 production resulting from organism growth. This reduced labor required and allowed for automated monitoring, but direct sampling was still necessary, resulting in potential cross contamination - not to mention that generation of large quantities of biohazardous radioactive material is an occupational health professional’s worst nightmare.

Time to Positivity

When the instrument detects growth, it records the time between the bottle’s start of incubation and end time, termed the time to positivity. Time to positivity is ideally reported as days and hours to provide maximal information to the clinician. Importantly, this can occur automatically and requires no technologist input, essentially making it a “free” piece of data.

End time should ideally be the time that the instrument flagged positive rather than the time the bottle was removed from the instrument to avoid artificially extending time to positivity due to delay in processing. This is particularly important point for laboratories that do not process bottles on a 24-hour basis.

Time to positivity is in part related to the number of organisms initially inoculated into the bottle. For example, assuming a doubling time of 20 minutes for Escherichia coli and a threshold of ~5 x 108 colony forming units/mL for a positive blood culture, one organism would be detected in about 10 hours and 100 cells in about 7.5 hours. Although the difference in time to positivity cannot quantitatively indicate the number of cells initially in the blood, a longer time to positivity suggests lower initial organism burden.

However, other factors affect time to positivity including organism identity, volume of blood collected, presence of antibiotics in the blood, and use of resin or non-resin bottles to sequester antibiotics. One particularly important consideration when looking at time to positivity is the processing time of a blood culture. Studies show that a delay in processing can impact time to positivity, presumptively due to an extended lag phase resulting from storage at suboptimal temperatures. This factor may be of particular importance for centralized laboratories with significant specimen transport times.

Contamination vs. True Infection

Blood culture contaminants are often skin flora introduced during the blood draw. Whether an organism is a contaminant is commonly defined by the number of blood culture sets (or bottles within a set) that flag positive with skin flora. Conceptually, if only a few contaminating organisms are present, they are not likely to grow in every bottle drawn. Although lower inocula are associated with longer time to positivity for coagulase-negative staphylococci (common skin contaminants) there is no consensus on whether there time to positivity is predictive of contamination versus true infection.

Correlation with Clinical Outcome

It is logical to think that a relatively higher organism burden in blood may be associated with poor clinical outcomes. Indeed, reduced time to positivity has been associated with increased mortality in Staphylococcus aureus (≤12 hours), E. coli (<7 hours) bacteremia and Candida albicans (<24 hours) fungemia. Likewise, mortality is increased in S. aureus bacteremia when time to positivity of subsequent blood cultures is similar to the original, implying lack of effective treatment and failure to reduce organism burden.

Detection of Central Line-Associated Bloodstream Infection (CLABSI)

Time to detection may also assist in determining the origin of organisms found in blood, especially in the case of CLABSI. To make this determination, cultures of equal volume are simultaneously drawn peripherally and through a catheter. In theory, if the catheter is colonized with a biofilm, more organisms will be recovered from a culture drawn through the line than from a culture drawn by peripheral venipuncture, yielding a higher inoculum than in the peripherally drawn bottle.

Subtracting the catheter from peripheral time to positivity yields a “differential time to positivity” (DTTP). A DTTP of ≥2 hours shows good sensitivity and specificity for detection of CLABSI. Use of DTTP as part of CLABSI diagnosis is supported by the Infectious Diseases Society of America guidelines.

Should Time to Positivity Always be Reported?

The only guidelines that incorporate time to positivity are the IDSA guidelines for detection of CLABSI. However, reporting time to positivity for all cultures may lead clinicians to believe that it has well-established diagnostic value in all cases. Risk of misinterpretation or overinterpretation of data - for example, a clinically significant commensal organism may be mistakenly called a contaminant due to extended time to positivity - should be balanced with potential benefits of reporting. Alternatively, time to positivity may be reported only on request, providing the laboratory an opportunity to explain the limitations of incorporating this metric into clinical decision making.


Modern automated blood culture instruments can automatically report time to positivity, an essentially “free” test which may reflect the abundance of organisms in the inoculum. This data may be part of the puzzle to help clinicians identify the source of a bloodstream infection and may also be predictive of clinical outcome.

Want more #ASMClinMicro news?


The Clinical and Publich Health Microbiology Newsletter is a biweekly newsletter sent to ASM members (not a member? CLICK HERE TO JOIN). It requires a subscription, free with membership, which you can adjust HERE (make sure you look under the "Newsletters" tab). The CPHM Newsletter provides updates on ASM's clinical and public health microbiology products and services, as well as advocacy updates. It's a great way to keep abreast of all that is going on at ASM!  

The above represent the opinions of the author and does not necessarily reflect those of ASM.

Author: Kenneth (K.P.) Smith, Ph.D.

Kenneth (K.P.) Smith, Ph.D.
Dr. Kenneth (K.P.) Smith is Assistant Director of the Infectious Disease Diagnostics Laboratory at Children's Hospital of Philadelphia.