Time to blood cultures positivity of microorganisms using a continuous-monitoring automated blood cultures system

This retrospective study was approved by the Research Ethics Committee Faculty of Medicine, Chiang Mai University (certificate of approval no. 272/2014).

A cross-sectional study was conducted among patients receiving care at Maharaj Nakorn Chiang Mai Hospital from October 1 to November 30, 2014. All patients who met the following criteria were consecutively enrolled: (1) age ≥15 years old, (2) had blood cultures drawn at the emergency department, outpatient department, and/or inpatient units, (3) blood culture grew organisms, and (4) medical records were available for review.

Demographic (e.g. age, sex, underlying diseases) and clinical characteristics including medical services, clinical of SIRS, source of infection, prior antimicrobial uses, organisms grew from blood cultures, and TTP for each blood culture bottle were collected. Patients with multiple episodes of infection were counted only once for analysis of demographic characteristics, but counted as the number of the episode for analysis of clinical characteristics.

Blood culture was obtained by trained health care personnel using 2% alcoholic chlorhexidine or 75% alcohol followed by 10% povidone–iodine as skin antisepsis for peripheral venipuncture. The aseptic technique was also performed in blood obtained from central venous/arterial catheters. At least 10 mL of blood was obtained for each bottle and inoculated into aerobic blood culture bottle using BD BACTEC™ Plus Aerobic/F culture bottles for bacteria processing on the BD BACTEC FX™ (Becton, Dickinson and Company, Franklin Lakes, New Jersey) at the central diagnostic laboratory, Maharaj Nakorn Chiang Mai Hospital. Anaerobic blood cultures were not routinely performed in our hospital. Each patient who had ≥2 blood culture bottles drawn should have been drawn at least 30 minutes apart, except for emergency conditions in which two blood cultures were drawn at the same time but from different venipuncture sites. Positive blood culture was reported immediately when the organism was detected, whereas negative blood culture was reported when no organism growth was observed after incubation for 5 days.

Episode of infection was defined as the first positive blood culture in a series and judged by investigators according to clinical information.

TTP is defined as the beginning time of the blood culture incubation until the positive result was detected by the automated system. It is the standard practice in our hospital that blood culture bottle must then be sent to the central diagnostic laboratory as soon as possible, and usually within 30 minutes after blood drawn.

True pathogen is considered for Staphylococcus aureus, Streptococcus pneumoniae, Streptococcus pyogenes, Streptococcus agalactiae, Listeria monocytogenes, Neisseria meningitidis, Neisseria gonorrhoeae, Haemophilus influenzae, Escherichia coli and other enterobacteriaceae, Pseudomonas aeruginosa, members of the Bacteroides fragilis group, all Candida species, and Cryptococcus neoformans [14, 15, 16].

Contaminant blood culture is considered if (1) 1 of 1 blood culture obtained or 1 of ≥2 blood cultures obtained grew coagulase-negative staphylococci, Corynebacterium spp., Micrococcus spp., Bacillus spp. not anthracis, or viridans group streptococci or (2) ≥2 blood cultures obtained within 48 hours grew coagulase-negative staphylococci, Coryne-bacterium spp., Micrococcus spp., Bacillus spp. not anthracis, or viridans group streptococci, in which the isolates were different, that is, different antibiograms [17]. However, in the presence of central venous or arterial catheters, the isolation of coagulase-negative staphylococci might be considered as a true pathogen by taking into account of the following factors: clinical manifestations, leukocyte count, imaging results, clinical course, response to therapy, and the absence or presence of other sites of infection.

Multiple isolates blood culture is defined if >1 microorganism was isolated from the same bottle or if two populations of the same microorganism with different antibiotic susceptibility patterns were identified [5]. Only cultures from which a single organism was recovered were used to analyze time-to-positivity.

SIRS, sepsis, and septic shock were followed those of Levy et al. and has been described elsewhere [18].

Data were presented in mean ± SD, median and IQR, number (%) as appropriate. Comparisons between groups were compared using Student’s t-test or Mann Whitney U test for continuous data, and c² test and Fisher’s exact test for categorical data. TTP between groups was analyzed using Kaplan–Meier estimates and log rank tests. Factors associated with TTP within 24 hours were tested in univariate models. Factors with the P < 0.10 from univariate analysis were then tested in a multivariate logistic regression model using backward stepwise procedure. Statistical significance was set as two-tailed, with P < 0.05. All statistical analyses were performed using Stata statistical software version 11.0 (StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX: Stata-Corp LP.).

There were 195 episodes of infection. The common clinical presentations were fever (73.3%), dyspnea (19%), and gastrointestinal symptoms (18%). Genitourinary tract infection was the most common source of bloodstream infection (15.4%), followed by gastrointestinal and hepatobiliary tract infection (14.9%) and catheter-related infection (12.8%). Blood cultures were taken while receiving antimicrobials in 55 episodes (28.2%). Clinical characteristics of 195 episodes of infections are shown in Table 2 and clinical parameters are shown in Table 3.

Among 506 blood cultures, 490 cultures grew single microorganism, 16 cultures contained 2 microorganisms. All 70 contaminant blood cultures contained Gram-positive bacteria which included coagulase-negative staphylococci (38 cultures), Bacillus spp. (22 cultures), Corynebacterium spp. (7 cultures), and Micrococcus spp. (3 cultures). The list of microorganisms and TTP for each microorganism is shown in Table 4. Sixteen multiple isolates blood cultures were excluded from the analysis of TTP.

Overall, the median TTP was 17 hours (IQR 11.5, 24.5), the medians TTP for Gram-positive and Gram-negative bacteria were 17.5 (IQR 12.5, 23) and 14 hours (IQR 10.5, 20), respectively (P = 0.004). The median TTP for yeast was 27.5 hours (IQR 22, 38). S. agalactiae had the shortest TTP, followed by Vibrio vulnificus, S. pneumoniae, and Streptococcus bovis with the medians TTP of 5.5 (IQR 5.5, 6), 8.5 (IQR 8.5, 8.5), 10 (IQR 5, 12), and 10 (8.5, 12.5) hours, respectively.

Among contaminant blood cultures, medians TTP are shown in Table 5. The medians TTP for true pathogens and contaminants were 16 (IQR 11.5, 23), and 23 (IQR 18.5, 28) hours, respectively (P < 0.001). The median TTP for coagulase-negative staphylococci which were true pathogens and contaminants were 18.5 (16.5, 20.5) and 22 (19.5, 25.5) hours, respectively.

Three-hundred and sixty-seven (74.9%) blood cultures grew within 1 day, 462 (94.3%) cultures grew within 2 days, 478 (97.6%) cultures grew within 3 days, 489 (99.8%) cultures grew within 4 days, and all 490 (100%) cultures grew within 5 days. Microorganisms with TTP >3 days were Micrococcus spp. (2 cultures), Microaerophillic streptococci (1 culture), Corynebacterium diphtheria (1 culture), Burkholderia cepacia (2 cultures), E. coli (1 culture), E. cloacae (1 culture), Candida albicans (1 culture), C. parapsilosis (1 culture), unidentified dematiaceous fungi (1 culture), and Talaromyces (Penicillium) marneffei (1 culture). The details of cumulative TTP for particular microorganisms are shown in Table 6. The median TTP for patients with and without clinical SIRS was 16 hours (IQR 11.5, 23.5), and 21 hours (16, 39.5), respectively (P = 0.002).

Comparing characteristics between blood cultures grew within 24 hours and after 24 hours were shown in Table 7. Blood cultures grew ≤24 hours were more likely drawn from patients who had hematologic malignancy, received immunosuppressive agents, altered mental status, endocarditis or vascular infection, clinical SIRS, high blood lactate, low white blood cell count, and low platelet count. On the other hand, blood culture grew >24 hours were more likely drawn from patients who had intravascular catheters, catheter-related infection, prior use of antimicrobials within 72 hours particularly carbapenems, vancomycin, and antifungals.

Multivariate analysis revealed that factors associated with TTP ≤24 hours were blood cultures drawn from patients with hematologic malignancy (OR 9.6, 95% CI 1.2, 74.3, P = 0.030), endocarditis and vascular infection (OR 8.7, 95% CI 1.1, 67.2, P = 0.038), thrombocytopenia (OR 2.4, 95% CI 1.3, 4.4, P = 0.004), had clinical signs of SIRS (OR 2.3, 95% CI 1.2, 4.5, P = 0.014), and not receiving antimicrobials within 72 hours before cultures were taken (OR 2.2, 95% CI 1.4, 3.6, P < 0.001).

TTP varied depends upon the pathogens and clinical settings. Although microorganisms were isolated from almost all of the blood cultures within 3 days of incubation, 2.4% of cultures would be missed. A 5-day incubation period per manufacturer instructions is still required.