The speed and accuracy of preliminary blood culture reports impacts patient management and outcomes.
To evaluate the accuracy and timeliness of preliminary blood culture results among multiple laboratories.
Q-Probes participants collected turnaround time (TAT) data on preliminary Gram stains, compared accuracy of up to 100 preliminary to final culture Gram stain results, and described blood culture laboratory practices.
Sixty-four laboratories and 5031 blood cultures were evaluated. All participants used continuously monitoring blood culture systems. Median TAT from initial growth detection to notification of results was 45 minutes, with the longest component being preparation of Gram stains (median time = 25 minutes). Participants (N = 40) reporting a continuous schedule for processing blood cultures had significantly lower overall TAT (medianm= 37 minutes) compared with 15 participants with intermittent processing schedules (medianm= 124 minutes),mPm= .003. Time to complete Gram stain processing was lower (median time = 21 minutes) for 39 participants using continuous processing schedule compared with 14 others (median timem= 67 minutes),mPm= .03. Goals for total TAT were used by 27 of 56 participants (48.2%). Having goals did not significantly affect TAT. A total of 4962 of 5021 Gram stain results (98.8%) agreed with final culture results. The highest discrepancy rates occurred among gram-positive bacilli (20 of 335; 6.0%) and mixed cultures (22 of 106; 20.8%).
This study provides benchmarks for assessing blood culture quality performance. Timeliness and accuracy of preliminary blood culture reports were excellent. However, nearly one-third of laboratories did not process blood cultures continuously. This significantly prolonged reporting results, which could affect patient outcomes.
Sepsis is an important cause of serious morbidity with a high risk of mortality. Nearly all cases are associated with bacteremia. Immediate intervention with broad-spectrum empirical antimicrobial therapy and supportive treatment has been shown to have the greatest impact on outcomes.1,2 The next most important factor is timeliness of reporting preliminary blood culture results. Faster results reporting has been shown to be associated with lower mortality,3 especially when combined with clinical consultation.4,5 Other benefits of rapid reporting include improved selection of antimicrobial therapy,6–9 lower overall health care costs,10 and reduced length of stay.8,11 Timely detection of bacteremia also facilitates initiation of rapid, direct methods for early identification and susceptibility testing.12–14 Final culture and susceptibility reporting, which ordinarily occurs several days after onset of bloodstream infection, has been shown to have less impact on patient management and outcomes than preliminary Gram stain results.15 Because rapid reporting of preliminary blood culture is linked to patient outcomes, it is generally considered standard laboratory practice to handle positive blood cultures as critical values that must be reported as soon as possible.16
Reporting preliminary blood culture results presents challenges not present in the handling of other types of critical results. Unlike most critical values, which are reported soon after specimen collection, blood culture results may be delayed for many hours or even days. Reporting preliminary positive blood cultures requires both extra processing (preparing Gram smear) and results interpretation (reading Gram smear) before critical results can be reported. With continuous-monitoring blood culture systems, a positive result may occur at any time of day or night. However, laboratory staff able to process blood cultures and interpret Gram stain results may not always be present when a positive blood culture is initially detected.17 Furthermore, careful preparation and accurate interpretation of the Gram stain is vital to avoid errors that may lead to inappropriate treatment.18 Also, particularly at night, the provider who ordered a blood culture may no longer be primarily responsible for the patient's care, so that reaching an accountable clinical staff member who can receive and act on results may be another potential obstacle affecting timeliness of reporting results.
Because of the established impact on outcomes associated with timely and accurate processing and reporting of preliminary blood culture Gram stain results, this study aimed to establish benchmarks for laboratories to apply and assess performance of this critical test.
This observational study was conducted according to the Q-Probes format as previously described, which involved laboratories that subscribe to and participate in the program.19 Instructions and data collection materials are provided by and returned to the College of American Pathologists by a specified time. After data analysis, subscribers receive an individual detailed report about their performance on quality indicators benchmarked against other participants. The range of performance among participants, summary data analysis of significant associations with demographics, and practice variables and critique are reported to each participating laboratory.
Participants were instructed to collect information on 100 consecutive positive blood culture cases or for 90 days, whichever came first. For each positive blood culture, patient locations (inpatient, outpatient, or emergency department) where the specimen was obtained were recorded, as well as patient locations when results were reported. Date and time of 4 processing events were recorded: (1) initial detection time (instrument signal or alarm), (2) Gram stain processing completed, (3) Gram stain microscopic examination completed and results ready to be reported, and (4) results successfully reported to appropriate clinical staff. The number of contacts or other separate actions required to complete final notification was collected. Finally, Gram stain and final culture results were recorded as one or more morphologic types that included gram-positive cocci, gram-positive bacilli, gram-negative cocci, gram-negative bacilli, and fungi.
Participants also provided information about various laboratory practices that potentially could affect reporting timeliness. These variables included blood culture system used; criteria, if any, for blood culture processing timeliness goals; type of personnel processing blood cultures during various times and days of the week; and locations where blood cultures are processed.
Cases involving a positive blood culture signal with negative culture, with negative preliminary Gram stain (no organisms identified), or when preliminary Gram stain was not performed were excluded. Cases were also excluded in which, by policy, notification of results from preliminary blood culture Gram stains was not required (eg, autopsy, repeat positives).
Participants also provided institutional demographic information that included occupied bed size, government affiliation, location, and type of institution, as well as accreditation and inspection status by the College of American Pathologists and the Joint Commission.
Data were tabulated and analyzed by the biostatistics department of the College of American Pathologists. A 2-phase approach was used to analyze the participant turnaround times. Individual associations between the turnaround times and the demographic and practice variables were analyzed using Kruskal-Wallis tests for discrete-valued independent variables and regression analysis for the continuous independent variables. Variables with significant associations (P < .10) were then included in a forward-selection multivariate regression model. A level of .05 was used for statistical significance. All statistical analyses were performed using SAS v9.2 (SAS Institute, Cary, North Carolina).
A total of 65 institutions submitted data involving 5108 positive blood cultures. One laboratory was excluded from analysis of turnaround time because of lack of complete data. Demographic diversity among participants is shown in Table 1. Most participants (60 of 65; 92.3%) were located in the United States, with 2 each from Canada and Saudi Arabia and 1 from Brazil. Within the last 2 years the College of American Pathologists had inspected 56 of 63 participating laboratories (88.9%) and the Joint Commission had conducted laboratory inspections at 11 of 63 (17.5%).
Of 56 laboratories that reported the type of blood culture method used, all used continuous monitoring systems. Most (51 of 56; 91.1%) blood culture systems were located in the microbiology laboratory section; 3 were located in the laboratory accessioning area, 1 in the core laboratory, and 1 in an off-site laboratory. Among 56 laboratories, 40 (71.4%) reported processing positive blood cultures 24 hours a day and 7 days per week. Among the remaining 16 laboratories, 8 processed positive blood cultures 7 to 9 hours per day and 8 processed positive blood cultures 14 to 18 hours per day. Among 56 laboratories, positive blood culture Gram stains were not performed by 5 laboratories (8.9%) and 6 laboratories (10.7%) during the weekday and weekend evening shifts, respectively. Eleven laboratories (19.6%) did not perform Gram stains during weekday or weekend night shifts. Preliminary testing on positive blood cultures, other than Gram staining, was done by 18 of 58 reporting laboratories (31.0%). Of these 18, two-thirds started testing after reporting Gram stain results, 1 before, and the remaining 5 variably began testing either before or after.
Gram stains were nearly always prepared in a laboratory's microbiology section, with 56 of 57 (98.2%) processed there during the day shift, 53 of 54 (98.1%) on weekday evening shifts, 50 of 53 (94.3%) on weekend evening shifts, and 43 of 47 (91.5%) during night shifts. In addition, Gram stains were nearly always read in a laboratory's microbiology section, with 54 of 56 Gram stains (96.4%) read there during the day shift, 44 of 50 (88.0%) during weekday evening shifts, 43 of 49 (87.8%) during weekend evening shifts, and 34 of 43 (79.1%) during night shifts. It was more common for microbiologists to prepare and read Gram stains during the day shift, whereas general medical technologists most often prepared and read Gram stains during the evening and night shifts (Table 2).
Among 5085 positive blood cultures, 2843 (55.9%) were collected from inpatients, 1951 (38.4%) from the emergency department, and 291 (5.7%) from outpatients. Information about patient location at the time of reporting results was available for 4976 blood cultures, of which 4236 (85.1%) were reported to inpatient locations and the remaining 740 (14.9%) to outpatient locations.
Timeliness of Reporting Blood Culture Results
Among all institutions, the median time from when the blood culture system produced a positive signal until preliminary Gram stain results were reported was 45 minutes. Within this time interval, preparing the Gram stain took the longest amount of time, followed by reading the Gram stain, then notification (Table 3).
The median overall time (37 minutes) among the 40 laboratories that process blood cultures continuously on a 24 hours per day, 7 days per week schedule was statistically different from the median turnaround time of 124 minutes for 15 laboratories that reported another processing schedule (P = .003). Likewise, for laboratories that continuously processed blood cultures on a 24-hour schedule, the median time required to finish processing Gram stains was significantly shorter (21 minutes) compared with the 14 laboratories that did not continuously process blood cultures (67 minutes), P = .03 (Table 4).
The time from initiating to completing notification of preliminary blood culture results was significantly longer for outpatient locations during weekday evening shifts (P = .002) and weekend night shifts (P = .02). Other factors, including use of timeliness goals, performing additional preliminary testing other than Gram stain, type of staff who prepare or read Gram stains, staff notified (nurse, primary caregiver), shift, and day of week (weekday, weekend), showed no statistical differences in timeliness of reporting results.
A total of 27 of 56 laboratories (48.2%) reported using a timeliness goal for time from initial instrument signal to notification; goals were used by 24 of 40 laboratories (60.0%) with continuous processing and notification practices compared with only 3 of 16 laboratories (18.8%) that did not continuously process positive blood cultures. Turnaround time goals were set at less than 30 minutes for 4 laboratories, less than 60 minutes for 19, and between 105 and 180 minutes for the remaining 4 laboratories. The median success rate with meeting goals was 80.9% (middle 80% range = 41%–96%). In addition, 11 of 56 laboratories (19.6%) reported using another turnaround time goal, which was from time of Gram stain completion to notification of results; for this goal, targets were set at 15, 20, and 30 minutes for each of 3 labs and 60 minutes for the remaining 8. The median compliance rate for this goal was 95.5% (middle 80% range = 85.9%–100%) for 10 reporting laboratories. Among the 39 participants that continuously processed positive blood cultures, the median time from initial positive to final notification was 35 minutes for the 24 participants that had defined turnaround time goals and 45 minutes for 15 participants who did not have goals. This difference did not reach statistical significance, P = .17. Finally, successful notification of positive blood cultures reached nearly 100% among all participants. Among 4650 notifications, 4359 (93.7%) were completed after 1 attempt, 197 (4.2%) after 2 attempts, and 63 (1.4%) after 3 attempts; 31 (0.7%) required more than 3 attempts.
Accuracy of Preliminary Gram Stain Results
Of 57 reporting participants, 55 (96.5%) reported routinely comparing preliminary blood culture Gram stain with final blood culture results. Concordance of preliminary Gram stain interpretation and final blood culture results from 5021 cases reported by participants is shown in Table 5. Gram stain interpretations were highly accurate, with only 1.2% discrepancies reported when compared with final culture results. The median discrepancy rate among 63 institutions was 1.0%, with 10th, 25th, and 75th percentiles of 3.2%, 2.0%, and 0%, respectively. Among the few discordant results between preliminary and final results, the most common errors were seen with misinterpreting gram-positive bacilli as gram-negative bacilli. Although 20.8% of mixed cultures showed discrepancies, 19 of 22 (86.4%) had one or more correct morphologies reported. The other 3 discrepancies with mixed cultures were reported to have gram-negative bacilli that were not identified by culture. There was no significant association between accuracy of preliminary results and shift or type of staff who prepared or read Gram stains.
This study examined 2 important quality measures of blood culture practice: timeliness and accuracy of preliminary Gram stain results. Both of these measures are important because blood culture results may immediately impact patient management, treatment, and outcomes. Few other microbiology tests are as important and have such immediate clinical impact as blood cultures. Previous studies have shown that timeliness of detecting bloodstream pathogens improves antimicrobial therapy, reduces costs, and decreases length of stay and patient mortality.3–7,9–11 Timeliness and accuracy of preliminary blood culture results represent 2 important, evidence-based quality indicators that are closely linked to clinical outcomes.
Automation has substantially improved the ability to continuously monitor specimens to detect bacterial growth as soon as possible. These systems were used by 100% of study participants. Continuously monitoring blood culture systems detect more than half of bacteremias in 24 hours or less, and the vast majority (85%) in less than 48 hours.20 Although certain advances in direct identification of microorganisms from blood cultures have been described,12,13,21 the current diagnostic mainstay in routine laboratory practice is performing Gram smears from broth bottles that have signaled positive growth. The Gram stain provides a preliminary result that is vital to effectively managing patients with serious infections and has more clinical importance than does final culture results.15 Communicating preliminary blood culture results to the patient's responsible health care provider, who is able to take action as soon as possible, is necessary to achieve the best clinical outcomes.
The median time to report preliminary Gram stain results for 63 participating laboratories ranged from 27 minutes (10th percentile) to more than 2.5 hours (90th percentile), with a 50th percentile of 45 minutes (Table 3). This wide range indicates substantial variation in laboratory performance. The most important factor shown in this study was the laboratory schedule used for processing blood cultures. Those laboratories that continuously processed positive cultures on a 24 hours per day, 7 days per week schedule performed significantly better. Continuously monitoring blood culture systems are designed to signal positive growth at any time. Therefore, the best turnaround time for reporting preliminary results will be attained by laboratories that are able to process cultures every day of the week throughout all shifts. Otherwise, the value of continuously monitoring blood culture systems is not fully realized.17,22 Nevertheless, about 29% of participating laboratories reported processing blood cultures on an interrupted schedule so that processing and reporting positive blood cultures were delayed during certain times, generally on the weekend and/or at night. In this study, use of continuously monitoring blood culture technology was found to be suboptimal among some laboratories. Because of the importance of rapidly reporting positive blood cultures, alternative methods such as telepathology23 or call-back arrangements may be considered when qualified laboratory staff are unavailable to immediately examine preliminary blood culture Gram smears. For example, an experienced technologist might be assigned on-call responsibility to read Gram smears either by on-site examination or remotely using digital imaging technology.
There was no significant association between timeliness goals or type of personnel who process blood cultures and the turnaround time for reporting preliminary blood culture results. However, reporting results on weekday evening and weekend night shifts took significantly longer than at other times, presumably because of variation in provider availability during certain times of the day. Although not statistically significant, there was a trend toward faster reporting times among the 59% of participants that had established turnaround goals when compared with other participants. Because of the important clinical significance of rapid processing and reporting of blood cultures, establishing and monitoring turnaround time goals should be considered for use as quality indicator. In addition, monitoring turnaround time goals has been shown to improve timeliness for other tests.24
It has been shown that delays in incubation of blood cultures after collection will increase time to initial detection of positive growth and decrease timeliness of making changes in antimicrobial therapy.22,25,26 Although time from specimen collection to positive blood culture was not evaluated in this study, laboratories that do not process positive blood cultures continuously should ensure that there is no similar interruption in initially loading or incubating bottles that would further prolong overall time for detecting and reporting bacteremia.
Positive blood cultures are generally treated as critical values for reporting purposes.16 Only about a third of laboratories report positive blood cultures directly to the patient's primary caregiver. In view of the Joint Commission's patient safety goal of reporting critical values to the patient's primary caregiver, it is presumed that nurses or other personnel notified by the laboratory of positive blood cultures notify the patient's primary caregivers. Although notification took, on average, about 1 minute longer among laboratories that directly notify primary caregivers of positive blood culture results, the overall time may be longer for laboratories that notify nurses or other personnel, who are then responsible for handing off this information to the patient's primary caregiver. It may be useful for laboratories that do not directly notify primary caregivers to check nursing or other staff procedures and practices to ensure that positive blood cultures are reported in a timely way to the patient's primary caregiver.
Organism morphology and Gram smear characteristics (positive/negative) were used as criteria to evaluate interpretive accuracy among all participants. Previous studies have used criteria that are more stringent to identify organisms such as Staphylococcus aureus or Candida albicans at the species level.27,28 In this study participants did not report the extent of Gram smear reporting or criteria used for identifying discrepancies. Although newer molecular-based methods can rapidly and accurately identify a wide range of different microbial species directly from positive blood cultures,12–14 the impact of this technology on management and outcomes of patients with bacteremia, beyond reporting preliminary Gram smear results, is yet to be determined.29
Previous studies have shown excellent concordance between preliminary blood culture Gram stain interpretations and final culture results. Rand and Tillan18 reported a discrepancy rate of 57 of 8253 cultures (0.7%), but they excluded cultures growing fungi or any discrepancies that were considered less likely to lead to changes in antibiotic coverage. In another study, Sogaard et al30 evaluated the accuracy of Gram smears in positive blood cultures from a regional laboratory. Mixed cultures, which occurred in 7% of initial samples, were excluded from analysis and gram-positive cocci were evaluated as 2 categories, based on clusters or chains/diplococcus morphology. Of the 5893 cases evaluated in the study of Sogaard et al,30 the discrepancy rate was 1.0%. Finally, Uehara et al6 reported 12 discrepancies (2.5%) among 482 cultures, which included 7.9% that were mixed. Uehara et al6 applied stricter criteria to differentiate Staphylococcus from Streptococcus spp and bacteria in the Enterobacteriaceae family from Pseudomonas spp. Despite differences in study design, criteria used for discrepancies, and exclusions, the results of these 3 studies all show a high degree of accuracy for preliminary Gram smear results that is comparable with the median laboratory discrepancy rate of 1.0% and aggregate rate of 59 of 5021 cultures (1.2%) observed in a diverse group of 63 laboratories in the Q-Probes study. Discrepancies would have likely been slightly higher in this study if more strict criteria had been applied. The causes of discrepancies, such as clerical entry error versus interpretive error, were not evaluated in this or the other studies. However based on previous studies and variation of results among various organism morphologies, interpretive errors were probably the most important cause of discrepancies.
The highest discrepancy rate (20.8%) in this Q-Probes study was observed among mixed cultures. These cultures accounted for only 106 of 5021 positive blood cultures (2.1%). There were only 3 discrepancies involving mixed cultures in which at least 1 morphologic type from culture was not correctly identified. All 3 were read as gram-negative bacilli, of which 2 grew gram-positive bacilli. The next most common discrepancy (6.0%) was seen among gram-positive bacilli that were initially identified as gram-negative. This is consistent with other studies and has been partially explained by propensity of Bacillus sp and Clostridium sp to convert to gram-negative during later stages of growth as well as the tendency for these organisms to decolorize during the Gram stain procedure.18,30 One of 13 (7.7%) gram-negative cocci were reported initially as gram-positive. In one previous study, misinterpretation of Acinetobacter sp as gram-positive in blood cultures was the most common cause for this type of discrepancy.18 Finally, it should be noted that nearly all participants, 55 of 57 (96.5%), monitored accuracy of blood culture Gram stains as a quality indicator. This likely contributes, in part, to the high accuracy rates seen with preliminary Gram stain interpretations.
In summary, this Q-Probes study provides information about current blood culture practices and quality performance. In some laboratories, use of continuously monitoring blood culture technology is suboptimal when processing is performed only intermittently and not on all shifts. Otherwise, results from this study show excellent overall performance in the timeliness and accuracy of preliminary blood culture reporting practices. In addition to monitoring frequencies of solitary blood cultures and blood culture contamination rates, the timeliness and accuracy of preliminary blood culture results is another core laboratory quality performance indicator for assessing blood culture practices, which can now be evaluated based on benchmarks established by this Q-Probes study.
The authors have no relevant financial interest in the products or companies described in this article.