- Email: [email protected]
Clinical impact of reducing routine blood culture incubation time from 7 to 5 days
Pathology (December 2014) 46(7), pp. 636–639
MICROBIOLOGY
Clinical impact of reducing routine blood culture incubation time from 7 to 5 days MICHAEL J. MARGINSON1, KATHRYN L. DAVESON1,2
AND
KARINA J. KENNEDY1,2
1Australian National University Medical School, and 2Canberra Hospital and Health Services, Canberra, ACT,
Australia
Summary The aim of this study was to determine the clinical impact of reducing the blood culture incubation protocol from 7 to 5 days. A laboratory data extraction identified positive blood cultures occurring after 5 or more days of incubation at Canberra Hospital, Australia between 1 January 2001 and 31 August 2011. Isolates were identified as clinically significant using a pre-existing prospective bacteraemia database. Medical records review determined whether the positive result affected clinical management. Positive blood cultures after 5 or more days of incubation accounted for 2.65% (423/15979) of all positive blood cultures, although the majority were false positives or contaminants. Eighty-five were significant/indeterminate, representing an average of eight cases per year or 0.47% (85/15979) of all positive blood cultures sets. Forty-three were isolated for the first time, representing 1.1% of all significant/indeterminate blood culture episodes. Fungi and anaerobic bacteria compromised over 50% of isolates. In 26 cases (2.4 cases per year), the culture result led to a change in patient management. A 7 day incubation protocol is preferable due to late isolation of organisms; however, if space is required to accommodate increasing blood culture numbers, reducing to a 5 or 6 day protocol would miss only a small percentage of clinically significant isolates. Key words: BACTEC, bacteraemia, blood culture, fungaemia, incubation duration. Received 9 March, revised 6 July, accepted 13 July 2014
INTRODUCTION Blood cultures are one of the most critical specimens in the microbiology laboratory; therefore, it is important to maximise the sensitivity of the blood culture method. Our laboratory uses the BACTEC 9000 series automated system (Becton Dickinson Diagnostic Instrument Systems, USA), and has traditionally incubated the bottles for 7 days. More recently, pressure for incubation space from increasing specimen numbers, has led us to evaluate the impact of reducing the routine incubation time to 5 days. Previous studies have shown that reducing the incubation time from 7 days to 4 or 5 days decreases the sensitivity of blood cultures only slightly.1–3 The aim of this study was to determine the potential impact of reducing the blood culture incubation time in our laboratory, by identifying the number of clinically significant positive blood culture isolates following 5 or more days of incubation, and then establishing those which resulted in a significant change in patient management. Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0000000000000167
#
MATERIALS AND METHODS The study was conducted at Canberra Hospital, a 600 bed tertiary referral hospital in the Australian Capital Territory. In 2011, this included a 28 bed adult intensive care and high dependency unit, a 24 bed neonatal intensive care unit, a general paediatrics service, neurosurgery, cardiac surgery, trauma surgery and autologous stem cell transplantation. At the time of the study the annual separations were 73,828 with 227,251 occupied bed days. The hospital microbiology laboratory processes all blood cultures on-site. Hospital policy indicates that blood cultures are collected using an aseptic technique, with 8–10 mL inoculated into each adult bottle and 1–3 mL into paediatric bottles. A blood culture set for adults consists of aerobic and anaerobic bottles (BD BACTEC Plus Aerobic/F and BD BACTEC Lytic/ 10 anaerobic/F). For children, a set is a single aerobic bottle (BD Bactec PEDS Plus/F). On occasions, adults may have only one blood culture bottle taken. Blood cultures are considered a priority specimen and entered into the BACTEC 9000 system 24 h daily. Time to positivity for a given bottle is recorded automatically by the BACTEC 9000 system as the time elapsed between entry into the system and the moment the bottle flagged positive. During the study period all blood cultures were incubated for at least 7 days. The incubation period is extended to 21 days in selected cases where the clinical history indicates a provisional diagnosis of infective endocarditis or fastidious organisms. Body fluids other than blood, and post-mortem specimens, were not included in the study. A data extraction from the laboratory information system was performed of all blood cultures that were positive at greater than or equal to 5 days incubation between 1 January 2001 and 31 August 2011 from patients treated at Canberra Hospital. Episodes were excluded if the partner bottle from the same set of blood cultures was positive with the same organism in less than 5 days. False positive results were defined as those that flagged positive but had a negative Gram stain and no growth on subculture. The Canberra Hospital Bacteraemia Surveillance Database was crossreferenced to define the clinical significance of the episode of bacteraemia. The database involves prospective analysis of all blood culture episodes, including the classification of clinical significance into significant, indeterminate or contaminant. The significance of bacteraemia was based on Australian Infection Control Association definitions.4 In short, bacteraemia was significant when (1) recognised pathogens (e.g., Staphylococcus aureus, Streptococcus pneumoniae, Escherichia coli, Candida sp) were cultured from one or more blood cultures, or (2) potential contaminants were isolated from a patient with signs/symptoms of sepsis AND there was isolation of the same potential contaminant for two or more blood culture sets OR isolation of a potential pathogen from a single blood culture drawn from a patient with an intravascular line and appropriate antimicrobial therapy commenced. The indeterminate result represented positive cultures that did not make the above case definition, but were treated as significant by the treating team. All episodes of bacteraemia were prospectively reviewed by an infection control nurse and discussed with an infectious disease physician/microbiologist. Repeated isolation of the same organism within a 14 day period was defined as a single episode of bacteraemia. The significant isolates were subclassified as either ‘first time isolated’ or ‘duplicate isolate’, indicating that the same organism had previously flagged positive from another set of blood cultures at less than 5 days incubation. A retrospective medical records review was performed to determine if the results of the positive blood cultures beyond
2014 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
CLINICAL IMPACT OF REDUCING ROUTINE BLOOD CULTURE INCUBATION TIME
5 days resulted in a significant change in patient management, defined as a change in antimicrobial choice or duration, or the requirement for ancillary diagnostic tests or additional invasive intervention/surgery. The ACT Health Human Research Ethics Committee approved the study on 15 August 2011.
RESULTS Between 1 January 2001 and 31 August 2011, 15,979 blood culture sets collected at the Canberra Hospital flagged positive. A total of 423 (2.65%) flagged positive after 5 or more days. Seventy-nine were false positives, giving a total of 344 (2.15%) true positives. The majority (259/344, 75%) were regarded as contaminants. The overall rate of contamination for all positive blood culture episodes is 30%. Ten (3%) were of indeterminate significance and 75 (22%) were significant, accounting for only 0.47% (85/15979) of all positive blood culture sets. Of the significant isolates, 33 (44%) were first time isolated. The remaining positives were duplicate isolates. All indeterminate isolates were first time isolates. During the study period there were 3982 episodes of significant/indeterminate bacteraemia. The 43 episodes which were first time isolated therefore represented 1.1% (43/3982) of significant/indeterminate bacteraemia episodes over the study period. However, there was a reduction in this percentage over the study period from 1.7% (2001–2005) to 0.7% (2006–2011). Forty percent of first time isolates were part of two or more blood culture sets collected within 24 h. The distribution of significant and indeterminate isolates (Table 1) was divided fairly evenly between aerobic Gram positive bacteria (24/87, 28%), aerobic Gram negative bacteria Table 1
637
(16/87, 18%), anaerobic bacteria (23/87, 26%) and fungi (24/87, 28%). This included two episodes where two different organisms were isolated from the same blood culture set after 5 or more days. Of the significant/indeterminate isolates, almost one-third (26/85) of results were considered to have changed patient management. This group comprised 10 first isolates [C. albicans (3), C. glabrata (3), C. neoformans (1), P. aeruginosa (1), P. acnes (1) and F. nucleatum (1)], all which resulted in the addition of a new antifungal/antibiotic, as well as a lumbar puncture in the case of C. neoformans. There were 14 duplicate isolates [S. aureus (10), C. albicans (2), C. glabrata (1) and P. acnes (1)], in which the duration of therapy was changed (9), imaging or transoesophageal echocardiogram (TOE) performed (5) or surgery undertaken (2). Two indeterminate isolates (F. incarnatum and Rhodotorula sp) resulted in the addition of an antifungal agent and imaging/TOE in both cases. The impact of 13 isolates was uncertain, and therefore considered unlikely to have changed patient management. Eleven patients were already on appropriate antibiotics, and two had the same organism previously isolated from a different site. Of the 46 isolates which had no clear effect on patient management, 16 (34.8%) of the patients were already on appropriate antibiotics, 11 (23.9%) of the patients were deceased, 7 (15.2%) of the patients had the same organism isolated from another site, 7 (15.2%) were duplicate isolates, and 3 (6.5%) patients with cryptococcosis had been diagnosed by non-culture techniques (cryptococcal antigen). There were two cases (E. faecium and aerobic Gram-negative bacillus) where the management was not changed, but on retrospective review it
Identification of clinically significant or indeterminate blood culture isolates following 5 or more days of incubation, 1 January 2001–31 August 2012 Significant isolate
Aerobic Gram positive Staphylococcus aureus Streptococcus milleri group Enterococcus faecium Corynebacterium accolens Rhodococcus sp. Aerobic Gram negative Escherichia coli Pseudomonas aeruginosa Proteus mirabilis Serratia marcescens Morganella morganii Enterobacter cloacae Aerobic Gram negative bacillus Anaerobic bacteria Fusobacterium nucleatum Prevotella intermedia Bacteroides thetaiotamicron Veillonella sp Propionibacterium sp Peptostreptococcus sp Clostridium sp Mixed anaerobes Anaerobic Gram negative bacilli Fungi Candida albicans Candida glabrata Cryptococcus sp Rhodotorula sp Fusarium incarnatum
First time isolated
Duplicate isolate
Indeterminate isolate
3 1 2 0 0 0 7 3 3 1 0 0 0 0 13 4 1 1 1 2 0 1 1 2 10 4 4 2 0 0
17 15 0 1 1 0 7 1 1 1 1 1 1 1 6 1 0 0 0 4 0 0 0 1 12 6 4 2 0 0
4 1* 0 0 2* 1 2 0 0 0 0 0 0 2{ 4 0 0 0 0 0 2{ 0 1 1 2 0 0 0 1 1
*,{ Organisms isolated from same blood culture set.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
638
MARGINSON et al.
was felt that a management change would have been more appropriate.
DISCUSSION Blood culture positivity beyond five days incubation is not common, occurring in only 2.65% of positive blood culture sets in this study. Previous studies of the BACTEC 9000 series have had variable results. The greatest positivity from the fifth day of incubation was 10% from a paediatric hospital study.5 This higher late positivity rate may be related to the smaller volume used for paediatric blood cultures, as other studies in adults have found much lower rates. Two other studies had similar results of 1% (32/3127) of blood cultures positive on the fifth day of incubation3 and 1.4% (26/1848) after the fifth day.2 Another study had a positivity rate of only 0.11% after 5 days, however did not include anaerobic culture bottles.1 In the current study, false positives or contaminants comprised the majority of cases (2.18%), with only 85 (0.47%) of positive blood cultures regarded as either clinically significant or of indeterminate clinical significance. The 85 significant or indeterminate blood cultures represented an average of eight cases per year over the study period. Of these, 43 were first time isolates, accounting for 1.1% of all significant/indeterminate blood culture episodes, although this varied from 1.7% in the first half to 0.7% in the second half of the study period. This trend will be important to follow as it may indicate increasing sensitivity of the blood culture system. These rates are comparable with Reisner and Woods6 who found that 0.75% (21/2803) of normally pathogenic organisms were isolated for the first time after 5 days. Of most importance, the clinical management of the patient was affected in 26 cases, an average of 2.4 patients per year over the study period. This included 10 first time isolates, two indeterminate isolates and 14 duplicate isolates. The effect on management was uncertain on retrospective review in another 13 cases, and therefore considered unlikely to have impacted management. Previous studies, which have investigated this impact, have only identified 2–3 isolates that have changed patient management over a period of 1–5 years.2,3,6,7 Staphylococcus aureus accounted for almost half of the duplicate isolates, and over one-third of the isolates which resulted in a management change. This included prolongation of antibiotics, ancillary investigations, including TOE, and requirement for surgery, reflecting the nature of S. aureus in causing infective endocarditis and deep-seated infections. Although aerobic Gram negative and Gram positive bacteria are the most common significant blood culture isolates overall, this is not true for positive blood cultures after 5 days of incubation. Fungi and anaerobic bacteria comprised over 50% of significant or indeterminate isolates, and over twothirds of the significant first isolates in this study. Twenty-eight percent of isolates resulting in a change in clinical management, and 70% of first time isolates affecting clinical management were fungi. The predominance of fungi as clinically significant isolates occurring following blood culture incubation beyond 4 days has been highlighted in previous studies.1,3,6,8 and has been the main reason to support longer incubation times. It would be logical to suggest stratification of incubation times based on the clinical likelihood of bacteraemia versus fungaemia, however in practice there are not only difficulties for the clinician in recognising patients at risk for fungaemia, but also problems with communicating this to the
Pathology (2014), 46(7), December
laboratory to allow adjustment of protocols. Other recommendations, such as the use of specialised blood culture media (BACTEC MYCO/F Lytic)9,10 and terminal subculturing of negative blood cultures for suspected fungaemia,11 also encounter similar obstacles. As noted in previous studies,1,5,12,13 the rate of blood culture contamination increases with increasing duration of incubation. The overall contaminant rate at our institution is 39% of positive blood culture episodes, however it was 75% of positive blood cultures after 5 or more days of incubation. Blood culture contamination increases the cost of laboratory processing, for a specimen with no clinical significance, and also has the potential to lead to unnecessary clinical interventions and costs. Results from this study suggest that after five days incubation, approximately five contaminants must be processed to gain one significant isolate, or 16 contaminants processed to ensure one patient’s change in management. A limitation of this study is that the clinical impact of these contaminants, such as unnecessary antibiotics or prolongation of hospital stay, was not assessed. There are several other limitations in the present study. Although the clinical significance of the isolates was determined prospectively, the impact on the clinical management was assessed retrospectively by a medical record review, with inherent problems in interpretation of the clinical decision making at the time. Secondly, it was assumed that the change in management equated to improved patient outcome, however as there is no control group, there is no way of determining whether the change had a negative, positive or neutral effect on the patient’s condition. Thirdly, compliance with blood culture collection policies, including volume of blood inoculated and aseptic technique, which may have affected the contamination rate and time to positivity, was unable to be assessed. In addition, although hospital protocol suggests two sets of blood cultures are taken, 60% of first time isolates were from a single set. The overall reduced blood volume may contribute to reduced sensitivity and delays in any positive result. A reduction in the blood culture incubation protocol in our laboratory to 5 days would have resulted in the failure to detect 1.1% of clinically significant/indeterminate bacteraemia/fungaemia episodes, in addition to a similar number of duplicate isolates that had previously been isolated. In just over one-third of these isolates, or 2.4 patients per year, the clinical management was affected as a result of the blood culture result, with fungi representing a quarter of these cases. In view of these results, laboratories may wish to consider whether reducing the detection rate by 1.1% is acceptable to patient care outcomes in the context of limited laboratory resources. Locally, a decision was made based on these findings that whilst conditions may intermittently arise requiring the shortening of incubation time, maintaining a 7 day protocol where possible would continue with intermittent retrospective review. Acknowledgements: Angelique Clyde-Smith from ACT Pathology for providing the laboratory data extractions. Jan Roberts and Fiona Kimber of the Canberra Hospital Bacteraemia Surveillance Database. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Address for correspondence: Dr K. J. Kennedy, Department of Microbiology and Infectious Diseases, Canberra Hospital and Health Services, Level 4
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
CLINICAL IMPACT OF REDUCING ROUTINE BLOOD CULTURE INCUBATION TIME
Building 10, Canberra Hospital, PO Box 11, Woden, ACT 2606, Australia. E-mail: [email protected]
References 1. Durmaz G, Us T, Aydinli A, et al. Optimum detection times for bacteria and yeast species with the BACTEC 9120 aerobic blood culture system: evaluation for a 5-year period in a Turkish university hospital. J Clin Microbiol 2003; 41: 819–21. 2. Huang AH, Yan JJ, Wu JJ. Comparison of five days versus seven days of incubation for detection of positive blood cultures by Bactec 9240 system. Eur J Clin Microbiol Infect Dis 1998; 17: 637–41. 3. Johnson AS, Touchie C, Haldane DJ, Forward KR. Four-day incubation for detection of bacteremia using the BACTEC 9240. Diagn Microbiol Infect Dis 2000; 38: 195–9. 4. Australian Infection Control Association. Blood Stream Infection (BSI) Definition. Sept 2004; cited 28 June 2014. http://www.agargroup.org/files/ Blood%20stream%20defintions.pdf 5. Kara A, Kanra G, Cengiz AB, et al. Pediatric blood culture: time to positivity. Turkish J Ped 2004; 46: 251–5. 6. Reisner BS, Woods GL. Times to detection of bacteria and yeasts in BACTEC 9240 blood culture bottles. J Clin Microbiol 1999; 37: 2024–6.
639
7. Nolte FS, Williams JM, Jerris RC, et al. Multicenter clinical evaluation of a continuous monitoring blood culture system using fluorescent-sensor technology (BACTEC 9240). J Clin Microbiol 1993; 31: 552–7. 8. Pal N, Sharma R, Rishi S, Vyas L. Optimum time to detection of bacteria and yeast species with BACTEC 9120 culture system from blood and sterile body fluids. J Lab Physicians 2009; 1: 69–72. 9. Arendrup MC, Bruun B, Christensen JJ, et al. National surveillance of fungemia in Denmark (2004–2009). J Clin Microbiol 2011; 49: 325– 34. 10. Vetter E, Torgerson C, Feuker A, et al. Comparison of the BACTEC MYCO/F Lytic bottle to the isolator tube, BACTEC Plus Aerobic F/bottle, and BACTEC Anaerobic Lytic/10 bottle and comparison of the BACTEC Plus Aerobic F/bottle to the isolator tube for recovery of bacteria, mycobacteria, and fungi from blood. J Clin Microbiol 2001; 39: 4380–6. 11. Horvath LL, Hospenthal DR, Murray CK, Dooley DP. Detection of simulated candidemia by the BACTEC 9240 system with plus aerobic/F and anaerobic/F blood culture bottles. J Clin Microbiol 2003; 41: 4714–7. 12. Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006; 19: 788–802. 13. Haimi-Cohen Y, Shafinoori S, Tucci V, Rubin LG. Use of incubation time to detection in BACTEC 9240 to distinguish coagulase-negative staphylococcal contamination from infection in pediatric blood cultures. Pediatr Infect Dis J 2003; 22: 968–74.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
MICROBIOLOGY
Clinical impact of reducing routine blood culture incubation time from 7 to 5 days MICHAEL J. MARGINSON1, KATHRYN L. DAVESON1,2
AND
KARINA J. KENNEDY1,2
1Australian National University Medical School, and 2Canberra Hospital and Health Services, Canberra, ACT,
Australia
Summary The aim of this study was to determine the clinical impact of reducing the blood culture incubation protocol from 7 to 5 days. A laboratory data extraction identified positive blood cultures occurring after 5 or more days of incubation at Canberra Hospital, Australia between 1 January 2001 and 31 August 2011. Isolates were identified as clinically significant using a pre-existing prospective bacteraemia database. Medical records review determined whether the positive result affected clinical management. Positive blood cultures after 5 or more days of incubation accounted for 2.65% (423/15979) of all positive blood cultures, although the majority were false positives or contaminants. Eighty-five were significant/indeterminate, representing an average of eight cases per year or 0.47% (85/15979) of all positive blood cultures sets. Forty-three were isolated for the first time, representing 1.1% of all significant/indeterminate blood culture episodes. Fungi and anaerobic bacteria compromised over 50% of isolates. In 26 cases (2.4 cases per year), the culture result led to a change in patient management. A 7 day incubation protocol is preferable due to late isolation of organisms; however, if space is required to accommodate increasing blood culture numbers, reducing to a 5 or 6 day protocol would miss only a small percentage of clinically significant isolates. Key words: BACTEC, bacteraemia, blood culture, fungaemia, incubation duration. Received 9 March, revised 6 July, accepted 13 July 2014
INTRODUCTION Blood cultures are one of the most critical specimens in the microbiology laboratory; therefore, it is important to maximise the sensitivity of the blood culture method. Our laboratory uses the BACTEC 9000 series automated system (Becton Dickinson Diagnostic Instrument Systems, USA), and has traditionally incubated the bottles for 7 days. More recently, pressure for incubation space from increasing specimen numbers, has led us to evaluate the impact of reducing the routine incubation time to 5 days. Previous studies have shown that reducing the incubation time from 7 days to 4 or 5 days decreases the sensitivity of blood cultures only slightly.1–3 The aim of this study was to determine the potential impact of reducing the blood culture incubation time in our laboratory, by identifying the number of clinically significant positive blood culture isolates following 5 or more days of incubation, and then establishing those which resulted in a significant change in patient management. Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0000000000000167
#
MATERIALS AND METHODS The study was conducted at Canberra Hospital, a 600 bed tertiary referral hospital in the Australian Capital Territory. In 2011, this included a 28 bed adult intensive care and high dependency unit, a 24 bed neonatal intensive care unit, a general paediatrics service, neurosurgery, cardiac surgery, trauma surgery and autologous stem cell transplantation. At the time of the study the annual separations were 73,828 with 227,251 occupied bed days. The hospital microbiology laboratory processes all blood cultures on-site. Hospital policy indicates that blood cultures are collected using an aseptic technique, with 8–10 mL inoculated into each adult bottle and 1–3 mL into paediatric bottles. A blood culture set for adults consists of aerobic and anaerobic bottles (BD BACTEC Plus Aerobic/F and BD BACTEC Lytic/ 10 anaerobic/F). For children, a set is a single aerobic bottle (BD Bactec PEDS Plus/F). On occasions, adults may have only one blood culture bottle taken. Blood cultures are considered a priority specimen and entered into the BACTEC 9000 system 24 h daily. Time to positivity for a given bottle is recorded automatically by the BACTEC 9000 system as the time elapsed between entry into the system and the moment the bottle flagged positive. During the study period all blood cultures were incubated for at least 7 days. The incubation period is extended to 21 days in selected cases where the clinical history indicates a provisional diagnosis of infective endocarditis or fastidious organisms. Body fluids other than blood, and post-mortem specimens, were not included in the study. A data extraction from the laboratory information system was performed of all blood cultures that were positive at greater than or equal to 5 days incubation between 1 January 2001 and 31 August 2011 from patients treated at Canberra Hospital. Episodes were excluded if the partner bottle from the same set of blood cultures was positive with the same organism in less than 5 days. False positive results were defined as those that flagged positive but had a negative Gram stain and no growth on subculture. The Canberra Hospital Bacteraemia Surveillance Database was crossreferenced to define the clinical significance of the episode of bacteraemia. The database involves prospective analysis of all blood culture episodes, including the classification of clinical significance into significant, indeterminate or contaminant. The significance of bacteraemia was based on Australian Infection Control Association definitions.4 In short, bacteraemia was significant when (1) recognised pathogens (e.g., Staphylococcus aureus, Streptococcus pneumoniae, Escherichia coli, Candida sp) were cultured from one or more blood cultures, or (2) potential contaminants were isolated from a patient with signs/symptoms of sepsis AND there was isolation of the same potential contaminant for two or more blood culture sets OR isolation of a potential pathogen from a single blood culture drawn from a patient with an intravascular line and appropriate antimicrobial therapy commenced. The indeterminate result represented positive cultures that did not make the above case definition, but were treated as significant by the treating team. All episodes of bacteraemia were prospectively reviewed by an infection control nurse and discussed with an infectious disease physician/microbiologist. Repeated isolation of the same organism within a 14 day period was defined as a single episode of bacteraemia. The significant isolates were subclassified as either ‘first time isolated’ or ‘duplicate isolate’, indicating that the same organism had previously flagged positive from another set of blood cultures at less than 5 days incubation. A retrospective medical records review was performed to determine if the results of the positive blood cultures beyond
2014 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
CLINICAL IMPACT OF REDUCING ROUTINE BLOOD CULTURE INCUBATION TIME
5 days resulted in a significant change in patient management, defined as a change in antimicrobial choice or duration, or the requirement for ancillary diagnostic tests or additional invasive intervention/surgery. The ACT Health Human Research Ethics Committee approved the study on 15 August 2011.
RESULTS Between 1 January 2001 and 31 August 2011, 15,979 blood culture sets collected at the Canberra Hospital flagged positive. A total of 423 (2.65%) flagged positive after 5 or more days. Seventy-nine were false positives, giving a total of 344 (2.15%) true positives. The majority (259/344, 75%) were regarded as contaminants. The overall rate of contamination for all positive blood culture episodes is 30%. Ten (3%) were of indeterminate significance and 75 (22%) were significant, accounting for only 0.47% (85/15979) of all positive blood culture sets. Of the significant isolates, 33 (44%) were first time isolated. The remaining positives were duplicate isolates. All indeterminate isolates were first time isolates. During the study period there were 3982 episodes of significant/indeterminate bacteraemia. The 43 episodes which were first time isolated therefore represented 1.1% (43/3982) of significant/indeterminate bacteraemia episodes over the study period. However, there was a reduction in this percentage over the study period from 1.7% (2001–2005) to 0.7% (2006–2011). Forty percent of first time isolates were part of two or more blood culture sets collected within 24 h. The distribution of significant and indeterminate isolates (Table 1) was divided fairly evenly between aerobic Gram positive bacteria (24/87, 28%), aerobic Gram negative bacteria Table 1
637
(16/87, 18%), anaerobic bacteria (23/87, 26%) and fungi (24/87, 28%). This included two episodes where two different organisms were isolated from the same blood culture set after 5 or more days. Of the significant/indeterminate isolates, almost one-third (26/85) of results were considered to have changed patient management. This group comprised 10 first isolates [C. albicans (3), C. glabrata (3), C. neoformans (1), P. aeruginosa (1), P. acnes (1) and F. nucleatum (1)], all which resulted in the addition of a new antifungal/antibiotic, as well as a lumbar puncture in the case of C. neoformans. There were 14 duplicate isolates [S. aureus (10), C. albicans (2), C. glabrata (1) and P. acnes (1)], in which the duration of therapy was changed (9), imaging or transoesophageal echocardiogram (TOE) performed (5) or surgery undertaken (2). Two indeterminate isolates (F. incarnatum and Rhodotorula sp) resulted in the addition of an antifungal agent and imaging/TOE in both cases. The impact of 13 isolates was uncertain, and therefore considered unlikely to have changed patient management. Eleven patients were already on appropriate antibiotics, and two had the same organism previously isolated from a different site. Of the 46 isolates which had no clear effect on patient management, 16 (34.8%) of the patients were already on appropriate antibiotics, 11 (23.9%) of the patients were deceased, 7 (15.2%) of the patients had the same organism isolated from another site, 7 (15.2%) were duplicate isolates, and 3 (6.5%) patients with cryptococcosis had been diagnosed by non-culture techniques (cryptococcal antigen). There were two cases (E. faecium and aerobic Gram-negative bacillus) where the management was not changed, but on retrospective review it
Identification of clinically significant or indeterminate blood culture isolates following 5 or more days of incubation, 1 January 2001–31 August 2012 Significant isolate
Aerobic Gram positive Staphylococcus aureus Streptococcus milleri group Enterococcus faecium Corynebacterium accolens Rhodococcus sp. Aerobic Gram negative Escherichia coli Pseudomonas aeruginosa Proteus mirabilis Serratia marcescens Morganella morganii Enterobacter cloacae Aerobic Gram negative bacillus Anaerobic bacteria Fusobacterium nucleatum Prevotella intermedia Bacteroides thetaiotamicron Veillonella sp Propionibacterium sp Peptostreptococcus sp Clostridium sp Mixed anaerobes Anaerobic Gram negative bacilli Fungi Candida albicans Candida glabrata Cryptococcus sp Rhodotorula sp Fusarium incarnatum
First time isolated
Duplicate isolate
Indeterminate isolate
3 1 2 0 0 0 7 3 3 1 0 0 0 0 13 4 1 1 1 2 0 1 1 2 10 4 4 2 0 0
17 15 0 1 1 0 7 1 1 1 1 1 1 1 6 1 0 0 0 4 0 0 0 1 12 6 4 2 0 0
4 1* 0 0 2* 1 2 0 0 0 0 0 0 2{ 4 0 0 0 0 0 2{ 0 1 1 2 0 0 0 1 1
*,{ Organisms isolated from same blood culture set.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
638
MARGINSON et al.
was felt that a management change would have been more appropriate.
DISCUSSION Blood culture positivity beyond five days incubation is not common, occurring in only 2.65% of positive blood culture sets in this study. Previous studies of the BACTEC 9000 series have had variable results. The greatest positivity from the fifth day of incubation was 10% from a paediatric hospital study.5 This higher late positivity rate may be related to the smaller volume used for paediatric blood cultures, as other studies in adults have found much lower rates. Two other studies had similar results of 1% (32/3127) of blood cultures positive on the fifth day of incubation3 and 1.4% (26/1848) after the fifth day.2 Another study had a positivity rate of only 0.11% after 5 days, however did not include anaerobic culture bottles.1 In the current study, false positives or contaminants comprised the majority of cases (2.18%), with only 85 (0.47%) of positive blood cultures regarded as either clinically significant or of indeterminate clinical significance. The 85 significant or indeterminate blood cultures represented an average of eight cases per year over the study period. Of these, 43 were first time isolates, accounting for 1.1% of all significant/indeterminate blood culture episodes, although this varied from 1.7% in the first half to 0.7% in the second half of the study period. This trend will be important to follow as it may indicate increasing sensitivity of the blood culture system. These rates are comparable with Reisner and Woods6 who found that 0.75% (21/2803) of normally pathogenic organisms were isolated for the first time after 5 days. Of most importance, the clinical management of the patient was affected in 26 cases, an average of 2.4 patients per year over the study period. This included 10 first time isolates, two indeterminate isolates and 14 duplicate isolates. The effect on management was uncertain on retrospective review in another 13 cases, and therefore considered unlikely to have impacted management. Previous studies, which have investigated this impact, have only identified 2–3 isolates that have changed patient management over a period of 1–5 years.2,3,6,7 Staphylococcus aureus accounted for almost half of the duplicate isolates, and over one-third of the isolates which resulted in a management change. This included prolongation of antibiotics, ancillary investigations, including TOE, and requirement for surgery, reflecting the nature of S. aureus in causing infective endocarditis and deep-seated infections. Although aerobic Gram negative and Gram positive bacteria are the most common significant blood culture isolates overall, this is not true for positive blood cultures after 5 days of incubation. Fungi and anaerobic bacteria comprised over 50% of significant or indeterminate isolates, and over twothirds of the significant first isolates in this study. Twenty-eight percent of isolates resulting in a change in clinical management, and 70% of first time isolates affecting clinical management were fungi. The predominance of fungi as clinically significant isolates occurring following blood culture incubation beyond 4 days has been highlighted in previous studies.1,3,6,8 and has been the main reason to support longer incubation times. It would be logical to suggest stratification of incubation times based on the clinical likelihood of bacteraemia versus fungaemia, however in practice there are not only difficulties for the clinician in recognising patients at risk for fungaemia, but also problems with communicating this to the
Pathology (2014), 46(7), December
laboratory to allow adjustment of protocols. Other recommendations, such as the use of specialised blood culture media (BACTEC MYCO/F Lytic)9,10 and terminal subculturing of negative blood cultures for suspected fungaemia,11 also encounter similar obstacles. As noted in previous studies,1,5,12,13 the rate of blood culture contamination increases with increasing duration of incubation. The overall contaminant rate at our institution is 39% of positive blood culture episodes, however it was 75% of positive blood cultures after 5 or more days of incubation. Blood culture contamination increases the cost of laboratory processing, for a specimen with no clinical significance, and also has the potential to lead to unnecessary clinical interventions and costs. Results from this study suggest that after five days incubation, approximately five contaminants must be processed to gain one significant isolate, or 16 contaminants processed to ensure one patient’s change in management. A limitation of this study is that the clinical impact of these contaminants, such as unnecessary antibiotics or prolongation of hospital stay, was not assessed. There are several other limitations in the present study. Although the clinical significance of the isolates was determined prospectively, the impact on the clinical management was assessed retrospectively by a medical record review, with inherent problems in interpretation of the clinical decision making at the time. Secondly, it was assumed that the change in management equated to improved patient outcome, however as there is no control group, there is no way of determining whether the change had a negative, positive or neutral effect on the patient’s condition. Thirdly, compliance with blood culture collection policies, including volume of blood inoculated and aseptic technique, which may have affected the contamination rate and time to positivity, was unable to be assessed. In addition, although hospital protocol suggests two sets of blood cultures are taken, 60% of first time isolates were from a single set. The overall reduced blood volume may contribute to reduced sensitivity and delays in any positive result. A reduction in the blood culture incubation protocol in our laboratory to 5 days would have resulted in the failure to detect 1.1% of clinically significant/indeterminate bacteraemia/fungaemia episodes, in addition to a similar number of duplicate isolates that had previously been isolated. In just over one-third of these isolates, or 2.4 patients per year, the clinical management was affected as a result of the blood culture result, with fungi representing a quarter of these cases. In view of these results, laboratories may wish to consider whether reducing the detection rate by 1.1% is acceptable to patient care outcomes in the context of limited laboratory resources. Locally, a decision was made based on these findings that whilst conditions may intermittently arise requiring the shortening of incubation time, maintaining a 7 day protocol where possible would continue with intermittent retrospective review. Acknowledgements: Angelique Clyde-Smith from ACT Pathology for providing the laboratory data extractions. Jan Roberts and Fiona Kimber of the Canberra Hospital Bacteraemia Surveillance Database. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Address for correspondence: Dr K. J. Kennedy, Department of Microbiology and Infectious Diseases, Canberra Hospital and Health Services, Level 4
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
CLINICAL IMPACT OF REDUCING ROUTINE BLOOD CULTURE INCUBATION TIME
Building 10, Canberra Hospital, PO Box 11, Woden, ACT 2606, Australia. E-mail: [email protected]
References 1. Durmaz G, Us T, Aydinli A, et al. Optimum detection times for bacteria and yeast species with the BACTEC 9120 aerobic blood culture system: evaluation for a 5-year period in a Turkish university hospital. J Clin Microbiol 2003; 41: 819–21. 2. Huang AH, Yan JJ, Wu JJ. Comparison of five days versus seven days of incubation for detection of positive blood cultures by Bactec 9240 system. Eur J Clin Microbiol Infect Dis 1998; 17: 637–41. 3. Johnson AS, Touchie C, Haldane DJ, Forward KR. Four-day incubation for detection of bacteremia using the BACTEC 9240. Diagn Microbiol Infect Dis 2000; 38: 195–9. 4. Australian Infection Control Association. Blood Stream Infection (BSI) Definition. Sept 2004; cited 28 June 2014. http://www.agargroup.org/files/ Blood%20stream%20defintions.pdf 5. Kara A, Kanra G, Cengiz AB, et al. Pediatric blood culture: time to positivity. Turkish J Ped 2004; 46: 251–5. 6. Reisner BS, Woods GL. Times to detection of bacteria and yeasts in BACTEC 9240 blood culture bottles. J Clin Microbiol 1999; 37: 2024–6.
639
7. Nolte FS, Williams JM, Jerris RC, et al. Multicenter clinical evaluation of a continuous monitoring blood culture system using fluorescent-sensor technology (BACTEC 9240). J Clin Microbiol 1993; 31: 552–7. 8. Pal N, Sharma R, Rishi S, Vyas L. Optimum time to detection of bacteria and yeast species with BACTEC 9120 culture system from blood and sterile body fluids. J Lab Physicians 2009; 1: 69–72. 9. Arendrup MC, Bruun B, Christensen JJ, et al. National surveillance of fungemia in Denmark (2004–2009). J Clin Microbiol 2011; 49: 325– 34. 10. Vetter E, Torgerson C, Feuker A, et al. Comparison of the BACTEC MYCO/F Lytic bottle to the isolator tube, BACTEC Plus Aerobic F/bottle, and BACTEC Anaerobic Lytic/10 bottle and comparison of the BACTEC Plus Aerobic F/bottle to the isolator tube for recovery of bacteria, mycobacteria, and fungi from blood. J Clin Microbiol 2001; 39: 4380–6. 11. Horvath LL, Hospenthal DR, Murray CK, Dooley DP. Detection of simulated candidemia by the BACTEC 9240 system with plus aerobic/F and anaerobic/F blood culture bottles. J Clin Microbiol 2003; 41: 4714–7. 12. Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006; 19: 788–802. 13. Haimi-Cohen Y, Shafinoori S, Tucci V, Rubin LG. Use of incubation time to detection in BACTEC 9240 to distinguish coagulase-negative staphylococcal contamination from infection in pediatric blood cultures. Pediatr Infect Dis J 2003; 22: 968–74.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.