- Email: [email protected]
Blood culture series benefit may be limited to selected clinical conditions: time to reassess
Accepted Manuscript Blood culture series benefit may be limited to selected clinical conditions: time to reassess Riad Khatib , MD, Gordana Simeunovic , MD, Mamta Sharma , MD, Mohamad G. Fakih , MD, MBA, Leonard B. Johnson , MD, Laurence Briski , MD, William Lebar , PhD PII:
S1198-743X(14)00105-0
DOI:
10.1016/j.cmi.2014.11.019
Reference:
CMI 104
To appear in:
Clinical Microbiology and Infection
Received Date: 22 September 2014 Revised Date:
18 November 2014
Accepted Date: 19 November 2014
Please cite this article as: Khatib R, Simeunovic G, Sharma M, Fakih MG, Johnson LB, Briski L, Lebar W, Blood culture series benefit may be limited to selected clinical conditions: time to reassess, Clinical Microbiology and Infection (2014), doi: 10.1016/j.cmi.2014.11.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Original Article
RI PT
Blood culture series benefit may be limited to selected clinical conditions: time to reassess.
SC
Riad Khatib, MDa, Gordana Simeunovic, MDa, Mamta Sharma, MDa, Mohamad G. Fakih, MD, MBAa, Leonard B. Johnson, MDa, Laurence Briski, MDb, William
M AN U
Lebar, PhDb,c
Departments of Medicinea and Pathologyb St. John Hospital and Medical Center,
TE D
Grosse Pointe Woods, MI USA.
EP
Running title: Limited benefit of blood culture series: time to reassess. Corresponding author
AC C
Riad Khatib, MD.
19251 Mack Avenue, Grosse Pointe Woods, MI 48236. email: [email protected]. Fax:01-313-3437840; Tel:0-1313-3433802
1
ACCEPTED MANUSCRIPT
Abstract
Blood cultures are often submitted as series (2-3 sets/24 h) to maximize
RI PT
recovery. We assessed the actual benefit of additional sets. Methods: Blood cultures submitted from adults (≥ 18 years-old) over one year (2/1/2012-1/31/2013) were examined. The medical records of patients with
SC
positive cultures were reviewed. Cultures with commensal organisms were
considered contamination in the absence of a source and clinical findings. The
M AN U
impact of additional sets on antibiotic therapy was estimated. We evaluated 15394 blood cultures. They were submitted as 2-5 set/24 hour in 12236 (79.5%) instances. Pathogens were detected in 1227 sets, representing 741 bacteremias of which 618 (83.4%) were detected in the first set and 123 (16.6%) in the additional sets. Pathogens missed in the first set were recovered from patients
TE D
on antibiotics (n=72; 58.5%) and following a procedure (n=54; 43.9%). The additional sets result could have influenced antibiotic therapy in 76/6235 (1.2%)
EP
instances including 40 (0.6%) antibiotic switch and 36 (0.6%) possible extension of therapy. The potential impact of detection of missed pathogen on antibiotic
AC C
therapy was mot apparent in patients with an endovascular focus (26/27; 96.3%) and those without an obvious source (10/10; 100%). These findings suggest that one blood culture is probably adequate in patients with an obvious source. Blood culture series are beneficial in patients without an obvious source and with those with endovascular infections. It is time to reassess blood culture series benefit and may be limiting them to selected conditions. Keywords: Blood culture; bacteremia; contaminant; culture series; limitations.
2
ACCEPTED MANUSCRIPT
Introduction
RI PT
Detection of bacteremia or fungemia by blood culture is essential for proper
management of patients with infection. Current guidelines advocate 2-3 sets
obtained simultaneously from different venipunctures or serially within 0-24 hours
SC
for optimal detection of bacteremia and interpretation of blood culture results [15]. The potential benefits of multiple sets include better detection of bacteremia,
M AN U
distinguishing contaminants and the recognition of persistent bacteremia [4]. The actual benefit was attributed to simple increase in the blood volume rather than numbers of venipunctures [6, 7]. These guidelines are based on several studies that illustrate better detection of bacteremia with more cultures and larger blood
TE D
volume [2, 8, 9-11]. Most of these studies focus on the detection of bacteremia without assessing the overall benefit of additional sets, whether they alter the patient management and if they are more applicable to selected clinical settings.
EP
With the increasing workload in the microbiology laboratory, cost containment efforts, and the drive for early initiation of antibiotics, we assessed the benefit of
AC C
more than one culture/24 hour.
Material and Methods:
Facilities
3
ACCEPTED MANUSCRIPT
Our hospital is an 804-bed teaching hospital in Detroit Michigan. Our microbiology laboratory receives blood cultures from the inpatients, emergency department, affiliated outpatient settings, as well from other affiliated chronic and
RI PT
acute care facilities. We selected cultures submitted from our adult (≥18 years
old) inpatients, in the emergency departments and affiliated outpatient settings. We excluded cultures from chronic care facilities and other hospitals. We
SC
performed a retrospective evaluation of the results of blood cultures processed
Blood culture processing
M AN U
between February 1st 2012 and January 31st, 2013.
The blood culture system used in our hospital during the study period was the BacTAlert with FA FAN® Aerobic and Anaerobic bottles (bioMerieux) with instructions for 10 ml blood draw/bottle. The phlebotomists are expected to mark
TE D
the level of media in each bottle and the desired fill volume, which is 10 ml in adults, using the pre-marked five ml increments on the bottle label prior to blood collection. Additionally, about 20% of blood culture bottles are randomly selected
EP
periodically and examined to assure compliance with proper blood volume fill; and, feedback is provided to the phlebotomists.
AC C
2.3 Assessment of blood cultures We reviewed the medical records of all patients with positive cultures. We excluded line draws and patients with inadequate medical record documentation for assessment. We abstracted the clinical findings, the condition at the time of culture, the source of bacteremia with the source culture results, and the clinician
4
ACCEPTED MANUSCRIPT
approach to the blood culture result. The reviewers were not blinded to the blood culture results. Definitions
RI PT
A blood culture set was defined as aerobic and anaerobic blood culture bottles
obtained from one venipuncture site. A single set was defined as one set/24 h. Culture series represented two or more sets/24 h. The first and additional sets
SC
referred to the first and subsequently drawn sets within 24 h. In simultaneously drawn sets, the set entered first in the record was considered the first.
M AN U
Commensal organisms signified organisms recognized as potential contaminants such as coagulase-negative staphylococci, viridans group streptococci, Bacillus spp., Corynobacterium spp., Propionibacterium spp., Aerococcus spp., and Micrococcus spp. Isolation of one of these organisms was considered true
TE D
bacteremia in the presence of a source and clinical signs, contamination in the absence of a source and clinical signs and of uncertain status in the presence of a potential source or clinical signs but not both [10-12]. Cases with uncertain
EP
status were counted as bacteremias. An episode of bacteremia was defined as the recovery of the same pathogen from single or multiple sets within the same
AC C
day and the recovery of a different pathogen by the additional sets within the same day.
The impact of the additional set result on antibiotic treatment was estimated based on whether the usual empirical therapy for the disease process was adequate to cover the yielded pathogens, if treatment duration needed to be
5
ACCEPTED MANUSCRIPT
extended based on documenting persistent bacteremia and if treatment could have been discontinued.
RI PT
Statistical methods
Chi square test was used to assess the significance of differences in categorical variables and the significance of trends was assessed by the Extended Mantel-
SC
Haenszel chi square test for linear trend, utilizing the computer software SPSS
M AN U
release 20. P value <0 .05 was considered to indicate statistical significance.
The study was approved by the St John Hospital and Medical Center Institutional
TE D
Review Board.
EP
Results
Study population
AC C
A total of 18507 blood cultures were reviewed, 3113 were excluded including 2311 from pediatric cases, 788 from chronic care facilities, four line draws and ten with inadequate medical records for assessment. The 15394 included cultures were obtained from 5251 patients encompassing 6854 BC (44.5%) from the general wards, 6006 (39.0%) the emergency department, 1999 (13.0%) intensive care units, 459 (3.0%) outpatients and 76 (0.5%) labor and delivery.
6
ACCEPTED MANUSCRIPT
The cultures were submitted as a single set/24h in 3158 (20.5%) and series of 25 sets/24 h in 12236 (79.5%) instances, corresponding to 6001 series with 6235 additional sets. The interval between blood cultures sets was ≤30 minutes in
RI PT
4759 (76.3%) of the cases including 162 (2.6%) simultaneous submissions.
Blood culture results
SC
A total of 2072 organisms were recovered from 1887 (12.3%) blood culture sets including 1601 pathogens, 404 contaminants and 67 organisms of uncertain
M AN U
status that were considered pathogens. The organisms recovered are shown in table 1. Detection of pathogens was similar in single set and culture series (337 [10.7%] vs. 1227 [10.0%]; p=0.3).
TE D
The value of additional blood culture sets
Detection of bacteremia
EP
The majority of additional sets (n=5522; 88.6%) was negative (Figure 1). Pathogens were detected in 1227 sets, corresponding to 741 episodes of
AC C
bacteremias among 239 patients. These episodes were detected by the first set in 618 (83.4%) and the additional sets in 123 (16.6%) instances. The characteristics of bacteremia missed by the first set are shown in table 2. Most patients were on antibiotic therapy and the additional sets were obtained after a procedure such as exchange of obstructed indwelling urinary catheter, deep respiratory suctioning in an intubated patient, incision and drainage of an abscess or debridement of necrotic tissues. They included follow up cultures
7
ACCEPTED MANUSCRIPT
from patients with known bacteremia and instances of polymicrobial bacteremia with an extra pathogen identified in the additional sets. Many of these patients had a focus of infection with defined microbiology, such as pneumonia, urinary
RI PT
tract infection and soft tissue/bone infection that would have warranted
appropriate antibiotic treatment. The additional culture set result could have
influenced antibiotic therapy in 76/6235 (1.2%) instances including antibiotic
SC
switch (n=40; 0.7%) and extension of therapy (n=36; 0.6%).
M AN U
The pathogens missed by the first set were detected by the second set in 119 (96.8%) instances, the third set in three (2.4%) and the fourth set in one (0.8%). Stratifying culture results according to the interval between serial cultures revealed that the missed pathogens were least likely to be recovered in additional sets obtained 0-30 minutes (72/4759 [1.5%]) compared to 31-60
TE D
minutes (14/506 [2.8%]) and > 61 minutes (31/970 [3.2%]) after the first set (p=0.0008; extended Mantel-Haenszel chi square for linear trend).
AC C
EP
Discussion
Current guidelines for blood cultures advocate 2-3 sets obtained simultaneously or within 24 h to maximize recovery of bacteremia and help in interpreting the isolation of common contaminants [1, 4, 5]. For detection of bacteremia, comparing single to multiple blood cultures is likely to reveal that the more cultures and/or the larger blood volume the better detection of bloodstream
8
ACCEPTED MANUSCRIPT
invasion [2, 4, 5, 8-10]. The major drawback of these studies is the testing of the same cultures that are being used to define bacteremia [6] and focusing on the diagnosis of bacteremia without addressing the actual impact of blood culture
RI PT
series on antibiotic therapy and patient outcome. In fact, the usefulness of blood culture in certain conditions such as urinary tract infection, cellulitis, and
pneumonia has been questioned [12-20]. Blood culture results in these settings
SC
may help optimizing antibiotic therapy but the benefit of more than one culture
M AN U
has not been substantiated.
Another potential value of multiple blood culture sets is distinguishing between persistent and transient bacteremia. This distinction may be valuable in suggesting an intravascular focus [10] and in modifying the treatment duration in Staphylococcus aureus bacteremia [21] and potentially other bacteremias but its
TE D
relevance in most clinical conditions is largely unknown.
Our findings support that detection of bacteremia was improved by a second set.
EP
However, the percentage of additional sets with missed pathogens was much higher in sets obtained after 30 minutes from the first culture. They may have
AC C
been obtained because of a change in the patient condition but determining the actual reason for late submission is difficult to determine retrospectively. Additionally, our findings show that the impact of the improved detection by the additional sets was most apparent in patients with suspected endovascular foci and those without an obvious source of infection.
9
ACCEPTED MANUSCRIPT
With respect to recognizing contamination, several tools exist [10, 22, 23-26]. They are based on the type of organism, the clinical features and the proportion of positive cultures. However, the value of these guidelines in patients with high
RI PT
risk for infection with common contaminants, such as those with devices or osteomyelitis, and patients on antimicrobial therapy has not been clearly
substantiated. Additionally, detecting different strains in multiple sets may
SC
represent bacteremia in one culture and a contaminant in the other. Finally,
represent bacteremia [26].
M AN U
isolating the same contaminant in two blood culture sets may not always
Limitations of our study include the retrospective review and the uncertainty of potential differences in blood draw volume in the first and additional cultures. We also did not examine blood culture media effects. A significant difference in
TE D
bacterial recovery was noted based on the type of media or bottles [27, 28]. Finally, our calculation of the impact of culture results on patient management
EP
and antibiotic therapy was imprecise. It was estimated based on the presumed empiric antibiotic therapy and the missed pathogen susceptibility and if detecting
AC C
bacteremia could have lead to extension of the duration of antibiotic therapy such as the finding of Staphylococcus aureus bacteremia [21]. Despite these limitations, our findings suggest that one blood culture is likely adequate in many conditions. Additional studies are needed to verify this observation and determine whether culture series benefit is more apparent in selected clinical settings such as endovascular infections and cases without an obvious source.
10
ACCEPTED MANUSCRIPT
Acknowledgement
RI PT
The study was supported by St John Hospital and Medical Center Department of
AC C
EP
TE D
M AN U
SC
Medical Education.
11
ACCEPTED MANUSCRIPT
References
Ntusi N, Aubin L, Oliver S et al. Guideline for the optimal use of blood cultures. S Afr Med J 2010; 100: 839-843.
2
RI PT
1
Riedel S, Bourbeau P, Swartz B, et al. Timing of specimen collection
2008; 46: 1381-1385. 3
Washington JA 2nd. Blood cultures: principles and techniques. Mayo
M AN U
Clinic Proceed 1975; 50: 91-98. 4
SC
for blood cultures from febrile patients with bacteremia. J Clin Microbiol
Washington JA, Ilstrup DM. Blood cultures: Issues and controversies. Rev Infect Dis 1986; 8: 892-902.
5
Wilson ML, Mitchell M, Morris AJ et al. Principles and procedures for
TE D
blood cultures; approved guidelines. Clinical Laboratory Standards Institute 2007; 27: M47-A. 6
James LI, Plorde JJ, Carlson LG. Effects of blood volume and
7
EP
periodicity on blood cultures. J Clin Microbiol 1994; 32: 2829-2831. Tenney JH, Reller L B, Mirrett S et al. Controlled evaluation of the
AC C
volume of blood cultured in detection of bacteremia and fungemia. J Clin Microb 1982; 15: 558-5561.
8
Cockerill FR, Wilson JW, Vetter EA et al. Optimal testing parameters
for blood cultures. Clin Infect Dis 2004; 38: 1724-1730.
12
ACCEPTED MANUSCRIPT
9
Lee A, Mirrett S, Reller LB et al. Detection of bloodstream infections in adults: how many blood cultures are needed? J Clin Microbiol 2007; 45: 3546-3548. Weinstein MP. Blood culture contamination: persisting problems and
RI PT
10
partial progress. J Clin Microbiol 2003; 41: 2275-2278. 11
Riedel S, Carroll KC. Blood cultures: key elements for best practices
12
SC
and future directions. J Infect Chemother 2010; 16: 301-316.
Afshar N, Tabas J, Afshar K et al. Blood cultures for community-
M AN U
acquired pneumonia: are they worthy of two quality measures? A systematic review. J Hosp Med 2009; 4: 112-113. 13
Campbell SG, Marrie TJ, Anstey R et al. The contribution of blood cultures to the clinical management of adult patients admitted to the
TE D
hospital with community-acquired pneumonia: a prospective observational study. Chest 2003; 123: 1142-1150. 14
Chen Y, Nitzan O, Saliba B et al. Are blood cultures necessary in the
EP
management of women with complicated pyelonephritis? J Infect 2006; 53: 235-240.
Coburn B, Morris AM, Tomlinson G et al. Does this adult patient with
AC C
15
suspected bacteremia require blood cultures? JAMA 2012; 308: 502511.
16
Kang CK, Kim ES, SONG k-h et al. Can a routine follow-up blood
culture be justified in Klebsiella pneumoniae bacteremia? a retrospective case-control study. BMC Infect Dis 2013; 13: 365.
13
ACCEPTED MANUSCRIPT
17
Makam AN, Auerbach AD, Steinman MA. Blood culture use in the emergency department in patients hospitalized with respiratory symptoms due to a nonpneumonia illness. JAMA 2014; Intern Med
18
RI PT
174: 803-806.
Velasco M, Martinez JA, Moreno-Martinez A et al. Blood cultures for
Clin Infect Dis 2003; 37: 1127-1130. 19
Parikh K,. Davis AB, Pavuluri P. Do we need this blood culture?
M AN U
Hospital Pediatrics 2014; 4: 78-84. 20
SC
women with uncomplicated acute pyelonephritis: are they necessary?
Schroeder AR. The never-ending quest to detect bacteremia: time for a culture change. Hospital Pediatrics 2014; 4: 85-87.
21
Chong YP, Moon SM, Bang KM, et al. Treatment duration for
TE D
uncomplicated Staphylococcus aureus bacteremia to prevent relapse: analysis of a prospective observational cohort study. Antimicrob Agents Chemother 2013; 57: 1150-1156. Richter SS, Beekmann SE, Croco JL et al. Minimizing the work up of
EP
22
blood culture contaminants: implementation and evaluation of a
AC C
laboratory-based algorithm. J Clin Microbiol 2002; 40: 2437-2444.
23
MacGregor RR, Beaty HN. Evaluation of positive blood cultures: guidelines for early differentiation of contamination from valid positive culture. Arch Intern Med 1972; 130: 84-87.
24
Mirrett S, Weinstein MP, Reimer LG et al. Relevance of the number of positive bottles in determining clinical significance of coagulase-
14
ACCEPTED MANUSCRIPT
negative staphylococci in blood cultures. J Clin Microbiol 2001; 39: 3279-3281. 25
Domingguez-de VE, Algora-Webber A, Millen I et al. Early evaluation
RI PT
of coagulase negative Staphylococcus in blood samples of intensive
care unit patients. A clinically uncertain judgment. Intensive Care Med 1987; 13: 390-394.
Khatib R, Riederer KM, Clark JA et al. Coagulase-negative
SC
26
staphylococci in multiple blood cultures: strain relatedness and
M AN U
determinants of same-strain bacteremia. J Clin Microbiol 1995; 33: 816-820. 27
Lee DH, Kim SC, Bae IG, Koh EH, Kim S. Clinical evaluation of BacT/Alert FA plus and FN plus bottles compared with standard
28
TE D
bottles. J Clin Microbiol 2013; 51: 4150-4155.
Zardroga R, Williams DN, Gottschall R et al. Comparison of 2 blood culture media shows significant differences in bacterial recovery for
AC C
EP
patients on antimicrobial therapy. Clin Infect Dis 2013; 56: 790-797.
15
ACCEPTED MANUSCRIPT
Table 1: The organisms recovered from 1887/15394 blood cultures during the study period, stratified into bacteremia and contaminations: n (%) Bacteremia Contaminant N=1668 N=404 The organism (N) Staphylococcus aureus (531) 530 (31.5) 1 (0.2) 143 (8.6)
Coagulase-negative staphylococci 398
128 (7.3)
Streptococcus pneumoniae (36)
36 (2.2)
Streptococcus milleri (21)
21 (1.3)
Viridans group streptococci (41)
23 (1.3)
β hemolytic streptococci (49)
49 (2.9)
270 (68.1) 0 0
SC
M AN U
Bacillus spp. (37)
21 (5.0)
RI PT
Enterococcus spp. (164)
18 (4.5) 0
11 (0.6)
26 (6.0)
7 (0.7)
37 (8.5)
Other gram-positive organisms (37)
15 (1.2)
22 (5.2)
Escherichia coli (244)
244 (14.6)
0
Other Enterobacteriaceae (250)
250 (15.1)
0
Pseudomonas aeruginosa (56)
56 (3.4)
0
Other gram-negative bacilli (46)
46 (2.6)
0
Anaerobes (79)
71 (4.3)
8 (2.5)
Yeast (38)
38 (2.4)
0
Mycobacterium mucogen (1)
0
1
AC C
EP
TE D
Corynobacterium spp. (44)
16
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 2: Potential impact of bacteremia missed by the first set on antibiotic treatment stratified according to the clinical setting: n (%). The clinical setting Impact on antibiotic therapy No change Change Extension Additional set >1 hour (n=31) 13 (41.9) 13 (41.9) 5 (16.1) Following a procedure (n=56)a 15 (26.8) 16 (28.6) 25 (44.6) Known bacteremic (n=23) 1 (4.3) 4 (17.4) 18 (78.3) Polymicrobial (n=19) 9 (47.4) 8 (42.1) 2 (10.5) Antibiotics (n=72) 22 (30.6) 24 (33.3) 26 (36.1) Source Vascular (n=26) 1 (3.8) 8 (30.8) 17 (65.4) Respiratory (n=17) 7 (41.2) 3 (17.6) 7 (41.2) Urine (n=27) 16 (59.3) 11 (40.7) 0 Other tissues (n=43) 23 (53.5) 10 (23.2) 10 (23.2) Unknown (n=10) 0 8 (80.0) 2 (20.0) Over all impact (n=123) 47 (38.2) 40 (32.5) 36 (29.3) a: exchange of blocked indwelling catheters, endotracheal suctioning, incision and drainage of an abscess and debridement of necrotic tissues.
17
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Additional culture results in blood culture series.
RI PT
Legend to figure 1
18
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1
19
S1198-743X(14)00105-0
DOI:
10.1016/j.cmi.2014.11.019
Reference:
CMI 104
To appear in:
Clinical Microbiology and Infection
Received Date: 22 September 2014 Revised Date:
18 November 2014
Accepted Date: 19 November 2014
Please cite this article as: Khatib R, Simeunovic G, Sharma M, Fakih MG, Johnson LB, Briski L, Lebar W, Blood culture series benefit may be limited to selected clinical conditions: time to reassess, Clinical Microbiology and Infection (2014), doi: 10.1016/j.cmi.2014.11.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Original Article
RI PT
Blood culture series benefit may be limited to selected clinical conditions: time to reassess.
SC
Riad Khatib, MDa, Gordana Simeunovic, MDa, Mamta Sharma, MDa, Mohamad G. Fakih, MD, MBAa, Leonard B. Johnson, MDa, Laurence Briski, MDb, William
M AN U
Lebar, PhDb,c
Departments of Medicinea and Pathologyb St. John Hospital and Medical Center,
TE D
Grosse Pointe Woods, MI USA.
EP
Running title: Limited benefit of blood culture series: time to reassess. Corresponding author
AC C
Riad Khatib, MD.
19251 Mack Avenue, Grosse Pointe Woods, MI 48236. email: [email protected]. Fax:01-313-3437840; Tel:0-1313-3433802
1
ACCEPTED MANUSCRIPT
Abstract
Blood cultures are often submitted as series (2-3 sets/24 h) to maximize
RI PT
recovery. We assessed the actual benefit of additional sets. Methods: Blood cultures submitted from adults (≥ 18 years-old) over one year (2/1/2012-1/31/2013) were examined. The medical records of patients with
SC
positive cultures were reviewed. Cultures with commensal organisms were
considered contamination in the absence of a source and clinical findings. The
M AN U
impact of additional sets on antibiotic therapy was estimated. We evaluated 15394 blood cultures. They were submitted as 2-5 set/24 hour in 12236 (79.5%) instances. Pathogens were detected in 1227 sets, representing 741 bacteremias of which 618 (83.4%) were detected in the first set and 123 (16.6%) in the additional sets. Pathogens missed in the first set were recovered from patients
TE D
on antibiotics (n=72; 58.5%) and following a procedure (n=54; 43.9%). The additional sets result could have influenced antibiotic therapy in 76/6235 (1.2%)
EP
instances including 40 (0.6%) antibiotic switch and 36 (0.6%) possible extension of therapy. The potential impact of detection of missed pathogen on antibiotic
AC C
therapy was mot apparent in patients with an endovascular focus (26/27; 96.3%) and those without an obvious source (10/10; 100%). These findings suggest that one blood culture is probably adequate in patients with an obvious source. Blood culture series are beneficial in patients without an obvious source and with those with endovascular infections. It is time to reassess blood culture series benefit and may be limiting them to selected conditions. Keywords: Blood culture; bacteremia; contaminant; culture series; limitations.
2
ACCEPTED MANUSCRIPT
Introduction
RI PT
Detection of bacteremia or fungemia by blood culture is essential for proper
management of patients with infection. Current guidelines advocate 2-3 sets
obtained simultaneously from different venipunctures or serially within 0-24 hours
SC
for optimal detection of bacteremia and interpretation of blood culture results [15]. The potential benefits of multiple sets include better detection of bacteremia,
M AN U
distinguishing contaminants and the recognition of persistent bacteremia [4]. The actual benefit was attributed to simple increase in the blood volume rather than numbers of venipunctures [6, 7]. These guidelines are based on several studies that illustrate better detection of bacteremia with more cultures and larger blood
TE D
volume [2, 8, 9-11]. Most of these studies focus on the detection of bacteremia without assessing the overall benefit of additional sets, whether they alter the patient management and if they are more applicable to selected clinical settings.
EP
With the increasing workload in the microbiology laboratory, cost containment efforts, and the drive for early initiation of antibiotics, we assessed the benefit of
AC C
more than one culture/24 hour.
Material and Methods:
Facilities
3
ACCEPTED MANUSCRIPT
Our hospital is an 804-bed teaching hospital in Detroit Michigan. Our microbiology laboratory receives blood cultures from the inpatients, emergency department, affiliated outpatient settings, as well from other affiliated chronic and
RI PT
acute care facilities. We selected cultures submitted from our adult (≥18 years
old) inpatients, in the emergency departments and affiliated outpatient settings. We excluded cultures from chronic care facilities and other hospitals. We
SC
performed a retrospective evaluation of the results of blood cultures processed
Blood culture processing
M AN U
between February 1st 2012 and January 31st, 2013.
The blood culture system used in our hospital during the study period was the BacTAlert with FA FAN® Aerobic and Anaerobic bottles (bioMerieux) with instructions for 10 ml blood draw/bottle. The phlebotomists are expected to mark
TE D
the level of media in each bottle and the desired fill volume, which is 10 ml in adults, using the pre-marked five ml increments on the bottle label prior to blood collection. Additionally, about 20% of blood culture bottles are randomly selected
EP
periodically and examined to assure compliance with proper blood volume fill; and, feedback is provided to the phlebotomists.
AC C
2.3 Assessment of blood cultures We reviewed the medical records of all patients with positive cultures. We excluded line draws and patients with inadequate medical record documentation for assessment. We abstracted the clinical findings, the condition at the time of culture, the source of bacteremia with the source culture results, and the clinician
4
ACCEPTED MANUSCRIPT
approach to the blood culture result. The reviewers were not blinded to the blood culture results. Definitions
RI PT
A blood culture set was defined as aerobic and anaerobic blood culture bottles
obtained from one venipuncture site. A single set was defined as one set/24 h. Culture series represented two or more sets/24 h. The first and additional sets
SC
referred to the first and subsequently drawn sets within 24 h. In simultaneously drawn sets, the set entered first in the record was considered the first.
M AN U
Commensal organisms signified organisms recognized as potential contaminants such as coagulase-negative staphylococci, viridans group streptococci, Bacillus spp., Corynobacterium spp., Propionibacterium spp., Aerococcus spp., and Micrococcus spp. Isolation of one of these organisms was considered true
TE D
bacteremia in the presence of a source and clinical signs, contamination in the absence of a source and clinical signs and of uncertain status in the presence of a potential source or clinical signs but not both [10-12]. Cases with uncertain
EP
status were counted as bacteremias. An episode of bacteremia was defined as the recovery of the same pathogen from single or multiple sets within the same
AC C
day and the recovery of a different pathogen by the additional sets within the same day.
The impact of the additional set result on antibiotic treatment was estimated based on whether the usual empirical therapy for the disease process was adequate to cover the yielded pathogens, if treatment duration needed to be
5
ACCEPTED MANUSCRIPT
extended based on documenting persistent bacteremia and if treatment could have been discontinued.
RI PT
Statistical methods
Chi square test was used to assess the significance of differences in categorical variables and the significance of trends was assessed by the Extended Mantel-
SC
Haenszel chi square test for linear trend, utilizing the computer software SPSS
M AN U
release 20. P value <0 .05 was considered to indicate statistical significance.
The study was approved by the St John Hospital and Medical Center Institutional
TE D
Review Board.
EP
Results
Study population
AC C
A total of 18507 blood cultures were reviewed, 3113 were excluded including 2311 from pediatric cases, 788 from chronic care facilities, four line draws and ten with inadequate medical records for assessment. The 15394 included cultures were obtained from 5251 patients encompassing 6854 BC (44.5%) from the general wards, 6006 (39.0%) the emergency department, 1999 (13.0%) intensive care units, 459 (3.0%) outpatients and 76 (0.5%) labor and delivery.
6
ACCEPTED MANUSCRIPT
The cultures were submitted as a single set/24h in 3158 (20.5%) and series of 25 sets/24 h in 12236 (79.5%) instances, corresponding to 6001 series with 6235 additional sets. The interval between blood cultures sets was ≤30 minutes in
RI PT
4759 (76.3%) of the cases including 162 (2.6%) simultaneous submissions.
Blood culture results
SC
A total of 2072 organisms were recovered from 1887 (12.3%) blood culture sets including 1601 pathogens, 404 contaminants and 67 organisms of uncertain
M AN U
status that were considered pathogens. The organisms recovered are shown in table 1. Detection of pathogens was similar in single set and culture series (337 [10.7%] vs. 1227 [10.0%]; p=0.3).
TE D
The value of additional blood culture sets
Detection of bacteremia
EP
The majority of additional sets (n=5522; 88.6%) was negative (Figure 1). Pathogens were detected in 1227 sets, corresponding to 741 episodes of
AC C
bacteremias among 239 patients. These episodes were detected by the first set in 618 (83.4%) and the additional sets in 123 (16.6%) instances. The characteristics of bacteremia missed by the first set are shown in table 2. Most patients were on antibiotic therapy and the additional sets were obtained after a procedure such as exchange of obstructed indwelling urinary catheter, deep respiratory suctioning in an intubated patient, incision and drainage of an abscess or debridement of necrotic tissues. They included follow up cultures
7
ACCEPTED MANUSCRIPT
from patients with known bacteremia and instances of polymicrobial bacteremia with an extra pathogen identified in the additional sets. Many of these patients had a focus of infection with defined microbiology, such as pneumonia, urinary
RI PT
tract infection and soft tissue/bone infection that would have warranted
appropriate antibiotic treatment. The additional culture set result could have
influenced antibiotic therapy in 76/6235 (1.2%) instances including antibiotic
SC
switch (n=40; 0.7%) and extension of therapy (n=36; 0.6%).
M AN U
The pathogens missed by the first set were detected by the second set in 119 (96.8%) instances, the third set in three (2.4%) and the fourth set in one (0.8%). Stratifying culture results according to the interval between serial cultures revealed that the missed pathogens were least likely to be recovered in additional sets obtained 0-30 minutes (72/4759 [1.5%]) compared to 31-60
TE D
minutes (14/506 [2.8%]) and > 61 minutes (31/970 [3.2%]) after the first set (p=0.0008; extended Mantel-Haenszel chi square for linear trend).
AC C
EP
Discussion
Current guidelines for blood cultures advocate 2-3 sets obtained simultaneously or within 24 h to maximize recovery of bacteremia and help in interpreting the isolation of common contaminants [1, 4, 5]. For detection of bacteremia, comparing single to multiple blood cultures is likely to reveal that the more cultures and/or the larger blood volume the better detection of bloodstream
8
ACCEPTED MANUSCRIPT
invasion [2, 4, 5, 8-10]. The major drawback of these studies is the testing of the same cultures that are being used to define bacteremia [6] and focusing on the diagnosis of bacteremia without addressing the actual impact of blood culture
RI PT
series on antibiotic therapy and patient outcome. In fact, the usefulness of blood culture in certain conditions such as urinary tract infection, cellulitis, and
pneumonia has been questioned [12-20]. Blood culture results in these settings
SC
may help optimizing antibiotic therapy but the benefit of more than one culture
M AN U
has not been substantiated.
Another potential value of multiple blood culture sets is distinguishing between persistent and transient bacteremia. This distinction may be valuable in suggesting an intravascular focus [10] and in modifying the treatment duration in Staphylococcus aureus bacteremia [21] and potentially other bacteremias but its
TE D
relevance in most clinical conditions is largely unknown.
Our findings support that detection of bacteremia was improved by a second set.
EP
However, the percentage of additional sets with missed pathogens was much higher in sets obtained after 30 minutes from the first culture. They may have
AC C
been obtained because of a change in the patient condition but determining the actual reason for late submission is difficult to determine retrospectively. Additionally, our findings show that the impact of the improved detection by the additional sets was most apparent in patients with suspected endovascular foci and those without an obvious source of infection.
9
ACCEPTED MANUSCRIPT
With respect to recognizing contamination, several tools exist [10, 22, 23-26]. They are based on the type of organism, the clinical features and the proportion of positive cultures. However, the value of these guidelines in patients with high
RI PT
risk for infection with common contaminants, such as those with devices or osteomyelitis, and patients on antimicrobial therapy has not been clearly
substantiated. Additionally, detecting different strains in multiple sets may
SC
represent bacteremia in one culture and a contaminant in the other. Finally,
represent bacteremia [26].
M AN U
isolating the same contaminant in two blood culture sets may not always
Limitations of our study include the retrospective review and the uncertainty of potential differences in blood draw volume in the first and additional cultures. We also did not examine blood culture media effects. A significant difference in
TE D
bacterial recovery was noted based on the type of media or bottles [27, 28]. Finally, our calculation of the impact of culture results on patient management
EP
and antibiotic therapy was imprecise. It was estimated based on the presumed empiric antibiotic therapy and the missed pathogen susceptibility and if detecting
AC C
bacteremia could have lead to extension of the duration of antibiotic therapy such as the finding of Staphylococcus aureus bacteremia [21]. Despite these limitations, our findings suggest that one blood culture is likely adequate in many conditions. Additional studies are needed to verify this observation and determine whether culture series benefit is more apparent in selected clinical settings such as endovascular infections and cases without an obvious source.
10
ACCEPTED MANUSCRIPT
Acknowledgement
RI PT
The study was supported by St John Hospital and Medical Center Department of
AC C
EP
TE D
M AN U
SC
Medical Education.
11
ACCEPTED MANUSCRIPT
References
Ntusi N, Aubin L, Oliver S et al. Guideline for the optimal use of blood cultures. S Afr Med J 2010; 100: 839-843.
2
RI PT
1
Riedel S, Bourbeau P, Swartz B, et al. Timing of specimen collection
2008; 46: 1381-1385. 3
Washington JA 2nd. Blood cultures: principles and techniques. Mayo
M AN U
Clinic Proceed 1975; 50: 91-98. 4
SC
for blood cultures from febrile patients with bacteremia. J Clin Microbiol
Washington JA, Ilstrup DM. Blood cultures: Issues and controversies. Rev Infect Dis 1986; 8: 892-902.
5
Wilson ML, Mitchell M, Morris AJ et al. Principles and procedures for
TE D
blood cultures; approved guidelines. Clinical Laboratory Standards Institute 2007; 27: M47-A. 6
James LI, Plorde JJ, Carlson LG. Effects of blood volume and
7
EP
periodicity on blood cultures. J Clin Microbiol 1994; 32: 2829-2831. Tenney JH, Reller L B, Mirrett S et al. Controlled evaluation of the
AC C
volume of blood cultured in detection of bacteremia and fungemia. J Clin Microb 1982; 15: 558-5561.
8
Cockerill FR, Wilson JW, Vetter EA et al. Optimal testing parameters
for blood cultures. Clin Infect Dis 2004; 38: 1724-1730.
12
ACCEPTED MANUSCRIPT
9
Lee A, Mirrett S, Reller LB et al. Detection of bloodstream infections in adults: how many blood cultures are needed? J Clin Microbiol 2007; 45: 3546-3548. Weinstein MP. Blood culture contamination: persisting problems and
RI PT
10
partial progress. J Clin Microbiol 2003; 41: 2275-2278. 11
Riedel S, Carroll KC. Blood cultures: key elements for best practices
12
SC
and future directions. J Infect Chemother 2010; 16: 301-316.
Afshar N, Tabas J, Afshar K et al. Blood cultures for community-
M AN U
acquired pneumonia: are they worthy of two quality measures? A systematic review. J Hosp Med 2009; 4: 112-113. 13
Campbell SG, Marrie TJ, Anstey R et al. The contribution of blood cultures to the clinical management of adult patients admitted to the
TE D
hospital with community-acquired pneumonia: a prospective observational study. Chest 2003; 123: 1142-1150. 14
Chen Y, Nitzan O, Saliba B et al. Are blood cultures necessary in the
EP
management of women with complicated pyelonephritis? J Infect 2006; 53: 235-240.
Coburn B, Morris AM, Tomlinson G et al. Does this adult patient with
AC C
15
suspected bacteremia require blood cultures? JAMA 2012; 308: 502511.
16
Kang CK, Kim ES, SONG k-h et al. Can a routine follow-up blood
culture be justified in Klebsiella pneumoniae bacteremia? a retrospective case-control study. BMC Infect Dis 2013; 13: 365.
13
ACCEPTED MANUSCRIPT
17
Makam AN, Auerbach AD, Steinman MA. Blood culture use in the emergency department in patients hospitalized with respiratory symptoms due to a nonpneumonia illness. JAMA 2014; Intern Med
18
RI PT
174: 803-806.
Velasco M, Martinez JA, Moreno-Martinez A et al. Blood cultures for
Clin Infect Dis 2003; 37: 1127-1130. 19
Parikh K,. Davis AB, Pavuluri P. Do we need this blood culture?
M AN U
Hospital Pediatrics 2014; 4: 78-84. 20
SC
women with uncomplicated acute pyelonephritis: are they necessary?
Schroeder AR. The never-ending quest to detect bacteremia: time for a culture change. Hospital Pediatrics 2014; 4: 85-87.
21
Chong YP, Moon SM, Bang KM, et al. Treatment duration for
TE D
uncomplicated Staphylococcus aureus bacteremia to prevent relapse: analysis of a prospective observational cohort study. Antimicrob Agents Chemother 2013; 57: 1150-1156. Richter SS, Beekmann SE, Croco JL et al. Minimizing the work up of
EP
22
blood culture contaminants: implementation and evaluation of a
AC C
laboratory-based algorithm. J Clin Microbiol 2002; 40: 2437-2444.
23
MacGregor RR, Beaty HN. Evaluation of positive blood cultures: guidelines for early differentiation of contamination from valid positive culture. Arch Intern Med 1972; 130: 84-87.
24
Mirrett S, Weinstein MP, Reimer LG et al. Relevance of the number of positive bottles in determining clinical significance of coagulase-
14
ACCEPTED MANUSCRIPT
negative staphylococci in blood cultures. J Clin Microbiol 2001; 39: 3279-3281. 25
Domingguez-de VE, Algora-Webber A, Millen I et al. Early evaluation
RI PT
of coagulase negative Staphylococcus in blood samples of intensive
care unit patients. A clinically uncertain judgment. Intensive Care Med 1987; 13: 390-394.
Khatib R, Riederer KM, Clark JA et al. Coagulase-negative
SC
26
staphylococci in multiple blood cultures: strain relatedness and
M AN U
determinants of same-strain bacteremia. J Clin Microbiol 1995; 33: 816-820. 27
Lee DH, Kim SC, Bae IG, Koh EH, Kim S. Clinical evaluation of BacT/Alert FA plus and FN plus bottles compared with standard
28
TE D
bottles. J Clin Microbiol 2013; 51: 4150-4155.
Zardroga R, Williams DN, Gottschall R et al. Comparison of 2 blood culture media shows significant differences in bacterial recovery for
AC C
EP
patients on antimicrobial therapy. Clin Infect Dis 2013; 56: 790-797.
15
ACCEPTED MANUSCRIPT
Table 1: The organisms recovered from 1887/15394 blood cultures during the study period, stratified into bacteremia and contaminations: n (%) Bacteremia Contaminant N=1668 N=404 The organism (N) Staphylococcus aureus (531) 530 (31.5) 1 (0.2) 143 (8.6)
Coagulase-negative staphylococci 398
128 (7.3)
Streptococcus pneumoniae (36)
36 (2.2)
Streptococcus milleri (21)
21 (1.3)
Viridans group streptococci (41)
23 (1.3)
β hemolytic streptococci (49)
49 (2.9)
270 (68.1) 0 0
SC
M AN U
Bacillus spp. (37)
21 (5.0)
RI PT
Enterococcus spp. (164)
18 (4.5) 0
11 (0.6)
26 (6.0)
7 (0.7)
37 (8.5)
Other gram-positive organisms (37)
15 (1.2)
22 (5.2)
Escherichia coli (244)
244 (14.6)
0
Other Enterobacteriaceae (250)
250 (15.1)
0
Pseudomonas aeruginosa (56)
56 (3.4)
0
Other gram-negative bacilli (46)
46 (2.6)
0
Anaerobes (79)
71 (4.3)
8 (2.5)
Yeast (38)
38 (2.4)
0
Mycobacterium mucogen (1)
0
1
AC C
EP
TE D
Corynobacterium spp. (44)
16
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 2: Potential impact of bacteremia missed by the first set on antibiotic treatment stratified according to the clinical setting: n (%). The clinical setting Impact on antibiotic therapy No change Change Extension Additional set >1 hour (n=31) 13 (41.9) 13 (41.9) 5 (16.1) Following a procedure (n=56)a 15 (26.8) 16 (28.6) 25 (44.6) Known bacteremic (n=23) 1 (4.3) 4 (17.4) 18 (78.3) Polymicrobial (n=19) 9 (47.4) 8 (42.1) 2 (10.5) Antibiotics (n=72) 22 (30.6) 24 (33.3) 26 (36.1) Source Vascular (n=26) 1 (3.8) 8 (30.8) 17 (65.4) Respiratory (n=17) 7 (41.2) 3 (17.6) 7 (41.2) Urine (n=27) 16 (59.3) 11 (40.7) 0 Other tissues (n=43) 23 (53.5) 10 (23.2) 10 (23.2) Unknown (n=10) 0 8 (80.0) 2 (20.0) Over all impact (n=123) 47 (38.2) 40 (32.5) 36 (29.3) a: exchange of blocked indwelling catheters, endotracheal suctioning, incision and drainage of an abscess and debridement of necrotic tissues.
17
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Additional culture results in blood culture series.
RI PT
Legend to figure 1
18
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1
19