Current approaches to the diagnosis of bacterial and fungal bloodstream infections in the intensive care unit

Patrick R. Murray, PhD; Henry Masur, MD

Healthcare systems spend considerable resources collecting of organism detection, thus providing valuable guidance to clini- and processing blood cultures for the detection of blood stream cians who strive to initiate the appropriate antimicrobial therapy pathogens. The process is initiated with the collection of blood as rapidly as possible, and who wish to discontinue unnecessary cultures that depend upon proper skin disinfection, collection drugs expeditiously. Molecular and mass spectroscopy techniques of an adequate number of specimens and volume of blood, and are changing sepsis diagnosis rapidly and will provide far more prompt processing in a sensitive culture system. Complement- specific information far more quickly, but the performance char- ing blood cultures and gaining in use are techniques such as acteristics of these systems must be understood by intensivists nucleic acid amplification tests and mass spectroscopy that allow who use such information to guide their patient management. (Crit clinical laboratories to detect and identify organisms from blood Care Med 2012; 00:0–0) cultures substantially faster than conventional systems. Further- KEY WORDS: bacteremia; blood culture; fungemia; matrix-assist- more, certain resistance mutations can be detected within hours ed laser desorption/ionization; molecular; sepsis

rompt, accurate detection and must recognize that technology is revo- Instruments were introduced in the early identification of bloodstream lutionizing the tools that are being used. 1970s that could automatically monitor pathogens are essential for opti- For the past 30 yrs, healthcare sys- the bottles for microbial growth (e.g., mal management of intensive tems have relied on techniques in which production of carbon dioxide) and alert staff Pcare unit patients with sepsis syndromes. blood is incubated in various media, when growth was detected. Subsequent Healthcare facilities spend considerable semiautomated instruments are used to refinements in both culture media and resources in terms of labor and equip- monitor microbial growth, and organ- detection systems have improved the overall ment drawing, processing and analyzing isms are identified by Gram stain and recovery and time to detection of organisms blood cultures. Missed opportunities to biochemical tests. In recent years, clinical in septic patients while reducing lab-related document the true cause of blood stream laboratories have begun to move to novel contamination of cultures. infections can adversely affect patient approaches: nucleic acid amplification The value of blood cultures for con- outcome if the true causative organism is tests and mass spectrometry are two of firming the clinical diagnosis of sepsis, not identified and treated with an active the most common approaches that have severe sepsis, and septic shock (i.e., dis- antibiotic. Conversely, the identification been introduced into clinical laboratories seminated infection from a localized of organisms in blood cultures that are for routine detection and identification focus such as meninges, lungs, abdomen, contaminants can lead to the unnecessary of organisms, and molecular tests are urinary tract, or from febrile neutropenia) administration of antimicrobial agents being used increasingly to rapidly iden- is suboptimal. Although most untreated that will increase cost, increase toxicity, tify microbial sequences that confer drug patients with bacterial meningitis have positive blood cultures, only 30% of and distract the clinical team from treat- resistance. Clinicians must understand patients with bacterial pneumonia and ing the true causative agent. Thus, tech- the sensitivities and specificities of results intra-abdominal infections have positive niques that are sensitive, specific, and derived from both conventional diagnos- cultures, and positive blood cultures in rapid for identifying the microbial cause tics and newer molecular approaches. The patients with urologic disease are primar- of sepsis are major operational tools for implications of using these technologies for the assessment of septic patients will ily restricted to those with acute pyelone- all critical care units, and intensivists be evaluated in this review. phritis. Only 5% to 15% of the all cultures drawn for any reason, and only 50% of From the Becton Dickinson Diagnostic Systems patients with septic shock, are positive. (PRM), Sparks, MD; and Department of Critical Care Evolution of Conventional Blood Medicine (HM), Clinical Center, National Institutes of Culture Whether the low rate for positive blood Health, Bethesda, MD. cultures is related to the sensitivity of the Patrick Murray is an employee of Becton Dickinson Historically, blood cultures were per- diagnostic techniques or the biology of and has equity interest. Dr. Masur has not disclosed any formed by inoculating a large volume of the infectious process is unclear. potential conflicts of interest. blood into one or more bottles of a nutrient Contaminants represent 15% to 30% For information regarding this article, E-mail: hma- [email protected] medium after which the bottles were of the isolated organisms in some hos- Copyright © 2012 by the Society of Critical Care examined visually each day for evidence pitals. Overall, the success of recovering Medicine and Lippincott Williams and Wilkins of microbial growth (e.g., visualization of pathogens and eliminating contaminants DOI: 10.1097/CCM.0b013e318270e771 discrete colonies, turbidity, gas production). is directly related to the techniques used

Crit Care Med 2012 Vol. 40, No. 12 1

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. to collect and process blood cultures, and multiple bottles or in multiple blood cul- the higher the yield of the process (10– the patient population being evaluated. ture sets, they need to be considered seri- 15). Cockerill et al (13) documented a Blood Culture Contaminants. Careful ously as true pathogens, especially if they 29.8% increase in positive cultures when attention to the blood drawing and bottle are detected in <24–48 hrs (time of blood 20 mL of blood (divided into two bottles) inoculation techniques are important to draw to time of laboratory detection). were cultured compared with 10 mL of maximize culture specificity for the true Cultures Drawn From Catheters. blood. Additional positive cultures were causative organism. This begins with Collection of blood cultures through observed when 30 mL (13.4% increase careful disinfection of the phlebotomy intravascular devices was traditionally vs. 20 mL) and 40 mL of blood (7.2% in- site (1, 2) with 70% ethanol followed by discouraged because the prevalence of crease vs. 30 mL) were cultured. The application of chlorhexidine (30 secs) or contaminated cultures is slightly higher blood culture yield also increases with tincture of iodine (1 min). Betadine prep- compared with venipuncture. However, the collection of additional blood cultures arations are not recommended because as more and more patients have indwell- (consisting of 20 mL of blood divided into they must be applied for 1.5–2 mins to be ing intravascular devices, clinicians have two bottles). Cockerill et al (13) also re- effective, and clinical staff are unlikely to recognized that blood draws through ported that when a minimum of four wait for this long (3). Catheter access sites catheters can increase the likelihood of blood culture sets were collected within as well as the rubber diaphragm on blood identifying the cause of sepsis because a 24-hr period, the yield increased with culture bottles should also be disinfected the catheter is often the infected nidus, each additional culture drawn: 61.4% of with 70% ethanol. The contamination and because patient phlebotomy can be the patients with blood stream pathogens rate should not exceed 2% to 3% of blood difficult and painful when patients are in had the causative organism detected with culture sets (a set consists of two to three intensive care units for many days, have the first collected culture, 78.2% with the bottles inoculated with a single blood col- coagulation disorders, and are subjected first two cultures, and 93.1% with the lection). Focused training for those draw- to many intravascular accesses. In fact, first three cultures. Lee et al (14) reported ing such cultures is likely to result in the likelihood of obtaining contaminants very similar data. substantial savings in terms of reducing by drawing blood through intravascular Whereas patients with catheter-related the prevalence and consequences of con- catheters is only slightly higher than the sepsis, endocarditis, or other intravas- taminated cultures. risk associated with venipuncture (5, 6). cular infections may be persistently It is no longer recommended that the An important concept for interpreting bacteremic, most other infections are needle inserted at the phlebotomy site be the source of bacteremia and fungemia is associated with intermittent bacteremia replaced with a second needle before the the recognition that the “time to positiv- or fungemia. Although it is commonly blood is transferred to the blood culture ity” (time from when the blood culture believed that high-grade seeding of the bottles. Although there is a small decrease was drawn until the positive result was blood corresponds to temperature eleva- in contaminated cultures with such needle detected by the automated system) is tions, Riedel et al (16) demonstrated in a exchange (4), this benefit does not out- meaningful. If a culture drawn through multicenter study that timing collection weigh the risk to the phlebotomist of nee- one lumen is positive at least 90–120 of blood cultures with temperature eleva- dle-associated injury (e.g., transmission of mins before a culture drawn from another tions did not increase the yield of blood human immunodeficiency virus, hepatitis site (percutaneously or another catheter cultures. B virus, hepatitis C virus, or rarely other or lumen), the results suggest that the Because clinical signs including fever pathogens), should there be a mishap in lumen with the earlier report of positivity and symptoms cannot be used to pre- which the phlebotomist sustains a sharp has a higher concentration of organisms dict the optimum time for specimen injury while changing needles. and is the source of the infection (7). collection, the Society for Critical Care Current blood culture instruments are Recent studies (8, 9) documented the Medicine, Infectious Diseases Society of “closed” systems that detect microbial need to culture all lumens in multilu- America, Surviving Sepsis Campaign, and growth with external monitoring devices. men catheters to avoid missing a signifi- Clinical and Laboratory Standards Com- This means that virtually all contami- cant number of catheter-related septic mittee recommend that a minimum of nants in blood cultures originate from the events. However, for patients with mul- two blood cultures consisting of 20–30 mL skin or intravenous catheter surface when tiple lumens and/or multiple catheters, per culture (ideally one peripheral draw the blood is collected and inoculated into drawing cultures from each lumen may and one drawn through the catheter most bottles. Most contaminants are coagu- not be feasible due to considerations suspicious of being infected if line sepsis lase-negative Staphylococci, Bacillus, of volume of blood required and cost of is suspected) should be collected within Corynebacterium, or Propionibacterium. numerous cultures. Clinicians must then a 30-min period when a septic patient When these organisms are identified make a judgment as to which sites most is first evaluated, before antibiotics are by the laboratory, clinicians should be merit culture. Such decisions are influ- administered or changed, and additional suspicious that they represent contami- enced by which lumens are accessed most cultures should be collected over a 24-hr nants, especially if they take >48–72 hrs often, which lumens have recently failed period. to grow (suggesting they were present in to function optimally, physical findings of Recovery of Anaerobic Bacteria. small numbers) and are present in only erythema or tenderness or exudates, and Historically it was recommended that one bottle or one set of bottles. However, knowledge about the conditions under blood should be subdivided into an aero- these organisms can occasionally be true which the catheter was placed. bic bottle (supports the growth of strict pathogens, especially in patients with Effect of Blood Volume. The volume of aerobic and facultatively anaerobic [grows implanted hardware (prosthetic valves, blood cultured is a pivotal variable for the aerobically or anaerobically] bacteria as implanted cardiac devices, or mechani- successful recovery of bloodstream patho- well as yeasts) and an anaerobic bottle cal joints). When these pathogens grow in gens: the more volume that is cultured, (supports growth of strict anaerobic

2 Crit Care Med 2012 Vol. 40, No. 12

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. and facultatively anaerobic bacteria). Table 1. Bacteria unlikely to grow in standard blood culture broth systems Retrospective studies analyzing positive blood cultures in the 1970s and 1980s Bacteria Detection Method documented poor recovery of anaerobic Anaplasma species, Ehrlichia species Giemsa stain of peripheral blood; PCR available in bacteria. Although laboratory recommen- reference labs dations have evolved as various media Bartonella species Lysis centrifugation to chocolate agar incubated in have been developed and assessed, most capnophilic atmosphere at 37°C for 4 wks laboratories currently favor use of an Borrelia species Serology; culture methods and PCR are insensitive aerobic bottle and an anaerobic bottle for Brucella species Extended incubation in conventional culture system Campylobacter species, Helicobacter Lysis centrifugation to nonselective blood agar incubated in optimum recovery of a broad spectrum of species microaerobic atmosphere at 37°C bacteria and fungi (17–20). Coxiella burnetii Serology; culture insensitive Recovery of Fungi and Fastidious Francisella tularensis Extended incubation in conventional culture system Bacteria. The use of special medium Legionella pneumophila Lysis centrifugation to buffered charcoal yeast extract formulations for the recovery of yeasts agar incubated aerobically at 37°C Leptospira species ° is generally not necessary because Lysis centrifugation to EMJH broth incubated in air at 30 C; PCR available in reference labs most grow well in conventional aerobic Mycobacterium species Lysis centrifugation to Middlebrook agar incubated in blood culture broths within 2–3 days. capnophilic atmosphere at 37°C for 4–6 wks Exceptions to this rule include Candida Nocardia species Lysis centrifugation to nonselective blood agar incubated in glabrata and Cryptococcus neoformans, capnophilic atmosphere at 37°C for 2 wks which typically require 3–5 days of incu- Rickettsia species Serology bation. Fusarium and Paecilomyces can PCR, polymerase chain reaction. be recovered in conventional blood cul- ture broth, but most other filamentous fungi are not detected. Dimorphic molds incubation period is early detection of detection of resistance genes with results such as Histoplasma and Blastomyces positive cultures. More than 90% of all available in a few hours after specimen can grow in blood culture broths al- positive blood cultures are detected with- acquisition. However, such techniques though incubation for >2 weeks is re- in the first 48 hrs of incubation (13, 25), are not yet optimally developed: current quired which is generally impractical. and extended incubation beyond 5–7 days techniques for amplification of microbial Use of supplementary systems such as is rarely indicated unless the pathogens DNA for the detection and identification the lysis-centrifugation system (Isolator, listed in Table 1 are suspected (26, 27). of microorganisms in blood samples such Wampole Laboratories, Cranbury, NY) is One under-appreciated fact is that sig- as the SeptiFast system (Roche Molecular recommended for isolation of slow-grow- nificant delays between collection of blood Systems) have poor sensitivity and speci- ing molds and fastidious bacteria. cultures and initiation of incubation will ficity, are technically cumbersome requir- Table 1 lists organisms that are unlikely prolong detection times. These delays can ing specimen batching and a minimum to grow in standard blood culture sets and occur both at the patient’s bedside and of 6 hrs processing time, and provide no that require clinicians to alert the labo- in the laboratory. Kerremans et al (28, information about antimicrobial suscep- ratory for special processing in terms of 29) and van der Velden et al (30) demon- tibility results (33–36). It can be argued selective media, different incubation tem- strated in a series of elegant studies that that these tests should be used as comple- peratures, or extended incubation times. incubation delays for almost half of all mentary tests to traditional culture, and Antibiotic Inactivation Systems. blood cultures exceeded 4 hrs (including economic models have been developed Patients frequently receive antibiotics median transport times of 3.9 hrs and 16.0 that purport rapid polymerase chain that suppress the growth of bacteria and hrs for specimens from the intensive care reaction identification of microorgan- fungi. Manufacturers of most blood cul- unit and emergency department, respec- isms that has the potential to be a cost- ture systems supplement their media tively), preincubation at the collection site effective component for managing sepsis with proprietary formulations of anti- significantly reduced the time to detection (37); however, these models assume the biotic-binding resin beads or absorbent of positive cultures, and this resulted in molecular tests have sufficient sensitivity charcoal and Fuller’s earth. Analysis of more rapid adjustment of antibiotic ther- to detect all significant organisms with a the performance of these compounds has apy. Although blood culture instruments single test, are specific and detection of demonstrated superior performance of are rarely used outside the clinical lab, the microbial DNA in blood is always clini- the resins for removal of antibacterial and installation of such instruments proxi- cally significant, and are performed in a antifungal antibiotics, improved recovery mate to intensive care units should be hospital population where there is a high of bacteria and fungi, and decreased time considered in high volume settings where proportion of inadequate empirical ther- to detection of positive cultures (21–24). transport delays are likely. apy (38). Despite the current limitations Thus, because these substances are now of molecular sepsis tests, we believe it is routinely used in culture systems, clini- Molecular-Based Techniques appropriate to be optimistic that the rapid cians do not need to request “antibiotic progress in technology development will removal systems” for cultures drawn in Detection of Bacteremia and make such direct sample testing feasible patients receiving antimicrobial therapy. Fungemia. There is great interest in mo- and useful in the near future, potentially Detection Time. One significant ad- lecular techniques to diagnose sepsis in as point of care testing in emergency de- vantage with the use of automated blood blood samples taken directly from patients partments and intensive care units. culture systems that continuously moni- (31, 32). The goal is diagnosis of bacte- Microbial Identification. New ap- tor microbial growth throughout the remia or fungemia with simultaneous proaches are having impact on the

Crit Care Med 2012 Vol. 40, No. 12 3

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. identification of organisms once the or- of a positive blood culture isolate within antibiotics,” one of the most pivotal ganism is growing in the blood culture 1 hr of detection can be used to guide parameters in improving the prognosis of broth. The traditional approach for pro- empiric therapy. Although the results of patients with life-threatening infections. cessing a positive blood culture is to re- most antimicrobial susceptibility tests are move a portion of the broth, subculture not available for 8–24 hrs, 1- to 2-hr poly- REFERENCES it to agar media, and after overnight in- merase chain reaction tests for the pres- cubation select isolated colonies for iden- ence of genes that encode resistance to 1. Little JR, Murray PR, Traynor PS, et al: A ran- domized trial of povidone-iodine compared tification and antimicrobial susceptibility oxacillin, vancomycin, and the carbapen- with iodine tincture for venipuncture site tests. This process requires 1–3 days be- emases are commercially available and disinfection: Effects on rates of blood culture fore definitive results are available. The used in many clinical laboratories. These contamination. Am J Med 1999; 107:119–125 use of fluorescence in situ hybridization tests are useful; however, the tests may 2. Mimoz O, Karim A, Mercat A, et al: Chlorhexi- using peptide nucleic acid probes (PNA- provide misleading results because they dine compared with povidone-iodine as skin FISH) has been used for direct identifi- do not measure if the resistance gene is preparation before blood culture. A random- cation of Staphylococcus, Enterococcus, expressed. For example, presence of the ized, controlled trial. Ann Intern Med 1999; Klebsiella, and Candida species in posi- mecA gene is associated with methicil- 131:834–837 tive blood cultures (39–42) in <2 hrs. lin resistance in Staphylococcus aureus, 3. King TC, Price PB: An evaluation of iodo- phors as skin antiseptics. Surg Gynecol This identification approach is laborious and the vanA gene is associated with van- Obstet 1963; 116:361–365 because individual probes have to be cre- comycin resistance in Enterococcus fae- 4. Spitalnic SJ, Woolard RH, Mermel LA: The ated for each species. cium. If the regulatory genes that control significance of changing needles when inocu- More promising is mass spectrom- expression of these resistance markers are lating blood cultures: A meta-analysis. Clin etry that can identify organisms (but not inoperable, then the bacteria will remain Infect Dis 1995; 21:1103–1106 detect antibiotic susceptibility) within drug susceptible, and the gene detection 5. DesJardin JA, Falagas ME, Ruthazer R, et al: an hour or less from the culture broth, assay would mislead the clinician (55, 56). Clinical utility of blood cultures drawn from that is, from the time the system alarm Likewise, the absence of mecA or vanA indwelling central venous catheters in hospi- alerts the technician of growth. Mass cannot be used to predict susceptibility to talized patients with cancer. Ann Intern Med 1999; 131:641–647 spectrometry, specifically matrix-assisted methicillin or vancomycin because resis- laser desorption/ionization time-of-flight 6. Martinez JA, DesJardin JA, Aronoff M, et al: tance to the antibiotic may be the result Clinical utility of blood cultures drawn from (MALDI-TOF) mass spectrometry, has of another mechanism. For this reason, central venous or arterial catheters in criti- been used for identification of isolated assessment of antibiotic susceptibility cally ill surgical patients. Crit Care Med 2002; colonies of bacteria and fungi and is by genomic techniques should be con- 30:7–13 rapidly replacing biochemical and gene sidered a presumptive test that must be 7. Seifert H, Cornely O, Seggewiss K, et al: sequencing methods for organism iden- confirmed by the current phenotypic tests Bloodstream infection in neutropenic can- tification because it is highly accurate, that assess the growth of bacteria in the cer patients related to short-term nontun- inexpensive, and results are available in presence of the test antibiotic. It is likely nelled catheters determined by quantitative <1 hr. MALDI-TOF can also be used for that rapid susceptibility tests will con- blood cultures, differential time to positivity, and molecular epidemiological typing with the direct identification of bacteria and tinue to evolve beyond the detection of yeasts isolated in blood culture broths pulsed-field gel electrophoresis. J Clin Micro- resistance gene sequences or detection of biol 2003; 41:118–123 (43–49). Processing these specimens is growing organisms by visual methods and 8. Cuellar-Rodriguez J, Connor D, Murray P, et more complex because the nonmicrobial ultimately rely on early detection of the al: Discrepant results from sampling different cells, serum proteins, and broth culture response to antibiotics by measuring gene lumens of multilumen catheters: The case for nutrients must be removed before the expression or metabolic activity without sampling all lumens. In: IDSA meeting. Bos- microbial cells are evaluated. However, the need for cell division. ton, MA, 2011 definitive identification results from posi- 9. Robinson JL: Sensitivity of a blood culture tive blood culture broths are generally CONCLUSIONS drawn through a single lumen of a multilu- available in <1 hr after the technician is men, long-term, indwelling, central venous alerted to growth. Approximately 15% Molecular and mass spectroscopy catheter in pediatric oncology patients. J to 20% of the isolates are not initially techniques are changing sepsis diagnosis Pediatr Hematol Oncol 2002; 24:72–74 rapidly. These techniques provide sub- 10. Hall MM, Ilstrup DM, Washington JA 2nd: identified primarily because insufficient Effect of volume of blood cultured on detec- numbers of cells are in the positive blood stantially more rapid and more specific tion of bacteremia. J Clin Microbiol 1976; culture broth (e.g., skin bacteria such as information on organism identification 3:643–645 coagulase-negative Staphylococci and and on the presence of resistance mech- 11. Tenney JH, Reller LB, Mirrett S, et al: Con- Corynebacteria). However, modification anisms than conventional broth-based trolled evaluation of the volume of blood of the extraction procedures is improv- techniques. Critical care physicians will cultured in detection of bacteremia and ing the test sensitivity (50, 51). It should have to expeditiously learn what infor- fungemia. J Clin Microbiol 1982; 15:558–561 be noted that not all blood culture broth mation to expect from such systems, and 12. Mermel LA, Maki DG: Detection of bactere- formulations produce adequate results, how such information can be used to mia in adults: Consequences of culturing an particularly media supplemented with assure that initial antimicrobial regimens inadequate volume of blood. Ann Intern Med 1993; 119:270–272 charcoal (52–54). This technique has not are appropriate, and the unnecessary 13. Cockerill FR 3rd, Wilson JW, Vetter EA, et been used reliably on direct patient speci- drugs are discontinued. These techniques al: Optimal testing parameters for blood cul- mens, for example, blood samples. are expected to contribute substantially tures. Clin Infect Dis 2004; 38:1724–1730 Antimicrobial Susceptibility Tests. The to improving antibiotic stewardship 14. Lee A, Mirrett S, Reller LB, et al: Detection of ability to obtain a definitive identification and to improving “time to appropriate bloodstream infections in adults: How many

4 Crit Care Med 2012 Vol. 40, No. 12

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. blood cultures are needed? J Clin Microbiol Cardiobacterium, Eikenella, and Kingella for simultaneous dual-color identification 2007; 45:3546–3548 organisms: A retrospective multicenter evalu- of C. albicans and C. glabrata directly from 15. Patel R, Vetter EA, Harmsen WS, et al: Opti- ation. J Clin Microbiol 2006; 44:257–259 blood culture bottles. J Clin Microbiol 2008; mized pathogen detection with 30- compared 28. Kerremans JJ, van der Bij AK, Goessens W, 46:50–55 to 20-milliliter blood culture draws. J Clin et al: Needle-to-incubator transport time: 42. Søgaard M, Hansen DS, Fiandaca MJ, et al: Microbiol 2011; 49:4047–4051 Logistic factors influencing transport time Peptide nucleic acid fluorescence in situ 16. Riedel S, Bourbeau P, Swartz B, et al: Tim- for blood culture specimens. J Clin Microbiol hybridization for rapid detection of Klebsiella ing of specimen collection for blood cultures 2009; 47:819–822 pneumoniae from positive blood cultures. J from febrile patients with bacteremia. J Clin 29. Kerremans JJ, van der Bij AK, Goessens W, et Med Microbiol 2007; 56(Pt 7):914–917 Microbiol 2008; 46:1381–1385 al: Immediate incubation of blood cultures 43. La Scola B, Raoult D: Direct identification of 17. Cockerill FR 3rd, Hughes JG, Vetter EA, et outside routine laboratory hours of operation bacteria in positive blood culture bottles by al: Analysis of 281,797 consecutive blood cul- accelerates antibiotic switching. J Clin Micro- matrix-assisted laser desorption ionisation tures performed over an eight-year period: biol 2009; 47:3520–3523 time-of-flight mass spectrometry. PLoS ONE Trends in microorganisms isolated and the 30. van der Velden LB, Vos FJ, Mouton JW, et 2009; 4:e8041 value of anaerobic culture of blood. Clin al: Clinical impact of preincubation of blood 44. Stevenson LG, Drake SK, Murray PR: Infect Dis 1997; 24:403–418 cultures at 37°C. J Clin Microbiol 2011; Rapid identification of bacteria in posi- 18. Dorsher CW, Rosenblatt JE, Wilson WR, et 49:275–280 tive blood culture broths by matrix-assisted al: Anaerobic bacteremia: Decreasing rate 31. Bauer M, Reinhart K: Molecular diagnos- laser desorption ionization-time of flight over a 15-year period. Rev Infect Dis 1991; tics of sepsis–where are we today? Int J Med mass spectrometry. J Clin Microbiol 2010; 13:633–636 Microbiol 2010; 300:411–413 48:444–447 19. Lassmann B, Gustafson DR, Wood CM, et al: 32. Mancini N, Carletti S, Ghidoli N, et al: The 45. Ferroni A, Suarez S, Beretti JL, et al: Real- Reemergence of anaerobic bacteremia. Clin era of molecular and other non-culture-based time identification of bacteria and Candida Infect Dis 2007; 44:895–900 methods in diagnosis of sepsis. Clin Microbiol species in positive blood culture broths by 20. Murray PR, Traynor P, Hopson D: Critical Rev 2010; 23:235–251 matrix-assisted laser desorption ionization- assessment of blood culture techniques: 33. Josefson P, Strålin K, Ohlin A, et al: Evalu- time of flight mass spectrometry. J Clin Analysis of recovery of obligate and faculta- ation of a commercial multiplex PCR test Microbiol 2010; 48:1542–1548 tive anaerobes, strict aerobic bacteria, and (SeptiFast) in the etiological diagnosis 46. Christner M, Rohde H, Wolters M, et al: Rapid fungi in aerobic and anaerobic blood culture of community-onset bloodstream infec- identification of bacteria from positive blood bottles. J Clin Microbiol 1992; 30:1462–1468 tions. Eur J Clin Microbiol Infect Dis 2011; culture bottles by use of matrix-assisted laser 21. Viganò EF, Vasconi E, Agrappi C, et al: Use of 30:1127–1134 desorption-ionization time of flight mass simulated blood cultures for antibiotic effect 34. Varani S, Stanzani M, Paolucci M, et al: Diag- spectrometry fingerprinting. J Clin Microbiol on time to detection of the two blood culture nosis of bloodstream infections in immuno- 2010; 48:1584–1591 systems BacT/ALERT and BACTEC 9240. New compromised patients by real-time PCR. J 47. Prod’hom G, Bizzini A, Durussel C, et al: Microbiol 2004; 27:235–248 Infect 2009; 58:346–351 Matrix-assisted laser desorption ionization- 22. Flayhart D, Borek AP, Wakefield T, et al: 35. Dierkes C, Ehrenstein B, Siebig S, et al: time of flight mass spectrometry for direct Comparison of BACTEC PLUS blood culture Clinical impact of a commercially available bacterial identification from positive blood media to BacT/Alert FA blood culture media multiplex PCR system for rapid detection of culture pellets. J Clin Microbiol 2010; for detection of bacterial pathogens in sam- pathogens in patients with presumed sepsis. 48:1481–1483 ples containing therapeutic levels of antibiot- BMC Infect Dis 2009; 9:126 48. Ferreira L, Sánchez-Juanes F, Porras-Guerra ics. J Clin Microbiol 2007; 45:816–821 36. Wallet F, Nseir S, Baumann L, et al: Prelimi- I, et al: Microorganisms direct identification 23. Riedel S, Eisinger SW, Dam L, et al: Compari- nary clinical study using a multiplex real- from blood culture by matrix-assisted laser son of BD Bactec Plus Aerobic/F medium to time PCR test for the detection of bacterial desorption/ionization time-of-flight mass VersaTREK Redox 1 blood culture medium and fungal DNA directly in blood. Clin Micro- spectrometry. Clin Microbiol Infect 2011; for detection of Candida spp. in seeded blood biol Infect 2010; 16:774–779 17:546–551 culture specimens containing therapeutic 37. Lehmann LE, Herpichboehm B, Kost GJ, et 49. Kok J, Thomas LC, Olma T, et al: Identifica- levels of antifungal agents. J Clin Microbiol al: Cost and mortality prediction using poly- tion of bacteria in blood culture broths using 2011; 49:1524–1529 merase chain reaction pathogen detection in matrix-assisted laser desorption-ionization 24. Miller NS, Rogan D, Orr BL, et al: Compari- sepsis: Evidence from three observational tri- Sepsityper™ and time of flight mass spec- son of BD Bactec Plus blood culture media als. Crit Care 2010; 14:R186 trometry. PLoS ONE 2011; 6:e23285 to VersaTREK Redox blood culture media for 38. Schrenzel J: PCR for the diagnosis of sepsis: 50. Schubert S, Weinert K, Wagner C, et al: Novel, detection of bacterial pathogens in simulated Hope or hype? Crit Care 2011; 15:111 improved sample preparation for rapid, direct adult blood cultures containing therapeutic 39. Deck MK, Anderson ES, Buckner RJ, et al: identification from positive blood cultures concentrations of antibiotics. J Clin Microbiol Multicenter evaluation of the Staphylococcus using matrix-assisted laser desorption/ioniza- 2011; 49:1624–1627 QuickFISH method for simultaneous identi- tion time-of-flight (MALDI-TOF) mass spec- 25. Huang AH, Yan JJ, Wu JJ: Comparison of fication of Staphylococcus aureus and coag- trometry. J Mol Diagn 2011; 13:701–706 five days versus seven days of incubation for ulase-negative staphylococci directly from 51. Loonen AJ, Jansz AR, Stalpers J, et al: An detection of positive blood cultures by the blood culture bottles in less than 30 minutes. evaluation of three processing methods Bactec 9240 system. Eur J Clin Microbiol J Clin Microbiol 2012; 50:1994–1998 and the effect of reduced culture times for Infect Dis 1998; 17:637–641 40. Forrest GN, Roghmann MC, Toombs LS, et faster direct identification of pathogens from 26. Baron EJ, Scott JD, Tompkins LS: Prolonged al: Peptide nucleic acid fluorescent in situ BacT/ALERT blood cultures by MALDI-TOF incubation and extensive subculturing do hybridization for hospital-acquired entero- MS. Eur J Clin Microbiol Infect Dis 2012; not increase recovery of clinically signifi- coccal bacteremia: Delivering earlier effective 31:1575–1583 cant microorganisms from standard auto- antimicrobial therapy. Antimicrob Agents 52. Schmidt V, Jarosch A, März P, et al: Rapid mated blood cultures. Clin Infect Dis 2005; Chemother 2008; 52:3558–3563 identification of bacteria in positive blood 41:1677–1680 41. Shepard JR, Addison RM, Alexander BD, et culture by matrix-assisted laser desorption 27. Petti CA, Bhally HS, Weinstein MP, et al: Util- al: Multicenter evaluation of the Candida ionization time-of-flight mass spectrom- ity of extended blood culture incubation for albicans/Candida glabrata peptide nucleic etry. Eur J Clin Microbiol Infect Dis 2012; isolation of Haemophilus, Actinobacillus, acid fluorescent in situ hybridization method 31:311–317

Crit Care Med 2012 Vol. 40, No. 12 5

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. 53. Szabados F, Michels M, Kaase M, et al: The sen- rate of direct identification from positive mecA gene repressor: A surprising observa- sitivity of direct identification from positive cultures by matrix-assisted laser desorp- tion. PLoS ONE 2011; 6:e23287 BacT/ALERT™ (bioMérieux) blood culture tion/ionisation time-of-flight (MALDI- 56. Gagnon S, Lévesque S, Lefebvre B, et al: bottles by matrix-assisted laser desorption TOF) mass spectrometry. J Infect 2011; 62: vanA-containing Enterococcus faecium ionization time-of-flight mass spectrometry 251–253 susceptible to vancomycin and teicoplanin is low. Clin Microbiol Infect 2011; 17:192–195 55. Oliveira DC, de Lencastre H: Methicillin- because of major nucleotide deletions in 54. Romero-Gómez MP, Mingorance J: The resistance in Staphylococcus aureus is not Tn1546. J Antimicrob Chemother 2011; 66: effect of the blood culture bottle type in the affected by the overexpression in trans of the 2758–2762

6 Crit Care Med 2012 Vol. 40, No. 12

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.