Role of the Microbiology Laboratory and Molecular Diagnostics in Healthcare Epidemiology and Infection Prevention
Kaede V. Sullivan
Daniel J. Diekema
A primary goal of healthcare epidemiology is to protect patients and healthcare personnel (HCP) from infectious diseases and other microbial threats.1 The clinical microbiology laboratory contributes to this mission by detecting and reporting infectious agents and continually optimizing laboratory testing to ensure timely and accurate provision of key test results to the infection prevention (IP) and antimicrobial stewardship (AS) programs. In this chapter, we provide a general background on clinical microbiology laboratory operations and discuss the roles of the lab in IP surveillance, outbreak detection and management, and AS.
CLINICAL MICROBIOLOGY LABORATORY OPERATIONS
High-quality clinical microbiology testing requires optimizing activities at all phases of diagnostic testing, which begin at specimen ordering and end with result reporting. Figure 5-1 summarizes this process. Broadly, preanalytic activities are those that involve test ordering and specimen collection; analytic activities involve specimen processing and testing; and postanalytic activities encompass all aspects of test result reporting. We will discuss each category in turn.
One of the most important roles of the clinical microbiology laboratory is to clearly communicate correct specimen collection techniques and transportation requirements for laboratory submission. Incorrect specimen collection and transport can result in inaccurate test results, which may in turn lead to unnecessary labor in the laboratory, suboptimal clinical decisions by providers, and unnecessary patient charges. Best practices for specimen collection can be obtained from the literature and should be summarized in a specimen collection manual that is available hospitalwide.2,3 Further, to reduce inappropriate and unnecessary testing, microbiology leadership should partner with colleagues in IP, AS, infectious diseases, and other stakeholder groups to develop practice guidance for appropriate indications for ordering microbiology tests.
Once a test has been ordered and a specimen is collected and received by the laboratory, laboratory staff will assess specimen quality to ensure that it meets testing specifications. Clinical microbiology laboratories may cancel test requests for a variety of reasons. Common reasons include the following: aerobic swabs are received for anaerobic cultures; toxigenic Clostridioides difficile testing is requested on formed stool; and stool cultures are requested after 72 hours of hospitalization (because hospital-acquired gastroenteritis due to other bacterial pathogens is exceptionally rare). Further, sputum samples usually should not proceed to culture if a direct Gram stain shows large numbers of epithelial cells (eg, >25 per lowpower field under a light microscope). Finally, most laboratories also cancel test requests on unlabeled or incorrectly labeled specimens and specimens that have been received after excessively prolonged transport.3
If a specimen has been deemed to be appropriate for testing, the analytic activities begin. Cultivation of organisms from patient specimens through bacterial, fungal, and mycobacterial cultures is the bedrock of clinical microbiology services. Most organisms relevant to healthcare-associated infections (HAI; eg, Staphylococcus spp., Enterococcus spp., Enterobacteriaceae, Pseudomonas aeruginosa, and Candida spp.) grow easily and quickly on standard bacteriological media with colonies visible after 12-24 hours of incubation. Until recently, most laboratories assigned genus/species identifications to organisms using biochemical reactions that required at least 8-12 hours to complete. Automated biochemical identification instruments including BD Phoenix (BD, Sparks, MD), VITEK 2 (BioMérieux, Durham, NC), and MicroScan WalkAway (Beckman Coulter, Brea, CA) continue to be widely used throughout the world. These systems require preparation of an organism suspension from pure colonies for inoculation of a biochemical card or panel, which is then incubated in the instrument. Biochemical reactions are then read by the instrument and interpreted using proprietary software. In the last decade, however, automated biochemical identification has been increasingly supplanted by matrix-assisted laser desorption/ionization time-of-light mass spectrometry (MALDI-TOF MS), which has emerged as a faster, more accurate, and cost-efficient method of organism identification.4
“MALDI” refers to the matrix that assists in the desorption (lifting off of a surface) and ionization of microbial molecules through pulses of energy from a laser. The ionized proteins are directed through a positively charged electrostatic field into a time-of-flight (“TOF”) mass analyzer toward an ion detector with small molecules traveling fastest, followed by progressively larger ones. A mass spectrum is generated, representing the number of ions of a given mass impacting the ion detector over time. The spectrum is compared against those in a database containing spectra of known organisms.4 There are two U.S. Food and Drug Administration (FDA)-cleared MALDI-TOF MS systems that are commercially available in the United States: the Bruker Biotyper CA system (Bruker, Billerica, MA), and VITEK MS (BioMérieux). Both systems identify Gram-positive and Gram-negative organisms, anaerobes, and various yeast species within minutes. VITEK MS was also recently FDA-cleared for identification of various Nocardia spp., Mycobacterium spp., and molds.
“MALDI” refers to the matrix that assists in the desorption (lifting off of a surface) and ionization of microbial molecules through pulses of energy from a laser. The ionized proteins are directed through a positively charged electrostatic field into a time-of-flight (“TOF”) mass analyzer toward an ion detector with small molecules traveling fastest, followed by progressively larger ones. A mass spectrum is generated, representing the number of ions of a given mass impacting the ion detector over time. The spectrum is compared against those in a database containing spectra of known organisms.4 There are two U.S. Food and Drug Administration (FDA)-cleared MALDI-TOF MS systems that are commercially available in the United States: the Bruker Biotyper CA system (Bruker, Billerica, MA), and VITEK MS (BioMérieux). Both systems identify Gram-positive and Gram-negative organisms, anaerobes, and various yeast species within minutes. VITEK MS was also recently FDA-cleared for identification of various Nocardia spp., Mycobacterium spp., and molds.
After organism identification, antimicrobial susceptibility testing (AST) using methods such as disk diffusion, gradient diffusion (eg, E-test), broth dilution, and/or automated methods follows and typically requires an additional 18-24 hours to complete. BD Phoenix, VITEK 2, and MicroScan instruments also perform automated AST in addition to identification, but their performance may vary.5,6 Laboratory directors must therefore keep current with the literature pertaining to the performance of automated AST systems and be prepared to implement alternative AST methods when there are performance problems reported.
While most HAI-associated organisms are easily cultivated, Mycobacterium spp. and some fungi can require days to weeks to grow in culture. Viral cultures are also labor intensive, insensitive, and slow. For detection of these pathogens, many laboratories have implemented molecular methods such as polymerase chain reaction (PCR) that detect microbial nucleic acids, expedite testing turnaround time, and in some cases improve sensitivity. For example, molecular detection of respiratory viruses underwent early and rapid development. Today, rapid, multiplexed, fully automated respiratory pathogen panels are commercially available and are widely used. They require minimal specimen processing, can be performed on-demand 24/7, and are particularly helpful in the detection of nosocomial viral transmission. Mycobacterium tuberculosis complex has also been considered a logical candidate for rapid molecular diagnostics due to its slow growth in culture. In 2015, the FDA cleared Xpert MTB/RIF (Cepheid, Sunnyvale, CA) for testing of sputum specimens. Rapid diagnosis of MTB can expedite decisions related to application of airborne precautions in patients admitted with possible pulmonary tuberculosis and assist with bed management.
Finally, postanalytic activities focus on ensuring that test reports are accurate and clinically actionable. Reports should provide clear and understandable information about test results and indicate the testing method used (eg, “positive for influenza A by real-time PCR”). Reports may also communicate specimen quality problems (eg, on a urine culture report, “mixed growth suggesting contamination”) and/or clinical recommendations (eg, on a report of a multidrug-resistant organism (MDRO), “this patient should be placed in contact precautions”).
Laboratory personnel should call the IP program directly to report some results to ensure that appropriate control measures are implemented. Examples of organisms requiring immediate notification include Neisseria meningitidis, Bordetella pertussis, Legionella, acid-fast bacilli, and emerging threats such as carbapenem-resistant Enterobacteriaceae (CRE) and Candida auris (see below). In addition, new or unusual pathogens or potential agents of bioterrorism (eg, Bacillus anthracis, Yersinia pestis, orthopoxviruses) should be reported immediately to the IP program. The list of organisms for immediate notification may vary from one institution to another and should be developed in consultation with IP and bioemergency preparedness personnel.
SURVEILLANCE
Monitoring of clinical laboratory reports is essential to case finding for HAI surveillance, including those that are subject to public reporting. The Centers for Medicare and Medicaid Services (CMS) mandates the reporting of certain HAIs through the National Healthcare Safety Network (NHSN).7 States and counties may also mandate active surveillance for MDROs and/or reporting of certain organisms through legislation or municipal ordinances.
To ensure compliance with these reporting requirements, it is crucial for microbiology laboratories to promptly and accurately report positive test results and ensure that they are transmitted to the IP team. This communication may occur via phone calls, faxes, emails, or electronic surveillance systems that interface directly with the laboratory information system (LIS). Such systems allow IP teams to configure alerts and efficiently monitor laboratory results in real time. The following summarizes laboratory developments that impact HAI detection and surveillance.
Blood Cultures
Best practices for blood culture specimen collection are well established. To minimize contamination with skin flora at the time of blood collection, hand hygiene should be followed by skin disinfection ideally using alcoholic chlorhexidine gluconate or alcohol iodine (alcohol for rapidity and the other antiseptic for persistence), although alcohol, tincture of iodine, or chlorhexidine
gluconate solution alone may be used.8 After disinfecting the blood culture bottle septum, to maximize sensitivity, 20-30 mL of blood should be collected per septic episode, distributed to blood culture bottles according to manufacturer’s instructions, and the bottles should be transported to the laboratory within 2 hours of collection.3,9 To adjudicate situations in which possible contaminants are recovered (eg, coagulase-negative staphylococci, viridans group streptococci, Micrococcus, Bacillus spp.), a minimum of two blood culture sets should be obtained per septic episode prior to antibiotic administration.3 At least one of the blood cultures should be obtained through intact skin (eg, a “peripheral” blood culture) as blood obtained through an indwelling intravenous line is twice as likely to contain a surface contaminant.3 Use of blood diversion devices, which divert and sequester the first 1.5-2 mL of blood (likely carrying any contaminating skin flora present) during a blood draw, may also reduce blood culture contamination.10 Most clinical microbiology laboratories have implemented continuous monitoring blood culture systems (eg, BACTEC FX [BD], BACT/ALERT 3D [BioMérieux], VersaTREK [Thermo Fisher Scientific, Oakwood Village, OH]) that monitor for carbon dioxide production by bacteria and fungi proliferating in the blood culture broth. When carbon dioxide levels reach predefined thresholds, blood culture instruments alarm, prompting laboratory staff to use an aliquot of the blood culture broth to set up cultures on solid media. Organism identification and AST are performed on isolates recovered on those cultures using the methods previously described.
gluconate solution alone may be used.8 After disinfecting the blood culture bottle septum, to maximize sensitivity, 20-30 mL of blood should be collected per septic episode, distributed to blood culture bottles according to manufacturer’s instructions, and the bottles should be transported to the laboratory within 2 hours of collection.3,9 To adjudicate situations in which possible contaminants are recovered (eg, coagulase-negative staphylococci, viridans group streptococci, Micrococcus, Bacillus spp.), a minimum of two blood culture sets should be obtained per septic episode prior to antibiotic administration.3 At least one of the blood cultures should be obtained through intact skin (eg, a “peripheral” blood culture) as blood obtained through an indwelling intravenous line is twice as likely to contain a surface contaminant.3 Use of blood diversion devices, which divert and sequester the first 1.5-2 mL of blood (likely carrying any contaminating skin flora present) during a blood draw, may also reduce blood culture contamination.10 Most clinical microbiology laboratories have implemented continuous monitoring blood culture systems (eg, BACTEC FX [BD], BACT/ALERT 3D [BioMérieux], VersaTREK [Thermo Fisher Scientific, Oakwood Village, OH]) that monitor for carbon dioxide production by bacteria and fungi proliferating in the blood culture broth. When carbon dioxide levels reach predefined thresholds, blood culture instruments alarm, prompting laboratory staff to use an aliquot of the blood culture broth to set up cultures on solid media. Organism identification and AST are performed on isolates recovered on those cultures using the methods previously described.
The introduction of MALDI-TOF MS into microbiology laboratories has had some interesting consequences on the reporting of positive blood cultures due to its ability to accurately assign identifications to organisms that were previously difficult to identify to the genus and species level. Some species that were previously lumped into large categories (eg, “diphtheroids”) can now be reliably and accurately identified by MALDI-TOF MS (eg, Brevibacterium casei, Dermabacter hominis). These unfamiliar names can confuse IP teams when determining if patients with positive blood cultures meet CLABSI criteria as defined by NHSN. NHSN has updated their Commensal Organism list in an effort to help IP personnel to differentiate contaminants from pathogens when applying these criteria.11
Urine Cultures
Diagnosis of urinary tract infections (UTIs) is challenging primarily because of the lack of a gold-standard test. Further, asymptomatic bacteriuria is common, particularly in the elderly. Finally, contamination during urine collection is common. Contamination can be reduced using a number of different measures including collecting urine after cleansing of the vulva in women and the meatus in men. Midstream (as opposed to the first void) collection also reduces contamination in men.12 If collecting urine from a Foley catheter system, hand hygiene is essential, and the port should be cleansed prior to obtaining 20-30 mL of urine into a sterile container.13 Urine should not be collected from the collection bag. Regardless of the method of collection, urine should be submitted within 2 hours of collection and refrigerated until processing.3 Submission of urine in boric acid preservative can also stabilize colony counts and mitigate false-positive urine cultures if delayed transport (>2 hours) is anticipated.12 Once in the laboratory, a calibrated quantitative loop is used to plate 0.001 mL of urine on to solid agar. Following incubation, each colony grown in culture is interpreted as representing 1000 colony-forming units (CFU)/milliliter (mL). In most laboratories, 10 000-100 000 CFU/mL is considered significant in “clean-catch” and indwelling catheter urine cultures. Cultures with three or more different organisms should be reported as contaminated with no further workup.13
Recent efforts have focused on reducing unnecessary urine cultures in asymptomatic patients through educational programs that emphasize guideline recommendations to avoid routine urine cultures in hospitalized patients with fever, limiting such cultures to specific highrisk populations (eg, renal transplant, posturologic procedure, obstructive process).14 Others have described using alerts embedded in the ordering system and use of audit and feedback.14
For example, Trautner et al. implemented a project called “Kicking CAUTI: The No-Knee-Jerk Antimicrobials” where the study team developed and distributed a “CAUTI vs ASB” treatment algorithm. They provided feedback to clinical teams in the form of in-person presentations that reviewed cases that were appropriately and inappropriately managed. Urine culture orders decreased from 41.2/1000 bed days to 23.3/1000 bed days postintervention. The incidence of asymptomatic bacteriuria overtreatment also decreased significantly at the intervention site (1.6-0.6 episodes/1000 bed days, IRR = 0.35, 95% CI, 0.22-0.55).15
Another novel intervention is the implementation of reflex urine culturing, where urine samples are screened first with urinalysis and only proceed to culture if predefined criteria are met. In a quasi-experimental study, Epstein and colleagues implemented a urine reflex policy in which cultures were performed only when there were >10 WBC/hpf on urinalysis and measured the impact of the new policy on reported catheter-associated urinary tract infection (CAUTI) rates in five intensive care units. The authors reported a statistically significantly decrease in urine culture volumes (P = .0012) and CAUTIs per 1000 patient-days (P = .04).16
Clostridioides difficile Toxin Testing
Despite advances in laboratory detection of C difficile toxin (toxin A and toxin B) over the last two decades, the optimal testing method remains controversial. The cell cytotoxicity neutralization assay (CCNA), which is highly sensitive and specific, is slow and labor intensive. Enzyme immunoassays (EIA) are faster and less labor intensive but are associated with poor sensitivity. Nucleic acid amplification tests (NAATs) are highly sensitive and specific but tend to be costly. C difficile NAAT alone has been criticized to be too sensitive; because NAAT tests detect C difficile toxin genes, rather than toxin itself, they may be positive not only in those with disease (CDI) but also in those without disease who are colonized with toxigenic C difficile. In an attempt to reduce costs, a two-step algorithm has been implemented by many laboratories, in which specimens are initially tested with a combined glutamate dehydrogenase (GDH) and toxin EIA assay followed by testing of discordant samples (ie, GDH-positive, EIA-negative) with
NAAT or another method. In a prospective cohort study, Polage et al. reported that outcomes of patients who tested PCR-positive/toxin-negative were similar to those who tested PCR-negative and suggested that use of NAAT alone leads to overtreatment of CDI.17 This study gave support to the “reverse” algorithm in which positive NAAT results are followed by a toxin test as a way to adjudicate whether the patient has CDI or is merely a carrier of toxigenic C difficile.
NAAT or another method. In a prospective cohort study, Polage et al. reported that outcomes of patients who tested PCR-positive/toxin-negative were similar to those who tested PCR-negative and suggested that use of NAAT alone leads to overtreatment of CDI.17 This study gave support to the “reverse” algorithm in which positive NAAT results are followed by a toxin test as a way to adjudicate whether the patient has CDI or is merely a carrier of toxigenic C difficile.
However, regardless of the testing method chosen, inappropriate CDI testing and reports of positive CDI tests can be mitigated by reducing testing on formed stool and in patients taking laxatives. Truong et al. implemented a policy in which the laboratory cancelled CDI tests in patients who received laxatives within 48 hours of specimen receipt or had fewer than three unformed stools documented in their electronic health record.18 They cancelled 16% of CDI tests and reported a reduction of hospital-onset CDI from 13 to 9.7 cases/10 000 patient-days (P = .009).
Ventilator-Associated Events
Both clinical and surveillance definitions related to ventilator-associated events require evidence of respiratory deterioration, evidence of infection or inflammation, and laboratory evidence of respiratory infection. According to NHSN, infection-related ventilator-associated complication (IVAC) status requires one of the three criteria: (1) positive respiratory culture meeting defined quantitative or semiquantitative thresholds without requirement of purulent respiratory secretions; (2) purulent respiratory secretions plus recovery of an organism in culture from sputum, endotracheal aspirate bronchoalveolar lavage (BAL), lung tissue, or protected specimen brush in quantities that do not meet criterion (1); or (3) positive, organism recovery from pleural fluid; findings on lung histopathology; Legionella test results; and test for a respiratory virus. Practically speaking, microbiology laboratories may offer quantitative respiratory cultures (which involve inoculation of solid media in the same manner as with urine cultures) or semiquantitative cultures (which report light, moderate, heavy growth) in combination with a Gram stain that reports neutrophils.19
Two new tests that focus on diagnosis of nosocomial pneumonia were recently introduced. The Unyvero lower respiratory tract (LRT) Application (Curetis USA, San Diego, CA) is a qualitative, multiplex PCR test that is performed on endotracheal tube aspirates. It detects 19 bacterial species that are responsible for hospital-acquired pneumonia and a variety of antibiotic resistance determinants. The limit of detection of 104-106 CFU/mL for bacterial species is higher than that of a standard quantitative respiratory culture (103 CFU/mL per isolate). Also, the Bio-Fire FilmArray Pneumonia Plus was FDA-cleared for use on sputum, endotracheal tube aspirates, BAL, and mini-BAL specimens. This assay detects 15 bacterial species and assigns positive results to semiquantitative “bins” of 104, 105, 106, and >107 copies per milliliter. This test also detects 7 antibiotic resistance determinants, 8 respiratory viruses, Chlamydia pneumoniae, Mycoplasma pneumoniae, and Legionella pneumophila. For both systems, respiratory cultures are still required to link organism identification with AST results. Both manufacturers also state that detection of microbial nucleic acid may indicate colonizing flora that may not be the causative agent of pneumonia.
Multidrug-Resistant Organisms
Methicillin-Resistant Staphylococcus aureus and Vancomycin-Resistant Enterococci Methicillin-resistant Staphylococcus aureus (MRSA) may be recovered from patient specimens that are collected in the course of clinical care; or they may be recovered from active surveillance cultures. S aureus isolates recovered on standard bacteriological media are usually easily identified by their characteristic golden yellow color, beta-hemolysis on sheep’s blood agar, positive catalase reaction, and a positive coagulase test. S aureus is also easily identified using commercially available MALDI-TOF MS platforms. Methicillin resistance can be determined through a variety of AST methods.
Active surveillance for MRSA detects individuals who are colonized but not necessarily infected with MRSA. During outbreaks, MRSA screening of the nares, rectum, and other sites can help to identify asymptomatic individuals who may be reservoirs in the context of ongoing MRSA transmission. MRSA screening is also common in endemic (nonoutbreak) settings. The U.S. Veterans Administration healthcare system and some states have mandated routine active surveillance for MRSA in healthcare facilities. Other hospitals have also adopted this approach voluntarily to monitor and control MRSA transmission.
MRSA active surveillance is typically performed using either culture-based or molecular methods. Culture-based methods are common and are well standardized. Screening agars are usually “selective” (ie, inhibit the growth of organisms other than those of interest) and “differential” (ie, allow the laboratory staff to swiftly differentiate the organism of interest). Chromogenic media, for example, allow rapid visual identification of MRSA colonies with minimal additional workup. Preincubation of swabs in nutritive broth has been reported to increase sensitivity.20 The development of molecular MRSA screening assays has been surprisingly challenging for a number of reasons. The mecA gene found in MRSA is very similar to that carried by coagulase-negative staphylococci and lacks specificity. Commercial assays therefore opted to amplify and detect a region at the SCCmec-orfX junction, which is specific to mecA-mediated MRSA. Subsequently, so-called mecA dropouts or “empty cassette variants” were described in which methicillin-susceptible S aureus were observed to test positive for the sccmec target while having “lost” the mecA gene through excision or mutation. This phenomenon was soon recognized to be a relatively common cause of false-positive MRSA PCR tests.21 Inclusion of mecA primers largely addressed the problem. As an example, the Xpert MRSA NxG assay, which is FDA-cleared for MRSA screening of the nares, detects sccmec, mecA, and the spa gene (which is specific to S aureus). Nonetheless, the early experience with MRSA screening has elucidated some of the challenges inherent in molecular detection methods applied to bacteria that are constantly evolving.
Vancomycin-resistant enterococci (VRE) grow readily on standard bacteriological media. Most VRE are Enterococcus faecium, which appear as alpha-hemolytic gray colonies on sheep’s blood agar, are catalase negative, are
test positive by the PYR test, and are easily identified by MALDI-TOF MS. Vancomycin resistance in enterococci is readily detected using a variety of AST methods. Patients are typically screened using rectal swabs. Like MRSA, VRE can also be detected on screening rectal swabs using both culture-based and molecular-based assays. The only FDAcleared assay screening test available in the United States is the Xpert vanA/vanB (Cepheid).
test positive by the PYR test, and are easily identified by MALDI-TOF MS. Vancomycin resistance in enterococci is readily detected using a variety of AST methods. Patients are typically screened using rectal swabs. Like MRSA, VRE can also be detected on screening rectal swabs using both culture-based and molecular-based assays. The only FDAcleared assay screening test available in the United States is the Xpert vanA/vanB (Cepheid).
Active surveillance using commercially available molecular MRSA and VRE screening tests is more expensive to implement than culture-based screening, and clinical data supporting molecular over culture-based screening are lacking. Also, while culture-based and molecular-based screening methods are relatively easy to perform in the laboratory, large-scale screening in general can be labor intensive and costly.
Carbapenem-Resistant Enterobacteriaceae CRE are a substantial clinical problem worldwide. While carbapenem resistance is often due to AmpC or ESBL production coupled with porin mutations, reports of plasmid-mediated carbapenemase-producing CRE that harbor blaKPC and, increasingly, blaNDM have increased over time.22 There are a variety of phenotypic methods for detection of carbapenemase production in bacterial isolates.22 The modified carbapenem inactivation method (mCIM) has a sensitivity and specificity of 98% and 99%, respectively, but requires overnight incubation (22-2321-22). The Carba NP procedure provides a result in about 2 hours but has a lower sensitivity (84%) and requires labor-intensive reagent preparation (22-2321-22).22 The RAPIDEC Carba NP (BioMérieux) is an FDA-cleared kit that can be used with Enterobacteriaceae and P aeruginosa isolates, is based on the Carba NP test, and requires 2 hours to complete. The kit provides preprepared reagents, reducing the need to prepare reagents in-house. It detects carbapenemase activity with a sensitivity and specificity of 98% and 99%, respectively.23 In terms of molecular testing, the Xpert Carba-R (Cepheid) is an FDA-cleared test that detects blaKPC, blaNDM, blaVIM, blaIMP, and blaOXA-48 from pure colonies of Enterobacteriaceae, Acinetobacter baumannii complex, and P. aeruginosa. Carba-R is also FDA-cleared to be performed on rectal screening swabs. Table 5-1 summarizes the commonly used screening methods for various MDROs.