Role of the Microbiology Laboratory and Molecular Epidemiology in Healthcare Epidemiology and Infection Control
Charles W. Stratton IV
John N. Greene
In the late 19th century, early microbiologists such as Pasteur and Koch demonstrated that infections were due to specific microorganisms and that these microbes could be isolated by appropriate cultures. Another early microbiologist, Lister, recognized the principle that certain chemicals antagonized microbes. Lister then applied this principle to infection control by using phenol to sterilize surgical instruments and dressings to reduce the morbidity and mortality associated at that time with surgery (1). Thus began the association of microbiology with antibiosis and infection control.
The relationship of microbiology with infection control was formally recognized in the early 1970s by the Centers for Disease Control and Prevention (CDC), which developed standard definitions for nosocomial infections and methods for infection surveillance (2). Infection control committees incorporated these CDC recommendations into practice at that time. The result was a shift from unproductive environmental sampling (3) to more directed surveillance and intervention when established baseline endemic rates of infection were exceeded. However, unless these baseline endemic rates of infection were exceeded, the surveillance process was, for the most part, passive.
As humans and medicine complete the first decade of the 21st century, infection control strategies continue to evolve. First, microbiologic surveillance has shifted away from general categories of medical service, infection site, and hospital-wide infection rates and is focusing instead on problem categories (4,5). These focused categories include high-risk areas such as intensive care units (ICUs) (6,7), preventable high-risk infections such as intravascular device-related infections (8,9), the surveillance and control of microbial resistance (10,11,12,13 and 14,15,16), and emerging pathogens (10,17,18). Second, infection control strategies today are more proactive, which simply means that active intervention for prevention of infections and control of resistance has an equal priority to simply monitoring for changes in these parameters (6,8,19,20 and 21). Third, it is now recognized that a key component in this proactive strategy is the need for ongoing and constant education of healthcare and infection control personnel as well as education of residents, pathologists, infectious disease fellows, and clinical microbiology fellows (22,23,24,25 and 26). Such educational efforts are becoming important functions of an infection control committee with the assistance of the microbiology laboratory. Fourth, implementation of efficient infection control requires the construction of a computerized information network that ideally includes the hospital, the community, the state, the nation, and other countries (27,28,29,30,31,32 and 33). Such networks that eventually would include guidelines, microbiologic surveillance data, and full-text references (i.e., PDFs) available on the Internet ultimately will become the cornerstone of infection control.
The interaction of the microbiology laboratory with healthcare epidemiology and infection control continues to evolve as an integral part of a nationwide concerted effort to develop and improve infection control practices and programs. This process began with the National Nosocomial Infections Surveillance (now the National Healthcare Safety Network) system developed by the CDC. This system provides risk-specific infection rates for use by hospitals and national healthcare planners to set priorities for their infection control programs and to evaluate the effectiveness of their effort (31). The Division of Healthcare Quality Promotion at the CDC through National Healthcare Safety Network continues to provide relevant surveillance information on healthcare-associated infections (34,35). In addition, the Division of Healthcare Quality Promotion is expanding to provide relevant information for other healthcare facilities such as dialysis centers (29). The Study on the Efficacy of Nosocomial Infection Control conducted by the CDC in the 1970s found that hospitals had lower rates of healthcare-associated infections if levels of surveillance activities were increased (36). Thus, many infection control programs received additional support to increase the number of infection preventionists (IPs).
Meanwhile, the focus and procedures of microbiology laboratories were changing because of multiple factors that included increasing resistance, emerging pathogens, and new technology (10,17,37). For example, the need for clinical microbiology laboratories to detect emerging antimicrobial resistance (38,39,40,41) has resulted in new approaches and technology for this purpose (42,43 and 44). All of these factors have resulted in important changes in the role of the microbiology laboratory in healthcare epidemiology and infection control.
The microbiology laboratory has always been recognized as an essential element in the control of healthcare-associated infection (3) and has long served as an early warning system for healthcare-associated infections by identifying clusters of microbes with unique phenotypic characteristics and communicating this information to IPs (45,46). In the past, such healthcare epidemiology and infection control activities did not place a great demand on the microbiology laboratory.
Today, however, the work done by the microbiology laboratory is increasingly complex and demanding. Much of this has direct implications on healthcare epidemiology and infection control. Microbiology laboratories now must be able to detect, identify, and characterize an expanded array of microbes, including newly emerging pathogens (10,40). Some of these pathogens, such as fungal microorganisms, may be important causes of healthcare-associated infections but difficult to detect (47,48 and 49). Fortunately, traditional methods using cultures for isolation, identification, and susceptibility testing of pathogens have been supplemented by highly sensitive, rapid, and specific molecular biologic techniques in which unique DNA or RNA sequences can be directly detected (10,12,17,18,43,44,50,51,52,53,54,55,56,57,58). These and other molecular techniques have enabled microbiology laboratories to “fingerprint” microbes, thereby facilitating studies of healthcare-associated transmission (58,59). Finally, the microbiology laboratory’s role in monitoring and controlling resistance has become critical because of the increasing frequency with which resistant pathogens are causing healthcare-associated infections (10,12,13 and 14,21,38,42,60). This role today may include not only the accurate detection of resistance per se but also the determination of the molecular epidemiology of the resistant isolates. The amount of work by the microbiology laboratory to support healthcare epidemiology and infection control has greatly increased.
The role of the microbiology laboratory in healthcare epidemiology and infection control continues to expand. For example, IPs today often augment their surveillance efforts by the use of computer-generated focused microbiologic surveillance reports from the microbiology laboratory. Problems thus detected may require molecular methods as a part of their evaluation. If the problems involve resistance, additional susceptibility testing and molecular methods may be required. Finally, the microbiology laboratory has become recognized as an important resource for the microbiologic training and education of healthcare and infection control personnel. Indeed, the interactions of infection control committees with the microbiology laboratory are now so complex and important that most committees require that a representative of the microbiology laboratory serve as an active member to ensure the appropriate advice, education, coordination, and technical support. This chapter examines these various facets of the changing and increasingly critical role of the microbiology laboratory in healthcare epidemiology and infection control.
SURVEILLANCE
The key to an effective infection control program continues to be effective surveillance, which the Study on the Efficacy of Nosocomial Infection Control has defined as an IP using basic epidemiologic techniques to perform surveillance on clinical ward rounds, to analyze rates of infection, and to incorporate the data generated in decision making (31). Such surveillance for healthcare-associated infections involves identifying patients who are colonized or infected, assessing the risk of transmission of infection between patients, proving transmission of a given strain from one patient to another, and, more generally, detecting healthcare outbreaks (61). However, to recognize the existence of an outbreak, baseline endemic rates of infection must be determined for each type of infection within a given institution.
Defining endemic rates (the number of infections divided by the number of patient-days or patients at risk) for services, sites of infection, microorganisms, and procedures can be accomplished in each hospital by an active surveillance system coordinated by the IP and the microbiology laboratory. Clusters and epidemics can be investigated when endemic threshold rates are exceeded, when unusual or new microorganisms are isolated, and when new sites of infection are identified. Collection of surveillance data, usually by the IP, consists of reviewing microbiology reports generated by the laboratory. If trends of increasing or unusual infection rates are discovered, then chart review and discussion with personnel involved in patient care should follow to determine the significance of these isolates. The importance of active surveillance is seen with the recent outbreak of the pandemic H1N1 influenza (62). Pandemic influenza is an example of emerging and reemerging infectious diseases that must be monitored with ongoing surveillance strategies and new diagnostic methods (40,43,55,56,57,63,64) (see also Chapters 101 and 102).
With increasing resistance and the fact that many healthcare-associated infections are caused by resistant microbes, surveillance and control of resistance have become critical (10,12,13 and 14,21,38,42,60). Susceptibility patterns can be monitored for emergence of resistant microorganisms; when resistant microorganisms are identified, appropriate isolation precautions should be instituted. Moreover, control of antimicrobial use has become important for controlling resistance (10,12,13,14,15 and 16). For this reason, the antibiotic subcommittee of the pharmacy and therapeutics committee should be included as a part of the infection control program for preventing resistance. One practical way to do this is for a representative of the microbiology laboratory to be a voting member of both the infection control committee and the antibiotic subcommittee. In addition, one or more members of the antibiotic subcommittee should be a member(s) of the infection control committee.
Today, all microbiology laboratories have a computerized reporting system known as the laboratory information system (65). Computer-generated microbiology reports are usually sorted by site of isolation, type of microorganism, and location of the patient, but they can be programmed to focus on any particular problem. Reports are generated daily and cumulatively. These reports are used to detect trends of increasing infection rates or increasing resistance and are reviewed daily by the IP. In addition, the IP often participates in daily clinical microbiology rounds in which new positive cultures at each bench station are reviewed.
The microbiology laboratory receives appropriate hospital demographic information on any culture request and often is able to use this information to recognize clusters of similar isolates. In addition, the availability of laboratory computer systems allows specific types of patients (e.g., transplant patients) or specific locations (e.g., ICUs) to be easily grouped and reviewed. When such focused microbiologic surveillance is desired, the microbiology laboratory should have the capability to provide such reports. In the past, a computerized reporting system did not necessarily mean that focused surveillance reports could be easily obtained. Often, some degree of computer programming was needed; therefore, this programming capability should be readily available. Once obtained, these focused surveillance reports for specific units should be incorporated as a routine surveillance method with these reports also provided to the medical director of the specific unit(s) (5,6,7 and 8).
Once the microbiology reports have been reviewed and prioritized, charts of the patients with the microorganisms of interest should be analyzed to evaluate the significance of the isolates as potential causes of healthcare-associated colonization and infection. Susceptibility trends should be analyzed. By defining baseline endemic rates for various infections and resistance problems through effective surveillance, unusual disease and resistance activity will trigger disease control and prevention efforts (10,12,13 and 14,21,38,42,60). In summary, an active surveillance system assists the clinician in making an accurate diagnosis and prescribing therapy by providing the knowledge of disease occurrence and antibiotic resistance patterns.
IDENTIFICATION OF OUTBREAKS
An investigation of a potential outbreak of healthcareassociated infections must first determine if these infections are related in any way (31,41,46). Most often, this determination involves recognition of the microbial pathogen causing the outbreak and differentiation from those microorganisms of the same genus or species that, although isolated from some patients, are not involved in the outbreak (31,41,46,55,59,66). However, an outbreak may involve resistance rather than an increased incidence of infections. For example, an outbreak might actually consist of only one strain of vancomycin-resistant Staphylococcus aureus because of the implications of such an isolate (67). If the outbreak can be linked to infection by a single strain (also called a clone), exposure to a common source or reservoir or transmission from patient to patient would be inferred.
Traditionally, the epidemic strain has been defined with phenotypic methods, which include genus, species, biotype, serotype, phage type, bacteriocin production, and antimicrobial susceptibility patterns (68). Phenotypic methods reflect genetic traits and may be quite specific. When a given phenotype is rarely found in a microbial strain, that phenotype alone may provide convincing evidence of transmission between patients (e.g., Escherichia coli O157:H7). However, microorganisms with commonly expressed phenotypic characteristics may require additional subtyping (37). Sometimes, isolates share phenotypic markers but are actually genotypically different; this implies the presence of two separate strains and infection from two different sources. The limitations of phenotypic techniques are presented in Table 95-1.
TABLE 95-1 Limitations of Phenotypic Methods | |||||||
|---|---|---|---|---|---|---|---|
|
When microbial pathogens are nontypeable by phenotypic methods or have only a few types, the poor discriminatory power precludes the use of these typing methods. This has led to the use of genotypic methods for typing. This approach has been extremely successful and is now termed molecular epidemiology (37,46,55,56,57,58,59,66,69). These molecular epidemiologic methods most often involve genotyping of microbial plasmid or chromosomal DNA and go far beyond the current limitations of phenotyping and provide more accurate data during outbreak investigation (37). Moreover, outbreaks of viruses (70) and free-living microorganisms can now be adequately studied with current molecular epidemiologic methods.
However, combining methods of microorganism identification provides stronger evidence for the presumed relationship between isolates. Such was the case with an outbreak of neonatal meningitis caused by Enterobacter sakazakii (71). Biotypes, plasmid DNA profiles, and antibiograms of isolates from patients and the environment were identical,
establishing the means of transmission from a powdered milk preparation. On the other hand, multiple typing systems may show dissimilarity among strains, casting uncertainty on the relatedness of isolates. This was illustrated when widespread colonization of personnel with methicillin-resistant coagulase-negative staphylococci (MRCNS) at a Veterans Affairs hospital was investigated (72). Antimicrobial susceptibility profiles, biotyping, phage typing, plasmid profiles, restriction fragment length polymorphism (RFLP), and plasmid hybridization with a DNA probe showed dissimilarity among strains. Because of the absence of strain similarity that has been found using the various methods, the role of human reservoirs of MRCNS as a source for infections in hospitalized patients remains obscure (72).
establishing the means of transmission from a powdered milk preparation. On the other hand, multiple typing systems may show dissimilarity among strains, casting uncertainty on the relatedness of isolates. This was illustrated when widespread colonization of personnel with methicillin-resistant coagulase-negative staphylococci (MRCNS) at a Veterans Affairs hospital was investigated (72). Antimicrobial susceptibility profiles, biotyping, phage typing, plasmid profiles, restriction fragment length polymorphism (RFLP), and plasmid hybridization with a DNA probe showed dissimilarity among strains. Because of the absence of strain similarity that has been found using the various methods, the role of human reservoirs of MRCNS as a source for infections in hospitalized patients remains obscure (72).
Identification of a microorganism by any means requires thorough knowledge of the unique attributes of the microorganism to distinguish it from the large background of nonepidemic, nonpathogenic strains (55,56,57,58,59,69). Specific strain identification can be critical in identifying outbreaks of infection (55,56,57,58,59,69). This can be seen with the speciation of coagulase-negative staphylococci. S. schleiferi is a species of coagulase-negative staphylococci that is pathogenic in animals and humans, causes pyoderma and abscesses, and has been described in an outbreak of wound infections (73). Isolation and speciation of this pathogen from a cluster of surgical site infections would have far greater impact than the isolation and report of coagulase-negative staphylococci from the same cluster, as the latter would be interpreted as likely representing various coagulase-negative species and thus skin contaminants. As the ability to characterize strains improves, the number of differences detected between strains will likely increase. This will allow better characterization of the pathogenesis of coagulase-negative staphylococci (74).
An important feature of an epidemiologic evaluation is the determination of clonality of the suspected pathogen regardless of the mode of transmission. A clone is a set of isolates that have been recovered independently from different sources, in different locations, and possibly at different times but that show so many identical phenotypic and genetic traits that the most likely explanation for this identity is a common origin (69,75). Clonality among isolates in an outbreak must be established before it can be concluded that the outbreak originated from a common source (69). Successful clone identification requires knowledge of the genetic stability of the microorganism, the selective pressure of the environment, and the discriminatory power of the given procedure used to characterize the isolate (69). The judgment of nonclonality eliminates an isolate from consideration as one involved in a particular chain of transmission (69). A judgment of probable clonality strengthens the case for either a common-source outbreak or an outbreak resulting from person-to-person transmission in proportion to the rareness of that clone in the environment (69). Following a given clone throughout its travels by surveillance methods has documented the worldwide spread of multiresistant strains of penicillin-resistant Streptococcus pneumoniae (76), methicillin-resistant S. aureus (MRSA) (77), and community-acquired MRSA (78).
Host responses to invading microorganisms may also be used to identify and track infections that are difficult to investigate using current phenotypic and genotypic methods. For instance, serology may be used to determine infection rates during outbreaks, particularly when cultures have not been obtained or are obtained after initiating treatment or when routine cultures may not detect infection (e.g., pneumonia). This was seen with group A Streptococcus (GAS) infections in a nursing home (79). Nine (56%) of the 16 cases of GAS disease or infection in residents were confirmed by serologic testing (anti-DNase B titers) alone (79). The identification of a single serotype (M-1, T-1) from the four available isolates and epidemiologic correlation suggested that a single strain of GAS was introduced into the nursing home by the index patient, with subsequent person-to-person transmission. Similarly, pulsed-field gel electrophoresis (PFGE) has been used to document a community outbreak of invasive GAS infection in Minnesota (80). Field inversion gel electrophoresis is another electrophoretic typing method similar to PFGE that has been developed for GAS (81). These electrophoretic methods are able to identify differences between and within M types of GAS. Another molecular method for distinguishing GAS is fluorescent amplified fragment length polymorphism (AFLP) analysis (82). Finally, the emm gene for the M protein has proven useful for typing GAS (83).
When investigating a possible outbreak, the healthcare epidemiologist or IP, who formulates a hypothesis based on clinical and epidemiologic evidence, must collaborate with a microbiology laboratory to provide microbiologic data to either support or refute the hypothesis (27,45,46,59). Isolates from multiple patients are examined to determine whether the infections are related. Establishing similarities or differences among epidemic isolates is not always sufficient to determine the source or the mode of dissemination (84). Data derived from epidemiologic studies are also needed. Cultures and molecular typing without an epidemiologic study often lead to uninterpretable results. However, when molecular typing is combined with an epidemiologic study, the two methods are complementary in confirming transmission of a single or multiple strains (85).
EPIDEMIOLOGIC TYPING
Currently, there are a vast number of epidemiologic typing systems available (37,55,56,57,58,59,69,86,87 and 88,89,90,91,92). These include molecular methods that are clearly useful for the epidemiologic analysis of infectious disease outbreaks (52,89,90,91,92). However, to gain acceptance and be routinely applied in clinical situations, molecular epidemiologic methods must be easy to perform, rapid, reproducible, and cost-effective and provide additional information not obtained from traditional typing techniques (84,85,87,89,90,91,92). Also, it is important with high-resolution typing systems to distinguish between comparative epidemiologic typing systems that are used in outbreak investigations and library epidemiologic typing systems that are used in surveillance systems (87). Most of the currently available molecular typing systems are comparative methods that are reproducible in single assay, have high discrimination (D > .95), and are used to compare isolates from a suspected outbreak and distinguish them from sporadic isolates. Such comparative methods include RFLP, PFGE, and arbitrarily primed and randomly
amplified polymorphic polymerase chain reaction (PCR) analysis. Library typing systems, in contrast, are reproducible over time and between laboratories, have discrimination power balanced against evolutionary stability, and are used for long-term surveillance. Library methods include serotyping, insertion sequence fingerprinting, ribotyping, PFGE, AFLP, infrequent-restriction-site amplification PCR, interrepetitive element PCR typing (rep-PCR), and PCRRFLP of polymorphic loci. Finally, a typing method cannot be considered valid unless it is capable of discriminating among randomly chosen isolates (84,85,86,87,88,89,90 and 91,92).
amplified polymorphic polymerase chain reaction (PCR) analysis. Library typing systems, in contrast, are reproducible over time and between laboratories, have discrimination power balanced against evolutionary stability, and are used for long-term surveillance. Library methods include serotyping, insertion sequence fingerprinting, ribotyping, PFGE, AFLP, infrequent-restriction-site amplification PCR, interrepetitive element PCR typing (rep-PCR), and PCRRFLP of polymorphic loci. Finally, a typing method cannot be considered valid unless it is capable of discriminating among randomly chosen isolates (84,85,86,87,88,89,90 and 91,92).
The basic premise inherent in any typing system is that epidemiologically related isolates are derived from the clonal expression of a single precursor and share characteristics that differ from epidemiologically unrelated isolates (52). The utility of a particular characteristic for typing is related to its stability within a strain and its diversity within the species (93). The most clinically relevant isolates are those with characteristics that provide for increased virulence or resistance and are often the most difficult to differentiate (93). The strength of typing depends on the discriminatory power of the method used (93). When strains are nontypeable or have only a few serotypes, such poor discriminatory power precludes the use of certain typing methods. Ideally, a typing method will recognize each unrelated isolate as unique. In practice, the technique is considered useful if the most common type it detects occurs in <5% of the population (93). There is currently no gold standard or definitive typing system or even an authoritatively validated collection of isolates against which a new method can be evaluated (93). Nevertheless, bacterial typing systems are applied clinically to address one fundamental question: Are two isolates the same or different? (93)
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
