William R. Jarvis
Healthcare-associated infections (HAIs) continue to present a major problem in hospitals today. Because of the importance of this subject, each hospital laboratory has the responsibility of supporting activities related to HAI surveillance, control, and prevention [1,2]. Each laboratory can make major contributions toward infection control as long as the people responsible for infection control or the clinical microbiology laboratory cooperate closely to attack this problem. Often the same people have both of these responsibilities; nearly half of those who chair infection control committees are laboratory per-sonnel [3].
Laboratory personnel attempt to minimize the occurrence of HAIs in the following seven ways: (1) participating in hospitalwide infection control activities, especially those of the hospital infection control committee or service team, (2) recovering and accurately identifying responsible organisms, (3) determining antimicrobial susceptibility of selected HAI pathogens, (4) reporting in timely fashion laboratory data relevant to infection control and participation in HAI surveillance, (5) providing additional studies, when necessary, to establish the similarity or difference of organisms, (6) providing, on occasion, microbiologic studies of the hospital environment, and (7) training infection control personnel.
Since 1990, improvements in laboratory instrumentation and procedures have provided dramatic aid to infection control efforts in several ways [4]. Among these are techniques for more rapid detection and differentiation of organisms and improved systems of reporting for both patient data and trend analysis. Perhaps the most dramatic advances have come in special procedures for examining (“typing”) hospital organisms for similarity or difference; in this area, molecular and other techniques have permitted more definitive examination of a wider range of organisms than was possible before [5,6].
Participation in Hospitalwide Infection Control Activities
Relationship of the Laboratory to the Infection Control Committee
A clinically oriented member of the laboratory staff can contribute substantially by serving on the infection control committee or service team, as it may be called
P.130
today. Such participation is essential in contributing to a harmonious relationship among clinical, infection control, and microbiology personnel [7].
In the typical hospital, most members of the infection control committee do not have a background in microbiology [8]. Thus, it is of great importance for the representative of the laboratory to provide the microbiologic expertise that is critical to many decisions of the group. This knowledge may be required for assessing the importance of culture data, determining the validity of laboratory techniques used to identify HAIs, and designing and implementing investigations and survey projects.
The diagnostic microbiology laboratory is engaged primarily in the evaluation of cultures related to infection. Because these are crucial data for successful infection control, the laboratory activities should be closely coordinated with the infection control committee. For example, the adequacy of the basic techniques for primary isolation, speciation, and antimicrobial susceptibility testing should be discussed by the microbiologists and the infection control committee. Laboratory resources often are stretched by patient-care requirements, especially in smaller hospitals [9]. Laboratory support for infection control activities must be given with discretion. For example, the use of laboratory resources to assess culture personnel or the environment should never be permitted when the epidemiologic indications are unclear [9].
Major changes in reimbursement methods for U.S. hospitals have occurred as managed care has become popular [10]. In view of these changes, it seems that the laboratory microbiologist can provide an added service to the infection control committee and other hospital committees concerned with infection control. Under managed care and pay for performance, new and intensive attempts will continue to evaluate the validity and usefulness of hospital programs, such as infection control [11]. Techniques for assessment are familiar to clinical microbiology personnel, who routinely have to make similar cost–benefit judgments for laboratory equipment, instruments, and procedures [12]. The insights and methods used for such laboratory activities should be helpful to the infection control team and the various committees (e.g., infection control, quality assurance, pharmacy, and therapeutics) as they review the benefit of their activities and attempt to improve the productivity of the program [13].
Budgetary Considerations
Costs for laboratory procedures that are not related directly to the care of patients (e.g., bacteriologic sampling of personnel or the environment) should be borne by a budget separate from that of the laboratory. To facilitate all of the microbiologic activities necessitated by an outbreak, the laboratory (or the hospital epidemiologist or the infection control committee, depending on the hospital's organizational structure) should have a contingency fund to enable personnel, materials, and space to be temporarily assigned to epidemic aid support [14]. An investigation of an outbreak should not be financed by charging individual patients for cultures taken during the study. This will become less an issue as capitated care becomes more prominent and direct charges to patients for specific services decline.
Accurate Identification of Organisms Involved in Healthcare-Associated Infection
Infection control personnel search constantly for evidence that a common organism has spread from patient to patient or from staff to patient. Thus, information permitting the successful tracing of organism movements within the hospital may be of value to the hospital infection control team, whether the positive cultures represent episodes of HAI or indicate colonization of the patient [15]. Although some clinical features of illness provide information about etiology, the main sources for this determination usually are the data provided by the clinical laboratory. Hence, the ability of the laboratory staff to isolate and identify responsible microorganisms is crucial to infection control [16].
The spectrum of organisms causing HAIs has changed dramatically since the mid-1980s [17,18]. Enterobacteriaceae, Staphylococcus aureus, Pseudomonas aeruginosa, and coagulase-negative staphylococci remain frequently associated with HAIs [2,19]. Among the patterns of special concern are the appearance of fungi and viruses as they become more frequent HAI agents [19]. Fortunately, technologic developments in the laboratory over the same period have continued to increase the efficiency with which HAI organisms can be recognized and recovered [4]. There are three main aspects to this. First, new instruments and devices have become widely available. These permit easier detection of the presence of organisms in blood cultures, identification of organisms, and testing of susceptibility to antimicrobials. Some of these devices are automated, permitting the laboratory to provide these improved services with the same or fewer personnel [20]. Many of the instruments can be cost effective for limited numbers of specimens; as a result, small and large laboratories can now include some of these methods in their programs. Use of these instruments and devices also has led to a more standard approach throughout the United States to the identification and susceptibility testing of HAI pathogens. Thus, most of the organisms causing HAI outbreaks or special endemic problems can be identified in the hospital laboratory.
Second, nonculture tests have permitted identification of HAI agents that would not have been recognized in earlier years [21]. Immunologic and nucleic acid testing methods have added to our ability to recognize viruses and other organisms that are difficult or impossible to grow in culture; some of these are involved in HAI. Amplification techniques, such as the polymerase chain reaction (PCR), and other gene-based methods are making
P.131
tests for diagnosis of infection or colonization even more sensitive [22,23]. Some of these tests are useful for rapid identification of isolates even if they cannot be characterized by routine biochemical methods [24].
Third, these newer tests and instruments not only permit identification of additional agents but also allow more rapid diagnosis of both new and old pathogens. This speedier testing should provide earlier recognition of outbreaks and more efficient handling of organisms in endemic nosocomial or community-acquired infections [25], which should reduce the likelihood of community-acquired organisms becoming HAIs.
Even with these new technologic developments, certain basic principles of operation remain crucial to producing reliable microbiologic data. Several of these are discussed here.
Collection and Transport of Specimens
Specimen collection, transport, and handling must be of sufficiently high quality to provide valid data [26]. Specimens that are not collected or transported properly may give inaccurate results, even when handled as well as possible once they reach the laboratory. In turn, these inaccurate results may lead to improper clinical decisions by physicians, unnecessary labor by laboratory personnel, and unnecessary patient charges. This is especially true for the newer molecular techniques [27].
The laboratory must monitor specimen handling continually and work closely with both inpatient and ambulatory care units to ensure minimization of the possibility of contaminated specimens. This is necessary to ensure that laboratory information presented to the hospital epidemiologist reports organisms actually associated with the patient's site of culture rather than contaminants.
Certain laboratory findings suggest specific handling errors [1]. For example, a frequent failure to isolate organisms from deep wounds or abscesses of patients who are not on antibiotics or the inability to recover pathogens seen on Gram stain in episodes of presumed anaerobic infections suggests inadequate anaerobic transport media, delay or inappropriate refrigeration of specimens in transit, or use of inadequate techniques for isolating anaerobes. The frequent recovery of three or more different organisms in clean-voided, midstream urine specimens suggests unsatisfactory technique in collecting specimens, a delay in transporting specimens to the laboratory, or a delay in culturing them. The finding of negative cultures from a high percentage of patients with positive smears for bacteria suggests unsatisfactory specimen collection or handling, errors in staining, contaminated reagents, or errors in culture techniques [28].
Specimen collection and handling should be assessed regularly to detect and correct such problems; the frequency with which probable contaminants are isolated from clinical specimens can be a measure of the quality of specimen collection in a specific hospital area. For example, determining frequency of urine specimens with characteristics that suggest specimen contamination permits wards with high rates to be singled out for evaluation and, if necessary, for in-service education programs instituted by laboratory or infection control personnel. In addition, identifying people who draw blood cultures that frequently contain diphtheroids, coagulase-negative staphylococci, or other probable skin contaminants may permit re-instruction of these personnel in aseptic technique. Periodic review of the relative incidence of false-positive smears for acid-fast bacilli or of specimens with heavy bacterial contamination may highlight problems in sputum collection and processing [26].
Many hospitals record both the time the specimen was collected and the time the laboratory received it so that transport time can be monitored periodically or continuously and the culturing of old specimens avoided. Evaluation of turnaround time has become an important element of laboratory quality assurance [29].
Initial Evaluation of Specimens
Assessing specimens at the time they are received in the laboratory is one of the best ways to evaluate their suitability. For example, microscopic review of Gram stain of sputum specimens remains the best way to determine whether these specimens are contaminated [30]; specimens identified as inadequate are not processed further and do not confuse either clinician or epidemiologist. A new specimen should be requested unless the clinician provides notice of special circumstances (e.g., immunosuppression) that might make it worthwhile to proceed with culture [25].
Culture or smear results for other types of specimens also may suggest contamination at the time of collection. For example, urine specimens with ≥3 different organisms present ordinarily suggest contamination in patients without chronic indwelling urine catheters. Such specimens should be held for 2 to 3 days without further processing. The patient's physician should be notified so that unusual clinical situations requiring further identification of the specimen can be recognized. Scoring systems for use in determining acceptable wound, vaginal, cervical, or other specimens also have been described [31]. Application of such criteria ensures that the information generated from the specimens that are processed completely will more likely correlate with true infecting organisms and will reduce unnecessary laboratory costs. Repeat specimen collection should be requested for these inadequate specimens, and additional processing of organisms isolated from poor specimens (e.g., speciation, susceptibility testing) should be delayed or eliminated. The culture report should alert the clinician about the questionable value of the specimen so that results will be used cautiously, if at all, for guidance in diagnosis and therapy.
For specimens from sputum and wounds, reporting the morphologic characteristics of bacteria seen on Gram stain may be misleading if no statement is made regarding the presence or absence of white blood cells. Both sites may
P.132
be extensively contaminated with skin, oropharyngeal, or intestinal bacterial flora. When organisms are found only in the presence of abundant squamous epithelial cells, it is unlikely that they are the causative agents, and reports of such mixed flora without qualification about the accompanying cells may lead the clinician falsely to assume mixed flora as the cause of the infection. Substantial effort will be conserved and superior information ultimately provided if repeat collection is requested for such specimens.
Microscopy at the time of specimen submission can help other aspects of microbiologic diagnosis. For example, examination on Gram stain for morphology can identify organisms that might be epidemiologically important but not reflected by culture. Thus, presence of a mixed flora on Gram stain of a sputum specimen, when coupled with an aerobic culture yielding only Hemophilus influenzae, may indicate possible mixed aerobic–anaerobic infection rather than pneumonia due to Hemophilus. Because infection control implications of these two causes may differ, evaluations of this type by the laboratory can be important. Similarly, nonculture methods for identifying the presence of parvovirus B19 (e.g., demonstration by electron microscopy or gene probe) have helped us learn more about this organism as a cause of HAIs [32].
Anaerobic culture of specimens should be limited to (1) those that show leukocytes on Gram stain, (2) those with no evidence of contaminating squamous cells and organisms suggestive of anaerobic species, and (3) specimens from patients whose unusual circumstances suggest the need for anaerobic culture. These limitations result in the reporting of isolates that have a much higher probability of association with infection. The application of sensitive techniques for culturing and identifying anaerobes to specimens containing endogenous flora is costly and produces misleading information [33]. Large numbers of anaerobic organisms are present in the normal flora of skin, oral cavity, and genital and gastrointestinal tracts. Therefore, swabs from superficial portions of skin or mucous membrane lesions, specimens of expectorated sputum, and any materials contaminated with feces should be considered inappropriate for anaerobic culture [33]. Submission for anaerobic culture of such specimens or of specimens from sites that are rarely infected by anaerobes (e.g., urine) suggests the need for in-service education of hospital personnel.
Efforts such as those outlined substantially reduce errors in the diagnosis and use of unnecessary antimicrobial therapy. Such an approach also improves the specificity of infection surveillance data that otherwise might include isolates of questionable etiologic significance.
Identification of Isolates
Once a specimen has been received in the laboratory, it must be processed in a way that maximizes the likelihood of recovering older and newer agents causing HAIs. Often it is difficult to determine the causative HAI agents. Recovery of an organism does not ensure that it is the causative HAI agent [25]; thus, etiologic diagnosis cannot be made with certainty in many instances. Most episodes today for which the cause is known involve gram-positive cocci or gram-negative aerobic bacilli [17,18]. Most frequent among these gram-negative rods are Klebsiella, Enterobacter, Pseudomonas, Serratia, Proteus, or Escherichia coli (in approximately that order) (see Chapter 29). More recently, organisms such asAcinetobacter Flavobacterium, Legionella, and Pseudomonas species other than P. aeruginosa have become increasingly prominent.
Anaerobic bacterial organisms (usually found in mixed aerobic–anaerobic infections) have become less frequent in HAIs since the mid-1980s. However, viral agents (e.g., rotavirus), fungi, and parasites, such as Pneumocystis and Toxoplasma, have been identified as important causes of HAIs [17,18]. This expansion of the list of possible microbial pathogens for hospitalized patients has made it more difficult for both microbiologists and clinicians to deal effectively with HAIs. Effective handling of such problems requires the laboratory staff to keep up with the steadily emerging oganisms important in cross-infection and to implement and maintain culture and other techniques that bring these to light.
Need for Complete Identification
The degree to which organism identification routinely is carried can be important to HAI control efforts. Infection control personnel constantly are searching for evidence that a common organism has spread from patient to patient [4]. The ability to detect such an event is enhanced by identifying the organism at least to the level of species. Reporting of “biotyping” information (i.e., pattern of response to biochemical testing) on occasion can be of value in differentiating organisms that are frequently encountered, but this identification is not needed on a routine basis [16].
In today's environment of cost controls, the value of complete identification of all isolates is questionable [34]. Regardless of the extent to which full identification is conducted, it is important that standard criteria and nomenclature be consistently applied. Otherwise, attack rates for HAIs with various species may identify false problems (e.g., because of previously unreported species or strains) or fail to identify true problems. Furthermore, such surveillance data may not be comparable to data developed in other institutions or in cooperative surveillance programs.
Even more important, incomplete or incorrect identification of organisms may obscure real problems and make retrospective epidemiologic investigation impossible. For example, a report of “Klebsiella Enterobacter group” fails to distinguish between two species (Klebsiella or Enterobacter spp.) that have different epidemiologic patterns of infection within the hospital [35]. Similarly, identifying an isolate as Burkholderia cepacia (formerly Pseudomonas cepacia), an organism frequently associated with illness
P.133
or pseudoepidemics caused by contaminated water or other solutions [36], provides more useful epidemiologic information than identifying the organism only as “nonfermenting gram-negative bacillus,” in which the strain is lumped with a group of organisms that may not have as characteristic a hospital reservoir.
Because of these considerations, laboratories should maintain the capability to identify gram-negative aerobic bacilli to the genus level. The laboratory also should have the capability to identify organisms to the species level when special or recurring problems in a given institution make such information useful for dealing with HAI problems.
Many hospitals find it advantageous to use commercial, multiple test media for biochemical testing that provide this degree of characterization. Acceptable methods for microbiologic identification procedures are described in detail elsewhere [37]; additional assistance in identifying unusual isolates beyond the stated expertise of an individual laboratory is available from state, national, or private reference laboratories.
Sometimes it is the pattern of susceptibility to antimicrobials that discriminates epidemiologically significant organisms from other apparently similar hospital organisms. For example, many U.S. hospitals encounter HAIs due to Enterococcus strains resistant to vancomycin [38]. Such organisms can be the subject of infection control activities only if the laboratory maintains effective and efficient means for their identification [39].
Need for Accuracy and Consistency
Many spurious outbreaks have been traced to inaccurate or inconsistent microbiologic procedures. An “outbreak” of S. aureus infection, for example, may be caused by delayed reading of coagulase tests, resulting in misidentification of coagulase-negative organisms as coagulase positive. Unfortunately, most of the available rapid tests identify organisms that are not common HAI pathogens. The challenge, therefore, remains to develop rapid testing methods for the organisms closely associated with HAI (especially staphylococci, enterococci, and gram-negative aerobic bacilli). Recently, such tests for these organisms using rapid selective media or PCR have become available.
Performance characteristics (e.g., sensitivity, specificity, reproducibility) of some of the rapid tests for identification of hospital pathogens are not good [40]. This means that the rapid tests are used only as an adjunct to other testing. Some of these tests tend to increase care costs rather than decrease them, and their utility is not clear. Improving test methodology will be essential if tests such as these are to assume a strong role in infection control [41].
The renaming of organisms that result from more precise knowledge of organism relationships also can cause confusion for HAI personnel. For example, the renaming as Xanthomonas maltophilia of the emerging nosocomial pathogen formerly called Pseudomonas maltophilia gave false alarm to institutions not used to seeing or dealing with what appeared to be a new intruder [42]. That organism now has again been renamed as Stenotrophomonas maltophilia. This constant effort to be more and more precise in taxonomy is less desirable in the era of managed care, and some now question the practice of slavish adherence to taxonomic and nomenclatural changes in the clinical setting [34].
Introduction of New Procedures
The laboratory also must consider whether additional laboratory techniques can make testing results more relevant. For example, cultures of intravenous catheter tips may become positive because of contamination at the time of catheter removal or from the intravascular device becoming contaminated. Several semiquantitative and quantitative methods for culture of intravenous catheters [43] have been shown to be useful in distinguishing between these possibilities (see Chapter 37). Similar claims of usefulness have been made for cultures of other fluids, burn wounds (see Chapter 36), or surgical wounds (see Chapter 35). It is not clear that these special techniques generate useful information.
Quality Control
Just as an effective clinical microbiology laboratory is essential to an effective infection control program, adequate quality control is essential to the practice of good clinical microbiology [31]. Such a quality control program begins with a comprehensive procedure manual that establishes standards for performance, including the definition of acceptable and unacceptable quality of specimens and specimen containers, permissible delay between collection and receipt of the specimen in the laboratory, and times during which specimens are accepted for processing. The action to be taken by workers when specimens are not in accord with these standards also must be defined. These standards should be communicated to clinicians and nurses and to laboratory personnel.
The procedure manual also should cover administrative aspects of laboratory operation related to infection control and employee safety [44]. Minimum standards for identification of isolates, including a list of the equipment and reagents to be monitored and the measures to be made to ensure reproducible and accurate performance, should be provided. The periodic evaluation of skills of all employees, including evening, night, and weekend workers, should be included in the program.
Participating in proficiency testing programs helps the laboratory maintain competence, particularly if proficiency test specimens are submitted to the laboratory in a blinded fashion and are handled by routine procedures [45]. If problems develop with such an evaluation, the identity of the problem specimens should be made known and the personnel challenged to deal with the specimen in a
P.134
fashion as careful as possible to ensure that the laboratory actually has within its capability the correct handling and identification procedures for the organisms.
In addition to such outcome-oriented projects, periodic review of selected laboratory materials, media, and other equipment should be performed. On occasion, erroneous microbiologic results related to the inadvertent use of contaminated or faulty materials may occur. For example, an epidemic of pseudomeningitis was traced to contamination of funnels and an automated Gram-staining apparatus [46]. Such “pseudo-outbreaks” must be considered when laboratory culture or stain results do not correlate with clinical or epidemiologic findings.
Hospital-supported continuing education is essential for high-quality work in the microbiology laboratory. It is especially important for personnel in smaller hospital laboratories to stay abreast of technologic advances and trends in HAI occurrence and diagnosis [9]. Fortunately, a number of organizations, including the American Society for Microbiology, the Association for Professionals in Infection Control, the Society for Healthcare Epidemiology of America, and the American Society of Clinical Pathologists, provide frequent programs on HAI topics.
Accurate Characterization of Antimicrobial Susceptibility of Healthcare-Associated Infection Pathogens
A standardized method of antimicrobial susceptibility testing subject to quality control evaluation is essential in any clinical microbiology laboratory and is equally critical to infection control studies. Occasionally, an epidemiologist suspects that a group of HAIs with organisms of the same species have a common origin. To investigate whether strains in this cluster are common or different, the usual practice is to examine results of speciation, biochemical tests, and the pattern of susceptibility to antimicrobial agents [4]. Often these results answer the question of relationships. Occasionally, additional tests are needed; these are described in a later section of this chapter.
New patterns of antimicrobial resistance have been characteristic of the organisms causing HAIs in the recent past (see Chapters 14, 15, 40). Organisms that had been consistently susceptible to older antimicrobials have developed resistance to these drugs, and some HAI pathogens have developed resistance to new antimicrobials almost as soon as the drugs were marketed [47]. Methicillin-resistant S. aureus (see Chapter 14, 40) or coagulase-negative strains are involved in HAIs nationwide [48]. Enterococci have increased in importance as HAI pathogens; some of these strains have become resistant to aminoglycoside and β-lactam drugs that had been the drugs of choice for treatment of serious infection due to this organism. Since the 1990s, strains of enterococci resistant to vancomycin (see Chapter 14, 40) have become widespread in many healthcare settings [49]. Enterobacteriaceae, a common source of HAIs, have developed resistance to some of the newer β-lactam and fluoroquinolone antibiotics. The sequential appearance and persistence of these resistant organisms suggests spread of the resistant organisms within the hospital [35].
Several of these current resistance patterns require new or modified laboratory techniques for detection. For example, detection of vancomycin-resistant S. aureus requires several modifications of susceptibility testing techniques and is especially a problem for automated systems of detection [50]. Vancomycin-resistant enterococcal (VRE) strains also require special means for identifying resistance, and for these organisms, automated detection systems are less reliable [49]. Detection of high-level aminoglycoside resistance in enterococci also presents challenges [51]. Detection of resistance to newer cephalosporins, ureidopenicillins, or β-lactamase inhibitor compounds in enterobacteriaceal strains poses special problems as well, although the role of automated testing systems seems more secure [52].
Many laboratories use the Kirby-Bauer disk agar-diffusion method or an equivalent test system for routine testing of antimicrobial susceptibility of bacteria [53]. However, many other laboratories routinely perform a more quantitative evaluation of sensitivity, using broth-dilution or agar-dilution test methods [53]. In addition, tube dilution, E-test, or other methods of establishing minimum inhibitory concentration or susceptibility to “gate” concentrations of antibiotics must be used for testing organisms that have not been standardized for testing by a disk method. The latter include a number of anaerobic bacteria, fungi, or yeasts. Other sources may be consulted for detailed discussion of the performance and quality control of these procedures [37,54].
Some microorganisms can be additionally differentiated by indicating the relative degree to which susceptibility or resistance to antimicrobials is present [4]. This can be done by noting the absolute value of zone size in agar-diffusion testing or by providing assessment of minimum inhibitory concentration. Situations in which this more quantitative information would be useful should be delineated jointly by the laboratory and infection control team.
Selection of Strains for Susceptibility Testing
Applications of susceptibility tests to bacteria that are doubtfully related to infection must be avoided, and the laboratory should establish specific guidelines for the selection of isolates for susceptibility determination. For example, the request for testing of susceptibility should be carefully evaluated when the organisms isolated are endogenous flora present at sites in which they are not normally pathogens. Similarly, testing organisms from mixed culture should be avoided in most instances because of the unclear role of the various isolates [33]. Direct testing
P.135
of urine and spinal fluid is not essential in most instances. Potential pathogens with well-established susceptibility to antimicrobials (e.g., Streptococcus pyogenes to penicillin) should not be tested routinely.
Selection of Drugs for Routine and Special Testing
The laboratory should undertake the selection of drugs for routine testing after consultation with the infection control committee and the pharmacy and therapeutics committee; the chosen agents should reflect both the common usage practices of physicians in the hospital and the spectrum of pathogens that are frequently encountered [55]. Occasionally, testing of susceptibility to certain drugs is performed for epidemiologic purposes; results of such testing may be omitted from routine clinical reporting. Similarly, certain antimicrobials for which the hospital wishes to control usage may be tested but not reported routinely or tested only after consultation.
Different groups of antimicrobials often are used for gram-negative or gram-positive aerobic organisms. Drugs included in each panel should be periodically evaluated and updated. The epidemiologic value of susceptibility patterns may be enhanced by inclusion of certain antibiotics that are not in routine clinical use. Such additional information also can provide valuable taxonomic and quality-control information, but these benefits must be weighed against the extra time and cost required for testing and recording of the additional studies.
For some of the newer HAI pathogens, susceptibility testing methods are not very good. For example, susceptibility testing methods for fungi do not correlate well with clinical outcome [56]. Thus, another challenge is the need to develop susceptibility testing methods for some of the newer organisms closely associated with HAIs, especially gram-negative nonfermentive bacilli, fungi, or viruses.
Quality Control
Consistent and accurate identification of organisms over time is necessary for susceptibility data to be useful for clinical and epidemiologic purposes. In addition, errors in performance of susceptibility tests may result in information that is misleading about therapy. To minimize this possibility, detailed quality control procedures must be maintained for all elements of the susceptibility testing process [53,54]. Special attention must be given to the storage of reagents, control of batch-to-batch variation in media, use of control strains for testing, and monitoring of incubation temperatures and atmosphere. When results of these quality control tests exceed acceptable limits, reports on clinical isolates should be withheld until satisfactory control results are obtained. The reproducibility of susceptibility tests also can be assessed by participation in quality-control programs of groups such as the College of American Pathologists and the Centers for Disease Control and Prevention (CDC), which periodically distribute unknown specimens for evaluation. Such testing programs focus on clinically and epidemiologically important strains; correct identification assures the laboratory and infection control personnel that the laboratory applies proper techniques and skills.
Timely Reporting of Laboratory Data and Participation in Surveillance of Healthcare-Associated Infection
To deal with individual problems of HAIs in the hospital as they arise, control measures must be taken as quickly as possible and must be based on accurate assessments of the problems and their causes [4]. Without rapid identification and reporting of the organisms involved, control measures cannot be efficiently designed and implemented.
Laboratory records represent an important tool for infection control professionals (ICPs) [7]. Development of computerized laboratory information systems has progressed rapidly, and this has led to major improvements in several ways that the laboratory can provide infection control information [57].
Surveillance
Laboratory records are an important tool for the surveillance of HAIs [7]. More than 80% of infections defined by other criteria as nosocomial may be identified by review of positive cultures from the microbiology laboratory. Review of laboratory records is the most common method for surveillance of hospital infection carried out in the United States. Therefore, data gathered by ICPs during laboratory visits form an important base to which additional surveillance data from clinical rounds must be added. Both sources must be used to obtain an accurate estimate of the true rate of HAI occurrence in a given hospital (see Chapter 5, 6).
For both endemic and epidemic HAIs, microbiologic and immunologic reports may be the starting point for additional epidemiologic investigations. These investigations often require information about attributes of the patient, the personnel involved in care, and/or the diagnostic and therapeutic procedures provided to the patient. Obtaining these nonlaboratory data usually is easier when the patient is still present in the hospital or at least is fresh in the minds of hospital personnel. Promptly reporting pertinent laboratory results facilitates information retrieval of this type.
Computer programs have been developed to identify clusters of infections with the same organism and susceptibilities that occur at the same time in the same patient care area (ward or service) [58]. Such programs have permitted identification of outbreaks; whether they provide such information in rapid enough fashion to permit the use of control measures remains open to question [59].
P.136
The laboratory can indicate only which organisms were present in culture. The epidemiologist or ICP must supplement this information with clinical data to determine whether organisms found in culture indicate infection or colonization [15]. If colonization is present, the identification of organisms in the culture may be of little help to the clinician (but important for infection control). To the epidemiologist, however, both the organisms involved in episodes of exogenous colonization and those of infection are of interest. Either may be evidence of the spread of organisms from one site to another, indicating an area in which control measures may halt transmission. Data used for this purpose must be accurate, which emphasizes the role of continuing quality control studies in the laboratory.
Reporting of Results
To facilitate HAI surveillance of all infections requiring isolation or notification of public health authorities, a copy of positive culture results should be provided to the infection control personnel. Physicians or nurses sometimes are lax about notifying public health authorities of reportable diseases. Isolation of such organisms is reported to health authorities more efficiently if the responsibility for reporting is delegated to the ICP or some other person designated by the infection control committee (see Chapter 5, 6, 12).
Availability of a computerized laboratory information system can make it possible for the laboratory routinely to produce frequent and tailored reports for the ICP (see Chapter 5). For example, such a report is generated at the start of each day at Grady Memorial Hospital [7]. The report lists selected positive cultures and immunologic tests from the previous day sorted by ward in the order in which each practitioner makes daily rounds. The only culture results selected and printed on this report are those specified as relevant by the ICP. This maintains a relatively concise report while ensuring that the ICP has access to all information of current interest. The list of results to be selected is changed on a periodic basis to make sure that current needs are addressed (e.g., when the name of an organism has changed or when a new HAI pathogen has arisen at the institution).
Prompt reporting by telephone to both clinicians and infection control personnel is essential when presumptive identification of isolates of nosocomial significance is made; this is the only way to ensure proper treatment of the patient and the application of proper infection control precautions [60]. Occasions for reporting include incidents such as the presumptive identification of certain agents in meningitis, isolation of Salmonellae or Shigellae spp. from stool specimens, positive smears or cultures of Mycobacterium tuberculosis bacilli from any patient or employee, or isolation of S. aureus from lesions of a newborn or other nursery patient.
Laboratory studies may provide early warning of the emergence within a hospital of highly infectious microorganisms, multidrug-resistant organisms, or clusters of unusual infections. In some hospitals, laboratory workers may be the first to detect these and other trends of infection. When findings suggest a possible outbreak, notification requires quicker action than a final report because rapid epidemiologic investigations triggered by the first preliminary data from the laboratory often are profitable. The major elements needed in any early warning are the interest and expertise of the laboratory worker in calling results to the attention of infection control colleagues. This may be done by telephone, pager, computer, or e-mail if urgent; if not, discussion during the daily visit of the infection control staff usually suffices.
Such reporting facilitates the efforts of the infection control personnel. At the same time, early warning must not be requested for so many situations that this becomes an unreasonable burden on laboratory personnel. The key is consultation between laboratory and infection control personnel to establish which findings need to be given critical-value status [60].
Laboratory Records
In addition to instituting control measures, infection control workers often need to analyze laboratory data from various periods to try to detect patterns of infection [4]. To assist in this effort, it is helpful if the laboratory can provide an archival summary of organisms on a periodic basis. Data of particular usefulness might include compiled listings of organisms by culture site, date, patient, and ward; a summary of susceptibility testing results for various species of organisms for given time periods might also be helpful. Computer storage and retrieval of all results can aid this process considerably. Some newer laboratory information systems permit downloading of information pertinent to infection control to a desktop computer [59]. Then these data can be used both by the laboratory and in infection control if these departments have a compatible desktop computer. The specific laboratory data that can aid epidemiologic analyses vary from hospital to hospital. The information to be included and the frequency with which such summaries are made should be determined by the people providing and working with the data in each hospital.
Laboratory records should be retained in such a way that they facilitate such retrospective epidemiologic investigations and quality control activities. The source of each specimen, date of collection, patient identification, hospital number, hospital service, ward, and organisms identified in the final report should be recorded. Records of results of antimicrobial sensitivity tests and of any special biochemical or typing reactions also should be kept.
All cultures should be recorded so that results are readily available by date, type of specimen, and pathogens isolated. Culture data on inpatients and outpatients should
P.137
be maintained separately. Computer storage and retrieval of all results is optimal [59]. These records also can be maintained in simple, inexpensive, and epidemiologically useful bound log books that are kept chronologically for each major type of specimen (e.g., blood, wound, skin, cerebrospinal fluid, urine, stool, sputum). Sole reliance on a filing system of loose laboratory slips is not desirable because specific data are difficult to retrieve and easily lost.
The permanent records of the microbiology laboratory should include dates and other details of any major changes in culturing techniques, tests used, or laboratory procedures. Dates of changes in the criteria for identification and taxonomic designations applied to isolates should be recorded as well.
Retention Period for Records
No analysis of previous data can be made if the records are not available. Thus, it is incumbent on the laboratory staff to maintain the microbiologic records in some accessible format (e.g., computer file, final report sheets, disk or tape storage) for a reasonable period. The length of time such records can be maintained depends on hospital size, laboratory work volume, available storage facilities, and infection control needs. Thus, storage time should be determined by laboratory personnel after consultation with the hospital infection control staff. One author considers 18 months to be a reasonable minimum [16]. With computer storage, it may be possible to maintain data for long periods.
Summary Reports for Clinical Use
Development of profiles for susceptibility of frequently tested pathogens to drugs commonly in use can be of considerable assistance in guiding therapy for sepsis of unclear cause and other infections. Testing other organisms (e.g., slow-growing bacteria or organisms requiring special test procedures) may be performed at intervals to develop a profile of their susceptibility. As long as susceptibility patterns can be presumed to remain stable, such testing may be a useful substitute for testing each isolate at the time of recovery.
These summaries of susceptibility patterns should be available to the medical staff on at least an annual basis [7]. In addition to the hospital epidemiologist (see Chapter 2), any or all of the infection control committee, medical staff committees, and quality assurance committee also should receive susceptibility summaries to guide their review of antibiotic use. The use of a laboratory information system to tabulate the data directly or to download the raw data to a desktop computer for calculation of results has eased the burden of performing this task by hand. The laboratory information system also can be used to provide information about proper use of antimicrobial agents [47]. Increasingly, laboratory data can be downloaded from computer files to infection control or other computer systems.
Tabulations that may be of particular use include frequency of susceptibility to individual drugs by site of infection or ward (which may provide guidance to the clinician for empiric therapy of infection before the causative organism has been identified) and tabulation of frequency of susceptibility to individual antimicrobials by pathogen (which may be used to direct therapy after an organism has been identified but before susceptibility tests have been completed) [58].
A list of the relative costs of the currently used antimicrobials may be developed with cooperation of the pharmacy; inclusion of this information with susceptibility summaries may increase the incentive to reduce costs of antimicrobial use [58].
Additional Studies to Establish Similarity or Difference of Organisms
On occasion, the epidemiologist suspects that a group of HAIs with organisms of the same species have a common origin. Determining the features of the epidemiologic problem or testing certain hypotheses about reservoir or mode of spread may be aided if the laboratory can define whether the individual strains are related or unrelated to each other [4]. To investigate whether strains in this cluster are common or different, the usual initial practice is to examine results of speciation, biochemical tests, and pattern of susceptibility to antimicrobial agents. However, for organisms commonly encountered in the hospital (e.g., S. aureus or Klebsiella), the general pattern of these results may be similar by chance alone. Conversely, for other organisms (e.g., P. aeruginosa), the variation in these characteristics from strain to strain is so small that the tests provide little information about similarity or difference of tested strains. Testing of additional antimicrobials not ordinarily included or of susceptibility to other antibacterial substances (e.g., silver) may differentiate strains in some instances. Many times, however, these tests can show no differences. In these situations, examination (“typing”) of additional organism characteristics (markers) can be of great assistance [61].
Although hospital laboratory personnel may not have the facilities to perform specialized typing procedures, they should know which organisms can be typed and which cannot and where specific procedures can be obtained. When epidemiologically important isolates require special typing, it may be necessary to forward them to public health or private reference laboratories. Potentially pathogenic materials should be packaged for air transport in conformance with federal regulations [62].
Methods for Typing of Isolates
A variety of techniques has been used for typing isolates [61]. Selected typing systems of special value in
P.138
investigating HAI problems are summarized in Tables 10-1 and 10-2. So many typing systems are being used today for so many organisms that only a few examples can be provided in the tables. Several reviews provide detailed descriptions of the procedures used for each [1,61]. Many are beyond the capabilities of the usual clinical laboratory, but a number can be conducted when circumstances dictate.
|
TABLE 10-1 |
||
|
||
|
TABLE 10-2 |
||
|
Although routine typing of all strains is not cost effective, some laboratories store selected isolates for later typing should it prove desirable. Some typing systems monitor changes in characteristics expressed by the microorganism (phenotype; see Table 10-1), and others analyze changes from organism to organism in chromosomal or extrachromosomal genetic elements (genotype; see Table 10-2). Each is considered in turn.
Phenotypic Techniques
A number of organisms involved in HAIs have been differentiated successfully by antimicrobial susceptibility testing. The pattern of resistance or susceptibility to several different antimicrobials often is referred to as the “antibiogram” of the organism [106]. Test antimicrobials not in routine clinical use can be of assistance here [107]. However, the antibiogram of strains in the same clone can
P.139
change over time [108]. Susceptibility to various chemicals, especially heavy metals (i.e., resistotyping), also has proved valuable in selected instances [67].
Biotyping is the use of certain characteristic biochemical reactions to identify subgroups of bacteria. Typing schemes using this method have been devised for a variety of bacterial organisms, both anaerobes and aerobes, and found to be useful in infection control investigations. The process can be simplified by automated performance and analysis by use of computerized methods [109]. Unfortunately, the biotype codes generated by many of the commercial identification systems are poorly reproducible [108]. The method is most useful when unusual biotypes characterize the HAI pathogen.
Bacteriophages are viruses that can kill certain bacteria by lysis. Susceptibility to bacteriophages (i.e., phage typing) is a characteristic used for typing a number of organisms of nosocomial importance. The technique has been useful for grouping strains of S. aureus. However, this procedure usually is available only in reference laboratories, and plasmid transfer of phage characteristics apparently can occur, so it has been replaced by genotypic techniques in most situations [72].
Serotyping is a technique used for typing many gram-negative aerobic bacilli, especially Klebsiella pneumoniae isolates [73]. The technique can be helpful for other organisms as well, in both outbreak situations and research investigations. However, when reagents are not readily available or when typing procedures are complex, the techniques are available only in referral laboratories.
Many bacteria produce products that can kill or inhibit the growth of other organisms. Production of such bacteriocins by an organism, or susceptibility of the organism in question to those produced by other bacteria, can be used as a typing tool for a number of organisms. However, the method requires careful use of controls; widespread agreement on standards for reagents and interpretation is unusual. In addition, reproducibility of results is a problem; all strains should be tested at the same time [110]. Thus, the method also has been supplanted in favor of genotypic procedures for most situations.
Electrophoretic typing of proteins results from isolating relevant proteins, separating them by protein agar gel electrophoresis, and staining the gel so that the resulting pattern can be observed. The individual enzymes present in a given organism vary in structure from those present in other organisms; this is reflected in their different mobility when subjected to electrophoresis. Separating these compounds by electrophoresis on a starch gel can provide information on variation in a number of different enzymes at the same time. Such testing is called multilocus enzyme electrophoresis. A variant of this is immunoblotting, in which a nitrocellulose membrane is used as a site for the application of antisera of various sorts.
Genotypic Techniques
Molecular and other, newer techniques have permitted more definitive typing of a wider range of organisms than ever before [61,111,112]. These methods have helped immeasurably in defining mode of spread, reservoirs, and asymptomatic and unsuspected sources of infection [113]. It is now clear that these techniques can provide an epidemiologic picture different from that of prior methods. In many instances, these techniques have supplanted the traditional phenotypic methods [114].
The simplest of these genetic techniques is the analysis of the plasmid profile of a given organism or group of organisms. Most bacterial species carry plasmids, which are extrachromosomal pieces of DNA that encode a variety of genes. After isolation, plasmids are separated by electrophoresis and the pattern (number and size) of the plasmids from different organisms are compared (so-called fingerprinting). Of course, if an organism has few or no plasmids, this technique provides little assistance. In addition, the rapid pace of gain or loss of plasmids by a single organism and the frequent exchange of plasmids between different organisms makes interpretation of this technique difficult [61,115].
Both plasmids and chromosomal DNA can be subdivided further for typing purposes by using restriction endonuclease enzymes that divide the genetic material into smaller fragments. The fragments then can be separated by agar gel electrophoresis and compared as a further fingerprinting step. The bacterial chromosome usually contains some regions that are variable and other regions that are quite similar among different strains of the same species (“conserved”). The regions with different nucleotide sequence result in different patterns in the agarose gel for different organisms. These are known as restriction fragment length polymorphisms. Plasmid evaluation by this method usually is simpler than that for chromosomal DNA. The utility of this technique for analysis of chromosomal elements is subject to confounding by presence of DNA in plasmids and may be difficult to interpret because of the number and variety of patterns produced.
Ribotyping is the use of ribosomal RNA as a probe to detect variations in the DNA sequences associated with ribosomal operons [111]. Although most genes are present in bacteria in single copies, ribosomal operon genes are unique in that they are present in multiple copies. This makes analysis easier. In addition, only a small section of the genome is being examined, so a small number of bands are produced in the analysis. These are easier to analyze than the huge number of fragments that result from analysis of the whole chromosome. However, ribosomal gene patterns are relatively stable within a given species, so the ability to discriminate between or among epidemiologic isolates may be hindered to a certain degree [61].
Pulsed-field gel electrophoresis also begins with lysis of organisms and digestion of their chromosomal DNA with restriction endonucleases [114]. However, the restriction
P.140
fragments are resolved into a pattern of discrete bands in the gel by switching the direction of current. The number and size of fragments from different organisms then are compared. This alteration of electric field more readily allows resolution of large DNA molecules. The method allows typing of a broad range of bacteria. A hindrance is that there are no standardized criteria for analyzing the fragment patterns; however, consensus approaches are being proposed [114].
A variant of pulsed-field gel electrophoresis is contour-clamped homogeneous electric field electrophoresis. It is used primarily to compare large chromosomal DNA fragments [1]. DNA is separated by subjecting the molecules alternately to perpendicular electric fields. The high-molecular-weight DNA then can be subjected to agarose gel electrophoresis or digested by restriction endonucleases. Bacteria, yeast, and mycobacteria can be studied using this approach.
Amplification techniques like PCR are widely used in characterizing HAI pathogens [116]. Analysis of the products of such amplification can provide information on organism typing. By enhancing the number of target molecules by several logs of concentration, the study of the variability of the individual targets is enhanced. However, specialized equipment and personnel are needed to conduct this procedure, and this is a major problem for many hospitals [117].
Variations using amplification methods are emerging constantly. One is the use of repeated PCR assay aimed at different parts of small sections of a genome. Arbitrary primed PCR assays are variations of the amplification technique using short primers whose nucleotide sequence is not directed at a known genetic site [100]. These result in amplification of unpredictable loci and generate a set of restriction fragments that can be compared for different isolates. Because the discriminatory power of the assay is not clear, interpretation of these tests depends even more strongly than usual on the correlation with epidemiologic data.
There has been a dramatic recent increase in methods and equipment for analyzing the exact nucleotide sequence of microbes. It now is technically possible to compare multiple isolates by sequencing the same genetic site from all [61]. Differences in content and arrangement of nucleotides are being used for epidemiologic typing studies [106]. Further work is needed to define the exact role that such techniques should play in cost-effective epidemiologic analysis.
Choosing Typing Systems
Although the potential benefit of these new techniques for hospital infection control is great, some cautions are needed regarding their use [4,114,118]. Sometimes the new systems for analysis have impeded rather than aided in outbreak investigations. For example, some investigations produce conflicting results from typing studies carried out by different techniques [72]. Incorrect results also may arise from contamination of reagents, a special concern with use of PCR, with which sensitivity is so great that the chance of contamination is high [119]. Some typing methods provide data of doubtful value or results that are difficult to interpret for a given epidemiologic situation [120]. To improve these, further information is needed to evaluate their validity as a tool for infection control [4]. Some have suggested that in analyzing subgroups of clonal organisms, it is necessary to use at least two typing methods [72,115].
Even if the methods are valid, overinterpretation, or underinterpretation of the results is a potential problem. Typing never proves that organisms are the same [61]. On the other hand, the use of too many individual markers in several typing methods can cause problems as well [121]. For example, use of 20 different antibiotics in typing a series of specimens in which organisms are defined as different if their pattern of susceptibility varies by more than one drug almost guarantees that organisms that actually are part of the outbreak will be identified as different and thus considered unrelated [4].
Results from some typing methods may be redundant; if so, this needs to be discovered so that the less expensive methods can be used. An important question for current investigation should be not whether new tests provide additional discrimination but how much discriminating ability they add to readily available tests [122]. For example, in a study of infections with coagulase-negative staphylococci, antibiogram, biotyping, phage typing, and plasmid profiling all were performed. The antibiogram, selected as the first stage of the scheme because it was the simplest and cheapest test, proved to be the most discriminatory stage, providing 66% of the discriminating ability between strains [123].
A related issue is how often most hospitals really need these newer techniques. Most of them are used for research (detailing routes of transmission or reservoirs of organisms) rather than for acting on outbreaks. For most of the newer molecular testing methods, instruments and reagents are not readily available and have yet to be widely automated [117]. The expertise to perform and interpret the tests is not widespread, and standards are lacking for most assays [117]. In addition, outbreak organisms rarely are all isolated at once, so typing procedures are performed at separate times; the need for control of batch-to-batch variation in media, reagents, phages, and so on is especially crucial with many of these typing procedures. These considerations explain why many newer typing procedures should not be performed in the routine clinical laboratory [117]. Thus, not every laboratory must be able to perform every, or even most, of these tests. The laboratory director must be aware of the available tests, determine which can be performed on-site when needed, and identify referral laboratories (e.g., colleague, state, private) where other methods are available on request.
P.141
Storage of Strains
For supplemental tests to be performed, such as those described in the preceding section, the organisms must be available. Thus, a related duty of the clinical laboratory staff is to retain strains that may relate to HAIs for a given period while it is determined whether additional testing is needed. In cooperation with infection control personnel, the laboratory staff should subculture and save epidemiologically important isolates, whether such isolates are from outbreaks or from single instances of unusual or potentially epidemic diseases. Isolates may be conveniently and inexpensively stored by placing a small amount of growth on a blank paper disk that is placed in a 2-ml glass screw-cap vial containing a few granules of silica gel. If the vial is kept tightly closed, isolates may be held up to 6 months; they can then be easily retrieved by placing the disk in broth.
A system for reviewing and periodically discarding these isolates also must be established. How long a storage period is required for this purpose varies from hospital to hospital and should be agreed on between epidemiologist and clinical laboratory director. The technique used to ensure the viability of the organisms (e.g., freezing, lyophilization) should be determined by the laboratory staff after considering the equipment and personnel that are available for this task.
Occasional Microbiologic Studies of Hospital Personnel or Environment
In some situations, HAIs may result from environmental or personnel sources. In evaluating such episodes, the infection control staff may ask the laboratory staff to process specimens from employees, environmental sources, or hospital equipment (e.g., respiratory therapy machines) as part of the investigation [4]. Such environmental culturing, however, should focus on investigation of documented infections in patients (see Chapters 7, 19) [2]. When such culturing is necessary, its costs should be considered part of the hospitals' infection control program, and charges for the cultures should not be billed to the patients involved in the outbreak.
A few procedures of this type should be done routinely (Table 10-3). Others are elective, performed in association with episodes of patient illness or as part of an educational program. A third group is specifically not recommended. Each is dealt with in turn.
|
TABLE 10-3 |
||||
|
Routine Environmental Sampling
Surveillance cultures of the hospital environment and personnel once were advocated on a routine basis. During the 1970s, studies found these programs to be of minimal value in infection control; by 1980, most institutions took the approach that routine environmental culturing should be severely limited. Since then, limiting cultures of the environment has become even more imperative because of changes in the economics of healthcare in the United States. Fortunately, further studies have supported this selective approach [124]. Close communication between epidemiologist and microbiologist about the need for such cultures continues to be essential to keep from wasting valuable resources.
In the absence of an epidemic, sampling should be minimal; microbiology and infection control personnel should be firm in not conducting indiscriminate routine microbiologic sampling and testing (see Chapter 7, 19) [2]. However, routine checks on the adequacy of sterilizer function, culture of dialysis infusates and the water used to prepare them, and culture of infant formula and some other products prepared in the hospital may help prevent infections from these sources (Table 10-3).
Monitoring of Sterilization
All steam sterilizers should be checked at least once each week with a suitable live-spore preparation [125]; if sterilization is performed less often, testing should be done on each day that sterilization is done (see Chapters 19, 20). Ethylene oxide gas sterilizers should be checked with each load of items that come into contact with blood or other tissues. In addition, each load in either type of sterilizer
P.142
should be monitored with a spore test if it contains implantable objects. These implantable objects should not be used until the spore test is reported as negative, usually at 48 hours of incubation. Dry-heat sterilizers should be monitored at least once each month [125].
The process of flash sterilization is used when contaminated items will be needed again in a short period of time. This time requirement makes the use of biologic indicator testing impractical [126]. This means that this method of sterilization is not feasible for items that are to be implanted in patients. Products for monitoring flash sterilization are being evaluated, but no consensus exists for the way they should be used. Likewise, low-temperature sterilization technologies are being developed to replace ethylene oxide use; these include vaporized hydrogen peroxide, gas plasmas, ozone, and chlorine dioxide [127]. Some progress has been made in developing biologic indicators for these newer systems [128], but no consensus yet exists. Developing these standard testing systems should remain a high priority.
All sterilizers should be equipped with time–temperature recorders to provide evidence of adequate exposure for each load. However, evidence that a sterilizing temperature has been held for an adequate time does not prove that sterilization took place; the temperature is measured at the outlet valve and does not reflect whether adequate sterilization occurred within dense volumes of fluid or large, dense, fabric-wrapped packs. The use of chemical monitors (e.g., test tapes or heat-sensitive color indicators) within the autoclave is recommended for the outside of each package sterilized [125]. This provides an indication that a sterilizing temperature may have been reached, although such monitors do not show whether there was an adequate duration of exposure. Therefore, additional monitoring systems are required, which ordinarily involve the laboratory: Biologic monitoring with spore strips generally has been accepted as the most effective way to determine successful sterilization (see Chapters 19, 20).
Microorganisms chosen for spore strip tests are more resistant to sterilization than are most naturally occurring pathogens. The test organisms are provided in relatively high concentrations to ensure a margin of safety. The spores may be provided either in impregnated filter paper strips or in solution in glass ampules. For steam sterilization, the thermophile Bacillus stearothermophilus is used, and for ethylene oxide and dry-heat sterilizers, Bacillus subtilis (var. niger) is used. Both species frequently are incorporated simultaneously in the test strips, and these can be used to test for adequate sterilization with either procedure.
Most spore strip preparations are packaged in envelopes that contain one or two test strips and a control strip. The test strips are packaged in separate envelopes and are removed and sterilized at the time other material is processed. Subsequently, the test strips and control strip are cultured by placing the strips in tubes of tryptic-digest casein-soy (TS) broth that are incubated at 37°C (99°F) for B. subtilis and 56°C (133°F) for B. stearothermophilus. It is not necessary to culture a positive control strip for each test; if strips are obtained from a single lot, only 10% of the positive control strips need be tested.
Spore solutions are prepared in sealed glass ampules for testing the adequacy of sterilization fluids. These ampules should be incubated at 56°C (133°F) in a water bath. If there is no change in the indicator by 7 days, the test is reported as negative. Alternatively, the fluid may be inoculated with a test culture, which may be subcultured after autoclaving.
Other types of spore preparations are commercially available and require different handling. In each case, the manufacturer's directions should be followed closely.
Test strips or spore solutions should always be placed in the center of the specimen to be tested, never on an open shelf in the autoclave. The center of a pack located near the bottom front exhaust valve will be exposed for the least adequate duration and temperature of sterilization and thus provides the best location for a test measurement. Testing the sterilization of fluids is accomplished by placing an ampule containing a spore solution in the largest vessel.
Use of ampules containing spore solutions is not appropriate for checking sterilization of microbiologic culture media because these media do not require the duration of exposure that is required for sterilization of material known to contain large populations of bacteria. Heating bacteriologic culture media to a temperature sufficient to ensure the sterilization of a test strip or spore ampule results in damage to the medium through overheating.
The likelihood of cross-contamination can be reduced by minimizing the handling of the strip after sterilization. Test strips can be removed from their envelopes and placed in sterile glass tubes before sterilization. The tube then is sterilized with the screw-cap removed or with other closures permeable to steam in place. After sterilization, the tube is sent to the laboratory, where the nutrient broth is added.
The handling of spore strips in the laboratory requires considerable care to prevent secondary contamination. The transfer should be made in a laminar-flow cabinet if available using sterile forceps and scissors. The forceps and scissors are common sources of contamination, which may be insufficiently sterilized by flaming or wiping with alcohol. Alcohol may contain viable spores that might not be killed by flaming, and flaming may be insufficient to heat instruments to a temperature that destroys viable spores. Care should be taken not to cross-contaminate the sterilized spore strips with the control strips.
Condensation on the cover of a 56°C (133°F) water bath may cause contamination of the caps and closures of tubes. A heating block may be used to avoid this, or the bath may be left uncovered; the latter makes it necessary to provide a reservoir to maintain the water level because the evaporation rate is high at this temperature. Uninoculated culture media should be incubated at 35°C (95°F) or at
P.143
56°C (133°F) to ensure that contamination does not yield false-positive reports.
Gram staining and subculturing should be performed to detect secondary contamination of test cultures. If organisms other than gram-positive bacilli are observed, the test should be repeated and reported as “possible laboratory contamination, test being repeated.”
Whenever positive results are obtained, the sterilizers should be checked immediately for proper use and function. Careful examination must be made of thermometer and pressure gauge readings, and recent time and temperature records must be reviewed. If any deficiency is observed or if the repeat sterility test still results in growth, engineering personnel and experts in autoclave maintenance and function should be consulted promptly. Objects other than those used for implants do not need to be recalled at this point unless defects are discovered in the sterilizer or its use; if spore tests remain positive after proper use of the sterilizer is documented, the machine should be removed from service until the defects are corrected (see Chapter 20).
Sampling of Infant Formula and Other Food Products Prepared in the Hospital
Few hospitals still prepare their own infant formula. However, infant formula prepared in the hospital kitchen should be monitored on a weekly basis [129]. A guideline for interpreting culture suggests that <25 organisms per milliliter be present and that no virulent bacteria, such as Salmonella or Shigella spp., be present [129]. The guideline is an arbitrary one, however, and should be a matter of local preference.
On occasion, other food products prepared in hospitals are considered as potential sources for HAI. Extensive guidelines for appropriate culture techniques have been developed [130] and should be consulted as necessary.
Culture of Blood Components
The American Association of Blood Banks [131] does not recommend routine sterility culturing of transfusion components. Likewise, surveillance bacterial cultures are not recommended for specific transfusion practices (e.g., exchange transfusions for neonates) [132].
Culture of Dialysis Fluid
The water used for preparing dialysis fluid should be tested by colony count at least once per month. Guidelines of the Association for Advancement of Medical Instrumentation [133] specify that the water used to prepare dialysis fluid and water used to rinse and reprocess dialyzers should contain <200 viable organisms per milliliter, and the dialysate <2,000 colony-forming units (cfu) per milliliter. Defined methods for such testing vary from one regulatory body to the next; in general, the correlation between the specified levels and occurrence of patient disease is poor. Counts obtained vary markedly with media and conditions of incubation [134] (see Chapter 24).
Periodic Sampling of Disinfected Equipment
Any article that makes direct contact with the vascular system or tissue other than unbroken skin should be sterile. Whenever possible, steam or gas sterilization should be applied. If chemical disinfection or pasteurization, rather than sterilization, is used on equipment such as cystoscopes, other endoscopes, or anesthesia equipment, some authorities recommend (and some require) that periodic microbiologic sampling be done to ensure the absence of pathogens after processing. Methods and equipment for testing have been specified [134].
The frequency of sampling such disinfected devices depends on any evidence that HAIs are associated with their use and an assessment by the infection control committee of the adequacy of standards for control of contamination of such equipment. There is little agreement on which items of this type should be tested or how often. If such a monitoring program is begun, it may be possible to cut back the frequency of culturing after a period of time in which cultures are negative as long as no changes are made in equipment and techniques used.
Infections have resulted from contamination of transplant materials such as bone marrow during processing [135]. Thus, on occasion, local or regional regulations require routine culture of transplant organs or tissues (e.g., eye, bone, allografts, porcine heart valves), laminar flow hoods, or pharmacy admixture solutions.
Elective Environmental Monitoring
A wide variety of items and substances can be responsible for cross-infection. Thus, environmental surveys may be useful during the investigation of specific problems within a hospital and should be instituted in response to, and specifically address, epidemiologic findings [4]. For example, microbiologic assessment of foods may be needed during investigation of suspected foodborne illness in the hospital setting [136]. Elective culturing programs may also be instituted in association with educational efforts.
Support for Investigation of Specific Problems of Healthcare-Associated Infections
These must be dealt with as rapidly as possible [4] (see Chapter 7). This means that the laboratory may face exceptional demands for service at the beginning of and throughout an epidemic period [16]. Advance preparation for such situations makes response easier in the time of need. The laboratory personnel should prepare contingency plans for the types of outbreaks that have occurred most frequently in the past in the hospital so that they are ready to deal with these exceptional requests in a smooth fashion.
P.144
Investigation of an HAI outbreak may require isolation and identification of isolates in specimens not only from patients but also from personnel who might be colonized with the outbreak strain and from environmental objects that might be similarly contaminated [4]. Such activity may require the laboratory staff to process and evaluate large numbers of cultures, and special techniques may be necessary to accomplish such projects. For example, reliable detection of colonization with VRE requires the use of selective media such as Enterococcosel broth [137]. The laboratory staff and the infection control team can process this work efficiently by carefully assessing the sites to be cultured and determining which culture media and techniques will be used.
A detailed description of suitable culture techniques for every possible vehicle of cross-infection is beyond the space allocation and practical scope of this chapter. The interested reader is referred to detailed protocols available for culture survey of hemodialysis equipment, dialysis fluid, intravascular devices, air cultures, and environmental and medical device surfaces, blood bank products, and water for Legionella spp. [134]. The development of selective media and techniques for culture of environmental objects continues to evolve as new potential reservoirs and vectors of HAI have been recognized [16]. The infection control worker and laboratorian must be familiar with the general aspects of culture procedures discussed in the following sections, but they should not obtain or process such cultures unless surveillance of infections in patients specifically implicates these items as potential HAI sources.
Because standard methods for the microbiologic evaluation of such culture procedures do not exist or are of doubtful validity, considerable expense may be incurred in producing information that is worthless or misleading. Requests for such cultures, therefore, should be approached with caution, and the infection control staff should be clear regarding how the culture results will affect patient care or epidemiologic control measures before undertaking such tasks [4]. Specific areas of support are considered in turn.
Culture of Blood Products After a Transfusion Reaction
Bacteria present in blood components can cause a septic transfusion reaction [138]. Fortunately, this is a very rare occurrence. Such reactions usually are due to endotoxin produced by organisms that can grow in the cold. If a transfusion reaction is suspected by clinical signs, the transfusion should be halted immediately and the unit examined. If further evaluation of the signs, symptoms, and clinical course of the patient then suggests bacterial contamination of the blood product, cultures of the suspect component(s) may be indicated at various temperatures [131,134]. It also is desirable to collect blood culture specimens by venipuncture from the patient and from all intravenous solutions in use for that patient.
Cultures of Parenteral Fluids and Intravascular Therapy Equipment
The investigation of bacteremia associated with parenteral therapy may require investigation of the needle, hub or catheter, portions of the administration set, the fluid being administered, and portions of the cap or closure provided with the fluid [134,139] (see Chapter 37). Blood culture specimens should be collected simultaneously from the patient. It is especially important to keep careful track of lot numbers, which should be recorded on the patient's chart and on all subsequent laboratory records.
Needles and catheters must be submitted separately from the hubs and other portions of the administration set that may have been exposed to superficial contamination. If portions of the administration set are suspected, these must be received properly capped to exclude spurious contamination. The bottle and administration set should remain connected and be placed in a plastic bag to minimize contamination during delivery to the laboratory.
The standard method for culture of catheters and needles has been the semiquantitative method in which the catheter tip is rolled across a plate containing solid media. Some authors suggest flushing the inside with nutrient broth or a quantitative sonication method for culture of the catheter instead of (or in addition to) a semiquantitative culture [140]. Substitution of hub cultures and cultures of the skin around the insertion point has been suggested as an alternative sample not requiring removal of the catheter [141]. Quantitative cultures are time-consuming and are a drain on personnel resources; the need to use them rather than the more efficient semiquantitative methods has not been demonstrated [140].
Methods for Culturing Hands and Skin
Several methods are suitable for culture of hands [134]. The simplest is to take a sterile swab, moisten it with sterile saline or a culture broth appropriate to skin organisms (brain–heart infusion broth is one), and then swab the palmar surface of the hand or hands. A rapid and simple way to obtain these cultures also is provided by pressing the subject's palm gently on a large culture plate containing a suitable agar medium (or a Rodac plate; see the section on culture of floors and other surfaces). When a more quantitative estimate of the organisms present is desired, 50 ml of culture broth can be poured into a sterile container or plastic bag. The person to be cultured is asked to rub hands together in the broth for 30 seconds. An estimate of cfu per milliliter can be obtained [142]. A bag-wash method has been found to enhance recovery of yeasts from hands of healthcare workers (HCWs) [143].
In contrast to these simple methods, sampling of hands associated with testing of antiseptics or disinfectants is a more complicated process and requires special techniques [144].
On occasion, monitoring antibiotic-resistant organisms has involved investigation of the skin of patients or HCWs
P.145
as potential sources of resistant organisms. However, quantitative culturing of skin or gastrointestinal tract seldom is indicated [15]. To quantify skin microflora, several methods have been developed [134].
Methods for Culturing Tubes and Containers
Cultures of external surfaces or internal cavities (e.g., tubes and containers) may be conducted by a swab–rinse technique [134].
Sampling of Respiratory Therapy Equipment
Disinfection of respiratory therapy and anesthesia equipment is discussed in Chapter 21. In situations of high endemic or epidemic levels of occurrence of nosocomial respiratory infections, sampling respiratory therapy apparatus may be of value [145,146].
Sampling of Air
Airborne spread of nosocomial bacterial or viral infection is known to occur [147] but is probably uncommon; air sampling should be required infrequently [148]. Common recent indications for use of this technique in healthcare settings include monitoring production areas for pharmaceuticals in operating rooms, and in investigating suspected episodes of airborne infection [149]. Air sampling may be performed with either settling plates or more sophisticated equipment [134,150].
Particles suspended in hospital air vary greatly in size and in the number of microorganisms they contain. The average diameter of airborne microbial particles in ward air is approximately 13 µm, but 7% are <4 µm in size and 30% are >18 µm. Particles with a mean size of 13 µm settle at a rate of approximately 1 foot/minute. Because the surface of a standard 100-mm Petri dish represents an area of approximately 1/15 square foot and assuming that the air in the study area contains particles of average size, an open Petri dish in still air will sample microbial particles from nearly 1 cubic foot of air during 15 minutes of exposure [151].
Although this is an inexpensive way to evaluate airborne microbial contamination, quantitative results may correlate poorly with those obtained with mechanical, volumetric air samplers because of variation in particle size and unknown influences of air turbulence. Under low-humidity conditions, droplet nuclei of approximately 3 µm can remain suspended indefinitely and can be collected only with high-velocity, volumetric air samplers.
A slit sampler is suitable for many precise air sampling applications [149]. Brain–heart infusion or tryptic soy agar (TSA) media should be used in the sampling plates. A staged sampler [150] should be needed only when there is some reason to determine the size distribution of the particles, which should be an extremely infrequent event in most U.S. hospitals. Efficient vacuum sources must be used for both samplers, and the rate of flow of air must be properly calibrated to ensure accurate results.
The total number of airborne microbial particles is not measured precisely by observing growth after impaction on an agar plate because an airborne microbial particle may contain more than one viable cell. Air-sampling techniques in which volumetric samples are taken by bubbling air through collection fluid break up airborne particulate matter and better reflect the total number of organisms than air samplers that impinge contaminated particles on agar.
Culture of Floors and Other Surfaces
Methods for sampling floors and other surfaces have been described in detail [134]. Standards for acceptable levels of contamination of floors and bedside tables as sampled by the Rodac plate technique have been suggested by a committee of the American Public Health Association [152]. There is no evidence, however, that any particular level of contamination is directly correlated with an increased risk of infection, and such standards probably are useful only in assessing the adequacy of house-cleaning procedures.
Sampling Water and Ice
Water that meets U.S. Public Health Service standards for drinking water frequently contains up to 1 million or more microorganisms per milliliter, and some of these organisms are potential pathogens. Ice also can contain organisms that can pose a threat of infection, especially in patients with compromised host defenses. However, correlation of levels of microorganisms with occurrence of patient illness has been rare.
Samples of water or melted ice can be obtained by collection in a sterile container. If chlorine is present, it can be inactivated by thiosulfate. The specimen is cultured by passing large quantities through a 0.45-µm (or 0.22-µm) Millipore filter and culturing the filter in broth or directly on agar. More than 4 colonies per 100 ml is considered abnormal by this test [151].
When transmission of Legionella infection occurs in the hospital setting, culture of water is recommended. Guidelines for such cultures have been proposed [153].
Developing Selective Media for Surveys
To reduce the workload in the laboratory and to expedite the processing of specimens, selective survey media should be used whenever possible for culturing specimens during outbreak investigations. Susceptibility data on known or suspected epidemic strains may be used to identify an appropriate selective medium for use in surveys of the animate and inanimate environment of the hospital. Once the implicated organism is isolated and tested on appropriate media containing ≥1 antibiotics to which it is resistant, the media can be used to exclude numerous bacteria unrelated to the outbreak. This may accelerate the detection of contaminated equipment or infected patients. Pretesting of the media is essential because of possible synergy or antagonism between the added antimicrobials or between these drugs and the media; such interactions
P.146
could cause inhibition of growth of the epidemic strain or failure to inhibit growth of nonimplicated organisms.
Other selective media also may be useful. For example, cetrimide medium may be helpful in selectively isolating P. aeruginosa from contaminated material or mixed cultures. Similarly, tetrathionate broth is an excellent medium for selective pre-enrichment of Salmonella spp. cultures. Mueller-Hinton agar containing sorbitol, a pH indicator, and antibiotics (vancomycin, colistin, and nystatin) provides selective differentiation of Serratia spp. Many epidemic strains of S. aureus are resistant to mercuric chloride, and the incorporation of small amounts of this compound in TSA can be helpful in inhibiting nonepidemic strains of S. aureus, Staphylococcus epidermidis, and most gram-negative organisms exceptPseudomonas spp.
Vancomycin-resistant strains of enterococci have become a problem in many U.S. hospitals; agar containing small concentrations of vancomycin have been of use in investigating hospital problems due to these strains [39,137]. The resistance of epidemic microorganisms to heavy metals, dyes, disinfectants, and other antimicrobial substances also may be used to identify and construct selective media for surveys.
Surveys for Educational Purposes
Sampling techniques that are not directly related to epidemiologic surveys may prove useful in educational programs; visible evidence of contamination of hands, clothing, equipment, and surfaces may serve to teach the need for effective aseptic technique and sanitation.
Sampling That Is Not Recommended
Routine Culture Surveys of Patients and Personnel
Routine culturing of patients or hospital personnel is not recommended (see Chapter 7). Surveys may be useful during investigation of specific problems within a hospital and should be instituted in response to, and specifically address, epidemiologic findings.
Routine Culture of Commercial Products
Although commercial patient-care items that are labeled sterile (e.g., intravascular catheters and fluids) occasionally have been contaminated with viable organisms that can cause patient disease, routine sampling of these items is not recommended because the low frequency of contamination makes it difficult (because of the large number of specimens that would have to be taken) and expensive to perform adequate sterility testing.
Routine Testing of Antiseptics and Disinfectants
In-use testing of antiseptics and disinfectants should not be a routine procedure for hospital microbiology laboratories [134]. No consensus is available about the optimal methods for such testing against many organisms, including viruses [154]. If contamination of commercial products sold as sterile is suspected, infection control personnel should be notified and the nearest office of the U.S. Food and Drug Administration contacted immediately [151]. State regulations may require immediate notification of state health authorities as well.
Random Culture of Blood Units
The American Association of Blood Banks [131] does not recommend random culture of blood units to ensure sterility.
Routine Monitoring of Disinfection Process for Respiratory Therapy Equipment
Guidelines from the CDC advise to “not routinely perform surveillance cultures of patients or of equipment or devices used for respiratory therapy, pulmonary function testing, or delivery of inhalation anaesthesia” [146].
Teaching Microbiologic Aspects of Healthcare-Associated Infection to Infection Control Personnel
The people responsible for infection control usually are not trained in clinical laboratory procedures. Because the key to success in infection control efforts is communication, it is necessary that all involved speak the same language. For this purpose, training infection control personnel in the language of the clinical laboratory microbiologist is important. Training many infection control personnel in microbiology is inadequate or out of date. The goal of such teaching is not necessarily to make the infection control staff accomplished laboratory workers but to familiarize them with the procedures and practices of the laboratory, the microorganisms involved in HAIs, the validity of test procedures used in identifying these pathogens, and the strengths and weaknesses of the resulting data.
Similarly, it is important for the microbiologist to learn some of the concepts of infection control, because few laboratory directors or technologists have adequate grounding in epidemiology or infection control. Especially important is exposure to techniques used for measuring frequency of infection and the concept of colonization versus infection.
Such joint efforts permit ready communication between the two groups of colleagues. Teaching of this type can be done in a formal fashion but also is effective when included as part of the day-to-day informal contacts between the infection control staff and laboratory personnel (see Chapter 2).
References
P.147
P.148
P.149
P.150