Franklin R. Cockerill III MD
Nancy K. Henry PhD, MD
Throughout the world, bacteria and viruses are among the most commonly encountered microbial pathogens in humans. Dimorphic fungi (ie, Histoplasma capsulatum, Blastomyces dermatitidis, Coccidioides immitis, Sporothrix schenckii, and Paracoccidioidomyces brasiliensis) and most parasites are generally limited to certain geographic areas. Immunocompromised patients are susceptible to common bacterial, viral, and parasitic pathogens that cause disease in normal hosts and to unusual bacterial, viral, fungal, and parasitic pathogens that infrequently cause disease in normal hosts. These latter organisms are opportunists and, owing to humoral or cell-mediated immune dysfunction within the host, may produce life-threatening disease.
Other organisms may cause disease in patients as the result of anatomical-barrier disruption. Endotracheal intubation may permit colonization of the oropharynx with nosocomial pathogens. These include gram-negative bacteria, especially members of the family Enterobacteriaceae (eg, Escherichia coli, Klebsiella spp., Enterobacter spp., Proteus spp., and Serratia spp), Pseudomonas spp., Stenotrophomonas maltophilia, or Burkholderia cepacia; or gram-positive bacteria, especially methicillin-resistant staphylococci (methicillin-resistant Staphylococcus aureus and methicillin-resistant coagulase-negative staphylococci) and vancomycin-resistant Enterococcus spp. (VRE) and yeasts. Any of these nosocomial pathogens may colonize the upper respiratory tract and spread to the lower respiratory tract, secondary to endotracheal intubation. VRE may also colonize the lower gastrointestinal tract and be spread from one patient to another by fecal contamination. Mucositis, as the result of total body irradiation and/or chemotherapy, may be associated with septicemia caused by oral flora, especially Streptococcus spp. of the viridans group and Candida spp. Intravascular catheters may become infected with cutaneous flora or nosocomial pathogens. These organisms include Staphylococcus spp., Streptococcus spp., and Corynebacterium spp., including Corynebacterium jeikeium, the same nosocomial gram-positive and gram-negative bacteria described previously, and yeasts. All of these organisms may be introduced into the bloodstream directly if intravenous fluids or medications are injected through an infected catheter port. Alternatively, these organisms may first colonize a proximal port and over time ascend to the catheter tip.
Neutropenia secondary to chemotherapeutic agents is frequently associated with bacteremia. Enteric gram-negative bacteria, notably members of the family Enterobacteriaceae, Pseudomonas spp., and the gram-positive organisms Staphylococcus spp. or Enterococcus spp., account for the majority of these bacteremias. Disseminated fungal infections caused by Candida spp. or Aspergillus spp. may also occur.
Because of the wide variety of microorganisms (bacteria, fungi, viruses, and parasites) that can cause infections in immunocompromised patients, in some instances, test-ordering protocols may be useful to assure that appropriate specimen processing and analyses are performed. This is especially important for specimens that are obtained by using relatively invasive procedures. For example, a standard battery of microbiology tests should be considered for all specimens obtained by bronchoalveolar lavage or transbronchoscopic, transthorascopic, or open-lung biopsies from immunocompromised patients. For all patients (normal or immunocompromised), the development of disease management strategies for commonly encountered infections may result in more consistent and appropriate use of laboratory tests.
Diagnosis of Bacterial, Fungal, & Parasitic Infections
Franklin R. Cockerill III MD
Nancy K. Henry PhD, MD
Tables 6-1, 6-2, and 6-3 summarize currently available recommended test methods for the detection of bacterial, fungal, and parasitic pathogens from human specimens. Methods for detection of viral pathogens are discussed later in this chapter. For completeness, the tables include pathogens common to both normal and immunocompromised hosts. Selected specimen collection procedures are shown in Table 6-4. However, this important step in the diagnostic-testing process cannot be overstated. An adequate amount of specimen (using sterile techniques when appropriate) should be placed in the appropriate transport device, and the specimen should be transported to the laboratory under appropriate environmental conditions and within a reasonable period. Because a large number of transport devices are available, consultation with the laboratory personnel is important, especially when unusual bacteria are considered. Comprehensive reviews of the specifics of specimen collection, transport, and storage are described elsewhere. As a general point, specimens requiring strictly anaerobic conditions for viability must be sent in appropriate anaerobic transport devices. Depending on the volume, most other specimens can be sent in either a swab transport device (the swab is immersed in nutrient broth) or a sterile container. Some studies have demonstrated the utility of blood culture bottles for transporting as well as culturing sterile body fluids, especially peritoneal fluid.
|
Table 6-1. Recommended rapid direct tests for detecting bacteria, fungi, or parasites in human specimens. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Table 6-2. Recommended culture methods for detecting bacteria, fungi, and parasites in human specimens. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Table 6-3. Serologic methods currently available for detecting antibodiesto bacterial, fungal, or parasitic pathogens |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAPID, DIRECT TEST METHODS & CULTURE-BASED METHODS
Direct, rapid diagnostic-test methods if available should be used in addition to conventional culturing techniques for diagnosing bacterial or fungal infections in critically ill patients. Most parasitic infections can be diagnosed by direct staining methods, and these tests should be performed emergently, especially if life-threatening infections like malaria are suspected. Rapid diagnostic tests are provided in Table 6-1, and culture-based diagnostic tests, which require longer time periods to complete, are presented in Table 6-2.
Blood
Both direct and culture-based techniques may be useful for identifying common and uncommon bacterial, fungal, and parasitic pathogens in blood. Some of the specialized testing methods shown in Tables 6-1 and 6-2 are generally available only at reference laboratories. If the clinician feels strongly that one of these tests is useful, an aliquot of the specimen should be sent to a reference laboratory. Direct test methods for the detection of bacterial or fungal pathogens in blood are limited in the number of different bacteria that they can detect and are less sensitive than culture methods. Bacterial-antigen tests for commonly encountered encapsulated organisms (Haemophilus influenzae type b, Neisseria meningitidis, or Streptococcus pneumoniae) are easy to perform and available in most laboratories. A latex agglutination procedure can be used to detect the yeast Cryptococcus neoformans..
|
Table 6-4. Selected specimen considerations for bacterial, fungal, andparasitic agents. |
||||||||||||||||||||
|
Blood cultures should be obtained from all patients in whom sepsis is considered. The standard blood culture set for adults consists of 20–30 ml of blood equally distributed between two or among three culture receptacles. Blood volume for pediatric patients is less and is dependent on patient weight. When 20 ml of blood is drawn, the standard practice is to culture 10 ml in an aerobic atmosphere and 10 ml in an anaerobic atmosphere. If 30 ml of blood is drawn, the same practice is followed except that the additional 10 ml is cultured under aerobic conditions. In patients with suspected endocarditis or endovascular infections, conditions in which bacteremia is continuous, two or three separate blood cultures collected at various intervals over a 24-h period are sufficient. For other types of bacteremia, 99% will be detected by three separate blood cultures collected at various intervals over a 24-h period. Most conventional broth-based blood culture systems also have the ability to detect candidemias. Specialized broth or procedures (see lysis centrifugation below) may be required to detect fastidious bacteria and dimorphic fungi, especially Histoplasma capsulatum.
Recently some authorities have questioned whether anaerobic blood cultures should be performed routinely in all patients, especially those in whom anaerobic bacteremia is unlikely. However, it has recently been demonstrated that anaerobic blood cultures may recover some facultatively anaerobic bacteria (eg, Enterococcus spp. and viridans streptococci) more efficiently than aerobic blood cultures. The lysis centrifugation method is a specialized blood culture method which is useful for recovering common as well as many unusual bacterial pathogens and fungi. In this system, blood is inoculated into a test tube containing saponin, a chemical that lyses erythrocytes and leukocytes, thereby releasing intracellular bacteria. The tube is centrifuged, and the sediment is inoculated onto solid agar plates or into broth. Mycobacterium spp., especially Mycobacterium avium-intracellulare, occasionally produce bacteremias in severely immunocompromised patients. The inoculation of the sediment from a lysis centrifugation blood culture tube onto specialized mycobacteria agar or into BACTEC 13A broth bottles is useful for recovering mycobacteria. Controversy exists as to whether blood cultures can be obtained from intravascular lines. Several investigators have shown that such a practice results in the isolation of more contaminating microorganisms than if blood is obtained from peripheral veins. Whenever possible, blood for culturing should be obtained from peripheral veins; however, this recommendation must be considered in the context of the clinical situation of the patients. For example, phlebotomy via peripheral veins may be risky in severely thrombocytopenic bone marrow transplant patients.
Parasitemias caused by malarial or filarial organisms are diagnosed with thick and thin blood smears. Thick smears permit screening for malaria parasites, and speciation is possible by careful evaluation of thin smears. When parasitemias are suspected, the laboratory director should be notified so that proper processing and careful evaluation of blood specimens are undertaken. In some cases, referral of specimens to reference laboratories may be beneficial.
Other Sterile Body Fluids or Tissues
Direct antigen testing for H influenzae type b, N meningitidis, S pneumoniae, group B streptococci, E coli, and C neoformans can be performed on cerebrospinal fluid (CSF) or urine. Direct antigen testing for Legionella pneumophila and H capsulatum can be performed on urine. For staining and culturing methods, CSF or joint, peritoneal, or pleural fluids should be concentrated by filtering or centrifugation. In contrast, Gram stains of urine should be performed on unconcentrated specimens. The presence of ≥2 bacteria per oil immersion field (×1000) in a Gram-stained smear of a drop of unconcentrated urine should represent ~ 105 colony forming units of bacteria/ml of urine. Other rapid screening tests for bacteruria are commercially available. These tests, which detect either bacteria or leukocytes by direct or indirect methods, are generally no more accurate than the Gram stain method and may be more costly.
Cultures of bone marrow specimens may be particularly valuable for diagnosing Salmonella typhi (the agent of typhoid fever), Brucella spp., disseminated mycobacteria infections (M avium-intracellulare) or H capsulatum and can be processed by using the lysis centrifugation method. Granulomas surrounding small vessels in a bone marrow biopsy (ring granulomas) are associated with Coxiella burnetti (agent of Q fever) infection. Bone marrow specimens may be positive by staining methods in patients with disseminated infection caused by a variety of organisms, including H capsulatum, M avium-intracellulare, Trypanosoma spp. or Leishmania spp.
Respiratory Tract Specimens
All spontaneously produced sputa that are submitted for general bacteria culture should be screened for the presence of squamous epithelial cells. It has been demonstrated that expectorated sputum samples having > 25 squamous epithelial cells per low-power microscopic field are unacceptable for bacterial culture, because these samples likely are contaminated by oropharyngeal secretions. Common community-acquired respiratory bacterial pathogens include S pneumoniae, Streptococcus pyogenes (Lancefield group A β-hemolytic streptococci), Klebsiella pneumoniae, S aureus, Legionella spp., Chlamydia pneumoniae, and Mycoplasma pneumoniae. Gram-negative bacilli, especially Enterobacteriaceae and Pseudomonas spp., and VRE and staphylococci may cause nosocomial respiratory infection. Pseudomonas aeruginosa and B cepacia are frequently associated with pulmonary infection in patients with cystic fibrosis. All of the above bacteria, with the exception of C pneumoniae and M pneumoniae, are easily diagnosed by culture-based methods. Indirect serologic methods are the best methods for diagnosing the latter two pathogens.
Although Legionella spp., Mycobacteria spp., and Nocardia spp. can cause pulmonary disease in normal hosts, they may be a more frequent cause of pulmonary disease in immunocompromised hosts. Legionella spp. can be diagnosed by direct examination of pulmonary secretions or alveolar tissue, with a fluorescent antibody technique. Alternatively, acute infection with the most frequently encountered Legionella spp., L pneumophila, can be diagnosed by screening for antigen in the urine. Legionella antigenuria can persist for months after acute infection, a factor that may limit the usefulness of this direct test for diagnosing subsequent L pneumophila infections. Microorganisms that stain poorly by the Gram stain method and that appear to branch or are beaded in appearance should be suspect for Nocardia spp. These organisms frequently stain acid-fast by a modified acid-fast staining method. This method uses less intense decolorizing agents than those used for conventional acid-fast staining of mycobacteria. Nocardia spp. grow more slowly than other bacteria, but can be recovered on standard bacteriologic media. Nocardia spp. also grow well on media used for isolating fungi and mycobacteria and on media used to isolate Legionella spp. (buffered charcoal yeast extract). A recently described opportunist gram-positive bacillus, Rhodococcus equi, also may cause pulmonary infection in immunocompromised patients and, like Nocardia spp., may branch and stain acid-fast by a modified acid-fast staining method.
The dimorphic fungal pathogens (yeast and hyphal forms) H capsulatum, Blastomyces dermatitidis, Coccidioides immitis and Paracoccidioides brasiliensis and the monomorphic fungus (yeast form only) C neoformans may cause pulmonary or disseminated disease in both normal and immunocompromised hosts. The dimorphic fungal pathogen Sporothrix schenckii rarely produces respiratory disease and more frequently presents as cutaneous disease in normal hosts. Monomorphic fungi (hyphal form only) such as Aspergillus spp. generally cause respiratory or disseminated infection in immunocompromised hosts. Pneumocystis carinii has recently been reclassified as a fungus. It is an opportunistic fungal pathogen that causes disease only in immunocompromised hosts, especially patients coinfected with human immunodeficiency virus (HIV). All fungal pathogens, with the exception of P carinii, can be cultured from pulmomary secretions. Diagnosis of P carinii requires direct examination of pulmonary secretions or tissue. Some fungal infections can also be diagnosed by indirect serologic methods; additionally, H capsulatum antigen and C neoformans antigen can be detected in urine and in the serum in disseminated disease.
In certain areas of the world where helminths (Ascaris lumbricoides, Strongyloides stercoralis, Paragonimus westermani, and Echinococcus granulosa) are endemic, pulmonary infection with these parasites may occur. Disseminated infection with S stercoralis, including respiratory infection, may occur in immunocompromised hosts. Toxoplasma gondii may cause pulmonary disease in immunocompromised hosts. Direct examination of wet preps of pulmonary secretions may be useful to identify helminths; direct examination of tissue, culture, and indirect serologic methods are useful for diagnosing T gondii.
Feces
Enteric infections caused by Salmonella spp., Shigella spp., Yersinia enterocolitica, pathogenic E coli, and Campylobacter spp. are increasing in frequency in the United States. These infections are diagnosed by culture of feces, although direct detection of antigens by enzyme-linked immunoassays are sometimes useful for Salmonella spp. and E coli O157:H7. At present, Whipple's disease caused by the bacterium Tropheryma whippelii can be definitively diagnosed by evaluating small-bowel tissue for the presence of nucleic acid that is unique to this organism. Traditionally, periodic acid Schiff staining has been used to demonstrate the organism in tissue (see subsequent discussion).
Occasionally, in severely immunocompromised patients, acid-fast staining and culture may be useful for diagnosing enteric infection caused by Mycobacterium tuberculosis or M avium-intracellulare. However, cultures for mycobacteria should be performed only on stools with positive acid-fast stains. If mycobacteria are recovered from feces, disseminated disease is frequently present.
Enteric parasitic infections are diagnosed by direct examination of fresh or preserved stools; however, direct immunoassays may be useful for some parasitic organisms (Giardia lamblia and Cryptosporidium parvum). Extraintestinal parasitic infections may require other methods, including culture for diagnosis. Not shown in Table 6-2 are culture methods for Trypanosoma spp., Leishmania spp., T gondii, and Entamoeba histolytica. These specialized tests are available only at a few reference laboratories in North America. Culture methods for Acanthamoeba spp. and Naegleria spp. are included in Table 6-1. These methods are easy to perform and especially useful for diagnosing these parasitic infections in patients with corneal infections, which can occur with contact lens use.
Considerable attention has focused recently on newly discovered parasites that are opportunists in immunocompromised patients. These include species of four genera of coccidia (Isospora, Sarcocystis, Cryptosporidium, and Cyclospora) and species of five genera of microsporidia (Enterocytozoon, Septata, Nosema, Encephalitozoon, and Pleistophora). Infections caused by the coccidia or microsporidia have been reported in immunosuppressed patients, notably those patients coinfected with human immunodeficiency virus. Intestinal disease has been demonstrated to occur with all coccidia genera, Enterocytozoon spp., and Septata spp. Extraintestinal disease has been reported with Sarcocystis spp., Cryptosporidium spp., Septata spp., Encephalitozoon spp., Nosema spp., and Pleistophora spp. As previously mentioned, T gondii and the helminth S stercoralis can cause severe disseminated disease in immunocompromised hosts.
Intravascular Catheters
Patients who have indwelling central intravascular catheters for prolonged periods are susceptible to infection. If other sources for infection are ruled out, then infection related to the intravascular catheter must be considered. Diagnosing intravascular-catheter–associated infection can be challenging for the clinician, considering that a definitive diagnosis cannot be achieved unless the catheter is removed and a culture of the tip yields potentially pathogenic bacterium in sufficient quantity (ie, >15 colony forming units of bacteria). If there is evidence for a catheter tunnel infection (subcutaneous infection around the catheter), swabs of the affected area or pus if present should be stained and cultured for bacteria, mycobacteria, and fungi. Occasionally, the fungus Malassezia furfur can infect intravascular catheters and the blood in patients receiving intralipid infusions. Special culture techniques are required to isolate this organism (see Table 6-2).
Surveillance Cultures
In certain instances, surveillance cultures for bacteria and fungi may be useful. Patients who are immunocompromised, receiving broad-spectrum antimicrobial agents, or both may be surveyed for drug resistant bacteria or fungi. As part of infection control programs, institutionalized patients may be surveyed for VRE or methicillin-resistant S aureus carriage. In both of these examples, cultures of the upper airway, feces, or both may be useful to screen for carriage of these potential pathogens.
SEROLOGIC TEST METHODS
Indirect serologic methods may be useful for diagnosing infections caused by certain bacteria, fungi, and parasites and in some cases may be the only means by which a diagnosis is achieved. To diagnose infections by Rickettsia spp., for which alternative diagnostic test methods are limited (attempts at culturing these organisms should be avoided owing to their high infectivity), these tests may be the only means by which a diagnosis is established. Table 6-3 shows the serologic test methods currently available at most reference laboratories. Of note, these methods detect immunoglobulin G (IgG) antibody, IgM antibody, or both to specific bacteria pathogens and for the most part require that the infection has existed in a patient for a finite period so that detectable levels of antibody exist. For IgG analyses, a fourfold increase between baseline and convalescent antibody titers may be required to confirm infection. Because the demonstration of a fourfold rise in antibodies may require > 4 weeks, the diagnostic utility of IgG analyses may be limited, especially in the acute disease phase. However, in some situations, baseline IgG antibody levels may exceed a critical threshold, which is considered diagnostic for infection. Except for the last example, indirect serological methods cannot be considered as rapid diagnostic tests and therefore may be of limited utility. These tests may also be of limited value in patients who lack a humoral response, especially bone marrow transplant recipients.
MOLECULAR TEST METHODS
Molecular test methods, including nucleic acid-probing and -sequencing techniques, allow for the detection of pathogens directly from human specimens. Molecular test methods and nucleic acid amplification techniques, which are frequently available at most reference laboratories are shown in Table 6-5. These methods are potentially useful for fastidious or slowly growing organisms like Legionella spp., Bartonella spp., Mycobacterium spp., Borrelia burgdorferi, and dimorphic fungi. The bacterial agent of Whipple's disease, T whippelii, has never been recovered on culture and presently can be diagnosed only by nucleic acid testing methods. Quantities of parasites in blood or tissue might be sufficiently low (eg, T gondii or Babesia microtii) that direct examination does not provide a diagnosis. In these cases, molecular diagnostic tests may also be useful. It must be emphasized that some studies, particularly those that have evaluated molecular identification methods for group A streptococci and mycobacteria, have demonstrated that these methods, including those that use nucleic acid amplification techniques, are less sensitive than culture. Therefore, if the results for molecular test methods such as these are negative, other diagnostic test methods including culture techniques should be considered.
SUSCEPTIBILITY TESTING OF BACTERIAL & FUNGAL ISOLATES TO ANTIMICROBIAL AGENTS
The National Committee for Clinical Laboratory Standards (NCCLS) provides published guidelines for conventional susceptibility test methods for commonly encountered bacterial organisms that grow aerobically or anaerobically and for yeasts. For bacteria that are not easily cultured (eg, Bartonella spp., Ehrlichia spp., or T whippelii), antimicrobial susceptibility testing is currently not possible. Tentative NCCLS guidelines exist for mycobacteria. No standards exist for parasites. Antimicrobial susceptibility methods include disk diffusion, broth dilution, and agar dilution. For disk diffusion, the inhibition of growth of an organism on solid media is assessed around a paper disk from which an antimicrobial agent diffuses. The greater the zone of inhibition of bacterial growth, the more effective the antimicrobial agent. In the broth dilution procedure, the effect of a known concentration of antimicrobial agent dispersed along with the organism in liquid media is assessed. No growth of the organism (the broth remains clear) indicates that the organism is inhibited at the concentration of antimicrobial agent tested. For agar dilution, a known amount of antimicrobial agent is dispersed in solid medium, and its effect on growth of organisms that are spot inoculated onto the surface of the medium is assessed. No visible growth of the organism means that it is inhibited by the specific concentration of antimicrobial agent that is present in the solid medium.
|
Table 6-5. Molecular test methods currently available for direct detection of bacteriain human specimens. |
|||||||||||||||||||||||||||||||
|
Interpretation of the results for each of these methods may differ with the antimicrobial agent and organism tested, and guidelines for such are provided by the NCCLS. These interpretations are provided as the following categories: resistant, susceptible, or intermediately susceptible. If an organism is identified as resistant to a particular antimicrobial agent, that agent should not be used in the clinical setting.
Diagnosis of Viral Infections
It is becoming increasingly important for the practitioner to perform viral diagnostic studies because prognostic, epidemiologic, and therapeutic considerations may be greatly influenced by knowledge of the specific virus causing a given illness. Even if no therapy is available, the establishment of a definite diagnosis of viral infection is often beneficial in (1) epidemiologic monitoring, (2) educating physicians and patients, (3) defining the disease process, and (4) evaluating therapeutic implications, both positive and negative. Moreover, identification of a virus as the cause of a patient's illness may be cost effective, because expensive diagnostic procedures and antibiotic therapy may be avoided or discontinued.
The virology laboratory can confirm the suspected diagnosis by cytologic examination of clinical specimens; attempting to isolate the virus; detecting the presence of viral antigens, or nucleic acids or evaluating the patient's immune response to the virus (serology).
CYTOLOGY
The simplest technique for viral diagnosis is cytologic examination of specimens for the presence of characteristic viral inclusions, but this approach is insensitive and applicable to only a few viruses, especially herpes viruses. These intracellular structures may represent aggregates of virus within an infected cell or may be abnormal accumulations of cellular material resulting from the virus-induced metabolic disruption. Papanicolaou (Pap) smears may show these inclusions in single cells or in large syncytia (aggregates of cells containing more than one nucleus), as in a patient with herpes simplex infection of the cervix. Cytology can be used to detect infections with herpes simplex virus, varicella-zoster virus, cytomegalovirus, human papillomavirus, and adenoviruses. Rabies infection may also be detected by finding Negri bodies (rabies virus inclusions) in brain tissue.
CULTURE-BASED METHODS
The historical “gold standard” of viral diagnosis is recovery of the agent in tissue culture, embryonated eggs, or experimental animals. Embryonated eggs are still used for the growth of virus for some vaccines but have been replaced by cell cultures for routine virus isolation in clinical laboratories. Likewise, the use of experimental animals rarely occurs in most clinical laboratories. Just as multiple media are used in bacteriology, several different types of tissue culture cells (eg, monkey kidney, human fetal lung, human amnion, or human cancer cells) are inoculated with each viral specimen.
Most clinically significant viruses can be recovered in at least one of these cell cultures, but several clinically important viruses are not isolated in these cells. For example, specimens submitted for the identification of viruses such as human immunodeficiency virus, coxsackie A virus, and rubella virus require, respectively, cocultivation with normal human peripheral blood mononuclear cells, inoculation of suckling mice, and the use of specialized cell cultures, which are not generally available. Therefore infections caused by these and several other viruses are most frequently diagnosed serologically or by detection of virus-specific antigens or nucleic acids.
Detection of the growth of a virus is by observation of changes in the cell culture monolayer [cytopathic effect (CPE)]. Characteristic CPEs include changes in cell morphology, cell lysis, vacuolation, syncytia formation, and presence of inclusion bodies. Inclusion bodies are histologic changes in cells caused by the presence of viral components or changes in cell structures. With experience a technologist can distinguish CPE characteristics of the major virus groups. The observation of which cell culture exhibits CPEs and the rapidity of viral growth can be used for the presumptive identification of many clinically important viruses. This approach for identifying viruses is similar to bacterial identification based on growth and morphology of colonies on selective, differential media. Some viruses do not readily cause CPEs in cell lines typically used in clinical virology laboratories. However, some of these can be detected by other techniques, such as (1) erythrocyte hemadsorption onto cells infected with paramyxoviruses or mumps virus or (2) interference with the replication of other viruses (eg, picornaviruses cannot replicate in cells previously infected with rubella virus; this is known as heterologous interference).
In contrast with the intrinsic delay of antibody studies, the results of viral culture can be surprisingly rapid. Almost 50% of all viral isolates can be reported within 3 to 4 days of culture, with herpes simplex virus and influenza A virus usually detected within 1 to 3 days (Table 6-6).
The selection of the appropriate specimen for viral culture is complicated because several different viruses may cause the same clinical disease (Table 6-7). For example, several types of specimens should be submitted from patients with viral meningitis to enhance the recovery of the possible etiologic agents: CSF (enteroviruses, mumps virus, and herpes simplex virus), throat swabs and washings (enteroviruses), and stool or rectal swabs (enteroviruses). Also, serum should be collected as an acute-phase specimen in case subsequent serologic tests are indicated (eg, acute and convalescent sera for mumps virus or arbovirus infections). Many considerations, however, allow the physician to select the most appropriate specimens (Table 6-8). For example, during the summer, when enteroviral meningitis is prevalent, CSF, throat, and stool specimens should be submitted. On the other hand, the development of encephalitis in children after being bitten by mosquitos in wooded areas endemic for California encephalitis virus suggests that a serum specimen for antibody testing would be preferred. Central nervous system disease after parotitis would suggest collection of CSF and urine for the isolation of mumps virus. The specimens that should be collected for other viruses are summarized in Table 6-7.
|
Table 6-6. Average detection time for commonly isolated viruses. |
||||||||||||||||||||
|
Timing of Specimen Collection
Proper timing of specimen collection is essential for adequate recovery of viruses. Specimens should be collected early in the acute phase of infection. Studies with respiratory viruses indicate that the mean duration of viral shedding may be only 3–7 days. Also, herpes simplex virus and varicella-zoster virus may not be recovered from lesions beyond 5 days after onset. Isolation of an enterovirus from the CSF may be possible for only 2–3 days after onset of the central nervous system manifestations.
Transport to Laboratory
The shorter the interval between collection of a specimen and its delivery to the laboratory, the greater is the potential for isolating an agent. When feasible, all specimens other than blood, feces, urine, and tissue, which need special processing, should be inoculated directly onto cell cultures at the patient's bedside (Table 6-9). These should then be transported to the laboratory promptly.
For specimens that cannot be inoculated onto cell cultures immediately, several types of transport media have been used. It is generally believed that protein (serum, albumin, or gelatin) incorporated into a transport medium enhances survival of viruses.
Improper storage of specimens before processing can also adversely affect viral recovery. Significant losses in infective titer occur with enveloped viruses (eg, herpes simplex virus, varicella-zoster virus, or influenza virus) after specimens have been frozen and then thawed. This is not observed with nonenveloped viruses (eg, adenoviruses or enteroviruses). Therefore, when it is impossible to process a specimen immediately, it should be refrigerated but not frozen and packed in shaved ice for delivery to the laboratory if delays in transit are anticipated. Storage of specimens for the recovery of viruses at 4°C is far superior to storage at ambient temperature.
Interpretation of Culture Results
In general the detection of any virus in host tissues, CSF, blood, or vesicular fluid can be considered highly significant. Recovery of viruses other than cytomegalovirus in urine may be diagnostic of significant infection. For example, both mumps virus and adenovirus type 11 (associated with acute hemorrhagic cystitis) may be recovered in urine and indicate acute infection. However, the presence of cytomegalovirus in urine is difficult to interpret because this may reflect asymptomatic virus replication long after infection or indicate a significant active infection in the patient. In the newborn, viruria (isolation of virus in urine) in the first 3 weeks of life establishes a diagnosis of congenital cytomegalovirus infection, whereas the onset of viral excretion after 3–4 weeks of life reflects intrapartum or postpartum infection. Diagnosis of acquired cytomegalovirus in older patients usually requires a combination of findings, including positive cultures, illness compatible with cytomegalovirus disease, reasonable exclusion of other potential etiologic agents, and support by specific serologic or histologic data.
The significance of viruses isolated in upper respiratory tract, vaginal, or fecal specimens varies greatly. At one extreme, isolates such as measles, mumps, influenza, parainfluenza, and respiratory syncytial virus are significant because asymptomatic carriage and prolonged shedding of these viruses are unusual. Conversely, other viruses can be shed continually or intermittently without symptoms for periods ranging from several weeks (enteroviruses in feces) to many months or years (herpes simplex virus or cytomegalovirus in the oropharynx and genital tract; adenoviruses in the oropharynx and intestinal tract). Herpes simplex virus, cytomegalovirus, varicella zoster virus, and Epstein-Barr virus may remain latent for long periods and then become reactivated in response to a variety of stressful stimuli, including other infectious agents. In this setting their detection may not be significant, may merely represent a secondary problem complicating the primary infection (eg, herpes simplex virus “cold sores” in patients with bacterial sepsis), or may be associated with significant disease, especially in the immunocompromised patient.
|
Table 6-7. Appropriate specimens for viral isolation.1 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Based on the epidemiology of adenovirus infections and observed serologic responses, the simultaneous isolation of these viruses from throat and feces is significantly associated with febrile respiratory disease. Isolation of viruses from the throat alone is less frequently associated with disease, and isolates from feces alone are probably nondiagnostic in a patient with respiratory disease.
Enteroviruses are generally found in infants and children, particularly during the late summer and early autumn. A knowledge of the relative frequency of virus shedding among various age groups in a particular locale is extremely helpful in assessing the significance of results of throat or stool cultures. For example, the peak prevalence of enteroviruses in the stools of toddlers during the late summer may range from 5% in temperate zones to > 20% in subtropical climates. Even in temperate areas, rates may approach 30% in infants during periods of enterovirus activity. Shedding of enterovirus in the throat usually occurs for 1–2 weeks, whereas fecal shedding may last 4–16 weeks. Thus, in a clinically compatible illness, isolation of an enterovirus from the throat supports a stronger temporal relationship to the disease than does an isolate from only the feces.
|
Table 6-8. Specimen collection considerations for viral agents. |
||||||||||||||||||||
|
||||||||||||||||||||
Herpes simplex virus is unusual in a fecal culture. In such cases it usually represents either severe disseminated infection or local infection of the perianal areas. Detection of herpes simplex virus in the upper respiratory tract may mean nothing other than nonspecific reactivation of virus caused by fever unless typical vesicles or ulcers are also present. Because of the fever-related phenomenon, isolation of herpes simplex virus in the throat or mucocutaneous lesions of patients with encephalitis cannot be interpreted as causing the central nervous system disease. Currently the definitive way to establish a diagnosis of herpes simplex encephalitis is by direct demonstration of the virus in a brain biopsy or by polymerase chain reaction (PCR) on CSF. In neonates, however, isolation of the virus from any site should raise the possibility of severe infection.
|
Table 6-9. Summary of procedures for obtaining and transporting specimens for viral studies. |
||||||||||||||||||||||||
|
Isolation of adenoviruses, herpes simplex virus, varicella-zoster virus, and some enteroviruses from the cornea and conjunctiva in patients with inflammatory disease at these sites usually establishes the etiology of the infection.
DETECTION OF VIRAL ANTIGENS
Antibodies can be used as sensitive tools to detect, identify, and quantitate the presence of viral antigen in clinical specimens or cell culture. Monoclonal or polyclonal antibodies prepared in animals may be used. Viral antigens on the cell surface, within the cell, or released from infected cells can be detected by IF, EIA, RIA, and latex agglutination (LA). IF detects and locates cell-associated antigens, whereas RIA or different variations of enzyme-linked immunosorbent assay (ELISA) are used to detect and quantitate soluble antigens. LA is a rapid, easy assay for antigen; viruses or viral antigens in a sample cause the clumping of latex particles coated with specific antibody.
Virus-infected tissue or cell cultures can be detected by IF or EIA. By attaching a fluorescent signal to an antiviral antibody, and reacting it with the sample, viral antigen can be detected; this is called direct IF. A modification of this technique is the use of unlabeled antiviral antibodies and then a second antibody with a fluorescent label that will bind to IgM or IgG antibodies. EIA uses a second antibody conjugated to an enzyme, such as horseradish peroxidase or alkaline phosphatase, which releases a chromophore to mark the presence of antigen.
Direct IF assay is especially useful for (1) respiratory viruses (eg, respiratory syncytial virus or influenza A), (2) varicella-zoster virus and herpes simplex virus antigen in lung and visceral biopsies, and (3) cytomegalovirus in leukocytes from blood or CSF.
Soluble antigen can be quantitated by ELISA, RIA, and LA. The basis for these procedures is the separation and quantitation of antibody-bound and free antigen. Many of the ELISA and RIA techniques use an antibody immobilized to a solid support to capture soluble antigen and a labeled antibody to detect captured antigen.
Influenza, parainfluenza, and togaviruses produce a glycoprotein that binds erythrocytes. This property allows detection of free virus produced in cell culture by agglutination of erythrocytes, a process termed hemagglutination. The infected cells also adsorb erythrocytes to the surface by a process referred to as hemadsorption.
Detection and assay of characteristic enzymes can identify and quantitate specific viruses. For example, reverse transcriptase in cell culture is used as an indicator of infection by retroviruses.
SEROLOGIC TEST METHODS
Serology can be used to determine whether an infection is primary or a reinfection and if acute or chronic. The first antibodies to be produced by the immune system are directed against antigens on the virion or infected cell surfaces and are best detected by, eg, IF. Later in the infection, when cells have been lysed by the infecting virus or the cellular immune response, antibodies are directed against the cytoplasmic viral proteins and enzymes and can be detected by, eg, complement fixation. Seroconversion is characterized by a change from negative to positive antibody between serum taken during the acute phase of disease and that taken ≥ 2–3 weeks later; it is the best serologic marker of recent infection. A fourfold or greater rise in antibody titer in paired sera may also be significant but fluctuations in antibody titer do occur naturally. Detection of virus-specific IgM is theoretically associated with recent infection, but technical difficulty in measuring specific IgM responses may lead to false-positive and false-negative results.
For many viruses, culture or antigen detection is the best diagnostic test. However, certain viruses (eg, HIV, hepatitis A and B viruses, rubella virus, Epstein-Barr virus, measles virus, coronaviruses, and togaviruses) are difficult to isolate in cell culture, and infections are diagnosed most easily by serologic techniques. When a virologic workup is planned for a patient, it is generally useful to obtain ≥ 2–3 ml of serum during the acute phase of disease and store it at–20°C. This may become valuable, particularly if virus detection subsequently fails or if the significance of an isolate is uncertain. In these instances a convalescent-phase serum specimen may be requested 2–3 weeks later, and both the acute and convalescent sera may then be tested against appropriate viral antigens. In general, if a virus is isolated, the antibody titers need not be measured to confirm infection; for example, if a patient has aseptic meningitis and an enterovirus is recovered from the throat, that agent is probably responsible for the illness. Similarly, if influenza virus is recovered from the throat of a patient who has clinical influenza, no serologic confirmation of the etiology is necessary.
Complement fixation is a standard but technically difficult serologic test. The serum is first reacted with the suspected viral antigen and complement, and the residual complement is assayed by lysis of indicator antibody-coated erythrocytes. If the complement is used in the first reaction and is therefore not available for the second reaction, it indicates the presence of antibody to the suspected virus. Antibodies measured by this system generally develop slightly later in the course of an illness than those measured by other techniques. This delayed response is useful for documenting seroconversion when the initial serum specimen is collected late in the clinical course. Members of some virus groups (eg, enteroviruses) do not possess group-specific antigen and must be tested individually.
The neutralization test is essentially a protection test. When a virus is incubated with homologous type-specific antibody, the virus is rendered incapable of producing infection in an indicator cell culture system. A neutralization antibody response is virus type specific, with titers rising rapidly and persisting for long periods.
The hemagglutination inhibition test can be performed with a variety of viruses that can selectively agglutinate erythrocytes of various animal species (eg, chicken, guinea pig, or human). The hemagglutination capacity of a virus is inhibited by specific immune or convalescent sera. Hemagglutination-inhibiting antibody develops rapidly after the onset of symptoms, plateaus, declines slowly, and may last indefinitely at low levels. This test is useful for both the detection of acute viral infection and the determination of immunity.
For the indirect fluorescent antibody test, virus-infected cells are placed in prepared wells on microscope slides and then fixed in cold acetone and dried. Patient serum is applied and, after incubation, anti-human globulin conjugated with fluorescein is added. If fluorescence is observed, it indicates the presence of specific antiviral antibody.
Interpretation of Serologic Results
Virus-specific IgM antibody usually rises during the first 2–3 weeks of infection and persists for several weeks to months. Thus an elevated titer of specific IgM antibody suggests a recent primary infection, which may be further supported by demonstrating a fall in IgM antibody in subsequent sera. Detection of specific IgM has been used with success in the diagnosis of infections caused by cytomegalovirus and rubella virus and is currently the procedure of choice to establish a recent or acute infection from hepatitis A or B.
Several limitations of interpretation must be remembered. It is now recognized that IgM-specific antibody responses are not always restricted to primary infections. Reactivation or reinfection may result in IgM responses, particularly in herpes virus infection. In addition, patients may continue to produce IgM-specific antibody to rubella virus or cytomegalovirus for many months after a primary infection. Heterotypic IgM responses may also occur. For example, antibody responses to cytomegalovirus may develop in Epstein-Barr virus infections and vice versa. Other pitfalls include falsely low or negative IgM titers caused by competition from high-titer IgG antibody for antigen-binding sites and false-positive reactions resulting from rheumatoid factor. Both types of errors appear to occur most frequently in solid-phase assays with IF.
The serologic diagnosis of most viral infections is based on demonstration of a seroconversion or a rise (fourfold or greater) of IgG antibody. However, significant antibody titer rises may result from cross-reactions to related antigens; for example, an antibody rise to parainfluenza virus may actually result from infection with mumps virus. Furthermore, seroconversions may not be seen with some patient populations (eg, infants and immunocompromised patients) or when the initial serum is collected late in the course of disease.
A serum specimen can also be used for screening an infant's blood for certain antibodies of the IgG class. Antibodies to Toxoplasma spp., rubella virus, cytomegalovirus, and herpes simplex virus (known as the “TORCH” screen) may be measured to determine possible congenital infection with these agents. However, the value of these tests must be understood. They are useful in excluding a possible infection but not in proving an etiology. For example, if rubella antibody is absent, an infant almost certainly does not have congenital rubella infection. To diagnose active rubella infection in such a baby, viral cultures are required. Screening of blood supplies for cytomegalovirus antibody is used to eliminate transmission of antibody-positive blood to seronegative babies and other immunocompromised patients.
Serologic Panels
Selection of several antigens for testing with paired sera in cases in which a virus is suspected can usually be made based on clinical syndrome, the known local epidemiology of particular viruses, and the patient's age. This has led to the concept of serologic batteries of panels. Some examples of possible panels are included in Table 6-10.
Antigens from mumps, western equine encephalitis, eastern equine encephalitis, St. Louis encephalitis, and California encephalitis viruses—and perhaps lymphocytic choriomeningitis virus, Epstein-Barr virus, and HIV—may be included in a panel of tests for central nervous system diseases. Although herpes simplex antigen is sometimes included in such a panel, a rise in antibody titer is not sufficient to diagnose herpes encephalitis. Many viral central nervous system illnesses, especially aseptic meningitis, are caused by the enteroviruses; however, the many serotypes and the cumbersome serologic methods necessary for their diagnosis usually make it impractical to include them in a panel. When one or two enteroviruses have been shown to be epidemic in an area in one summer, one can pick up some additional cases by performing neutralization tests on paired sera by using only those specific enteroviruses that are endemic in the community.
|
Table 6-10. Serological panels. |
||||||||||||||
|
The viral antigen panel for testing respiratory syndromes might include influenza A and B; respiratory syncytial virus; parainfluenza types 1, 2, and 3; and adenoviruses.
To test for viral causes of exanthems, the panel would include measles and rubella. If the disease is vesicular, herpes simplex virus and varicella-zoster virus should be included, although the herpes viruses are best diagnosed by culture or antigen detection.
Antigens from group B coxsackie virus types 1–5 and perhaps influenza A and B viruses could make up the panel for myocarditis and pericarditis. Although numerous viruses have been implicated in inflammatory diseases of the heart and its covering membranes, the group B coxsackieviruses are believed to account for almost one half of the cases. Unfortunately, much of the clinical illness is expressed late in the infection, at the time when standard methods of virus detection are likely to fail and it is too late to demonstrate seroconversion or significant antibody titer rises.
MOLECULAR TEST METHODS
In recent years a variety of assays have been developed to detect viral nucleic acid—either DNA or RNA. The best known of these is PCR, an amplification technique that allows the detection and selective replication of a targeted portion of the genome. The technique uses special DNA polymerases that initiate replication in either the 3′ or 5′ direction. The specificity is provided by primers that recognize a pair of unique sites on the genome so that the DNA between them can be replicated by repetitive cycling of the test conditions. Because each newly synthesized fragment can serve as the template for its own replication, the amount of DNA doubles with each cycle. The amplification power of PCR offers a solution for the sensitivity problems inherent in the direct application of probes. Although the nucleic acid segment amplified by PCR can be seen directly on a gel, the greatest sensitivity and specificity are achieved when probe hybridization is carried out after PCR. A probe is a fragment of DNA that has been cloned or otherwise recovered from a genomic or plasmid source. In some cases the probe is synthesized as a single chain of nucleotides (oligonucleotide probe) from known sequence data. The probes are labeled with a radioisotope or other marker and used in hybridization reactions either to detect the homologous sequences in unknown specimens or in gel electrophoresis.
The diagnostic use of DNA probes is to detect or identify microorganisms by hybridization of the probe to homologous sequences in DNA extracted from the entire organism. A number of probes have been developed that will quickly and reliably identify organisms that have already been isolated in culture. The application of probes for detection of infectious agents directly in clinical specimens such as blood, urine and sputum is more difficult.
Recently, the branched DNA (bDNA) a rapid assay for direct quantification of viral nucleic acid has been developed for hepatitis B, hepatitis C, and HIV infections. Because the bDNA assay measures viral nucleic acids at physiological levels by boosting the reporter signal, rather than by amplifying target sequences, it is not subject to the errors inherent in the amplification steps of PCR-based methods. Inherently quantitative and amenable to routine use in a clinical setting, the bDNA assay may be useful in the management of patients with chronic viral diseases. Recent studies have illustrated the potential clinical utility of the bDNA assay in determining the prognosis and in therapeutic monitoring of infection. Additional nucleic acid tests include hybrid capture and nucleic acid sequence-based amplification.
REFERENCES
Diagnosis of Bacterial, Fungal, & Parasitic Infections
Bryant JK, Strand CL: Reliability of blood cultures collected from intravascular catheter versus venipuncture. Am J Clin Pathol 1987;88:113.
Cockerill FR III et al: Analysis of 281,797 consecutive blood cultures over an eight-year period: trends in microorganisms isolated and the value of anaerobic culture of blood. Clin Infect Dis 197;24:403.
Current WL, Garcia LS: Cryptosporidiosis. Clin Microbiol Rev 1991;4:325.
Dalovisio JR et al: Comparison of the amplified Mycobacterium tuberculosis (MTB) Direct Test, Amplicor MTB PCR, and IS6110-PCR for detection of MTB in respiratory specimens. Clin Infect Dis 1996;23:1099.
Douard MC et al: Quantitative blood cultures for diagnosis and management of catheter-related sepsis in pediatric hematology and oncology patients. Intensive Care Med 1991;17:30.
Entemadi HA et al: Isolation of Brucella spp. from clinical specimens. J Clin Microbiol 1984;20:586.
Felices FJ et al: Use of the central venous pressure catheter to obtain blood cultures. Crit Care Med 1979;7:78.
Havlik JA et al: A prospective evaluation of Mycobacterium avium-complex colonization of the respiratory and gastrointestinal tracts of persons with human immunodeficiency virus infection. J Infect Dis 1993; 168:1045.
Hay JE et al: Clinical comparison of Isolator, Septi-Chek, nonvented tryptic soy broth, and direct agar plating combined with thioglycollate broth for diagnosing spontaneous bacterial peritonitis. J Clin Microbiol 1996;34:34.
Kaditis A et al: Yield of positive blood cultures in pediatric oncology patients by a new method of blood culture collection. Pediatr Infect Dis J 1996;15:615.
Kiehn TE, Commarato R: Comparative recoveries of Mycobacterium avium–M. intracellulare from Isolator lysis-centrifugation and BACTEC 13A blood culture systems. J Clin Microbiol 1988;26:760.
Kohler RB, Winn WC Jr, Wheat LJ: Onset and duration of urinary antigen excretion in Legionnaires' disease. J Clin Microbiol 1984;20:605.
Maki DG, Weiss CE, Sarafin HW: A semiquantitative culture method for identifying intravenous catheter-related infection. N Engl J Med 1977;296:1305.
Miller JM: Specimen Management in Clinical Microbiology. American Society for Microbiology Press, 1999.
Mosca R et al: The benefits of Isolator cultures in the management of suspected catheter sepsis. Surgery 1987;718.
Murray PR, Washington JA: Microscopic and bacteriologic analysis of sputum. Mayo Clin Proc 1975;50: 339.
National Committee for Clinical Laboratory Standards: Performance Standards for Antimicrobial Susceptibility Testing—Ninth Informational Supplement; Approved Standard 100-59. National Committee for Clinical Laboratory Standards, 1999.
National Committee for Clinical Microbiology Standards: Antimycobacterial Susceptibility Testing: Publication No. NCCLS M24-P. National Committee for Clinical Laboratory Standards, 1990.
National Committee for Clinical Microbiology Standards: Reference Method for Broth Antifungal Susceptibility Testing of Yeasts: Approved Standard M27-A. Publication No. NCCLS M27-A. National Committee for Clinical Laboratory Standards, 1997.
Ortega Y et al: Cyclospora species: a new protozoan pathogen of humans. N Engl J Med 1993;328:1308.
Paya CV, Roberts GD, Cockerill FR III: Laboratory diagnosis of disseminated histoplasmosis—clinical importance of the lysis-centrifugation blood culture technique. Mayo Clinic Proc 1987;62:480.
Pokorski SJ et al: Comparison of Gen-Probe Group A Streptococcus Direct Test with culture for diagnosing streptococcal pharyngitis. J Clin Microbiol 1994;32: 1440.
Sharp SE: Routine anaerobic blood cultures: still appropriate today? Clin Microbiol Newslett 1991;13:179.
Tonnesen A, Peuler M, Lockwood WR: Cultures of blood drawn by catheters vs venipunctures. J Am Med Assoc 1976;235:1877.
Vallenas C et al: Efficacy of bone marrow, blood, stool, and duodenal content cultures for bacteriologic confirmation of typhoid fever in children. Pediatr Infect Dis 1985;4:496.
Voigtz J, Delsol G, Fabre J: Liver and bone marrow granulomas in Q fever. Gastroenterology 1983;84: 887.
Washington JA II: Blood cultures: principles and techniques. Mayo Clin Proc 1975;50:91.
Washington JA II: Collection, transport, and processing of blood cultures. Clinics Lab Med 1994;14:59.
Weber R et al: Human microsporidial infections. Clin Microbiol Rev 1994;7:426.
Diagnosis of Viral Infections
Tang Y-W, Procop GW, Persing DH: Molecular diagnostics of infectious diseases. Clin Chem 1997;43: 2021.