CLINICAL MICROBIOLOGY REVIEWS, July 2004, p. 571580 0893-8512/04/$08.000 DOI: 10.1128/CMR.17.3.571580.2004 Copyright 2004, American Society for Microbiology.

. All Rights Reserved.

Vol. 17, No. 3

Rapid Diagnosis of Pharyngitis Caused by Group A Streptococci

Michael A. Gerber1* and Stanford T. Shulman2
Division of Infectious Diseases, University of Cincinnati College of Medicine, Cincinnati Childrens Hospital Medical Center, Cincinnati, Ohio 45229,1 and Department of Pediatrics, The Feinberg School of Medicine, Northwestern University, and Division of Infectious Diseases, The Childrens Memorial Hospital, Chicago, Illinois 60614 2 INTRODUCTION .......................................................................................................................................................571 PERFORMANCE CHARACTERISTICS.................................................................................................................572 CRITERION STANDARD .........................................................................................................................................572 COST AND FEASIBILITY ........................................................................................................................................573 SENSITIVITY AND SPECIFICITY..........................................................................................................................573 REASONS FOR FALSE-POSITIVE AND APPARENT FALSE-POSITIVE AND FALSE-NEGATIVE RADT RESULTS.................................................................................................................................................573 RADT AND PHARYNGITIS DUE TO GROUP C AND GROUP G STREPTOCOCCI ..................................574 CLIA AND ITS IMPACT ON RADTS.....................................................................................................................574 COMPARATIVE DATA .............................................................................................................................................574 RECOMMENDATIONS OF ADVISORY COMMITTEES...................................................................................575 ACTUAL PRACTICES ...............................................................................................................................................575 MEDICAL DECISION MAKING.............................................................................................................................576 Cost-Effectiveness....................................................................................................................................................577 RADT AND SUSCEPTIBILITY TESTING .............................................................................................................578 GAPS IN KNOWLEDGE...........................................................................................................................................578 CONCLUSION............................................................................................................................................................579 REFERENCES ............................................................................................................................................................579 INTRODUCTION Because the signs and symptoms of pharyngitis due to group A beta-hemolytic streptococci (GABHS) pharyngitis can be nonspecic, establishing an accurate clinical diagnosis often is difcult even for experienced physicians. Therefore, it has been standard practice to seek laboratory conrmation of the diagnosis. Culturing a throat swab on a sheep blood agar plate (BAP), as rst described by Breese and Disney in 1954 (6), has been accepted as the standard for the diagnosis of GABHS pharyngitis for nearly ve decades (3). The major disadvantage of culturing throat swabs on BAPs is the delay (overnight or longer) in obtaining the results. In the early 1980s, commercial rapid antigen detection tests (RADTs) were developed for the rapid identication of GABHS directly from throat swabs. All such rapid tests involve an acid extraction step to solubilize GABHS cell wall carbohydrate and to identify its presence by an immunologic reaction. Although RADTs generally are more expensive than BAP cultures, the advantage they offer over the traditional procedure is the speed with which they can provide results. Rapid identication and consequent prompt treatment of patients with GABHS pharyngitis can reduce the risk of spread of GABHS, can allow the patient to return to school or work sooner, and may reduce the acute morbidity of this illness (23, 61). The rst RADTs used the latex agglutination technique. This is a relatively insensitive method with somewhat unclear end points. Newer tests based on enzyme immunoassay (EIA) techniques offered more sharply dened end points as well as increased sensitivity. More recently, a RADT using the optical immunoassay (OIA) technology has become available. Data suggest that this test may be more sensitive than other RADTs and perhaps may even be as sensitive as BAP cultures. However, there has been some reluctance to recommend the OIA for routine use without backing up negative test results with a conrmatory BAP culture. The latest commercial RADTs for the diagnosis of GABHS pharyngitis to be developed are two tests that employ molecular biology methods. The rst is a chemiluminescent singlestranded DNA probe that detects specic rRNA sequences unique to GABHS (Group A Streptococcus Direct Test; GenProbe, Inc., San Diego, Calif.) (10, 34, 35, 60, 73). Several investigations have demonstrated that this test has a sensitivity between 86 and 94.8% and a specicity between 95 and 100% compared to BAP cultures. This test takes approximately 2 h to complete; because of the multiple steps involved, the special equipment required, and the need to batch specimens, this is not a point-of-care test. The second test is a one-rapid-cycle, real-time PCR method (LightCycler Strep-A assay; Roche Applied Science, Indianapolis, Ind.) (79). In the only published report available, this test was demonstrated to have a sensitivity of 93% and a specicity of 98% compared to BAP culture. This test takes approximately 1.5 h to complete; because of the special equipment involved and the need to batch specimens, this also is not a point-of-care test. However, because only about 3 min of personnel time is required, this test would be
571

* Corresponding author. Mailing address: Division of Infectious Diseases, Cincinnati Childrens Hospital Medical Center, 3333 Burnet Ave. MC5019, Cincinnati, OH 45229-3039. Phone: (513) 636-5594. Fax: (513) 636-7039. E-mail: [email protected].

572

GERBER AND SHULMAN

CLIN. MICROBIOL. REV. TABLE 1. Sensitivity and specicity of RADTs compared with cultures
Culture medium RADT method Sample size (no. of patients) Sensitivity (%) Specicity (%) Reference

adaptable for use in smaller laboratories that are experienced in molecular biology-based testing. Currently a wide variety of RADTs are available for diagnosing GABHS pharyngitis. In this paper, we review the available data with respect to the accuracy of these tests and their utilization. We also identify areas where gaps in knowledge exist. PERFORMANCE CHARACTERISTICS The results obtained with RADTs for GABHS are inuenced by the skill, experience, and expertise of the individual who is actually obtaining the throat swab and performing the RADT (23, 43). In addition, with some RADTs that have a relatively subjective end point, strong observer bias can be apparent (36). When duplicate throat swabs obtained simultaneously from the same patient are studied, there is a relationship between the amount of growth on a BAP culture performed with one swab and the sensitivity of the RADT performed with the other (26, 46). That is, heavy growth is associated with very high RADT positivity rates while light growth is generally associated with lower RADT rates of positivity. Increasing the colony count and the quantity of antigen available by using two simultaneously collected throat swabs can increase the yield on the BAP culture signicantly and also increases the sensitivity of the RADT (45). Increasing the colony count and the quantity of antigen present by using a 2-h broth culture enhancement of the throat swab before performing the RADT also increases the sensitivity signicantly (5). In addition, it has been shown that the performance of RADTs increases as the pretest probability of GABHS being present (as predicted by the presence of specic clinical ndings) increases (17). There is no single xed value for the performance of a particular RADT for the diagnosis of GABHS pharyngitis; the performance is a function of the clinical characteristics of the illness of the patients selected for testing (17). Because of this spectrum bias, RADT performance is not an absolute feature of a given test. CRITERION STANDARD When one attempts to evaluate the sensitivity and specicity of a given RADT for GABHS by comparing the results obtained by the RADT to a criterion standard, the choice of standard is critical. If a relatively insensitive criterion standard is chosen, the RADT may be evaluated erroneously because the apparent sensitivity is elevated and the apparent specicity is lowered. Conversely, if the criterion standard is too sensitive (for example, one that is capable of detecting a single GABHS organism), the RADT may be evaluated erroneously because the apparent sensitivity is lowered and the apparent specicity is elevated. In most clinical studies, RADTs for conrming the diagnosis of GABHS pharyngitis have been compared with BAP cultures as the criterion standard (see Table 1). However, comparisons of a RADT with a BAP culture are complicated by the fact that there is no universally accepted procedure for performing a BAP culture (23). The yield of GABHS from a BAP culture may vary with the type of medium, the atmosphere of incubation, the duration of incubation, the number of swabs obtained,

BAP BAP BAP BAP BAP BAP BAP BAP THB BAP BAP BAP BAP THB THB THB THB BAP BAP BAP BAP THB BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP BAP
a

THB THB

THB THB THB THB THB THB THB THB

RADT

MMa MM MM MM MM MM OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA OIA EIA EIA EIA EIA EIA EIA EIA EIA EIA

767 277 1,103 318 520 384 200 301 301 505 505 262 690 690 475 800 424 233 86 363 263 413 801 500 77 302 2,113 520 302 182 777 1,449 258 258 322 351 454

88.6 86 92.4 91.4 94.8 93 80 91.4 90.4 93.2 94.8 77 97 95 97.4 98.9 84.2 79.5 85 89 77 89 91.5 83 78 75.5 84 86.1 92.6 95.5 89 52.6 95 87 80 96.9 89.9

97.8 95 99.6 97 100 98 89 95.6 94.1 95.0 98.8 97 94 97 95.6 98.4 95.7 96.9 95 96.5 62 93 94.8 89 90 97.1 93 97.1 92.8 96.3 96 98.2 100 100 97.4 95.8

60 73 35 34 10 79 42 62 62 22 22 68 16 16 32 32 15 59 71 45 33 75 35 63 1 30 24 10 30 41 20 86 70 70 64 18 47

MM, molecular biology-based method.

the specic grouping procedure used, and whether the culture is performed in a physicians ofce or in a diagnostic laboratory (72). Fundamental questions about whether a RADT performed in a practice setting should be compared with a BAP culture performed in a similar setting or with a BAP culture performed in a diagnostic microbiology laboratory still remain unanswered (23). It has also been suggested in several studies that culturing the pledget (the plug separating the transport medium and swab in a transport tube) in Todd-Hewitt broth (THB) signicantly increases the yield of GABHS compared with that obtained in standard BAP culture using the same throat swab (15, 16, 22, 32). However, in other studies, the yields of GABHS from BAP and THB cultures were comparable (62, 68). Because of the reported 5 to 10% discordance between duplicate throat cultures (23), the question also arises whether single or duplicate throat swab cultures should serve as the criterion standard against which a RADT is measured (23). Some have suggested that RADTs should be compared to a combination of an aerobically incubated BAP and an anaerobically incubated selective streptococcal agar plate (84, 85). Others, however, found that the use of an anaerobically incubated selective streptococcal agar plate added little to the yield of GABHS from a BAP culture incubated aerobically (18, 62). Although the sensitivity of a BAP culture performed in a

VOL. 17, 2004

RAPID DIAGNOSIS OF STREPTOCOCCAL PHARYNGITIS

573

clinical practice setting may be relatively low compared with a highly sensitive criterion standard (e.g., the BAP or THB culture performed in a diagnostic microbiology laboratory), it is important to consider how sensitive the BAP culture should be ideally to provide clinically valuable information. As an extreme example, it was proposed that a PCR procedure rather than a culture procedure be used as the criterion standard for RADT evaluations (42). However, it is not clear that identication of subjects with negative BAP cultures but positive RADT or PCR results would result in better clinical outcomes, since the clinical signicance of very small numbers of GABHS organisms is doubtful. BAP cultures are used frequently to conrm the clinical diagnosis of GABHS pharyngitis in U.S. ofce practice settings (37, 69). If a substantial number of children with GABHS pharyngitis were not being diagnosed and treated properly because of insufcient sensitivity of a single BAP culture, one would anticipate many cases of suppurative complications (e.g., peritonsillar abscess and retropharyngeal abscess) and many cases of one of the nonsuppurative complications of GABHS pharyngitis (e.g., acute rheumatic fever). Suppurative complications are rarely observed, and although there were several focal outbreaks of acute rheumatic fever in the mid to late 1980s in this country, there is little if any evidence to suggest that these outbreaks resulted from deciencies in BAP culture procedures. There is also no evidence that the recent, well-publicized cases of invasive GABHS disease and GABHS toxic shock syndrome, which are usually unrelated to pharyngeal infections (74), resulted from deciencies in BAP cultures. Because the sensitivity of an ofce BAP culture is generally considered sufcient for use alone in the diagnosis of GABHS pharyngitis (even with its variable methodology) (3, 12, 14), it follows that an alternate test with sensitivity approximating that of the ofce BAP and with adequate specicity should be equally acceptable when used alone for the diagnosis of GABHS pharyngitis. COST AND FEASIBILITY Charges for various diagnostic tests for GABHS pharyngitis vary widely. In some settings, BAP cultures may cost as little as $5 or as much as $50 while RADTs may cost as little as $10 or as much as $60. Therefore, physicians should be aware of how much patients are charged for BAP cultures, RADTs, and the combination. Medicare and Medicaid reimbursement for RADTs for GABHS pharyngitis (CPT code 87430) is $16.01, reimbursement for a GABHS direct probe (code 87650) is $26.90, and reimbursement for a GABHS amplied probe such as PCR (code 87651) is $47.00. Because OIA-based RADTs are more expensive (with a physicians cost of reagents from $4 to $6.95 per test), time-consuming (approximately 12 min per test), and labor-intensive than many EIA procedures, it may not be acceptable for all practice settings. The chemiluminescent DNA probe test has multiple steps, requires special equipment, and takes at least 2 h to perform. Therefore, it is unacceptable for an ofce laboratory or an emergency room/clinic laboratory, but it could be practical for a large hospital or commercial laboratory that can batch specimens (35, 60). The simplicity of some RADTs is exemplied by a study in which sixth- and seventh-grade students achieved a high level

of performance with a RADT that is a polyclonal antibody membrane lter separation EIA, despite having no formal laboratory background, no previous method-specic training, and limited self-training (19). The performance of RADTs by nurses and physicians in the ofce setting can decrease delays in specimen transport and shorten turnaround time. On the other hand, having experienced technologists performing RADTs may improve the accuracy of the results. SENSITIVITY AND SPECIFICITY The great majority of the RADTs that are currently available have a high specicity (i.e., 95% or greater) compared with BAP cultures (Table 1). This means that false-positive test results are unusual and therefore that therapeutic decisions can be made on the basis of a positive test with a great degree of condence and without need for BAP culture conrmation. The sensitivity of most of the currently available RADTs is between 70 and 90% compared with BAP culture. Because of concerns about the adequacy of the sensitivity of many RADTs, it has been recommended that a negative RADT be conrmed with a BAP culture (3, 12, 14). A representative sample of published reports that address the sensitivity and specicity of RADTs is presented in Table 1. Many patients with negative RADT results and a simultaneously obtained positive BAP culture have only small numbers of GABHS colonies on their BAP culture. Therefore, it has been suggested that most false-negative RADT results occur in patients who are merely chronic streptococcal carriers and not truly infected with GABHS. Because only patients with bona de acute streptococcal infections (positive culture for GABHS and evidence of a serologic response to GABHS antigens) are at risk of developing acute rheumatic fever, this suggestion, if correct, would have important implications for the role of RADTs in the primary prevention of rheumatic fever. However, studies have demonstrated that a large proportion of patients with false-negative RADT results are actually infected with GABHS and are not merely streptococcal carriers (25). Therefore, it is often recommended that a negative RADT result be conrmed by a conventional BAP culture (3, 12, 14). REASONS FOR FALSE-POSITIVE AND APPARENT FALSE-POSITIVE AND FALSE-NEGATIVE RADT RESULTS False-positive RADT results are unusual. They may be produced by the presence in the pharynx of Streptococcus milleri group (SMG) strains that express the group A carbohydrate antigen. SMG strains are commensals that are frequently alpha-hemolytic or nonhemolytic and therefore are difcult to identify on a BAP culture (40, 65). It is not clear how often this phenomenon occurs because there are no published data on the prevalence of SMG strains that express the group A carbohydrate antigen. A recent report suggested that the falsepositivity rate with RADTs may be as high as 15% compared with BAP cultures; however, the presence of an SMG strain expressing the group A carbohydrate or another explanation for the false-positive RADT results was identied in only a

574

GERBER AND SHULMAN

CLIN. MICROBIOL. REV.

very small proportion of these cases (40). Other reports have found much lower false-positivity rates (Table 1). Apparent (but not true) false-positive RADT results are produced on rare occasions by either nutritional variants of group A streptococci or nonhemolytic group A streptococci (40, 66). Nutritional variants of group A streptococci and nonhemolytic group A streptococci possess the group A carbohydrate and therefore can produce a positive RADT result. However, because of their unusual nutritional requirements or absence of beta-hemolysis, such strains are not apparent on a BAP culture. Therefore, the BAP culture will be read mistakenly as negative for group A streptococci. Apparent (but not true) false-negative RADT results may be produced on a BAP culture by the misidentication of betahemolytic organisms that lack group A carbohydrate antigen as group A streptococci. Large colonies of SMG that are betahemolytic, as well as large-colony, beta-hemolytic strains of group C or group G streptococci; may be bacitracin sensitive and thus may be misidentied as group A streptococci on a BAP culture if bacitracin sensitivity is utilized as the basis for categorizing organisms as members of group A. However, because of the absence of the group A carbohydrate antigen, these isolates produce a negative RADT result (31). RADT AND PHARYNGITIS DUE TO GROUP C AND GROUP G STREPTOCOCCI A potential concern about using RADT alone without BAP culture conrmation of negative results is that patients with acute pharyngitis due to group C or group G streptococci will not be identied. The clinical features of pharyngitis due to both group C and group G streptococci are similar to those of GABHS pharyngitis with fever, mild to moderate sore throat, pharyngeal exudate, and cervical adenitis. Usually group C pharyngitis affects an older population, particularly teenagers and young adults. Several studies have demonstrated that group C streptococci are a relatively common cause of acute pharyngitis among college students and among adults seeking care in an emergency room (54, 78). In addition to endemic pharyngitis, group C streptococci can cause epidemic foodborne pharyngitis after ingestion of contaminated products such as unpasteurized cows milk. Family and school outbreaks of pharyngitis due to group C streptococci also have been described (44). While there have been several well-documented food-borne outbreaks of pharyngitis due to group G streptococci and a single community-wide, respiratory outbreak in a pediatric population, the etiologic role of group G streptococci in acute, endemic pharyngitis remains unclear (29). Acute rheumatic fever has not been described as a complication of pharyngitis due to either group C or group G streptococci. Acute glomerulonephritis has been reported to be an extremely unusual complication of pharyngitis due to group C streptococci but a causal relationship between pharyngitis due to group G streptococci and acute glomerulonephritis has not been clearly established. Therefore, the primary reason for identication of either group C or group G streptococci as the etiologic agent of acute pharyngitis is to initiate antimicrobial therapy that may reduce the clinical impact of the illness. However, there is no convincing evidence from controlled studies of a clinical response to antimicrobial ther-

apy in patients with acute pharyngitis and either group C or group G streptococci isolated from their pharynx. Therefore, the clinical signicance of missing a case of pharyngitis due to group C or group G streptococci because of use of RADT appears to be limited. CLIA AND ITS IMPACT ON RADTS The Clinical Laboratory Improvement Act (CLIA) is federal legislation dating to 1967 but originally did not apply to physician ofce laboratories. CLIA 88 is a series of amendments to the previously existing law that were passed by Congress in 1988, to be effective in 1992 and to extend to physician ofce laboratories for the rst time. CLIA 88 was stimulated in large part by inaccurate laboratory results, especially related to Pap smears. Since 1992, physician ofce laboratories have been required to meet certain standards in order to be certied to perform all laboratory tests, except for tests considered to be sufciently uncomplicated to qualify as CLIA-waived tests. As noted above, a variety of different technologies have been employed in the various RADTs, including latex agglutination, EIA, OIA, and chemiluminescent gene probes. For at least the rst two technologies, many different congurations may be employed by various manufacturers. Some of these tests (primarily latex agglutination and EIA based) are CLIA waived, while the others are performed only in laboratories that have been certied to perform such tests. By denition, the latter tests are more complex and require more laboratory expertise while the former utilize few steps and require no special equipment or laboratory expertise. The impact of the CLIA 88 legislation on practices related to the diagnosis of GABHS pharyngitis was swift. A study by Schwartz et al., published in 1994, documented that 24% of questionnaire respondents had discontinued streptococcal testing in the ofce since the implementation of CLIA 88 in 1992 (69). The effect was particularly noticeable in rural areas and in pediatric ofces, with 72% of the latter indicating reduced test menus. Binns et al. surveyed Illinois pediatricians in 1988 and 1996 regarding the practices of their physician ofce laboratories (2). They identied a decline in the number of ofces offering throat cultures for GABHS from 63 to 34% (P 0.001) and a decline in the number performing RADTs from 77 to 64% (P 0.05). For solo practitioners, the latter decline was from 66 to 39% (P 0.001). COMPARATIVE DATA Remarkably few published studies have compared the performance of CLIA-waived tests for GABHS pharyngitis directly to that of CLIA-nonwaived tests. The published literature is composed almost exclusively of studies in which a specic test utilizing a specic method was compared to one or more culture techniques but not to other RADTs. In addition, relatively few studies have examined the performance of various RADTs in the ofce setting, as opposed to the hospital or diagnostic microbiology laboratory setting. In 1997, Schwartz published one of the few reports in which two RADTs, the OSOM Strep A test (Wyntek Diagnostics, San Diego, Calif.) and the QuickVue In-Line Strep A test (Quidel, San Diego, Calif.) were directly compared (70). Both tests were found to

VOL. 17, 2004

RAPID DIAGNOSIS OF STREPTOCOCCAL PHARYNGITIS

575

be 100% specic compared to BAP cultures, with sensitivities of 95 and 87%, respectively. Sensitivity was related directly to the intensity of growth on the culture plate. Pillai et al. presented data in abstract form in 1999 that compared the sensitivity, specicity, positive and negative predictive values, and overall accuracy of BAP cultures and RADTs as performed in 18 private pediatric ofces in the greater Chicago area (S. Pillai, L. Dennis, B. Sehwarte, C. Ciesielski, and S. Shulman, Prog. Abstr. XIV Lanceeld Int. Symp. Streptococci and Streptococcal Dis., abstr. 513, 1999). Results obtained in individual ofces (using the procedures then standard in these ofces) were compared with results for a simultaneously acquired throat swab that was shipped to a streptococcal research laboratory and processed by a broth culture technique. In this study, the performance of the RADTs appeared comparable to that of the BAP cultures when both were processed and interpreted in the ofce. Additionally, the CLIA-waived tests performed as well as the moderately complex tests, although the numbers were small and direct comparisons with duplicate specimens were not performed. Roosevelt et al. compared the performance of a CLIAwaived RADT performed by nurses in the emergency room of a large childrens hospital to a moderately complex test performed in the hospital microbiology laboratory (64). Duplicate throat swabs were obtained simultaneously for the two assays. In this study, there was a high degree of concordance of results (P 0.001) and there was no signicant difference in the performance of the tests. RECOMMENDATIONS OF ADVISORY COMMITTEES A number of advisory groups have formulated recent recommendations regarding the use of RADTs for the diagnosis of GABHS pharyngitis. For example, the Committee on Infectious Diseases of the American Academy of Pediatrics recommends that when a patient suspected of having GABHS pharyngitis has a negative RADT, a BAP culture should be performed to conrm the results of the RADT (12). However, because of the high specicity of RADTs, a positive test result does not require conrmation with a BAP culture. RADTs using newer technologies such as OIA or DNA probes may be more sensitive than earlier RADTs, and some experts think that these newer tests are sufciently sensitive to be used without BAP culture backup. The Committee on Infectious Diseases suggests that physicians who use such tests without BAP culture backup compare their results with those of BAP culture to validate the presence of adequate RADT sensitivity in their practice (12). The Practice Guidelines for the Diagnosis and Management of Group A Streptococcal Pharyngitis of the Infectious Diseases Society of America also states that positive RADT results do not require culture conrmation (3). However, for children and adolescents, they recommend that a negative RADT result be conrmed with a BAP culture unless the physician has determined in his or her own practice that the RADT utilized performs comparably to a BAP culture. The Infectious Diseases Society of America guidelines suggest that because adults have both a lower incidence of GABHS pharyngitis than children and adolescents and an extremely low risk

of contracting acute rheumatic fever, negative RADT results for adults need not be conrmed with BAP cultures (3). The Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the American Heart Association published recommendations in 1995 for the diagnosis and treatment of streptococcal pharyngitis (14). These recommendations indicate that RADTs are useful and can be utilized diagnostically, especially when positive results are obtained. These 8-year-old recommendations include the suggestion that negative RADTs be conrmed by BAP cultures. The Centers for Disease Control and Prevention, the American College of PhysiciansAmerican Society of Internal Medicine, and the American Academy of Family Physicians have recently published a clinical practice guideline for acute pharyngitis in adults (13). This guideline recommends that all adults with pharyngitis be screened for the presence of the four Centor criteria: history of fever, tonsillar exudate, absence of cough, and tender anterior cervical lymphadenopathy. Those with none or one of these criteria need not be tested or treated. The appropriate options for adults with two or more criteria are the following: (i) test by RADT and treat only those with positive results; (ii) use a RADT for those with two or three criteria and treat only those with a positive result or with four criteria; or (iii) use no RADT and treat only those with three or four criteria. In contrast to other guidelines directed primarily at children (12), but consistent with recent guidelines directed primarily at adults (3), this guideline recommends that BAP cultures not be required for conrmation of negative RADT results (13). Some have expressed concern that this approach might result in the administration of antimicrobial therapy to an unacceptably large number of adults with nonstreptococcal pharyngitis (4). In addition, it has been suggested that these recommendations should be conrmed by clinical trials before clinicians abandon long-held teachings regarding the need for bacteriologic conrmation in patients suspected to have GABHS pharyngitis on the basis of clinical and epidemiologic ndings before antimicrobial therapy is provided (4). ACTUAL PRACTICES There is relatively little published information about how physicians in practice actually use RADTs for the diagnosis of GABHS pharyngitis. In 1994, Schwartz et al. (69) sent surveys to 1,060 board-certied pediatricians in seven western states. There were 679 responses, of which 531 were appropriate for comparison with contemporaneous recommendations from the American Academy of Pediatrics. An optimal approach (dened as RADT alone when the results were positive and RADT backed up by BAP culture when the results were negative) for at least 80% of patients was employed by only 44% of the responding pediatricians. Of the pediatricians who responded, 17% used RADT in an appropriate manner for fewer than 50% of their patients. In 1997, Hofer et al. (37) reported the results of a survey sent to 1,000 board-certied pediatricians throughout the United States from September to November 1991. Of the 510 respondents, 20% used RADTs for the diagnosis of GABHS pharyngitis in a manner not consistent with contemporaneous recommendations from the American Academy of Pediatrics. These studies demonstrate that many physicians do not fol-

576

GERBER AND SHULMAN

CLIN. MICROBIOL. REV.

low recommended guidelines for using RADTs to diagnose GABHS pharyngitis, particularly including the conrmation of negative RADT results with BAP cultures. The implementation of CLIA in 1996 served as an incentive for manufacturers of diagnostic test to develop new RADTs that did not require CLIA certication (CLIA-waived tests). Over the past several years, numerous CLIA-waived RADTs for the diagnosis of GABHS pharyngitis have been marketed in the United States; however, there are virtually no published data about how physicians are using these CLIA-waived tests in practice. MEDICAL DECISION MAKING It has been estimated that approximately 18 million patients sought care for a sore throat in the United States in 1996, making it the sixth leading reason to visit a physician (67). Unfortunately, some physicians perform a BAP culture or RADT for almost every patient with an upper respiratory tract infection. Such indiscriminate use of these diagnostic tests not only contributes to the excessive cost of health care but also does not serve patients well. Whether one uses a BAP culture or a RADT, it is important that they be used selectively in order to remain useful and costeffective diagnostic procedures. When attempting to decide whether to perform a test such as a BAP culture or RADT on a particular patient, the epidemiologic and clinical ndings should be considered (80). For example, in temperate climates, GABHS pharyngitis is a disease of the winter and early spring and primarily affects children between 3 and 18 years of age. Therefore, physicians should have a lower threshold for performing throat cultures or RADTs in the winter and spring and for children aged 3 to 18 years. Awareness of much GABHS in the community or the presence of a specic viral agent of pharyngitis at a particular time can also be helpful in deciding whether to perform a diagnostic test for GABHS, as is a history of close contact with a well-documented case of GABHS pharyngitis. In addition, there are certain clinical ndings which strongly suggest either a GABHS (tender anterior cervical lymph node, palatal petechiae) or a nonstreptococcal (rhinorrhea, cough, hoarseness, diarrhea) etiology of pharyngitis (80). Selective use of these diagnostic tests not only increases their rate of positivity but also increases the percentage of test-positive patients who are truly infected while reducing the percentage of test-positive patients who are merely chronic streptococcal carriers. Consequently, there is an increase in the percentage of test-positive patients who would benet from antimicrobial therapy and a reduction in the percentage of those who, as carriers, would not benet from such therapy. In addition, because of the presence of a selection bias with RADTs, selective use of these tests may also increase their sensitivity. To evaluate the potential clinical impact of RADTs for the diagnosis of GABHS pharyngitis, one should consider carefully the specic reasons for diagnosing and treating GABHS pharyngitis. One important reason is to prevent acute rheumatic fever (81). Prevention of acute rheumatic fever is thought of in terms of primary prevention (prevention of the initial episode) and secondary prevention (prevention of recurrent episodes). Primary prevention depends on identication of all patients

with GABHS infections of the upper respiratory tract and their treatment with a course of antimicrobials that will eradicate GABHS (9). Rapid diagnosis is not essential; antimicrobial therapy can be initiated as late as 9 days after the onset of GABHS pharyngitis and still effectively prevents acute rheumatic fever (9). However, effective primary prevention requires that the diagnostic test used have a high degree of sensitivity in order to identify all patients with GABHS pharyngitis at risk of developing acute rheumatic fever. A second reason to diagnose and treat GABHS pharyngitis is to reduce the associated acute morbidity. Children with GABHS pharyngitis who have received immediate treatment with the appropriate antibiotic demonstrate signicantly faster resolution of clinical signs and symptoms than do those for whom antimicrobial therapy was temporarily withheld (61). Additional evidence suggests that the earlier in the course of GABHS pharyngitis that antimicrobial therapy is initiated, the greater is the potential impact (8). In contrast to their relatively low sensitivity, the currently available RADTs consistently show high specicities of about 95 to 100% (23, 26). At such high specicities, RADTs can identify patients with GABHS pharyngitis who would benet from early antimicrobial therapy and would consequently be able to return to school or work more quickly. GABHS are spread by respiratory secretions and droplets, and another reason for diagnosing and treating GABHS pharyngitis is to prevent the spread of GABHS to others. Even untreated patients become asymptomatic within a few days and, on return to normal activities, may serve as an occult source for spread of GABHS to others. In contrast, 95% of treated patients are culture negative within 24 h of the initiation of appropriate therapy and may return safely to school or work (61). Evidence from school-, military-, and family-based investigations suggests that the sooner antimicrobial therapy is initiated, the lower the risk of transmission of GABHS to others (23, 26). Because of their high specicity, RADTs can rapidly identify patients with GABHS pharyngitis who pose a transmission risk and who could return to normal activities sooner with timely initiation of antimicrobial therapy. Finally, diagnosing and treating GABHS pharyngitis can prevent the development of suppurative complications such as peritonsillar and retropharyngeal abscesses, as well as cervical lymphadenitis. Although the use of antimicrobials to treat GABHS pharyngitis clearly reduces the incidence of these suppurative sequelae, it is not clear that speed is essential (39). Therefore, there is little or no advantage to RADTs over BAP cultures in this respect. Are there any potential disadvantages to prompt diagnosis and early initiation of antimicrobial therapy for GABHS pharyngitis? The concern has been raised that this approach might interfere with the development of immunity and consequently could result in a higher recurrence rate as well as an older population of susceptible patients (52). Although antimicrobial therapy can interfere with the production of type-specic (anti-M protein) antibodies, there is no well-documented evidence that early therapy results in a shift in the age-related incidence of GABHS pharyngitis to an older population (7). Pichichero et al. reported that patients with GABHS pharyngitis who were treated immediately with penicillin had signicantly more recurrences than did patients treated more than 48 h after diag-

VOL. 17, 2004

RAPID DIAGNOSIS OF STREPTOCOCCAL PHARYNGITIS

577

nosis (57). However, this investigation was weakened by a lack of GABHS serotyping of isolates and an inability to distinguish recurrences from new infections. It was not conrmed by other investigators (27). In addition, there is some evidence that non-type-specic immunity may be important in preventing recurrent GABHS infections (21). Taking all these issues into consideration, physicians should examine their threshold for testing for GABHS in patients with acute pharyngitis. Both cost-effectiveness and positive predictive value may be unacceptably low when the prevalence of GABHS is low. Physicians also should note that the greater the likelihood of problems with follow-up, the lower the threshold should be for using a RADT. Even a relatively insensitive RADT may be superior to clinical judgment alone. Studies indicate that even experienced clinicians diagnose GABHS pharyngitis correctly only 75 to 80% of the time when the diagnosis is based on clinical assessment alone (6, 7). Despite their limitations, the use of RADTs in certain populations signicantly increases the treatment rate for patients with GABHS pharyngitis compared with BAP cultures. For example, in a report by Lieu et al. (48), 255 children who presented to a busy urban emergency room with acute pharyngitis were evaluated. A BAP culture and RADT were simultaneously performed for each patient. The RADT had a sensitivity of only 55% compared with the BAP culture. However, 80% of patients with positive BAP cultures received appropriate antimicrobial therapy when RADTs were used as an adjunct to the BAP cultures, compared with a 57% treatment rate among patients who had GABHS pharyngitis diagnosed on the basis of a positive BAP culture alone (P 0.05). This difference, at least in part, reects the fact that the need to call back patients after BAP cultures have been incubated is eliminated for many by the use of RADTs It is important to note that the majority of patients with acute pharyngitis have nonstreptococcal infection and therefore do not benet from antibiotic therapy (61). Nevertheless, many physicians treat pharyngitis patients with antibiotics while awaiting results of the BAP culture; in two surveys, approximately 40% of physicians admitted to continuing antibiotic treatment even after the BAP culture was reported as negative (11, 38). Rapid exclusion of GABHS pharyngitis by a negative RADT could avoid the use of unnecessary antibiotics in many patients. However, before a practice decides to use a RADT without BAP culture backup, it should compare the RADT and BAP culture results to validate adequate RADT sensitivity in the practice. This is particularly important in areas of the country that are experiencing a resurgence of acute rheumatic fever. In 1998, Needham et al. performed an investigation to determine whether the availability of results from a RADT for GABHS improved the management of patients with acute pharyngitis (55). Physicians intent to prescribe antimicrobial therapy based on clinical impression alone was compared to physicians intent once the result of the RADT was known for 465 consecutive pediatric and adult patients with acute pharyngitis. They found that use of the particular RADT they studied (which had a sensitivity and specicity of 87 and 97%, respectively, compared with BAP culture) substantially improved physicians decisions to prescribe appropriate therapy. Had the RADT alone guided therapeutic choice at the initial

encounter, physicians would have prescribed appropriate antimicrobial therapy for 95% of the patients. The investigators concluded that neither clinical assessment nor BAP culture would add signicantly to the outcome achieved when a RADT, with performance characteristics like those of the test used in this study, is used to assist in management of patients with acute pharyngitis. In 1990, Meier et al. investigated the impact of the introduction of a RADT in an inner-city community health center on the use of antimicrobials in patients with negative test results (53). They compared antimicrobial-prescribing behavior for a 7-month period during which BAP cultures were the primary diagnostic tests for patients with acute pharyngitis to a 7-month period during which a RADT was the primary diagnostic test. They found that use of the RADT signicantly reduced the rate of antimicrobial therapy in patients with negative test results. In a limited analysis, the investigators also found that the savings in antimicrobial therapy avoided in patients with negative test results offset the increased cost of RADT compared to BAP cultures. Webb et al. examined the incidence of suppurative and nonsuppurative complications of GABHS pharyngitis in a population served by a large suburban medical center during two different periods (83). During the rst 2-year period, diagnosis of GABHS pharyngitis was made almost exclusively on the basis of BAP cultures, while during the second 2-year period, this diagnosis was made almost exclusively on the basis of a sensitive RADT without BAP culture conrmation of negative results. A total of 30,036 patients were evaluated for pharyngitis during the 4-year period, and there was no difference in the incidence of either suppurative or nonsuppurative complications of GABHS pharyngitis between the two study periods. Cost-Effectiveness In 1998, Webb performed a decision analysis using conditions in a large, suburban, pediatric practice to determine which of four strategies for the management of children with acute pharyngitis (treat all, use RADT alone, use BAP culture alone, use RADT with BAP culture conrmation of negative results) was most cost-effective (82). Using data from 13 published studies in which this particular RADT (an OIA) and BAP culture were evaluated against a variety of criterion standards, this RADT was assumed to have a sensitivity and specicity of 89.1 and 95%, respectively, while the BAP culture was assumed to have a sensitivity and specicity of 83.4 and 99%, respectively. Although the strategy of treating all children with acute pharyngitis with antimicrobial therapy was the most costeffective, this strategy was not recommended because of concerns about the promotion of antimicrobial resistance (which could not be included in the model) and about the large number of allergic reactions induced in children who did not have conrmed GABHS pharyngitis. Of the strategies in which a diagnostic test was employed, use of a RADT with performance characteristics similar to those of the test used in the model without BAP culture conrmation of all negative RADT results was the most cost-effective. In 1999, Tsevat and Kotagal performed a decision analysis to determine the cost-effectiveness of seven different strategies for managing children with acute pharyngitis: no testing or

578

GERBER AND SHULMAN

CLIN. MICROBIOL. REV.

treatment, treat all, BAP culture only, RADT #1 with sensitivity of 85.9% and specicity of 94.3% only, RADT #2 with sensitivity of 80.8% and specicity of 89.5% only, RADT #1 with backup BAP culture for negative results, and RADT #2 with backup BAP culture for negative results (77). At a baseline prevalence of 20.8% for GABHS pharyngitis and with all tests performed at a local reference laboratory, BAP culture only was the most cost-effective strategy. When all testing was ofce based, BAP culture only was still the most cost-effective strategy. In 1993, Majeed et al. in Kuwait performed a decision analysis to determine the cost-effectiveness of four different strategies for managing children with acute pharyngitis: treat all with benzathine penicillin G, treat with benzathine penicillin G on the basis of clinical diagnosis, treat with benzathine penicillin G on the basis of RADT, and treat with oral penicillin on the basis of RADT (49). The last strategy, using RADT and treating positive patients with a 10-day course of oral penicillin, was determined to be the most cost-effective. Neuner et al. recently reported a detailed cost-effectiveness analysis of the diagnosis and management of adults with acute pharyngitis. The investigators compared ve management strategies: no diagnostic testing or treatment, empirical treatment with penicillin with no diagnostic testing, BAP culture using a two-plate selective technique, OIA testing with BAP culture conrmation of negative results, and OIA testing alone. Assuming a 10% prevalence of GABHS among adults with pharyngitis, empirical treatment without diagnostic testing was the least effective strategy. Although the other strategies were similarly effective, the BAP culture was the most cost-effective (56). In a recent investigation designed specically to address the cost-effectiveness of using a RADT without BAP culture conrmation of negative test results, Mayes and Pichichero retrospectively reviewed the experience in their private pediatric practice with two different RADTs during a 3-year period from January 1996 through May 1999 (51). During this 3-year period, 11,427 RADTs were performed: 8,385 were negative and 3,042 were positive. A conrmatory BAP culture was performed for 8,234 (98.2%) of the negative RADTs; only 200 (2.4%) of the 8,234 patients with negative RADTs had a positive conrmatory BAP culture. A cost analysis showed that while the cost of performing conrmatory BAP cultures was high, the cost of complications arising from missed GABHS infections would be very low because false-negative RADT results and sequelae are infrequent and rare, respectively. The investigators concluded that BAP culture conrmation of negative RADT results in patients with acute pharyngitis is costly and may not be medically necessary for most patients, given the current availability of highly sensitive RADTs. RADT AND SUSCEPTIBILITY TESTING One of the problems with using RADTs rather than BAP cultures to diagnose GABHS pharyngitis is that the isolates of GABHS causing the pharyngitis are not available to test for antibiotic susceptibility. Despite the use of penicillin for almost 50 years to treat GABHS infections, there has been no significant change in the in vitro susceptibility of GABHS to penicillin, nor has there been a single, well-documented report of a

clinical isolate of GABHS that is resistant to penicillin (28). Although erythromycin resistance among GABHS had been a major problem in Japan and in Finland, it has not been a problem in the United States (28). However, Martin et al. recently reported a clonal outbreak of erythromycin-resistant GABHS pharyngitis in the Pittsburgh area in which 48% of the isolates of GABHS from the children in one school were erythromycin resistant (50). While focal outbreaks of GABHS infections due to erythromycin-resistant strains have been described in the United States in the past, there is no evidence of widespread erythromycin resistance among strains of GABHS. In fact, Tanz et al. recently reported the results of a national surveillance of 972 pharyngeal isolates of GABHS collected during the 2000 to 2001 season and almost 2,000 isolates collected during the 2001 to 2002 season, in which only 4.4% of the isolates were erythromycin resistant (R. R. Tanz, W. Kabat, K. Kabat, B. Beall, and S. Shulman, Prog. Abstr. 40th Annu. Meet. Infect. Dis. Soc. Am., abstr. 110, 2002). Although it is important to continue national surveillance for resistance to macrolides among pharyngeal isolates of GABHS, susceptibility testing on a routine basis is not indicated. Therefore, the inability to perform antibiotic susceptibility testing in cases of GABHS pharyngitis diagnosed solely on the basis of a RADT result should not be clinically signicant. GAPS IN KNOWLEDGE Currently available published data are inadequate to allow denitive conclusions to be made regarding the relative performance characteristics of the various RADTs for GABHS pharyngitis. In the very few published studies available, striking differences between performance characteristics of CLIAwaived and moderately complex tests were not apparent. However, more ofce-based studies are needed to adequately address this issue. In addition to the paucity of peer-reviewed publications in this area, another major problem is the variability of the criterion standard against which individual RADTs have been compared. The clinical setting may also dictate the way in which comparative data should be considered. For example, the performance of a RADT compared to a BAP cultures in the setting of an emergency room of a large hospital with a microbiology laboratory that is fully staffed by certied microbiology technologists may be different from the performance in the setting of a private physician ofce that lacks such highly skilled laboratorians. Therefore, recommendations regarding back-up BAP cultures when RADTs are reported to be negative may need to vary with the setting in which care is delivered. This is also an area in need of additional well-designed studies. There is an obvious paucity of clinical studies that have directly compared various rapid tests and culture techniques for diagnosing GABHS pharyngitis. This has resulted in a vacuum of information, with resulting confusion about the relative merits of various diagnostic modalities. It is apparent that there is a need for direct comparative studies of CLIA-waived and moderately complex RADTs for GABHS pharyngitis by using a standardized method. In addition, the rarity of serious life-threatening consequences of acute GABHS pharyngitis in the United States today suggests

VOL. 17, 2004

RAPID DIAGNOSIS OF STREPTOCOCCAL PHARYNGITIS

579

the need for careful cost-benet and risk-benet analyses of different diagnostic strategies in various clinical settings. The individual ofce laboratory should take care to verify the performance in that ofce of its chosen diagnostic method compared to standard BAP cultures performed in that ofce. CONCLUSION The development of easy-to-perform RADTs for the diagnosis of GABHS pharyngitis has altered clinical practice substantially. Although very high specicity is apparent with virtually all of the various available RADTs, the sensitivities of these assays vary. Assessment of performance characteristics, especially sensitivity, is highly unstandardized, with variously performed culture techniques used as comparators, and this accounts for much of the reported variability in sensitivity among different RADTs. Studies in which CLIA-waived and moderately complex RADTs are compared directly are lacking. Recommendations for the clinical utilization of RADTs should reect the clinical setting in which they are performed (hospital, private ofce, or clinic), cost issues (managed care, cost of individual tests), and the availability and expense of non-RADT tests (throat cultures). Only limited cost-effectiveness data exist, and they are markedly inuenced by the assumptions made regarding test performance as well as the risks and costs associated with certain outcomes. It would be prudent for clinicians to establish that a specic RADT performs satisfactorily with respect to throat cultures in their own practices before relying solely on RADTs for the diagnosis of GABHS pharyngitis. This can be done by performing both culture and RADT on duplicate or the same specimens and assessing the level of concordant results. This is particularly important in communities in which cases of acute rheumatic fever continue to occur. Because many physicians apparently do not follow recommended guidelines for using RADTs to diagnose GABHS pharyngitis, particularly the conrmation of negative RADT results with BAP cultures, the availability of a RADT that does not require BAP culture conrmation could improve patient care.
REFERENCES 1. Baker, D. M., R. M. Cooper, C. Rhodes, L. A. Weymouth, and H. P. Dalton. 1995. Superiority of conventional culture technique over rapid detection of group A streptococcus by optical immunoassay. Diagn. Microbiol. Infect. Dis. 21:6164. 2. Binns, H. J., S. LeBailly, and H. D. Gardner. 1998. The physicians ofce laboratory: 1988 and 1996 survey of Illinois pediatricians. Arch. Pediatr. Adolesc. Med. 152:585592. 3. Bisno, A. L., M. A. Gerber, J. M. Gwaltney, Jr., E. L. Kaplan, and R. H. Schwartz. 2002. Practice guidelines for the diagnosis and management of group A streptococcal pharyngitis. Infectious Disease Society of America. Clin. Infect. Dis. 15:113125. 4. Bisno, A. L., G. S. Peter, and E. L. Kaplan. 2002. Diagnosis of strep throat in adults: are clinical criteria really good enough? Clin. Infect. Dis. 15:126 129. 5. Bourbeau, P. P., B. J. Heiter, J. P. Anhalt, and D. W. Naumovitz. 1993. Comparison of direct specimen testing utilizing TestPack Strep A with testing of specimens following a two-hour broth enrichment. Diagn. Microbiol. Infect. Dis. 17:9396. 6. Breese, B. B., and F. A. Disney. 1954. The accuracy of diagnosis of beta streptococcal infections on clinical grounds. J. Pediatr. 44:670673. 7. Breese, B. B., and C. B. Hall. 1978. Beta hemolytic streptococcal disease, p. 79. Houghton Mifin, Boston, Mass. 8. Brink, W. R., C. H. Rammelkamp, Jr., F. W. Denny, and L. W. Wannamaker. 1951. Effect of penicillin and aureomycin on natural course of streptococcal tonsillitis and pharyngitis. Am. J. Med. 10:300308. 9. Catanazaro, F. J., C. A. Stetson, A. J. Morris, R. Chamovitz, C. H. Ram-

10. 11.

12. 13. 14. 15.

16. 17.

18.

19. 20. 21. 22. 23. 24.

25. 26. 27. 28. 29. 30. 31. 32.

33.

34.

melkamp, Jr., B. L. Stolzer, and W. D. Perry. 1954. The role of the streptococcus in the pathogenesis of rheumatic fever. Am. J. Med. 17:749756. Chapin, K. C., P. Blake, and C. D. Wilson. 2002. Performance characteristics and utilization of rapid antigen test, DNA probe, and culture detection of group A streptococci in an acute care clinic. J. Clin. Microbiol. 40:42074210. Cochi, S. L. 1984. Diagnosis and treatment of streptococcal phargnigitis: survey of U.S. Practitioners, p. 7394. In S. T. Shulman (ed.), Pharyngitis: management in an era of declining rheumatic fever. Praeger, New York, N.Y. Committee on Infectious Diseases. 2003. Group A streptococcal infection, p. 573584. In L. K. Pickering (ed.), 2003 red book American Academy of Pediatrics, Elk Grove Village, IU. Cooper, R. J., J. R. Hoffman, J. G. Bartlett, R. E. Besser, R. Gonzales, J. M. Hickner, and M. A. Sande. 2001. Principles of appropriate antibiotic use for acute pharyngitis in adults: background. Ann. Intern. Med. 20:509517. Dajani, A., T. Taubert, P. Ferrier, G. Peter, and S. T. Shulman. 1995. Treatment of acute streptococcal pharyngitis and prevention of rheumatic fever: a statement for health professionals. Pediatrics 96:758764. Daly, J. A., E. K. Korgenski, A. C. Munson, and E. Llausas-Magna. 1994. Optical immunoassay for streptococcal pharyngitis: evaluation of accuracy with routine and mucoid strains associated with acute rheumatic fever outbreak in the intermountain area of the United States. J. Clin. Microbiol. 32:531532. Della-Latta, P., S. Whittier, M. Hosmer, and F. Agre. 1994. Rapid detection of group A streptococcal pharyngitis in a pediatric population with optical immunoassay. Pediatr. Infect. Dis. J. 13:742743. DiMatteo, L. A., S. R. Lowenstein, B. Brimhall, W. Reiquam, and R. Gonzales. 2001. The relationship between the clinical features of pharyngitis and the sensitivity of a rapid antigen test: evidence of spectrum bias. Ann. Emerg. Med. 38:648652. Ehrlich, T. P., R. H. Schwartz, R. Wientzen, and M. M. Thorne. 1993. Comparison of an immunochromatographic method for rapid identication of group A streptococcal antigen with culture method. Arch. Fam. Med. 2:866869. Ferris, D. G., and P. M. Fischer. 1992. Elementary school students performance with two ELISA test systems. JAMA 268:766770. Finger, R., S. H. Ho, T. T. Ngo, C. D. Ritchie, and T. N. Nguyen. 1999. Rapid streptococcal testing in Vietnamese children with pharyngitis. Asia Pac. J. Public Health 11:2629. Fischetti, V. A., W. M. Hodges, and D. Hruby. 1989. Protection against streptococcal pharyngeal colonization with a vaccine M protein recombinant. Science 244:14871490. Fries, S. M. 1995. Diagnosis of group A streptococcal pharyngitis in a private clinic: comparative evaluation of an optical immunoassay method and culture. J. Pediatr. 126:933936. Gerber, M. A. 1989. Comparison of throat culture and rapid strep tests for diagnosis of streptococcal pharyngitis. Pediatr. Infec. Dis. J. 8:820824. Gerber, M. A., R. R. Tanz, W. Kabat, E. Dennis, G. L. Bell, E. L. Kaplan, and S. T. Shulman. 1997. Optical immunoassay test for group A beta-hemolytic streptococcal pharyngitis. An ofce-based, multicenter investigation. JAMA 277:899903. Gerber, M. A., M. F. Randolph, J. Chanatry, et al. 1986. Antigen detection test for streptococcal pharyngitis: evaluation of sensitivity with respect to true infection. J. Pediatr. 108:654658. Gerber, M. A. 1998. Diagnosis of group A streptococcal pharyngitis. Pediatr. Ann. 27:269273. Gerber, M. A., M. F. Randolph, K. K. DeMeo, and E. L. Kaplan. 1990. Lack of impact of early antibiotic therapy for streptococcal pharyngitis on occurrence rates. J. Pediatr. 117:653658. Gerber, M. A. 1996. Antibiotic resistance: relationship to persistence of group A streptococci in the upper respiratory tract. Pediatrics 97(Suppl.): 971975. Gerber, M. A., M. F. Randolph, N. J. Martin, et. al. 1991. Community-wide outbreak of group G streptococcal pharyngitis. Pediatrics 87:598603. Gleseker, K. E., T. MacKenzie, M. H. Roe, and J. K. Todd. 2002. Comparison of two rapid Streptococcus pyogenes diagnostic tests with a rigorous culture standard. Pediatr. Infect. Dis. J. 21:922926. Hamrick, H. J., and M. E. Mangum. 1999. Beta-hemolytic Streptococcus milleri group misidentied as Streptococcus pyogenes on throat culture. Pediatr. Infect. Dis. J. 18:7576. Harbeck, R. J., J. Teague, G. R. Crossen, D. M. Maul, and P. L. Childers. 1993. Novel, rapid optical immunoassay technique for detection of group A streptococci from pharyngeal specimens: comparison with standard culture methods. J. Clin. Microbiol. 31:839844. Hart, A. P., L. L. Buck, S. Morgan, S. Saverio, and J. C. McLaughlin. 1997. A comparison of the BioStar Strep A OIA rapid antigen assay, group A Selective Strep Agar (SSA), and Todd-Hewitt broth cultures for the detection of group A Streptococcus in an outpatient family practice setting. Diagn. Microbiol. Infect. Dis. 29:139145. Heelan, J. S., S. Wilbur, G. Depetris, and C. Letorneau. 1996. Rapid antigen testing for group A Streptococcus by DNA probe. Diagn. Microbiol. Infect. Dis. 24:6569.

580

GERBER AND SHULMAN

CLIN. MICROBIOL. REV.

Comparison of Gene-Probe Group A Streptococcus direct test with culture for diagnosing streptococcal pharyngitis. J. Clin. Microbiol. 32:14401443. Randolph, M. E., M. A. Gerber, K. K. DeMeo, and L. Wright. 1985. Effect of antibiotic therapy on the clinical course of streptococcal pharyngitis. J. Pediatr. 106:870875. Roddey, O. F., Jr., H. W. Clegg, E. S. Martin, R. L. Swetenburg, and E. W. Koonce. 1995. Comparison of throat culture methods for the recovery of group A streptococci in a pediatric ofce setting. JAMA 274:18631865. Roe, M., C. Kishiyama, K. Davidson, L. Schaefer, and J. Todd. 1995. Comparison of BioStar A OIA optical immune assay, Abbott TestPack Plus Strep A, and culture with selective media for diagnosis of group A streptococcal pharyngitis. J. Clin. Microbiol. 33:15511553. Roosevelt, G. E., M. S. Kulkarni, and S. T. Shulman. 2001. Critical evaluation of a CLIA-waived streptococcal antigen detection test in the emergency department. Ann. Emerg. Med. 37:377381. Rubin, L. G., R. A. Kahn, E. M. Vellozzi, and H. D. Isenberg. 1996. False positive detection of group A Streptococcus antigen resulting from crossreacting Streptococcus intermedius (Streptococcus milleri group). Pediatr. Infect. Dis. J. 15:715717. Rubin, L. G., and G. S. Mirkin. 2000. Apparent false positive detection of group A Streptococcus antigen resulting from pharyngeal infection with nonhemolytic Streptococcus pyogenes. Pediatr. Infect. Dis. J. 19:672674. Schappert, S. M. 1998. Ambulatory care visits to physicians ofces, hospital outpatient departments, and emergency departments: United States, 1996. National Center for Health Statistics, Hyattsville, Md. Schlager, T. A., G. A. Hayden, W. A. Woods, S. M. Dudley, and J. O. Hendley. 1996. Optical immunoassay for rapid detection of group A beta-hemolytic streptococci. Should culture be replaced? Arch. Pediatr. Adolesc. Med. 150: 245248. Schwartz, B., S. Fries, A. M. Fitzgibbon, and H. Lipman. 1994. Pediatricians diagnostic approach to pharyngitis and impact of CLIA 1988 on ofce diagnostic tests. JAMA 271:234238. Schwartz, R. H. 1997. Evaluation of rapid streptococcal detection tests. Pediatr. Infect. Dis. J. 16:10991100. Seaberg, D. C., G. Gettings, B. Rosenthal, and T. Geiger. 1997. Rapid, optical immunoassay for streptococcal pharyngitis. Acad. Emerg. Med. 4:81 83. Shulman, S. T. 1995. Value of new rapid tests for the diagnosis of group A streptococcal pharyngitis. Pediatr. Infect. Dis. J. 14:923924. Steed, L. L., E. K. Korgenski, J. A. Daly. 1993. Rapid detection of Streptococcus pyogenes in pediatric patient specimens by DNA probe. J. Clin. Microbiol. 31:29963000. Stevens, D. L. 1995. Streptococcal toxic-shock syndrome spectrum of disease, pathogenesis, and new concepts of treatment. Emerg. Infect. Dis. 1:6978. Supon, P. A., S. Tunnell, M. Greene, and R. M. Ostroff. 1998. Rapid detection of group A streptococcal antigen with a new optical immunoassay. Pediatr. Infect. Dis. J. 17:349351. Reference deleted. Tsevat, J., and U. R. Kotagal. 1999. Management of sore throats in children: a cost-effectiveness analysis. Arch. Pediatr. Adolesc. Med. 153:681688. Turner, J. C., F. G. Hayden, M. C. Lobo, C. E. Ramirez, and D. Murren. 1997. Epidemiologic evidence for Lanceeld group C beta-hemolytic streptococci as a cause of exudative pharyngitis in college students. J. Clin. Microbiol. 35:14. Uhl, J. R., S. C. Adamson, E. A. Vetter, C. D. Schleck, W. S. Harmsen, L. K. Iverson, P. J. Santrach, N. K. Henry, and F. R. Cockerill. 2003. Comparison of LightCycler PCR, rapid antigen immunoassay, and culture for detection of group A streptococci from throat swabs. J. Clin. Microbiol. 41:242249. Wannamaker, L. W. 1984. Diagnosis of pharyngitis: clinical and epidemiological features, p. 3346. In S. T. Shulman (ed.), Pharyngitis: management in an era of declining rheumatic fever. Praeger, New York, N.Y. Wannamaker, L. W., C. H. Rammelkamp, Jr., F. W. Denny, W. R. Brink, H. B. Houser, and E. O. Hahn. 1951. Prophylaxis of acute rheumatic fever by treatment of the preceding streptococcal infection with various amounts of depot penicillin. Am. J. Med. 10:673695. Webb, K. H. 1998. Does culture conrmation of high-sensitivity rapid streptococcal tests make sense? A medical decision analysis. Pediatrics 101:E2. Webb, K. H., C. A. Needham, and S. R. Kurtz. 2000. Use of high sensitivity rapid strep tests without culture conrmation of negative results: two years experience. J. Fam. Pract. 49:3438. Wegner, D. L. 1996. Proper evaluation of rapid antigen detection methods for diagnosing streptococcal pharyngitis. Arch. Pediatr. Adolesc. Med. 150: 245248. Wegner, D. L., D. L. Witte, and R. D. Schrantz. 1992. Insensitivity of rapid antigen detection methods and single blood agar plate culture for diagnosing streptococcal pharyngitis. JAMA 267:695697. Wong, M. C., and C. H. Chung. 2002. Group A streptococcal infection in patients presenting with a sore throat at an accident and emergency department: prospective observational study. Hong Kong Med. J. 8:9298.

35. Heiter, B. J., and P. P. Bourbeau. 1993. Comparison of the Gen-Probe group A streptococcus direct test with culture and a rapid streptococcal antigen detection assay for diagnosis of streptococcal pharyngitis. J. Clin. Microbiol. 31:20702073. 36. Henderson, E. L., F. A. Meier, C. A. Fortner, and J. R. Zanga. 1988. Physicians bias and the interpretation of rapid tests for group A streptococcus pharyngitis. Am. J. Dis. Child. 142:405406. 37. Hofer, C., H. J. Binns, and R. R. Tanz. 1997. Strategies for managing group A streptococcal pharyngitis: a survey of board-certied pediatricians. Arch. Pediatr. Adolesc. Med. 151:824829. 38. Holmberg, S. D., and G. A. Faich. 1983. Streptococcal pharyngitis and acute rheumatic fever in Rhode Island. JAMA 250:23072312. 39. Jersild, T. 1948. Penicillin therapy in scarlet fever and complicating otitis. Lancet i:671673. 40. Johnson, D. R., and E. L. Kaplan. 2001. False-positive rapid antigen detection test results: reduced specicity in the absence of group A streptococci in the upper respiratory tract. J. Infect. Dis. 183:11351137. 41. Joslyn, S. A., G. L. Hoekstra, and J. E. Sutherland. 1995. Rapid antigen detection testing in diagnosing group A beta-hemolytic streptococcal pharyngitis. J Am. Board Fam. Pract. 8:177182. 42. Kaltwasser, G., J. Diego, P. L. Welby-Sellenriek, R. Ferrett, M. Caparon, and G. A. Storch. 1997. Polymerase chain reaction for Streptococcus pyogenes used to evaluate an optical immunoassay for the detection of group A streptococci in children with pharyngitis. Pediatr. Infect. Dis. J. 16:748753. 43. Kaplan, E. L., H. F. Reid, D. R. Johnson, et. al. 1989. Rapid antigen detection in the diagnosis of group A streptococcal pyoderma: inuence of a learning curve effect on sensitivity and specicity. Pediatr. Infect. Dis. J. 8:591593. 44. Kaplan, E. L., and M. A. Gerber. 2004. Group A, group C, and group G beta-hemolytic streptococcal infections. In R. D. Feigin and J. D. Cherry (ed.), Textbook of pediatric infectious diseases, 5th ed., p. 11421156. The W. B. Saunders Co., Philadelphia, Pa. 45. Kuhn, S., H. D. Davies, G. Katzko, T. Jadavji, and D. L. Church. 1999. Evaluation of the Strep A OIA assay versus culture methods: ability to detect different quantities of group A Streptococcus. Diagn. Mirobiol. Infect. Dis. 34:275280. 46. Kurtz, B., M. Kurtz, M. Roe, and J. Todd. 2000. Importance of inoculum size and sampling effect in rapid antigen detection for diagnosis of Streptococcus pyogenes pharyngitis. J. Clin. Microbiol. 38:279281. 47. Laubscher, B., G. van Melle, N. Dreyfuss, and H. de Crousaz. 1995. Evaluation of a new immunologic test kit for rapid detection of group A streptococci, the Abbot Testpack Strep A Plus. J. Clin. Microbiol. 33:260261. 48. Lieu, T. A., G. R. Fleisher, and J. S. Schwartz. 1988. Clinical evaluation of a latex agglutination test for streptococcal pharyngitis: performance and impact on treatment rate. Pediatr. Infect. Dis. J. 7:847854. 49. Majeed, H. A., L. al-Doussary, M. M. Moussa, A. R. Yusuf, and A. H. Suliman. 1993. Ofce diagnosis and management of group A streptococcal pharyngitis employing the rapid antigen detecting test. A 1-year prospective study of reliability and cost in primary care centres. Ann. Trop. Paediatr. 13:6572. 50. Martin, J. M., M. Green, K. A. Barbadora, and E. R. Wald. 2002. Erythromycin-resistant group A streptococci in school children in Pittsburgh. N. Engl. J. Med. 346:12001206. 51. Mayes, T., and M. E. Pichichero. 2001. Are follow-up throat cultures necessary when rapid antigen detection tests are negative for group A streptococci? Clin. Pediatr. 40:191195. 52. McCarty, M. 1954. Antibody response to streptococci, P. 139140. In M. McCarty (ed.), Streptococcal infections. Columbia University Press, New York, N.Y. 53. Meier, F. A., H. Howland, J. Johnson, and R. Poisson. 1990. Effects of a rapid antigen test for group A streptococcal pharyngitis on physician prescribing and antibiotic costs. Arch. Intern. Med. 150:16961700. 54. Meier, F. A., R. M. Centor, L. Graham, Jr., and H. P. Dalton. 1990. Clinical and microbiological evidence for endemic pharyngitis among adults due to group C streptococci. Arch. Intern. Med. 150:825829. 55. Needham, C. A., K. A. McPherson, and K. H. Webb. 1998. Streptococcal pharyngitis: impact of a high-sensitivity antigen test on physician outcome. J. Clin. Microbiol. 36:34683473. 56. Neuner, J. M., M. B. Hamel, R. S. Phillips, K. Bona, and M. D. Aronson. 2003. Diagnosis and management of adults with pharyngitis: a cost-effectiveness analysis. Ann. Intern. Med. 139:113122. 57. Pichichero, M. E., F. A. Disney, W. B. Talpey, and R. A. Hoekelman. 1987. Adverse and benecial effects of immediate treatment of group A betahemolytic streptococcal pharyngitis with penicillin. Pediatr. Infect. Dis. J. 6:635643. 58. Reference deleted. 59. Pitetti, R. D., S. D. Drenning, and E. R. Wald. 1998. Evaluation of a new rapid antigen detection kit for group A beta-hemolytic streptococci. Pediatr. Emerg. Care 14:396398. 60. Pokorski, S. J., E. A. Vetter, P. C. Wollan, and F. R. Cockerill. 1994.

61. 62. 63.

64. 65.

66. 67. 68.

69. 70. 71. 72. 73. 74. 75. 76. 77. 78.

79.

80. 81.

82. 83. 84. 85. 86.