OBJECTIVES. The goals were to establish performance characteristics of a rapid antigen-detection test and blood agar plate culture performed and interpreted in community pediatric offices and to assess the effect of the pretest likelihood of group A streptococcus pharyngitis on test performance (spectrum bias).
METHODS. Two throat swabs were collected from 1848 children 3 to 18 years of age who were evaluated for acute pharyngitis between November 15, 2004, and May 15, 2005, in 6 community pediatric offices. One swab was used to perform the rapid antigen-detection test and a blood agar plate culture in the office and the other was sent to our laboratory for blood agar plate culture. Clinical findings were used to calculate the McIsaac score for each patient. The sensitivities of the office tests were calculated, with the hospital laboratory culture results as the criterion standard.
RESULTS. Thirty percent of laboratory blood agar plate cultures yielded group A streptococcus (range among sites: 21%–36%). Rapid antigen-detection test sensitivity was 70% (range: 61%–80%). Office culture sensitivity was significantly greater, 81% (range: 71%–91%). Rapid antigen-detection test specificity was 98% (range: 98%–99.5%), and office culture specificity was 97% (range: 94%–99%), a difference that was not statistically significant. The sensitivity of a combined approach using the rapid antigen-detection test and back-up office culture was 85%. Among patients with McIsaac scores of >2, rapid antigen-detection test sensitivity was 78%, office culture sensitivity was 87%, and combined approach sensitivity was 91%. Positive diagnostic test results were significantly associated with McIsaac scores of >2.
CONCLUSIONS. The sensitivity of the office culture was significantly greater than the sensitivity of the rapid antigen-detection test, but neither test was highly sensitive. The sensitivities of each diagnostic modality and the recommended combined approach were best among patients with greater pretest likelihood of group A streptococcus pharyngitis.
Throat culture with a blood agar plate (BAP) is the standard method for establishing the diagnosis of pharyngitis caused by group A streptococcus (GAS) in children.1–4 Rapid diagnostic tests became available for office-based physicians in the 1980s. Compared with BAP cultures, the specificity of these rapid antigen-detection tests (RADTs) was considered excellent (at least 90%–95%), but the sensitivity was lower (often 75%–85%).1 Although RADT technology has evolved since these tests were first introduced,1 the American Academy of Pediatrics, the Infectious Diseases Society of America, and the American Heart Association continue to recommend confirmation of negative RADT results with a BAP culture.2–4
There is no consensus regarding BAP culture methods. BAP culture performance is affected by swab techniques,5 the skill of the personnel who process and interpret BAP cultures,6 whether the cultures are processed in physician office laboratories (POLs) or more-centralized laboratories,6 and the specific materials and conditions used for plating and incubation of the cultures.1,7 Discordance between the results of simultaneous cultures is well described.8–10 Assessment of RADT performance is affected by all aspects of BAP culture methods that affect culture results. The impact of culture methods and interpretation on RADT sensitivity was clearly demonstrated in a study of the optical immunoassay RADT.11 Moreover, GAS RADTs are more likely to yield positive results when a concurrent BAP culture has heavier growth of GAS12,13 or when a larger inoculum is used to perform the RADT.14
The sensitivity of RADTs also may be affected by the pretest likelihood of GAS pharyngitis.13,15–17 This phenomenon is called spectrum effect or spectrum bias and occurs when test performance is affected by variations in disease presentation or population subgroups.18,19 For example, a test may have low sensitivity for young children and much higher sensitivity for teenagers (eg, heterophil test for infectious mononucleosis), or sensitivity may vary during the course of a disease (eg, chest radiographs early or later in the course of pneumonia). In the case of pharyngitis, RADT results may be affected by clinical disease severity or presentation.13,15–17
The Clinical Laboratory Improvement Act (CLIA) of 1988 has affected the availability of BAP cultures in POLs.20–22 The CLIA requires that POLs meet standards for test performance and quality control based on test complexity. BAP cultures and some RADTs are considered moderately complex tests, and POLs must be certified to perform them; some RADTs that are easier to perform are classified as “CLIA-waived” and can be performed without certification. RADTs were performed in the office by 90% of surveyed physicians who cared for children, but only 13% of family practice physicians and 32% of pediatricians performed BAP cultures in the office.22 Although physicians may send throat swabs for BAP cultures to a hospital or commercial laboratory, some do not test for GAS at all or rely on RADTs alone.20,22,23 Failure to test for GAS was associated with a high rate of antibiotic prescribing for pharyngitis in the Kentucky Medicaid program,23 whereas testing was associated with lower rates of antibiotic prescribing for pharyngitis in national surveys.24
RADTs may have higher sensitivity than BAP cultures, when both are performed and interpreted in POLs and then compared with simultaneous cultures performed in a research laboratory.11,25 If RADT sensitivity approximates or exceeds the sensitivity of BAP cultures, then the need to confirm negative RADT results with BAP cultures would be eliminated.
We evaluated a CLIA-waived RADT for GAS pharyngitis in 6 community pediatric offices that perform and interpret routinely both RADTs and BAP cultures in the office. Our 2 main hypotheses were that RADTs performed in POLs are more sensitive than BAP cultures performed and interpreted in POLs, when each test is compared with a simultaneous BAP culture processed and interpreted in a hospital laboratory, and RADT sensitivity is subject to spectrum bias.13,15–17 Specifically, RADT sensitivity is greater when the pretest probability of GAS pharyngitis is greater.
Patients and Community-Based Practices
Eligible patients were 3 to 18 years of age and evaluated for pharyngitis by their physicians between November 15, 2004, and May 15, 2005. Six pediatric offices participated, 3 in the Chicago, Illinois, area and 3 in the Cincinnati, Ohio, area. Each practice participates in ≥1 office-based research network. The POLs were CLIA-certified for moderately complex tests, used a RADT for GAS pharyngitis, processed BAP cultures by using CLIA-approved methods, and were monitored through CLIA proficiency testing. The BAP culture procedures used in the offices were not reviewed or modified for this study, and we did not provide training in performance or interpretation of BAP cultures. Staff members in each office were trained in performance of the RADT used in this study, by representatives of the RADT manufacturer, before initiation of the study.
Patients deemed by their physicians to need evaluation for GAS pharyngitis had throat swabs obtained with 2 round-tipped rayon swabs. The swabs were rubbed together, and swab A was used to streak the BAP used by the office (office BAP culture) and then was used to perform the RADT (Quidel QuickVue dipstick [Quidel, San Diego, CA]). The result of the RADT was recorded on the top portion of a data collection form developed for this study, which was then separated from the bottom portion. Office BAP cultures were processed and interpreted by designated personnel in each office, following the office protocol that conformed to CLIA standards. The office BAP culture results were recorded on the bottom portion of the data form.
Swab B was labeled and refrigerated in the office for 1 to 3 days, in the transport medium contained in the culture tube. It was then sent to the hospital laboratory in a shipment that did not include the matching data forms. The swab was streaked on a plate containing trypticase soy agar with 5% sheep blood (Becton Dickinson, Sparks, MD), and the plate was incubated in room air at 35°C (laboratory BAP culture) and checked for growth of β-hemolytic colonies at 24 and 48 hours. β-Hemolytic colonies were confirmed as GAS with PathoDX (Los Angeles, CA) grouping reagents. Data forms were sent to the laboratory after the laboratory BAP culture results were recorded; laboratory personnel did not have knowledge of the office results before completion of the laboratory BAP culture.
Clinical Information and Clinical Scores
Patient gender and age (in years) were recorded. By using a checklist on the data form, limited clinical information (temperature of >38°C [>100.4°F] by history or in the office, tender anterior cervical lymphadenopathy, enlarged tonsils, tonsillar or pharyngeal exudates, and cough) was collected for all patients whose throat swabs were included in the study.
McIsaac scores were calculated to stratify patients according to their clinical presentation, a proxy for the likelihood that they had GAS pharyngitis.26 The scoring system was validated in a general medical practice that included children and adults.26–28 The 5 McIsaac score criteria are history of temperature of >38°C, absence of cough, tender anterior cervical adenopathy, tonsillar swelling or exudates, and age of <15 years.26 Although 1 point is awarded for each factor present, the maximal score is sometimes considered 4.26,27 A maximal score of 5 was used more recently by McIsaac et al,28 and we chose to conform to that approach. We defined patients with increased pretest likelihood of GAS pharyngitis as those with McIsaac scores of >2.26–28
Sample size was estimated by using a contingency table that varied the sensitivity of one test from 80% to 95% and the other from 75% to 90%. A 10% difference between the tests would be statistically significant at P < .001 across this range of sensitivities if there were 500 patients with GAS pharyngitis. Assuming the prevalence of GAS pharyngitis to be 25%, we sought a total sample of 2000 patients, and we asked each office to collect swabs from 325 to 350 patients.
Neither hospital laboratory personnel nor the investigators could trace laboratory results to individual patients. The results of the studies performed in the hospital laboratory were not available for clinical care of patients. The institutional review boards at both Children's Memorial Hospital and Cincinnati Children's Medical Center considered the study exempt from review under US 45CFR46.101(b)4.
Patient demographic features, results of all tests for GAS, and clinical information were entered into a computer database. The laboratory BAP culture was considered the criterion standard for diagnosis of GAS pharyngitis. The McIsaac score was calculated for each patient. The sensitivity, specificity, and accuracy (proportion of patients categorized correctly as being GAS-positive or GAS-negative) of the RADT and the office BAP culture were determined for the entire cohort of eligible patients and stratified according to clinical scores (0–2 or >2). The 95% confidence intervals (CIs) were calculated for sensitivity, specificity, and accuracy.
The sensitivities of RADT and office BAP culture were compared by using McNemar's test for paired categorical data. Receiver operating characteristic analyses were conducted to identify a clinical score value that was associated with good sensitivity and specificity for each diagnostic test (RADT, office BAP culture, and laboratory BAP culture). Logistic regression analyses were used to determine the odds (with 95% CIs) of positive test results when the McIsaac score was >2.
Swabs were collected from 2008 patients with pharyngitis, 1848 of whom were 3 to 18 years of age, with an average of 308 eligible patients with swabs per office (range: 288–348 patients per office). Clinical scores could be calculated for all eligible patients. RADTs were performed for 1843 patients, office BAP cultures were performed for 1842 patients, and both office tests were performed for 1839 patients; therefore, there are minor variations in the denominators in the tables. Data on the patient population are presented in Table 1. GAS was isolated from 30% of laboratory BAP cultures (553 of 1848 cultures; range among sites: 21%–36%). Table 2 presents the rates of recovery of GAS from laboratory BAP cultures according to calculated McIsaac scores. McIsaac scores were >2 for 56% of patients, and 38% of them had GAS-positive laboratory BAP culture results (positive predictive value), accounting for 71% of true-positive culture findings (395 of 553 cultures) (sensitivity). The negative predictive value of a score of ≤2 was 81%.
The performance of RADTs and office BAP cultures differed little in specificity and accuracy in the individual offices and overall (Table 3). For all office sites combined, the sensitivity of office BAP cultures was significantly higher than that of RADTs (80.8% vs 69.6%; P < .001). In individual offices, the sensitivity of office BAP cultures ranged from 70.6% to 90.5% and that of RADTs ranged from 60.5% to 80.2%. BAP culture sensitivity was higher than RADT sensitivity in 5 of 6 offices, and the difference was statistically significant in 4 cases. In 5 of 6 offices, however, the sensitivity of office BAP cultures was <90%; in 4 offices, the sensitivity was <85%. Forty-one patients had positive office BAP culture results and negative laboratory culture results, whereas 105 had negative office BAP culture results and positive laboratory BAP culture results.
Both office diagnostic tests exhibited spectrum bias (Table 4). RADT and office BAP culture sensitivities were significantly higher among patients with McIsaac scores of >2, compared with patients with McIsaac scores of ≤2, but the sensitivities of these office tests remained significantly different (P < .001).
Table 5 shows that McIsaac scores of >2 were significantly associated with positive results for each diagnostic test. However, the receiver operating characteristic analyses demonstrated that there was not a clinical score that discriminated between positive and negative results for any of the tests well enough to be used alone (area under the curve: 0.65–0.70). Therefore, McIsaac scores cannot be used in place of laboratory test results for diagnosis of GAS pharyngitis.
The strategy of confirming negative RADT results with office BAP cultures improved diagnostic sensitivity to 85.1% (95% CI: 81.8%–87.9%). Marked spectrum bias was evident with this combination strategy; sensitivity was 70.5% for patients with McIsaac scores of ≤2 and 90.9% for patients with scores of >2 (Table 4).
No office-based diagnostic modality (clinical score, office BAP culture, or RADT), when used alone in this study, was both highly sensitive and significantly superior to the others, with laboratory BAP culture being used as the criterion standard. The sensitivity of the office BAP culture was significantly greater than the sensitivity of the RADT when the 2 tests were compared with a simultaneously collected swab that was plated and processed in the hospital laboratory. Both office-based tests performed better among patients with symptoms characteristic of acute GAS pharyngitis (ie, those with McIsaac scores of >2). Spectrum bias of RADTs was noted previously.13,15–17 Pretest likelihood of GAS pharyngitis should now be added to the list of factors that affect office BAP culture sensitivity.
American Academy of Pediatrics, Infectious Diseases Society of America, and American Heart Association guidelines recommend that negative RADT results be confirmed with throat cultures, whereas office BAP cultures are considered adequate as the only test for diagnosis of GAS pharyngitis even when results are negative.2–4 The offices that participated in this study were all CLIA-certified to perform BAP cultures, and all belonged to ≥1 pediatric office-based research network. Their performance of BAP cultures in the office would be expected to be excellent; however, the sensitivity of office BAP cultures, although higher than the sensitivity of RADTs, was only ∼81% overall. RADT sensitivity in some published studies was within the 95% CI for office BAP culture sensitivity found in the present study (77%–84%) or higher.1 The sensitivity of office BAP cultures in this study is not unprecedented; it is higher than the value of 50% reported by Mondzac29 and similar to values reported by Wegner et al,30 Battle and Glasgow,31 and Rosenstein et al,32 that is, 72%, 80%, and 85%, respectively. In an earlier study, we found the overall sensitivity of office BAP culture to be 89% (95% CI: 87%–91%) when laboratory BAP culture alone was the criterion standard and 78% (95% CI: 75%–80%) when a more-rigorous research laboratory method (laboratory BAP culture plus broth culture) was used,11 which illustrates the impact of culture methods on assessment of sensitivity.
If the diagnostic goal is to maximize identification of patients infected with GAS, then our data indicate that negative office test results, including throat culture results, should be confirmed with BAP cultures processed in a hospital or commercial laboratory. We do not advocate this, because use of the office BAP culture as the definitive diagnostic test for GAS pharyngitis has not been implicated in missed treatment opportunities and subsequent outbreaks of acute rheumatic fever (ARF), invasive GAS disease, or increased prevalence of GAS pharyngitis and its suppurative complications. Moreover, it would increase costs, and the subsequent delay in receiving final results might result in increased prescribing of antibiotics for patients who ultimately have negative culture results,24 many of whom would complete a full course of antibiotic therapy.22
With the application of published guidelines to our data, the sensitivity of RADTs plus back-up office BAP cultures was 85.1% overall and >90% for patients with McIsaac scores of >2, which is another demonstration of spectrum bias. These results support current recommendations,2–4 including advice to avoid testing patients with symptoms more typical of viral illnesses.2,3
Inappropriate or unnecessary use of antibiotics is common and contributes to the development of resistance to antimicrobial agents by many species of bacteria.33–35 Rates of antibiotic prescribing for pharyngitis, that is, 53% to 69% among children24 and 62% among patients of all ages seen by primary care physicians,34 are much higher than would be expected if only laboratory-confirmed GAS pharyngitis were treated with antibiotics by US physicians. McCaig et al36 reported that the rate of antibiotic prescribing for pharyngitis in 1999–2000 was 738 prescriptions per 1000 physician visits. At best, 68% of surveyed physicians reported using an appropriate strategy for evaluation of children with pharyngitis, but 27% reported often or always treating before they had test results.22 Mangione-Smith et al37 found that clinical guidelines for evaluation were not followed for 25% of children treated with antibiotics for pharyngitis in 5 health plans. A study of adults with pharyngitis found that no established strategy was followed in 66% of visits; 80% of patients had a test performed and 17% had positive test results, but 47% received antibiotics.38 These studies indicate that many physicians are unable or unwilling to follow current pharyngitis management guidelines and apparently favor antibiotic treatment over diagnostic testing.
The currently recommended diagnostic approach is intended to identify accurately patients with GAS pharyngitis, so that they can be treated with antibiotics. Diagnosis of GAS pharyngitis has been driven for 6 decades by the desire to prevent ARF, which is still considered by expert panels and 95% of clinicians as an important reason to identify and to treat children.2–4,22 Prevention of suppurative complications, symptom relief, and reduction of contagiousness were each chosen as reasons by ∼70% of pediatricians and family practitioners.22 Physicians clearly have incorporated prevention of ARF and other sequelae into their thinking about pharyngitis management. However, rates of ARF, suppurative complications of GAS pharyngitis, and invasive disease are quite low in the United States.39,40 In fact, the decrease in ARF in the United States began before effective antibiotics were available,39 and GAS serotypes responsible for ARF outbreaks in past decades either are rare or have vanished from North America.39,41 Modification of the management paradigm in the United States and other countries with low rates of ARF and suppurative complications may be warranted.
A significant conceptual change would be to shift the clinical goal from preventing rare GAS complications to minimizing unnecessary or inappropriate use of antibiotics by recognizing patients unlikely to have a GAS infection and not testing them. Data from adults indicate that more-selective testing alone may reduce inappropriate use of antibiotics.38 Our data indicated that McIsaac scores of ≤2 were associated with decreased likelihood of GAS pharyngitis (determined on the basis of isolation of GAS in the laboratory BAP culture). Limiting throat swabs to patients with scores of >2 would eliminate testing for the large proportion of patients with viral pharyngitis and some patients more likely to be GAS carriers; these patient groups do not benefit from antibiotics commonly used for GAS pharyngitis42 but often receive them.22,24,34,36–38 Our data indicated that selective testing using an office-based RADT, with office-based BAP culture confirmation of negative test results, would increase diagnostic sensitivity to ∼90%, but we think that the emphasis should be on limiting antibiotic exposure, rather than maximizing GAS identification. Selective testing is currently suggested by expert bodies2,3 and has been emphasized by Linder et al38 and Centor and Cohen.43 It could be incorporated into community-based interventions aimed at decreasing inappropriate antibiotic prescribing,44 and we think that it should be a focus of physician education.
An additional dramatic (and likely controversial) change in management could be reliance on the result of the initial office test (whether RADT or BAP culture) even if it is negative, ignoring the relatively low sensitivity of both office-based diagnostic tests. In this paradigm, no back-up testing would be performed routinely and antibiotics would be prescribed only for patients with positive test results. Because 90% of US pediatricians and family physicians have access to RADTs and 93% perform them on site,22 the vast majority of patients tested would receive a diagnosis immediately. This approach would be a major departure from >20 years of recommended practice, and additional studies are needed before it can be endorsed.
The risk that either of these approaches would be associated with increased rates of ARF is likely low. The risk for an increase in invasive infections also seems to be low, because most cases of invasive disease in developed countries are not associated with pharyngitis.32,45 The effect on rates of suppurative complications is not known. In developing countries or areas with high rates of ARF and other GAS complications, the enormous burden of GAS disease alters diagnostic and treatment considerations.46
Our results may have been affected by several factors. The physicians in the participating pediatric practices were instructed to evaluate patients with pharyngitis in their ordinary manner. We did not attempt to alter their management strategies, swab techniques, or office laboratory procedures, except that the same CLIA-waived study RADT was used in all offices. We did not review or modify the specific BAP culture methods used in each office, because each was approved and monitored according to the provisions of the CLIA; the results reflect the “real world” of pediatric POL performance. Our results might have been different if another RADT had been used. There are few data that address the possibility that waived and nonwaived RADTs perform differently in POLs.1
Although the RADT manufacturer indicated that the RADT should not be affected, it is possible that streaking the BAP with swab A before the RADT was performed decreased the sensitivity of the RADT. Discordant BAP culture results can affect the apparent sensitivity of the office BAP culture. Random swab discordance, favoring neither swab A nor swab B, should have resulted in similar frequencies of office-negative/laboratory-positive and office-positive/laboratory-negative cases. Our findings do not seem to be attributable to random swab discordance.
The office RADT results were separated from the office BAP culture results to minimize the possibility of the office BAP culture readings being influenced by knowledge of the RADT results. We are not certain of the effectiveness of this effort. During the course of this investigation, we discovered notations on some data collection forms that indicated that the RADT was used to evaluate suspected colonies on the office BAP culture. Personnel from 5 offices later acknowledged that they sometimes used the RADT for this purpose, but usually they did not document this procedure. We do not know how often this was done, but it might have affected the number of office BAP cultures reported as positive. Such off-label use of the RADT has not been evaluated.
The assessment of spectrum bias is not straightforward. We used the McIsaac score to stratify our patients, but other clinical scores or other means of determining the pretest likelihood of GAS pharyngitis might have altered the results. A threshold McIsaac score of >2 identified 71% of our patients with true-positive results and might not be optimal. Decreasing the threshold score to >1 in this study would increase sensitivity to 93% by including 87% of patients (1606 of 1848 patients), without greatly improving the negative predictive value (83% for McIsaac scores of ≤1; data not shown). A laboratory test is required to diagnose GAS pharyngitis reliably.
This large prospective study found that the sensitivity of throat cultures performed in the office was significantly greater than the sensitivity of office RADTs. However, both the office RADT and the office BAP culture demonstrated relatively low sensitivity overall. The RADT, office BAP culture, and office BAP culture confirmation of negative RADT results all exhibited spectrum bias, which supports selective testing of patients with pharyngitis. In areas with low rates of GAS complications, such as North America and Western Europe, the diagnostic paradigm for pharyngitis should emphasize selective swabbing to avoid testing of patients who are unlikely to have GAS pharyngitis and avoidance of antimicrobial overuse through treatment only of patients with positive test results.
This work was supported by a grant from Quidel (San Diego, CA). Quidel did not design this study, collect or analyze the data, or participate in any way in manuscript preparation or review.
We thank the office staff members and pediatricians of the 6 participating pediatric practices: North Shore Pediatrics (Chicago, IL), Town and Country Pediatrics (Chicago, IL), DuPage Pediatrics (Darien, IL), Anderson Hills Pediatrics (Cincinnati, OH), Landen Lake Pediatrics (Mason, OH), and West Side Pediatrics (Cincinnati, OH).
- Accepted May 19, 2008.
- Address correspondence to Robert R. Tanz, MD, Children's Memorial Hospital, Division of General Academic Pediatrics, 2300 Children's Plaza, Box 16, Chicago, IL 60614. E-mail:
Financial Disclosure: Dr Shulman is on the medical advisory board of Quidel Corp; the other authors have no financial relationships relevant to this article to disclose.
This work was presented in part at the 4th World Congress of the World Society for Pediatric Infectious Diseases, September 6, 2005, Warsaw, Poland; the XVIth Lancefield International Symposium on Streptococci and Streptococcal Diseases, September 26, 2005, Cairns, Queensland, Australia; and the annual meeting of the Pediatric Academic Societies, April 29, 2006, San Francisco, CA.
What's Known on This Subject
RADTs for GAS are considered insufficiently sensitive to be used alone. Therefore, throat culture is recommended to confirm negative RADT results. Previous studies have demonstrated that RADT results are affected by the pretest likelihood of GAS pharyngitis.
What This Study Adds
The sensitivity of the RADT is fairly low, as is the sensitivity of the office BAP culture. Both tests perform better among patients with greater pretest likelihood of strep throat. The results support selective testing of patients with pharyngitis.
- ↵Gerber MA, Shulman ST. Rapid diagnosis of pharyngitis caused by group A streptococci. Clin Microbiol Rev.2004;17 (3):571– 580
- ↵American Academy of Pediatrics, Committee on Infectious Diseases. Group A streptococcal infections. In: Pickering LK, ed. Red Book: 2006 Report of the Committee on Infectious Diseases. 27th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2006:612–613
- ↵Bisno AL, Gerber MA, Gwaltney JM Jr, Kaplan EL, Schwartz RH. Practice guidelines for the diagnosis and management of group A streptococcal pharyngitis. Clin Infect Dis.2002;35 (2):113– 125
- ↵Dajani A, Tuabert K, Ferrieri P, et al. Treatment of acute streptococcal pharyngitis and prevention of rheumatic fever: a statement for health professionals. Pediatrics.1995;96 (4):758– 764
- ↵Gerber MA. Diagnosis of pharyngitis: methodology of throat culture. In: Shulman ST, ed. Pharyngitis: Management in an Era of Declining Rheumatic Fever. New York, NY: Praeger; 1984:61–72
- Kaplan EL, Top EH, Dudding BA, Wannamaker LW. Diagnosis of streptococcal pharyngitis: differentiation of active infection from the carrier state in symptomatic children. J Infect Dis.1971;123 (5):490– 501
- ↵Kaplan EL, Couser R, Huwe BB, McKay BB, Wannamaker LW. Significance of quantitative salivary cultures for group A and non-group A β-hemolytic streptococci in patients with pharyngitis and their family contacts. Pediatrics.1979;64 (6):904– 912
- ↵Edmonson MB, Farwell KR. Relationship between the clinical likelihood of group A streptococcal pharyngitis and the sensitivity of a rapid antigen-detection test in a pediatric practice. Pediatrics.2005;115 (2):280– 285
- ↵Kurtz B, Kurtz M, Roe M, Todd J. Importance of inoculum size and sampling effect in rapid antigen detection for diagnosis of Streptococcus pyogenes pharyngitis. J Clin Microbiol.2000;38 (1):279– 281
- ↵Hall MC, Kieke B, Gonzales R, Belongia EA. Spectrum bias of a rapid antigen detection test for group A β-hemolytic streptococcal pharyngitis in a pediatric population. Pediatrics.2004;114 (1):182– 186
- Gieseker KE, Roe MH, MacKenzie T, Todd JK. Evaluating the American Academy of Pediatrics diagnostic standard for Streptococcus pyogenes pharyngitis: backup culture versus repeat rapid antigen testing. Pediatrics.2003;111 (6). Available at: www.pediatrics.org/cgi/content/full/111/6/e666
- ↵Park SY, Gerber MA, Tanz RR, et al. Clinicians' management of children and adolescents with acute pharyngitis. Pediatrics.2006;117 (6):1871– 1878
- ↵Pillai S, Dennis L, Schwartz B, Ciesielski C, Shulman S. How accurate are pediatrician office laboratories (POLs) in identification of group A β streptococci (GABS) from throat swabs? In: Martin DR, Tagg JR, eds., Streptococci and Streptococcal Diseases: Entering the New Millennium: Proceedings of the XIVth Lancefield International Symposium on Streptococci and Streptococcal Diseases. Porirua, New Zealand: Securacopy; 2000:513–517
- ↵McIsaac WJ, White D, Tannenbaum D, Low DE. A clinical score to reduce unnecessary antibiotic use in patients with sore throat. CMAJ.1998;158 (1):75– 83
- ↵McIsaac WJ, Goel V, To T, Low DE. The validity of a sore throat score in family practice. CMAJ.2000;163 (7):811– 815
- ↵Gonzales R, Malone DC, Maselli JH, Sande MA. Excessive antibiotic use for acute respiratory infections in the United States. Clin Infect Dis.2001;33 (6):757– 762
- ↵Shulman ST, Stollerman G, Beall B, Dale JB, Tanz RR. Temporal changes in streptococcal M protein types and the near-disappearance of acute rheumatic fever in the US. Clin Infect Dis.2006;42 (4):441– 447
- ↵Stevens DL. Life-threatening streptococcal infections: scarlet fever, necrotizing fasciitis, myositis, bacteremia, and streptococcal toxic shock syndrome. In: Stevens DL, Kaplan EL, eds. Streptococcal Infections: Clinical Aspects, Microbiology, and Molecular Pathogenesis. New York, NY: Oxford University Press; 2000:163–179
- ↵Shulman ST, Tanz RR, Kabat W, et al. North American group A streptococcal pharyngitis serotype surveillance. Clin Infect Dis.2004;39 (3):325– 332
- ↵Gerber MA, Tanz RR, Kabat W, et al. Potential mechanisms for failure to eradicate Group A streptococci from the pharynx. Pediatrics.1999;104 (4):911– 917
- ↵Darenberg J, Luca-Harari B, Jasir A, et al. Molecular and clinical characteristics of invasive group A streptococcal infection in Sweden. Clin Infect Dis.2007;45 (4):450– 458
- Copyright © 2009 by the American Academy of Pediatrics