Potential Mechanisms for Failure to Eradicate Group A Streptococci From the Pharynx
Objective. To investigate the relative efficacy of orally administered cefadroxil and penicillin V in the treatment of group A streptococcal (GABHS) pharyngitis and the mechanism(s) responsible for failure of antimicrobial therapy to eradicate GABHS from the pharynx.
Study Design. A prospective, randomized clinical trial was conducted in four pediatric offices in which 462 patients with acute pharyngitis and positive culture for GABHS were randomly assigned to receive cefadroxil (n = 232) or penicillin V (n = 230).
Results. Bacteriologic treatment success rates for patients in cefadroxil and penicillin groups were 94% and 86%, respectively. However, among patients classified clinically as likely to have bona fide GABHS pharyngitis, there was no difference in bacteriologic treatment success rates in cefadroxil and penicillin groups (95% and 94%, respectively). Among patients classified clinically as likely to be streptococcal carriers, bacteriologic treatment success rates in cefadroxil and penicillin groups were 92% and 73%, respectively. The presence of β-lactamase and/or bacteriocin-producing pharyngeal flora had no consistent effect on bacteriologic eradication rates among patients in either penicillin or cefadroxil treatment groups or among patients classified as having either GABHS pharyngitis or streptococcal carriage.
Conclusions. Neither β-lactamase nor bacteriocin produced by normal pharyngeal flora are related to bacteriologic treatment failures in GABHS pharyngitis. Cefadroxil seems to be more effective than penicillin V in eradicating GABHS from patients classified as more likely to be streptococcal carriers. However, among patients we classified as more likely to have bona fide GABHS pharyngitis, the effectiveness of cefadroxil and penicillin V seems to be comparable.
- GABHS =
- group A β-hemolytic streptococcal (streptococcus) •
- MDR =
- medication dosing record
Penicillin has been used to treat group A β-hemolytic streptococcal (GABHS) pharyngitis for more than 40 years and is currently recommended by both the American Heart Association and the American Academy of Pediatrics as the drug of choice for this illness.1 ,2 However, in recent years, some have reported an increasing incidence of treatment failures with penicillin therapy, with failure rates as high as 25% with intramuscularly administered benzathine penicillin G and as high as 30% with orally administered penicillin.3–5 It has also been suggested that penicillin may now be less effective in treating GABHS pharyngitis than in the past and that cephalosporins may be more effective than orally administered penicillin in treating this illness.5 ,6
A number of mechanisms have been proposed to explain penicillin treatment failures in acute GABHS pharyngitis, including resistance or tolerance of GABHS to penicillin; reduced interference of GABHS by normal pharyngeal flora; and presence of β-lactamase-producing organisms in the upper respiratory tract.7 These penicillin treatment failures have also been attributed to streptococcal carriers inadvertently treated for acute GABHS pharyngitis.7 There is no evidence that GABHS have become more resistant to penicillin,8 ,9 and a role for penicillin tolerance in GABHS pharyngitis treatment failures has not been established clearly.10 ,11 Although some published reports have suggested that bacterial interference and β-lactamase produced by normal pharyngeal flora may influence the outcome of penicillin treatment of GABHS pharyngitis, these findings have been inconsistent and inconclusive.10–15 Because penicillin is not effective in eradicating GABHS from the upper respiratory tracts of chronic streptococcal carriers, it has been suggested that the apparently high treatment failure rates with orally administered penicillin therapy for GABHS pharyngitis reported in several recent studies may be the result of carrier contamination of the study population.16 ,17 However, this hypothesis requires additional corroboration.
Cefadroxil monohydrate is a first generation cephalosporin approved for use in the United States in 1979. It has been demonstrated to be safe and effective in the treatment of GABHS pharyngitis as a single daily dose of 30 mg/kg for 10 days.18 To investigate the potential superiority of orally administered cephalosporins over orally administered penicillin in the treatment of GABHS pharyngitis, we initiated a prospective, randomized, controlled study to compare the effectiveness of cefadroxil given once daily with penicillin V given three times daily. To clarify the mechanism(s) responsible for the failure of antimicrobial therapy to eradicate GABHS from the pharynx, we investigated the potential roles of β-lactamase and bacterial interference produced by normal pharyngeal flora on the ability of penicillin V or cefadroxil to eradicate GABHS from the upper respiratory tract. In addition, to clarify the possible role of chronic streptococcal carriage in penicillin treatment failures, we correlated the relative effectiveness of penicillin and cefadroxil in eradicating GABHS from the upper respiratory tract with whether the patient clinically was likely to have bona fide acute infection with GABHS or was more likely to be a chronic carrier of GABHS with intercurrent viral pharyngitis.
MATERIALS AND METHODS
Children between 3 and 18 years of age who presented to one of four private pediatric offices in Danbury CT, Bristol CT, Albany NY, or Valparaiso, IN, between March 1, 1992 and June 30, 1993 with clinical findings consistent with acute pharyngitis had a throat culture (or rapid antigen detection test) performed if their private pediatrician decided that such a test was clinically indicated. All patients with a positive antigen detection test or with GABHS isolated from a throat culture were considered for inclusion in the study unless: 1) they were known to be allergic to penicillins or cephalosporins; 2) they had taken antimicrobials within the previous 7 days; 3) they had a history of recurrent GABHS pharyngitis (3 or more microbiologically confirmed episodes of GABHS pharyngitis in the previous year); 4) they were known chronic carriers of GABHS; or 5) they were taking immunosuppressive therapy or were otherwise immunocompromised. After written informed consent was obtained, demographic and clinical data were collected and a physical examination was performed.
Each patient had two sterile, rayon-tipped swabs rubbed over the posterior portion of the pharynx and over both tonsils (tonsillar fossae). One swab was placed in an aerobic transport system (S/P Culturette, Marion Laboratories, Inc, Kansas City, MO) and the other swab was placed in an anaerobic transport system (Port-a-Cul, Becton Dickinson, Microbiology Systems, Cockeysville, MD). Both swabs were then transported by overnight courier to the infectious diseases laboratory at Children's Memorial Hospital (Chicago, IL) in the case of the specimens from Valparaiso, or to the University of Connecticut Health Center for specimens from Albany, Bristol, and Danbury. On arrival in the laboratory, both swabs were tested for the presence of GABHS as well as for β-lactamase and bacterial interference activity using modifications of previously described methods.12 ,19 ,20
Each swab (aerobic and anaerobic) was placed in 0.5 mL of brain-heart infusion broth and shaken vigorously for 3 minutes. Then 50 μL of the broth from the aerobic swab was placed on the surface of a 5% sheep blood agar plate and streaked for isolation of colonies. This plate was incubated overnight at 37°C in room air and examined for the presence of β-hemolytic streptococci. If no β-hemolytic streptococci were identified, the plate was reincubated overnight at room temperature. All β-hemolytic streptococci were grouped with the use of the Streptex test (Welcome Diagnostics, Dartford, England).
An additional 50 μL of the broth from the aerobic swab was streaked in a similar manner over each of two Levinthal agar plates. The two Levinthal agar plates were then incubated overnight at 37°C in room air and inspected for the presence of a uniform distribution of discrete colonies of the pharyngeal flora.
Bacterial Interference Assay
One of the Levinthal agar plates was then overlaid with 10 mL of 5% sheep's blood-brain-heart infusion agar to which a standardized suspension of GABHS (a stored isolate from a patient at Children's Memorial Hospital) had been added (∼5 × 103 cfu/mL) sufficient to completely hemolyze the blood in the overlay layer. This overlay plate was then incubated overnight at 37°C in room air and the inhibition of GABHS growth (as reflected by the amount of β-hemolysis in the overlay agar) was then compared with control plates (similar overlay plates with no organisms on underlying Levinthal agar plate). The same procedures were followed for the anaerobic swab except that all incubations took place for 48 hours in a GasPak pouch (Becton Dickinson, Cockeysville, MD).
The second Levinthal agar plate was overlaid with 10 mL of brain-heart infusion agar into which an alkaline phenol red-penicillin mixture had been incorporated and allowed to solidify. This overlay plate was then incubated for 20 minutes at 37°C in room air and any color change from purple/red to yellow when compared with a control plate (similar overlay plates with no organisms on underlying Levinthal agar plate) was assessed. The same procedures were followed for the anaerobic swab except that all incubations took place for 48 hours in a GasPak pouch.
The laboratory personnel performing the bacterial interference and β-lactamase assays were unaware of the treatment assignments.
Using a table of random numbers, patients were randomly assigned at each of the study centers to receive orally either cefadroxil monohydrate 30 mg/kg/d (maximum of 1000 mg/d) in a single daily dose for 10 days or penicillin V 25 to 50 mg/kg/d (maximum of 1000 mg/d) in three divided doses for 10 days. To assess compliance, the patient (or parent or guardian) was contacted by telephone on day 2 to 3 of treatment to verify that the correct amount of drug was being taken. The patient (or parent or guardian) was asked to dip a piece of filter paper into the patient's urine on the ninth day of antimicrobial therapy. This filter paper was allowed to air dry and was then mailed to either the Children's Memorial Hospital or the University of Connecticut Health Center Laboratory where it was assayed for antimicrobial activity using a modification of the technique of Markowitz and Gordis.21 In addition, each patient was given a medication dosing record (MDR) in which to record the date and time that each dose of antimicrobial therapy was given. Patients were instructed to return the MDR and any remaining medication (or empty medication bottles) to the investigator at the follow-up visit 2 to 12 days after completing antimicrobial therapy as an additional measure of compliance. For patients who did not return the MDR, the amount of drug taken was calculated by subtracting the amount of drug returned from the amount of drug dispensed. The primary measure of compliance was the presence or absence of antimicrobial activity in the urine-impregnated filter paper strips. For patients who did not return their filter paper strip, compliance was defined as at least 80% of the prescribed doses taken as indicated on the MDR. For those who did not return either their filter paper strip or their MDR, compliance was defined as at least 80% of the prescribed doses taken as determined by medication remaining in returned bottles. Patients who did not fulfill these criteria were considered noncompliant and were excluded from the analyses.
All patients were asked to return for a follow-up throat culture and physical examination 2 to 12 days after completing the antimicrobial therapy. At that time, duplicate throat swabs (aerobic and anaerobic) were again obtained from each patient and were processed in the same manner as at the initial visit. All isolates of GABHS from the initial and follow-up visits, regardless of the clinical status of the patient, were stored in Todd-Hewitt broth at −70°C and then sent to the World Health Organization Collaborating Center for Reference and Research on Streptococci at the University of Minnesota to be characterized by M serotyping, T agglutination pattern, and serum opacity reaction by established methods.22
Analysis of Outcomes
The most accurate way to distinguish children with acute GABHS pharyngitis from chronic streptococcal carriers with intercurrent viral pharyngitis is the analysis of acute and convalescent serum specimens for antibodies to group A streptococcal extracellular antigens (eg, antistreptolysin O).23 However, in the absence of serum specimens, the observation that certain clinical findings suggest GABHS as the cause of an episode of acute pharyngitis whereas other clinical findings suggest a viral agent allows classification of patients with acute pharyngitis as more likely to have acute GABHS pharyngitis or more likely to be a chronic streptococcal carrier.24–26Before the analysis of bacteriologic outcomes, enrolled patients were classified for the purposes of this study based on the presence or absence of certain clinical findings as being more likely to have bona fide acute GABHS pharyngitis or more likely to be a chronic carrier of GABHS with an intercurrent viral illness. GABHS-positive patients who had either tender cervical lymph nodes, tonsillar exudate, or tonsillar petechiae in the absence of cough, nasal congestion, and diarrhea at the initial visit were classified as more likely to be acutely infected with GABHS. GABHS-positive patients who did not have tender cervical nodes, tonsillar exudate, or tonsillar petechiae at the initial visit were classified as more likely to be streptococcal carriers. Patients who could not be placed into one of these groups based on their clinical findings were considered unclassified.
The analysis of outcomes in this investigation was limited to bacteriologic responses. Clinical response to therapy was not considered for several reasons: 1) reporting of clinical findings is often done in an inconsistent manner; 2) the physicians were not blinded to therapy allocation when performing the clinical evaluation, which is often subjective; 3) the clinical significance of persistent or recurrent clinical findings of GABHS pharyngitis, which are self-limited even without therapy, is difficult to determine; and 4) the clinical significance of persistent or recurrent clinical findings of pharyngitis in the absence of a positive throat culture, is also difficult to determine.
Only patients who satisfied all the inclusion and exclusion criteria for entry into the study, who had a pretreatment throat culture positive for GABHS, who were compliant with the assigned antimicrobial, and who had a valid follow-up bacteriologic assessment, were included in the evaluation of the bacteriologic response. Bacteriologic treatment failure was defined as the presence of the same serotype of GABHS on the follow-up and initial cultures. Isolation of a different serotype of GABHS on the follow-up culture was defined as a new acquisition. Eradication was defined as no GABHS present on the throat culture at the follow-up visit. Bacteriologic treatment success was defined as either eradication of GABHS or loss of the original GABHS serotype with acquisition of a new serotype of GABHS.
Assuming a bacteriologic treatment failure rate of 15% among the penicillin group and 5% among the cefadroxil group, we calculated a required sample size of at least 140 patients in each treatment group to have a β = 0.80 with an α = 0.05. Demographic and clinical characteristics of the sample groups were compared using analysis of variance for continuous variables and χ2statistics for categorical variables. Differences in the bacteriologic treatment success rates between sample groups were assessed using χ2 tests and Fisher's exact test. All tests were two-tailed with a significance level of 0.05. Ninety-five percent confidence limits are given for the observed percentages of patients responding.
A total of 462 patients were enrolled at the four study centers; 232 (50%) patients were randomized to receive cefadroxil and 230 (50%) to receive penicillin. Compliance was based on the results of the filter paper strip test for 86.6% of the patients, on the results of the MDR for 10.4% of the patients, and on the amount of drug returned for 0.6% of the patients. No compliance data were available for 2.4% of the patients and they were classified as noncompliant. The patients in the two treatment groups were similar with respect to compliance at each of the four study centers as well as for the entire study population.
The bacteriologic outcome was evaluable and compliance was confirmed in 374 (81%) of the enrolled patients (187 cefadroxil; 187 penicillin). See flow diagram (Fig 1) for details of patients excluded for noncompliance and for other reasons. The distribution of these patients by study center is shown in Table 1. The demographic characteristics (age, sex, race), weight, duration of illnesses, and the signs and symptoms present at the time of enrollment were similar for the bacteriologically evaluable compliant patients in the two treatment groups at each of the four study centers as well as for the entire study population.
The bacteriologically evaluable, compliant patients returned for a follow-up evaluation 2 to 12 days after completing antimicrobial therapy (86% returned between 3 and 7 days; penicillin group, 84% and cefadroxil group, 88%). Among these patients the overall bacteriologic treatment success rate was significantly higher for the patients in the cefadroxil group (94%) than for the patients in the penicillin group (86%) (P < .01; Table 2). Included in these treatment successes were 4 (2.1%) patients in the cefadroxil group and 3 (1.6%) patients in the penicillin group with the acquisition of new isolates of GABHS. However, when analyzed by study center, there were no differences in bacteriologic eradication rates at two centers (Danbury, Bristol), a significant difference (P < .05) at one center (Valparaiso), and a trend toward a significant difference (P < .07) at the fourth center (Albany) (Table 2). The bacteriologic treatment success rates in each treatment group were independent of the duration of illness at the time of enrollment for each of the four study centers as well as for the entire study population. When the same outcomes analyses were performed for all the eligible patients (intent-to-treat analysis), the results did not change.
There were 191 bacteriologically evaluable, compliant patients who were classified clinically as more likely to have bona fide acute infection with GABHS. Of these patients, 65% were so classified based on acute pharyngitis and a single additional clinical finding (ie, cervical adenitis, tonsillar exudate, or tonsillar petechiae) in the absence of cough, nasal congestion, and diarrhea; 32% were so classified based on two additional clinical findings; and 3% were so classified based on all three additional clinical findings. For patients who were classified clinically as likely to have bona fide acute infection with GABHS, there was no difference in the overall bacteriologic eradication rates for the patients in the cefadroxil and penicillin treatment groups (95% and 94%, respectively; Table 3). However, for the 152 patients clinically classified as likely to be streptococcal carriers, the overall bacteriologic eradication rate was significantly higher for the patients in the cefadroxil treatment group than for the patients in the penicillin treatment group (92% and 73%, respectively;P < .01; Table 3). Of the 374 evaluable, compliant patients, 31 (8%) could not be clinically classified as either likely acutely-infected or likely carriers; all 31 were treated successfully. A significantly greater proportion of the enrolled patients were classified as more likely to be acutely-infected at the two study centers where the efficacy of cefadroxil and penicillin seemed to be comparable (Danbury and Bristol, 77% and 75% more likely infected, respectively) compared with the two centers at which the efficacy of cefadroxil seemed to be significantly greater than penicillin (Valparaiso and Albany, 39% and 20% more likely infected, respectively; P < .01).
The differences in the bacteriologic eradication rates for the patients treated with cefadroxil compared with the patients treated with penicillin were independent of the presence of β-lactamase activity, the presence of bacteriocin activity, or the presence of both β-lactamase and bacteriocin activity in the normal pharyngeal flora before the initiation of antimicrobial therapy (Table 4). Similarly, the differences in the bacteriologic eradication rates for the patients treated with cefadroxil compared with the patients treated with penicillin were independent of the presence of β-lactamase activity, the presence of bacteriocin activity, or the presence of both β-lactamase and bacteriocin activity in the normal pharyngeal flora after completion of antimicrobial therapy (Table 5).
Among the patients within each treatment group (penicillin or cefadroxil), the presence of β-lactamase activity, the presence of bacteriocin activity, or the presence of both β-lactamase and bacteriocin activity either before or after antimicrobial therapy, had no consistent effect on the bacteriologic eradication rates. This was true for all patients as well as for the subset of patients likely to have bona fide GABHS pharyngitis and the subset likely to be streptococcal carriers (Tables 4 and 5).
There were no serious adverse events reported by the patients in either treatment group. The study drug was discontinued early because of adverse reactions in 7 patients, all treated with cefadroxil (vomiting 5, diarrhea 1, rash 1; P = .015 compared with penicillin).
Although penicillin is effective in the treatment of acute GABHS pharyngitis, it is less effective in eradicating GABHS from the upper respiratory tract of chronic streptococcal carriers; in contrast, cephalosporins seem to be effective.16 ,17 Because it is difficult to distinguish prospectively children with bona fide acute GABHS pharyngitis from chronic streptococcal carriers with an intercurrent viral pharyngitis, some streptococcal carriers are always included in clinical trials of antimicrobial therapy for GABHS pharyngitis. With less selective entry criteria a greater proportion of streptococcal carriers will be included in such studies. In many of the clinical trials in which cephalosporins seem to be more effective than penicillin in eradicating GABHS from the upper respiratory tract, relatively loose entry criteria may have led to considerable carrier contamination.17 Thus, the apparent superiority of cephalosporins when compared with penicillin in many of these studies may be a reflection of their greater ability to eradicate the streptococcal carrier state rather than actual superiority in the treatment of bona fide acute GABHS pharyngitis.17 Because the streptococcal carrier state is a relatively benign condition for both carriers as well as their contacts, it is difficult to justify the often substantial additional cost associated with the use of many cephalosporins merely to eradicate streptococcal carriage.17 ,23 In addition, many cephalosporins have a considerably broader antimicrobial spectrum than penicillin, and their use could select for increasing numbers of antimicrobial-resistant organisms.
There has been no consistent increase in the reported bacteriologic failure rate with orally administered penicillin therapy for GABHS pharyngitis.27 Some studies have reported bacteriologic failure rates with orally administered penicillin as low as 2% to 8%,15 ,28 whereas other studies have found rates as high as 20% to 35%.29 ,30 Although our results demonstrate that overall cefadroxil was significantly more effective than orally administered penicillin in eradicating GABHS from children with acute pharyngitis, at 2 of the 4 study centers, bacteriologic failure rates with orally administered penicillin therapy were only 2% and 8%, respectively, and similar to the rates with cefadroxil. Importantly, the 2 study centers with the lowest bacteriologic failure rates with orally administered penicillin therapy were the ones with the largest proportion of enrollees categorized as more likely to have had bona fide acute GABHS pharyngitis. The pediatricians at the 2 study centers with the highest treatment failure rates with orally administered penicillin therapy seemed to have had a lower threshold for obtaining throat swabs from patients presenting with acute pharyngitis.
Some normal pharyngeal flora (particularly Streptococcus salivarius) can interfere with colonization and growth of GABHS in the upper respiratory tract and might contribute to the natural defense against GABHS infections of the throat. Much of this interference seems to occur through the action of bacteriocins, which are protein or protein-complex antibiotics produced by a wide variety of bacterial species. Their activity is usually directed against strains of the same or closely related species.31 Crowe and co-workers12 showed that children who did not become colonized with GABHS had significantly greater anti-GABHS inhibitory activity in their pharyngeal flora and more often possessed inhibitor-producing pharyngeal organisms than children who became colonized with GABHS. In addition, they found that once children had ceased being colonized with GABHS, there was more inhibitory activity in their pharyngeal flora than either before or during colonization with GABHS. It was also shown that healthy young adults had a quantitative and qualitative decrease in their pharyngeal flora after orally administered penicillin therapy and an even more striking decrease in the anti-GABHS inhibitory activity of the flora.32 This decrease in interference activity persisted for >2 weeks after the pharyngeal flora had reverted to its pretreatment composition. These observations led to the suggestion that bacterial interference by the normal pharyngeal flora may influence the outcome of penicillin treatment of GABHS pharyngitis. However, the role of bacterial interference in treatment failures has not been clearly established.13 ,14 ,33 In this study, we found no evidence to indicate that bacterial interference produced by normal pharyngeal flora is related to bacteriologic treatment failures in GABHS pharyngitis.
Simon and Sukain34 first suggested thatStaphylococcus aureus in the normal pharyngeal flora might produce a β-lactamase that could interfere with the action of penicillin on GABHS and thus produce penicillin treatment failures. However, early investigations were unable to confirm this hypothesis.35 ,36 More recent studies demonstrated that a wide variety of anaerobic and aerobic constituents of the normal pharyngeal flora produce β-lactamases.37–39 However, the evidence that β-lactamase produced by these pharyngeal flora plays a significant role in treatment failures has been inconsistent and inconclusive.13 ,15 ,33 ,40 In this study, we found no evidence to indicate that β-lactamase produced by normal pharyngeal flora is related to bacteriologic treatment failures in GABHS pharyngitis.
Identification of the factor(s) responsible for treatment failures in GABHS pharyngitis will require further investigation. Our findings suggest that much of the apparent superiority of cefadroxil when compared with oral penicillin may reflect cefadroxil's greater ability to eradicate chronic streptococcal carriage rather than superiority in the treatment of bona fide acute GABHS pharyngitis. The variability in performance among centers in this as well as in earlier studies probably reflects the variable proportions of streptococcal carriers enrolled at each of these centers. The clinical scheme that we used to classify patients in this study as either more likely to have bona fide acute GABHS pharyngitis or more likely to be a chronic carrier of GABHS with an intercurrent viral illness can not be used to definitively make this distinction. However, we believe that use of this classification has been helpful in understanding our findings and in suggesting future areas of investigation. Additional research, using serologic data to distinguish more accurately streptococcal carriers from those with bona fide acute GABHS pharyngitis, is needed to corroborate our findings. Currently, the factors responsible for streptococcal carriage remain obscure, and additional research to elucidate these factors is also required.
At the present time, penicillin remains the drug of choice for the treatment of acute GABHS pharyngitis.1 ,2 In future comparative antimicrobial treatment trials for acute GABHS pharyngitis, care must be taken to minimize enrollment of streptococcal carriers by including only patients with clinical and epidemiologic features suggestive of GABHS as the etiology and excluding patients with clinical and epidemiologic features more suggestive of a viral etiology.
In summary, neither β-lactamase nor bacterial interference produced by normal pharyngeal flora are related to bacteriologic treatment failures for GABHS pharyngitis. Cefadroxil seems to be more effective than penicillin in eradicating GABHS from patients we classified as more likely to be chronic streptococcal carriers. However, among patients we classified as more likely to have bona fide acute GABHS pharyngitis, the effectiveness of cefadroxil and penicillin seem to be comparable.
This work was supported by a grant from Bristol-Myers Squibb.
We wish to acknowledge the following pediatricians without whose assistance this study would not have been possible: Martin Randolph, MD; Delbert Hodder, MD; Ranjit Pandit, MD; Nancy Holyst, MD; John Poncher, MD; Susan Bishop, MD; Alice Harrington, MD; Thomas Ludwig, MD; Jeffrey Miller, MD; Jeffrey Schumacher, MD; Herbert Abbott, MD; and Douglas Larsen, DO.
- Received December 21, 1998.
- Accepted April 8, 1999.
Reprint requests to (M.A.G.) NIAID/DMID/CRAB, Rm 3104, 6700-B Rockledge Dr, Bethesda, MD 20892-7630. E-mail
- Dajani A,
- Taubert K,
- Ferrieri P,
- Peter G,
- Shulman S
- ↵American Academy of Pediatrics, Committee on Infectious Diseases. 1997 Red Book. Report of the Committee on Infectious Diseases. 24th ed. Elk Grove Village, IL: American Academy of Pediatrics; 1997
- Smith TD,
- Huskins WC,
- Kim KS,
- Kaplan EL
- O'Callaghan CH,
- Morris A,
- Kirby SM,
- Shingler AH
- Rosen IG,
- Jacobson J,
- Rudderman R
- Markowitz M,
- Gordis L
- ↵Johnson DR, Kaplan EL, Srarmek J, et al. Laboratory Diagnosis of Group A Streptococcal Infections. Geneva, Switzerland: World Health Organization; 1996
- Govan JRW
- Sanders CC,
- Sanders WE Jr,
- Harrowe DJ
- Simon HJ,
- Sakain W
- Quie PG,
- Pierce HC,
- Wannamaker LW
- Reilly S,
- Timmis P,
- Beeden AG,
- Willis AT
- Copyright © 1999 American Academy of Pediatrics