Journal of Clinical LaboratoryAnalysis 5367-371 (1991)
Detection of Group A Streptococci by Aerobic Culture and a New Simplified lmmunoassay in Three Pediatric Practices and a Hospital Laboratory James A. Kellogg,' David A. Bankert,' Thomas D. Schonauer,* Robert C. L a n d i ~Allen ,~ S. Nus~baurn,~ and John S. Levisky'
'
Clinical Microbiology Laboratory, York Hospital, Springdale Pediatrics,2 York Pediatric Associates, Pediatric Health Care of Y ~ r kand , ~ Department of Behavioral Sciences, York College of Penn~ylvania,~ York, Pennsylvania Duplicate throat swabs for detection of group A streptococci were collected in three pediatric offices from 1,035 patients with symptoms of pharyngitis. In the collecting office and in the hospital laboratory, one swab from each patient was first inoculated to sheep blood agar (incubated at 35°C aerobically for 2 days) and then tested for group A streptococcal antigen by using the SMART enzyme immunoassay technique (New Horizons Diagnostics Corp) incubated for up to 24 hours. Group A streptococci were recovered in culture (from one or both swabs) and serologically identified from 444 (42.9%) of the patients. Pediatric offices numbers 1, 2, and 3 detected 84.47'0, 84.6%, and 82.2%, respectively, of their patients who had posiKey words:
direct detection. SMART, arouD A StreDtococci
INTRODUCTION Because of numerous recent reports of the reemergence of rheumatic fever, as well as the appearance of a toxic shocklike syndrome due to group A streptococci, ''. . . it is vitally important to make every effort to identify all streptococcal throat infections and treat them adequately" (1). The clinical value of a well-performed throat culture procedure is in the high predictive value of a negative result (PVN) and in the physician's ability to avoid indiscriminate use of antibiotics when Streptococcuspyogenes can be confidently ruled out as a cause of pharyngitis (2). Most of the rapid antigendetection methods studied to date, however, are insufficiently sensitive for detection of group A streptococci, resulting in the need to confirm each negative screen with culture (3,4). Because of a desire to find a more sensitive atternative to the previously reported group A streptococcal screens, a study of the new SMART (Sensitive Membrane Antigen Rapid Test, New Horizons Diagnostics Corp., Columbia, MD) colorimetric immunoassay was undertaken. Aerobically incubated sheep blood agar (SBA) cultures were used as ii reference standard during the current study both because this method has 0 1991 Wiley-Liss, Inc.
tive cultures (in the office and/or laboratory) by using their own culture system and 82.6%, 71 YO, and 84.9%, respectively, of these same patients by using the SMART technique. In the laboratory, SMART test sensitivity and specificity were 71.4% and 98.7%, respectively, after 5 minutes of test incubation. However, SMART test sensitivity improved to 86.5% after overnight incubation of the immunoassay and to 91.3% if the data from one defective lot of seven SMART production lots studied were excluded. SMART test results which are negative after 5 minutes of incubation should therefore be confirmed both by reincubationof the antigen test up to 24 hours and by culture.
been shown to detect at least 90-95% of the group A streptococci recovered from symptomatic patients (5-9) and because the procedure is preferred by many pediatricians for their office laboratories.
MATERIALS AND METHODS In three different pediatric offices, duplicate oropharyngeal specimens were collected simultaneously by the physicians on rayon-tipped swabs (Culturette, Marion Scientific Corp., Kansas City, MO) from 1,035 patients with symptoms of pharyngitis. The patients ranged in age from 1 to 40 years (mean = 8.0 years; median = 7 years) and included 1,030 children and five parents. One swab from each patient was processed within 4 hours of collection in the pediatric office and the remaining swab was sent by laboratory courier to the York Hospital Clinical Microbiology Laboratory for testing within 4 to 18 hours. Swabs were processed in a simReceived March 5 , 1991; accepted April 23, 1991. Address requests to Dr. Kellogg, Clinical Microbiology Laboratory, York Hospital, York,PA 17405.
368
Kellogg et at.
ilar fashion first for culture and then for antigen detection both by pediatric office personnel and microbiology laboratory technologists, with exceptions as noted. All culture media and reagents necessary for the study were supplied to the pediatric offices by the hospital laboratory. Training of pediatric office personnel for both culture and antigen assay technique was provided by two of us (J.A.K.; D.A.B.). The number of pediatricians was four in office number 1 and three each in offices numbers 2 and 3.
Culture Technique Each swab was inoculated to a 5% sheep blood agar plate (Baltimore Biological Laboratories, Cockeysville, MD) and the cultures were streaked for isolation. A 22 x 30 mm cover glass was then placed firmly onto the primary inoculum zone to reduce oxygen tension (10). Cultures were incubated aerobically (without increased carbon dioxide) at 35°C overnight, inspected for beta-hemolytic streptococcal colonies, and reincubated aerobically for an additional 24 hours if negative. Each isolate was tested with Streptex antisera (Wellcome Diagnostics, Research Triangle Park, NC) to streptococcal groups A and B. Both positive (group A and B streptococcus) and negative (group C or G streptococcus) cultures were used as serologic controls each day of testing. Recovery of group A streptococci was grouped semiquantitatively as 1 to 10, 11 to 50,51 to 100, or >100 colonieskulture. Aerobically incubated cultures had been performed in pediatric offices number 1,2, and 3 for 6,20, and 2 years, respectively. Pediatric office cultures were inoculated and read each day by the pediatricians. Streptex identification of isolates in their offices was performed by nurse practitioners (office 1) or the pediatricians (offices 2 and 3). The sensitivity of office or laboratory cultures for group A streptococci was defined as the percentage of the total number of culture-positive patients from an office whose cultures were positive at the office or at the laboratory.
Direct Antigen Detection After culture inoculation, each swab was tested for the presence of group A streptococcal antigen by using the two-step SMART enzyme immunoassay. The swab was first placed into a tube which contained 6 drops of a buffer, along with both an extraction enzyme (an n-acetyl hexoseaminidase) and a capture antibody bound to colloidal gold. The rabbit polyclonal antibody specifically recognizes the n-acetyl glucosamine determinant of the group A streptococcal cell wall. The antibody will form an immune complex with group A antigen (if present) within the swab. After incubation at 25°C for 5 minutes, the swab was removed from the tube and placed into the SMART test module for 5 more minutes at 25°C. During this time, solubilized immune complex could diffuse from the swab into the SMART module’s test disc, to which is bound a detector antibody to group A streptococcal
cell wall n-acetyl glucosamine. A positive result was indicated by the appearance of a pink or red color as the immune complex (bound to colloidal gold) bound to the detector antibody, The SMART test modules contained a negative control disc to help validate a positive result. However, they did not contain a positive control for validation of negative results. All SMART test modules with negative results after 5 minutes of incubation were reincubated at 25°C overnight (18 to 24 hours) and reread. SMART results were read in the pediatric offices by nurse practitioners (offices 1 and 3) or the pediatricians (office 2 ) . “False-positive” (culture-negative) SMART results found at one location (office or laboratory) were considered true positives if the culture results from duplicate swabs processed at the other location were positive for group A streptococci. In the hospital laboratory only, when the culture contained group A streptococci but the SMART result was negative, Streptex identification of the culture isolate was repeated. When the SMART result was positive but the culture was negative, the primary inoculum zone of the culture was blindly tested for group A streptococci with Streptex antisera and also subcultured to both an anaerobically incubated SBA plate and an aerobically incubated (5% to 10%C02) SBA medium containing trimethoprim-sulfamethoxazole. Results of SMART tests were unknown by hospital technologists reading the cultures. In addition, neither the pediatricians nor hospital laboratory personnel were aware of the other’s results on specimens from the same patient until all work on each specimen was completed. For statistical analysis of the results, the Z-test for differences between proportions was used (1 1). Directional (onetail) results were considered significant if the probability of the null hypothesis was 0.05 or less. Confidence intervals around the single-point proportion estimates were calculated from equation 1 of Simon ( 12).
RESULTS From September 6, 1989, until January 31, 1990, group A streptococci were recovered in culture on one or both of the duplicate specimens from 444 (42.9%) of 1,035 patients. Patients with positive cultures ranged in age from 1 to 40 years (mean = 7.4 years; median = 7 years).
Recovery in Culture Of 444 patients with cultures containing group A streptococci, 423 (95.3%) were found to have the organism on specimens cultured in the laboratory. Of these 423 positive cultures, 26 (6.1%) were detected only on the second day of culture incubation and 16 (3.8%) were recovered only from underneath the cover glass. Of the total number of patients from each pediatric practice with positive cultures either in the office, in the laboratory, or both, from 82.2% to 84.6% were detected on cultures performed in the three offices (no sig-
SMART lmmunoassay Performance
TABLE 1. Detection of Group A Streptococci on Duplicate Swabs in Pediatric Ofices and the Hospital Laboratory
Patients(n)
Total No. (%) patients with positive cultures”
Office l(290)
109(37.6)
Office 2(367)
149(40.6)
Office 3(378)
186(49.2)
Test Location Office Laboratory Office Laboratory Office Laboratory
No. (%)of patients with group A streptococci detected Culture
Antigen assayb
92(84.4) 109(100) 126(84.6) 142(95.3) 153(82.2) 172192.5)
90(82.6) 87(79.8) 106(71.1) 128(85.9) 158(84.9) 162(87.1)
a h t i e m were culture positive in the office, the laboratory. or both locations. bAntigen assays with negative results at 5 minutes were reread following overnight incubation. “False-positive’’SMART results found at one location (office or laboratory) were considered true-positive if the culture result on the duplicate swab at the other location was positive for group A streptococci.
369
TABLE 3. Performance of the SMART Antigen Assay Compared to Culture Performed on the Same Specimen Test location
No. Of positive cultures
(time of SMART incubation)
at test Positive Negative location Sensitivity Specificity result result
Laboratory (5 minutes) (1 to 24 hours) Office 1 (up to 24 hours) Office 2 (up to 24 hours) Office 3 (UD to 24 hours)
SMART immunoassay results, % Predictive values
423 423
71.4 86.5
98.7 89.7
91.4 85.3
83.3 90.6
92
93.5
92.9
86.0
96.8
126
80.2
90.9
82.1
89.8
153
88.9
79.1
74.3
91.3
Compared to the positive cultures detected in the pediatricians’ offices, SMART system sensitivity after overnight nificant differences, Table 1). However, there was a direct incubation of the device was 93.5% (86 of 92; 95% confirelationshipbetween the sensitivity of the office,culture method dence limit, 86.5% to 96.9%) in office 1, 80.2% (101 of for group A streptococci and the numbers of colonies of the 126; 95% confidence limit, 72.3% to 86.2%) in office 2, and organism recovered on laboratory cultures from the same 88.9% (136 of 153; 95% confidence limit, 82.9% to 92.9%) patients (Table 2). in office 3 (Table 3). SMART test specificity in the offices ranged from 79.1% to 92.9%. However, 4 (28.6%), 5 Detection Using Antigen Assay (22.7%), and 22 (46.8%) of the “false-positive” SMART results obtained from specimens with negative cultures in Of the 423 isolates of group A streptococci which were offices 1, 2, and 3, respectively, came from patients whose recovered on laboratory cultures, the SMART test detected duplicate swabs were culture-positive in the laboratory. As only 302 (sensitivity = 71.4%; 95% confidence limit, 66.9% previously mentioned, “false-positive’’ SMART results found to 75.5%) after 5 minutes of incubation of the immunoassay at one location were considered true-positives if the culture in the laboratory (Table 3). However, 366 of the streptococresults from duplicate swabs processed at the other location cal isolates were detected following additional (overnight) incuwere positive for group A streptococci. Accordingly, the relbation of the antigen-detection device (sensitivity = 86.5%; ative ability of culture and the SMART antigen assay, as per95% confidence limit, 82.9% to 89.5%). SMART test sensiformed in the pediatric offices or the laboratory, to detect group tivity was significantly increased ( P
Detection of Group A Streptococci by Aerobic Culture and a New Simplified lmmunoassay in Three Pediatric Practices and a Hospital Laboratory James A. Kellogg,' David A. Bankert,' Thomas D. Schonauer,* Robert C. L a n d i ~Allen ,~ S. Nus~baurn,~ and John S. Levisky'
'
Clinical Microbiology Laboratory, York Hospital, Springdale Pediatrics,2 York Pediatric Associates, Pediatric Health Care of Y ~ r kand , ~ Department of Behavioral Sciences, York College of Penn~ylvania,~ York, Pennsylvania Duplicate throat swabs for detection of group A streptococci were collected in three pediatric offices from 1,035 patients with symptoms of pharyngitis. In the collecting office and in the hospital laboratory, one swab from each patient was first inoculated to sheep blood agar (incubated at 35°C aerobically for 2 days) and then tested for group A streptococcal antigen by using the SMART enzyme immunoassay technique (New Horizons Diagnostics Corp) incubated for up to 24 hours. Group A streptococci were recovered in culture (from one or both swabs) and serologically identified from 444 (42.9%) of the patients. Pediatric offices numbers 1, 2, and 3 detected 84.47'0, 84.6%, and 82.2%, respectively, of their patients who had posiKey words:
direct detection. SMART, arouD A StreDtococci
INTRODUCTION Because of numerous recent reports of the reemergence of rheumatic fever, as well as the appearance of a toxic shocklike syndrome due to group A streptococci, ''. . . it is vitally important to make every effort to identify all streptococcal throat infections and treat them adequately" (1). The clinical value of a well-performed throat culture procedure is in the high predictive value of a negative result (PVN) and in the physician's ability to avoid indiscriminate use of antibiotics when Streptococcuspyogenes can be confidently ruled out as a cause of pharyngitis (2). Most of the rapid antigendetection methods studied to date, however, are insufficiently sensitive for detection of group A streptococci, resulting in the need to confirm each negative screen with culture (3,4). Because of a desire to find a more sensitive atternative to the previously reported group A streptococcal screens, a study of the new SMART (Sensitive Membrane Antigen Rapid Test, New Horizons Diagnostics Corp., Columbia, MD) colorimetric immunoassay was undertaken. Aerobically incubated sheep blood agar (SBA) cultures were used as ii reference standard during the current study both because this method has 0 1991 Wiley-Liss, Inc.
tive cultures (in the office and/or laboratory) by using their own culture system and 82.6%, 71 YO, and 84.9%, respectively, of these same patients by using the SMART technique. In the laboratory, SMART test sensitivity and specificity were 71.4% and 98.7%, respectively, after 5 minutes of test incubation. However, SMART test sensitivity improved to 86.5% after overnight incubation of the immunoassay and to 91.3% if the data from one defective lot of seven SMART production lots studied were excluded. SMART test results which are negative after 5 minutes of incubation should therefore be confirmed both by reincubationof the antigen test up to 24 hours and by culture.
been shown to detect at least 90-95% of the group A streptococci recovered from symptomatic patients (5-9) and because the procedure is preferred by many pediatricians for their office laboratories.
MATERIALS AND METHODS In three different pediatric offices, duplicate oropharyngeal specimens were collected simultaneously by the physicians on rayon-tipped swabs (Culturette, Marion Scientific Corp., Kansas City, MO) from 1,035 patients with symptoms of pharyngitis. The patients ranged in age from 1 to 40 years (mean = 8.0 years; median = 7 years) and included 1,030 children and five parents. One swab from each patient was processed within 4 hours of collection in the pediatric office and the remaining swab was sent by laboratory courier to the York Hospital Clinical Microbiology Laboratory for testing within 4 to 18 hours. Swabs were processed in a simReceived March 5 , 1991; accepted April 23, 1991. Address requests to Dr. Kellogg, Clinical Microbiology Laboratory, York Hospital, York,PA 17405.
368
Kellogg et at.
ilar fashion first for culture and then for antigen detection both by pediatric office personnel and microbiology laboratory technologists, with exceptions as noted. All culture media and reagents necessary for the study were supplied to the pediatric offices by the hospital laboratory. Training of pediatric office personnel for both culture and antigen assay technique was provided by two of us (J.A.K.; D.A.B.). The number of pediatricians was four in office number 1 and three each in offices numbers 2 and 3.
Culture Technique Each swab was inoculated to a 5% sheep blood agar plate (Baltimore Biological Laboratories, Cockeysville, MD) and the cultures were streaked for isolation. A 22 x 30 mm cover glass was then placed firmly onto the primary inoculum zone to reduce oxygen tension (10). Cultures were incubated aerobically (without increased carbon dioxide) at 35°C overnight, inspected for beta-hemolytic streptococcal colonies, and reincubated aerobically for an additional 24 hours if negative. Each isolate was tested with Streptex antisera (Wellcome Diagnostics, Research Triangle Park, NC) to streptococcal groups A and B. Both positive (group A and B streptococcus) and negative (group C or G streptococcus) cultures were used as serologic controls each day of testing. Recovery of group A streptococci was grouped semiquantitatively as 1 to 10, 11 to 50,51 to 100, or >100 colonieskulture. Aerobically incubated cultures had been performed in pediatric offices number 1,2, and 3 for 6,20, and 2 years, respectively. Pediatric office cultures were inoculated and read each day by the pediatricians. Streptex identification of isolates in their offices was performed by nurse practitioners (office 1) or the pediatricians (offices 2 and 3). The sensitivity of office or laboratory cultures for group A streptococci was defined as the percentage of the total number of culture-positive patients from an office whose cultures were positive at the office or at the laboratory.
Direct Antigen Detection After culture inoculation, each swab was tested for the presence of group A streptococcal antigen by using the two-step SMART enzyme immunoassay. The swab was first placed into a tube which contained 6 drops of a buffer, along with both an extraction enzyme (an n-acetyl hexoseaminidase) and a capture antibody bound to colloidal gold. The rabbit polyclonal antibody specifically recognizes the n-acetyl glucosamine determinant of the group A streptococcal cell wall. The antibody will form an immune complex with group A antigen (if present) within the swab. After incubation at 25°C for 5 minutes, the swab was removed from the tube and placed into the SMART test module for 5 more minutes at 25°C. During this time, solubilized immune complex could diffuse from the swab into the SMART module’s test disc, to which is bound a detector antibody to group A streptococcal
cell wall n-acetyl glucosamine. A positive result was indicated by the appearance of a pink or red color as the immune complex (bound to colloidal gold) bound to the detector antibody, The SMART test modules contained a negative control disc to help validate a positive result. However, they did not contain a positive control for validation of negative results. All SMART test modules with negative results after 5 minutes of incubation were reincubated at 25°C overnight (18 to 24 hours) and reread. SMART results were read in the pediatric offices by nurse practitioners (offices 1 and 3) or the pediatricians (office 2 ) . “False-positive” (culture-negative) SMART results found at one location (office or laboratory) were considered true positives if the culture results from duplicate swabs processed at the other location were positive for group A streptococci. In the hospital laboratory only, when the culture contained group A streptococci but the SMART result was negative, Streptex identification of the culture isolate was repeated. When the SMART result was positive but the culture was negative, the primary inoculum zone of the culture was blindly tested for group A streptococci with Streptex antisera and also subcultured to both an anaerobically incubated SBA plate and an aerobically incubated (5% to 10%C02) SBA medium containing trimethoprim-sulfamethoxazole. Results of SMART tests were unknown by hospital technologists reading the cultures. In addition, neither the pediatricians nor hospital laboratory personnel were aware of the other’s results on specimens from the same patient until all work on each specimen was completed. For statistical analysis of the results, the Z-test for differences between proportions was used (1 1). Directional (onetail) results were considered significant if the probability of the null hypothesis was 0.05 or less. Confidence intervals around the single-point proportion estimates were calculated from equation 1 of Simon ( 12).
RESULTS From September 6, 1989, until January 31, 1990, group A streptococci were recovered in culture on one or both of the duplicate specimens from 444 (42.9%) of 1,035 patients. Patients with positive cultures ranged in age from 1 to 40 years (mean = 7.4 years; median = 7 years).
Recovery in Culture Of 444 patients with cultures containing group A streptococci, 423 (95.3%) were found to have the organism on specimens cultured in the laboratory. Of these 423 positive cultures, 26 (6.1%) were detected only on the second day of culture incubation and 16 (3.8%) were recovered only from underneath the cover glass. Of the total number of patients from each pediatric practice with positive cultures either in the office, in the laboratory, or both, from 82.2% to 84.6% were detected on cultures performed in the three offices (no sig-
SMART lmmunoassay Performance
TABLE 1. Detection of Group A Streptococci on Duplicate Swabs in Pediatric Ofices and the Hospital Laboratory
Patients(n)
Total No. (%) patients with positive cultures”
Office l(290)
109(37.6)
Office 2(367)
149(40.6)
Office 3(378)
186(49.2)
Test Location Office Laboratory Office Laboratory Office Laboratory
No. (%)of patients with group A streptococci detected Culture
Antigen assayb
92(84.4) 109(100) 126(84.6) 142(95.3) 153(82.2) 172192.5)
90(82.6) 87(79.8) 106(71.1) 128(85.9) 158(84.9) 162(87.1)
a h t i e m were culture positive in the office, the laboratory. or both locations. bAntigen assays with negative results at 5 minutes were reread following overnight incubation. “False-positive’’SMART results found at one location (office or laboratory) were considered true-positive if the culture result on the duplicate swab at the other location was positive for group A streptococci.
369
TABLE 3. Performance of the SMART Antigen Assay Compared to Culture Performed on the Same Specimen Test location
No. Of positive cultures
(time of SMART incubation)
at test Positive Negative location Sensitivity Specificity result result
Laboratory (5 minutes) (1 to 24 hours) Office 1 (up to 24 hours) Office 2 (up to 24 hours) Office 3 (UD to 24 hours)
SMART immunoassay results, % Predictive values
423 423
71.4 86.5
98.7 89.7
91.4 85.3
83.3 90.6
92
93.5
92.9
86.0
96.8
126
80.2
90.9
82.1
89.8
153
88.9
79.1
74.3
91.3
Compared to the positive cultures detected in the pediatricians’ offices, SMART system sensitivity after overnight nificant differences, Table 1). However, there was a direct incubation of the device was 93.5% (86 of 92; 95% confirelationshipbetween the sensitivity of the office,culture method dence limit, 86.5% to 96.9%) in office 1, 80.2% (101 of for group A streptococci and the numbers of colonies of the 126; 95% confidence limit, 72.3% to 86.2%) in office 2, and organism recovered on laboratory cultures from the same 88.9% (136 of 153; 95% confidence limit, 82.9% to 92.9%) patients (Table 2). in office 3 (Table 3). SMART test specificity in the offices ranged from 79.1% to 92.9%. However, 4 (28.6%), 5 Detection Using Antigen Assay (22.7%), and 22 (46.8%) of the “false-positive” SMART results obtained from specimens with negative cultures in Of the 423 isolates of group A streptococci which were offices 1, 2, and 3, respectively, came from patients whose recovered on laboratory cultures, the SMART test detected duplicate swabs were culture-positive in the laboratory. As only 302 (sensitivity = 71.4%; 95% confidence limit, 66.9% previously mentioned, “false-positive’’ SMART results found to 75.5%) after 5 minutes of incubation of the immunoassay at one location were considered true-positives if the culture in the laboratory (Table 3). However, 366 of the streptococresults from duplicate swabs processed at the other location cal isolates were detected following additional (overnight) incuwere positive for group A streptococci. Accordingly, the relbation of the antigen-detection device (sensitivity = 86.5%; ative ability of culture and the SMART antigen assay, as per95% confidence limit, 82.9% to 89.5%). SMART test sensiformed in the pediatric offices or the laboratory, to detect group tivity was significantly increased ( P
