Antigen detection test for streptococcal pharyngitis: Evaluation of sensitivity with respect to true infections

Antigen detection test for streptococcal pharyngitis: Evaluation of sensitivity with respect to true infections The clinical significance of false-negative results on antigen detection tests for group A E-hemolytic streptococcal (GABHS) pharyngitis (negative test results and positive culture) has yet to be determined. We recently compared the Culturette Brand Ten-Minute Group A Strep ID Kit with blood agar cultures in 313 patients with pharyngitis, 257 (82%) of whom had positive throat cultures for GABHS. The Culturette Brand test had a sensitivity of 88%, specificity of 96%, a positive predictive value of 99%, and negative predictive value of 64%. More than half of the false-negative Culturette Brand test results occurred in children with <10 GABHS colonies on throat culture (1+ culture). When these 1+ cultures were not considered positive, the sensitivity of the Culturette Brand test was 93%. The sensitivity of the Culturette Brand test increased with an increased degree of positivity of the corresponding throat culture. Of the 31 children with false-negative Culturette Brand test results, 14 (45%) had a significant streptoc o c c a l antibody response; of the 224 children with true-positive Culturette Brand test results (positive test results and positive culture) from whom serologic data were available, 114 (51%) had a significant streptococcal antibody response. This difference is not statistically significant. These findings suggest that almost half of patients with false-negative results on antigen detection tests for GABHS pharyngitis have true infections (positive culture and antibody rise) and are not merely streptococcal carriers. (J PEDIArR1986;108(I):654-658) Michael A. Gerber, M.D., Martin F. Randolph, M.D., Julie C h a n a t r y , B,S., Laura L. Wright, B.S., Kathleen K. D e M e o , B.S., a n d Lisa R. Anderson, B.A. From Department of Pediatrics, University of Connecticut School of Medicine, Farmington

Recently, several commercial antigen detection tests for the rapid identification of group A ~3-hemolytic streptococci directly from throat swabs have been developed. 1 As yet, there are few published data regarding the accuracy of most of these tests. The commercial ADT that have received the most extensive evaluations are two latex agglutination procedures: the Directigen Group A Strep Test Kit (Hynson, Westcott, and Dunning, Baltimore)and the Culturette Brand Ten-Minute Group A Strep ID Kit (Marion Scientific, Kansas City, Mo,). Several investigaSupported by a grant from Marion Laboratories, Inc. Submitted for publication Oct. 22, 1985; accepted Dec. 17, 1985. Reprint requests: Michael A. Gerber, M.D., Department of Pediatrics, University of Connecticut Health Center, Farmington, CT 06032. 654

tions of the accuracy of these two tests have demonstrated that they have a sensitivity of 83% to 95%, a specificity of 98% to 100%, a positive predictive value of 88% to 100%, and a negative predictive value of 93% to 99% when compared with blood agar cultures. 2-8 / I

ADT GABHS

Antigen detection tests Group A/3-hemolytic streptococci

I

Of the false-negative Directigen and Culturette Brand test results (negative ADT and positive throat culture) that were observed, a large percentage occurred in patients with < 10 GABHS colonies on blood agar culture (1 + cultures). When these 1+ cultures were not considered positive, the sensitivity of both the Directigen and Culturette Brand tests increased to 95% to 100%. 2-6 There has been a great deal of debate about the significance of throat cultures

Volume 108 Number 5, Part 1

Antigen detection test f o r streptococcal pharyngitis

T a b l e I. Comparison of Culturette Brand test with

blood agar culture Culturette Brand test

Positive Negative Total

I.U FI-IM

655

I00" 90

Blood agar culture Positive

Negative

Total

226 31 257

2 54 56

228 85 313

Sensitivity= 88% (226/257); specificity= 96% (54/56); Positive predictivevalue= 99% (226/228); Negativepredictivevalue= 64% (54/85)

8070-

IJ_

__.z 40 --n.~m z l.iJ

with < 10 GABHS colonies per plate. Some have suggested that most of these patients are merely streptococcal carriers and not truly infected.9,~~ Others believe that the differentiation of patients with streptococcal infections from those who are carriers cannot be made on the basis of the degree of positivity of the throat culture alone. I~ If the majority of patients with false-negative ADT results were carriers, then the sensitivity of these tests, in terms of identifying bona fide streptococcal infections, would be much higher. In addition, the clinical significance of these false-negative ADT results would be reduced, because streptococcal carriers are not at risk for developing either suppurative or nonsuppurative sequelae or of transmitting GABHS to others, j2 We recently completed an investigation designed to determine whether the majority of patients with falsenegative Culturette Brand test results were streptococcal carriers or truly infected. METHODS During winter-spring 1984-1985, 313 consecutive patients seen in a private pediatric office (M.F.R.) with clinical findings suggestive of GABHS pharyngitis were enrolled in the investigation after written, informed consent had been obtained. Throat swabs were obtained by simultaneously rubbing two sterile rayon-tipped swabs (Culturette II, Marion Scientific) over the posterior pharynx and both tonsils (or tonsillar fossae). One swab from each pair was immediately streaked onto a blood agar plate (trypticase soy agar with 10% sheep blood; BBL Microbiology Systems, BectomDickinson, Cockeysville, Md.); a baeitracin disk (Taxo A Disc; BBL Microbiology Systems) was placed on the primary inoculum; and the agar was stabbed in several areas. After overnight incubation in room air at 37 ~ C, the plates were examined for the presence of B-hemolytic streptococci and quantified as follows: 1+ culture, <10 colonies of ~-hemolytic streptococci per plate; 2+ culture, ->10 but <50 colonies; 3+ culture, >--50 colonies but not a pure culture; 4+ culture, >__50 colonies in pure culture. Beta-hemolytic streptococci susceptible to bacitracin were presumptively identified as

60-

20 I0 I+

2+

3+

4+

DEGREE OF CULTURE POSITIVITY

gig. 4. Relationship between degree of culture positivity and sensitivity of ADT. 1+ Culture, <10 colonies of GABHS per plate; 2+ culture, ->10 coloniesbut <50 colonies;3+ culture, ->__50 colonies but not a pure culture; 4+ culture, _>50 colonies in pure culture. group A. Grouping was later confirmed by either Lancefield precipitin13 or Streptex (Wellcome Reagents, Dartford, England) testing. The second swab from each pair was used to perform the Culturette Brand test according to the manufacturer's instructions. The results of all the Culturette Brand tests were read by the same person (J.C.), who was unaware of the results of the corresponding throat culture. Blood specimens were drawn from patients with positive throat culture, who were instructed to return in 4 weeks, at which time convalescent blood specimens were obtained. All serum specimens were stored at - 7 0 ~ C and later analyzed simultaneously in pairs for antistreptolysin O and antideoxyribonuclease B titers according to established methods. 14'~5 Patients with a positive throat culture for GABHS and a significant rise (-->2 dilution increment rise/>-0.2 log rise) in ASO or ADB titers were considered to have true streptococcal infections, whereas patients with a positive throat culture for GABHS without a significant rise in ASO or ADB titers were considered streptococcal carriers. Data were analyzed using the chi-square test and the Student t test. RESULTS Study patients ranged in age from 2 to 25 years (mean age 9.6 years, median age 9 years). Of the 313 patients in whom throat cultures were obtained, GABHS was isolated from 257 (82%). The Culturette Brand test yielded positive results in 226 of the 257 patients with positive throat cultures and negative results in 54 of the 56 patients with

656

Gerber et al.

The Journal of Pediatrics May 1986

ASO

ADB TRUE-POSITIVES

F A L S E - NEGATIVES

->0.8.

FALSE-NEGATIVES

>0.8-

9149149149149149149149149

! 9 9 ii

T R U E - POSITIVES

:.:'::....

el 9 9

iooQ

eeeeee

:."i %9e

aDeeee,leeo

eee

eeeeee

o,

Qg

0.5

0.5

ee

/ <~

oee*

e%%%

(1::

0.2

___

o.i-

tl.I

9 o e ~ 1 4 9 194 991 499 9 9 9 9 "

~=+0.14 9 ,,,

kZ

O-

UJ (.9 Z

-0.1 -0.2 -

"*:.: ..: ".

!.!'i'.:!S:: :::.': 9 9149 loooe

9

~_ I--

0, 0-

9 9

:.:.:.:............. ~: +oo8 9 = +0.02 9

eooooe

~

oeooeoo ee eo% oo eeeeeoooe ee

o o 9 9 9 eoe~oo o 9 9 ,%%%% 9 9 9

eeeoe

-W

-OJ

-

o%%eo

(.9 z

-0.2

-

oeeeeoeo

D eeoe

oe

oa,

ale

-0.5

eO

9 9 9 9 9 9

,%%%%%%%

Ii

<( "r" r


-0.5

0.2

hi

~=+0.12

...'.'+...:.:"

ell

11:

eeeeo eeeeee

_>-o.~

->-0.8

F i g . 2 . Comparison of changes in ASO titers (A log) in patients with false-negative ADT results (negative ADT and positive culture) and patients with true-positive ADT results (positive ADT and positive culture). Dotted line, mean titer change.

I I . Comparison of patients with false-negative and true-positive antigen detection test results

Table

False-

True-

negative ADT ( n = 3'1)

positive ADT (n = 226)

n

Mean age (yr) Male/female Duration of illness <24 hours Clinical findings Fever Cervical lymphadenitis Headache Abdominal pain Antibody rise

%

n

%

10.4 18/13 25/31

81

9.5 l 10/116 183/226

81

31/31 22/31 29/31 14/31 14/31

100 71 94 45 45

225/226 185/226 201/226 81/226 114/224

100 82 89 36 51

P values not significantfor these data.

negative throat cultures (Table I). The Culturette Brand test therefore had a sensitivity of 88%, specificity of 96%, positive predictive value of 99%, and negative predictive value of 64% when compared with blood agar cultures. Of the 31 patients with false-negative Culturette Brand test results (negative A D T and positive throat culture), 16 (52%) had a 1 + throat culture; of the 226 patients with a

iI !

ii

eeI%oe%%%%%ee

Fig. 3. Comparison of changes in ADB titers (4 log) in patients with false-negative ADT results (negative ADT and positive culture) and patients with true-positive ADT results (positive ADT and positive culture). Dotted line, mean titer change.

true-positive Culturette Brand test (positive A D T and positive throat culture), only 13 (6%) had a 1+ throat culture (P <0.001). When cultures with <10 G A B H S colonies per plate (1+ cultures) were not considered positive, the sensitivity of the Culturette Brand test increased to 93%. In addition, the sensitivity of the Culturette Brand test increased with an increased degree of positivity of the corresponding throat culture (Fig. 1). There was no significant difference between the patients with false-negative and true-positive Culturette Brand test results with respect to age, sex, duration of illness prior to treatment, or clinical findings at the time of the initial throat culture (Table II). Of the 31 patients with falsenegative Culturette Brand test results, 14 (45%) demonstrated a significant rise in ASO or ADB titers, and of the 224 patients with true-positive Culturette Brand test results from whom acute and convalescent ASO and ADB titers were available, 114 (51%) had a significant rise in ASO or ADB titers. There was no significant difference in the percentage of patients with false-negative or truepositive Culturette Brand test results who had a significant rise in ASO or ADB titers. In addition, there was no significant difference in the mean ASO or ADB titer changes in the patients in these two groups (Figs. 2 and

3).

Volume 108 Number 5, Part 1

DISCUSSION We found that significantly more patients with falsenegative Culturette Brand test results had 1+ positive throat cultures than did patients with true-positive CuP turette Brand test results, and that the sensitiv!ty of .the Culturette Brand test increased with an increased degree of positivity of the corresponding throat culture. However, there was little correlation between the degree of posit.ivity of the throat culture and the changes in streptococcal antibody titers or between the sensitivity of the Cu!turette Brand test and the changes in streptococcal antibody titers. These findings support the concept that the differentiation of Patients with streptococcal infections (positive throat culture and rise in strep.toeoceal antibodies) from those who are streptococcal carriers (positive throat culture and no rise in streptococcal antibodies) cannot be made on the basis of the degree of posit!vity of the blood agar culture alone. 1~ Furthermore, they suggest that almost half of patient s with false-negative ADT results for GABHS may have true streptococcal infections and are not merely streptococcal carriers. Several studies have demonstrated that appropriate antibiotic therapy suppresses the formation of streptococcal antibodies. 16,~7 However, other studies have failed to demonstrate a significant effect of antibiotic therapy on the streptococcal antibody response in patients with GABHS pharyngitisJ ~ ~8Ethical considerations would proelude the performance of an investigation in which streptococcal antibody titers were measured in patients with GABHS pharyngitis while antibiotic therapy was withheld, and this controversy will probably never be completely resolved. However, strep.tococcal antibody titers measured in patients with GABHS pharyngitis who are receiving appropriate antibiotic therapy may not reflect the maximal potential antibody response in these patients. Thus the antibody response we observed may not reflect the maximal potential antibody response in the study population. However, there is no reason to expect that the effect of antibiotic therapy on the antibody response would be any different in patients with fa!se-negative ADT results than in those with true-positive ADT results. Therefore, the similarity of strePtoCocCal antibody responses in our two grouPs remains a valid observation. The question remains as to what potential risk falsenegative ADT results for GABHS pharyngitis pose for the ind!vidual patient. False-negative results could mean that a child with GABHS pharyngitis will not receive proper therapy and, as a consequence, have a prolonged clinical illness and period of infectivity. In addition, this child would be at risk for acute rheumatic fever. The current risk for rheumatic fever in a child with untreated G A B H S

Antigen detection test for streptococcal pharyngitis

657

pharyngitis is not known. However, with the incidence of rheumatic fever in the United States being approximately 0.6 cases per 100,000,19 the risk would appear to be extremely small. Nevertheless, until more is known about the true clinical significance of false-negative ADT results, negatiye ADT results in patients in whom, on the basis of clinical and epidemiologic findings, there is a high degree of suspicion o f GABHS pharyngitis, should be confirmed with a blood agar culture: In contrast, given th e infrequency of false-posltive ADT results for GABHS pharyngitis, and their relative insignificance compared with falsenegative test results, a patient with positive A D T results could be given treatment without a blood agar culture being performed. REFERENCES 1. Radetsky M, Wheeler RC, Roe MH, etal. Comparative evaluation of kits for rapid diagnosis of group A streptococcal disease. Pediatr Infect Dis 1985;4:274. 2. Gerber MA, Spadaceinl LJ, Wright LL, et al. Latex agglutination tests for the rapid identification of group A streptococci directly from throat swabs. J PEDIA'rR 1984;105:702. 3. McCusker J J, McCoy EL, Young CL, et al. Comparison of Directigen Group A Strop Test with a traditional culture technique for detect!on of group A beta-hemolytic streptococci. J Clin Microbiol 1984;20:824. 4. Miller JM, Phillips HL, Graves RK, et al. Evaluation of the Directigen Group A Strop Test kit. J Clin Microbi01 1984;20:846. 5. Slifkin M, Gil GM. Evaluation of the Cutturette Brand Ten-Minute Group A Strop ID technique. J Clin Microbiol 1984;20:12. 6. Berkowitz CD, Anthony BF, Kaplan EL. Cooperative study of latex agglutination to identify group A streptococcal antigen on throat swabs in patients with acute pharyngitis. J PEDIA'rR 1985; 107:89. 7. Chang MJ, Mohla C. Ten-minute detection of group A streptococci in pediatric throat swabs. J Clin Microbiol 1985;21:258. 8. Miceika BG, Vitous AS, Thompson KD. Detection of group A streptococcal antigen directly from throat swabs with a ten-minute latex agglutination test. J Clin Microbiol 1985;211467. 9. Breese BB, Disney FA, Talpey WB, eta!. Beta-hemolytic streptococca! pharyngitis: the clinical ~and epidemiologic importance of the number of organisms found in cultures. Am J Dis Child 1970;119:18. 10. Stillerman M, Bernstein SH. Streptococcal pharyngitis: evaluation of clinical syndromes in diagnosis. Am J Dis Child 1961;10I:476. 11. Kaplan EL, TOp FH Jr, Dudding BA, et al. Diagnosis of streptococcal pharyngitis: differentiation of active infection from the carrier state in the symptomatic child. J Infect Dis 1971;123:490. 12. Kaplan EL. The group A streptococcal upper respiratory tract carrier state: an enigma. J pEDIATR 1980;97:337. 13. Lancefield RC. A serologic differentiation of human and

658

Gerber et al.

The Journal o f Pediatrics May 1986

other groups of hemolytic streptococci. J Exp Med 1933; 57:571. 14. Edwards EA. Protocol for micro antistreptolysin O determination. J Bacteriol 1964;87:1254. 15. Nelson J, Ayoub EM, Wannamaker LW. Streptococcal antideoxyribonuclease B: microtechnique determination. J Lab C!in Med 1968;71:867. 16. Brink WR, Rammelkamp CH Jr, Denny FW, et al. Effect of penicillin and aureomycin on the natural course of streptococcal tonsillitis and pharyngitis. Am J Med !951;10:300.

17. Kilbourne ED, Loge JP. The comparative effect of continuous and intermittent penicillin therapy on the formation of antistreptolysin in hemolytic streptococcal plaaryngitis. J Clin Invest 1948;27:418. 18. Siegel AC, Johnson EE, Stollerman GH. Controlled studies of streptococcal pharyngitis in a pediatric population. I. Factors related to the attack rate of rheumatic fever. N Engl J Med 1961;265:559. 19. Land MA, Bisno AL. Acute rheumatic fever: a vanishing disease in suburbia. JAMA 1983;249:895.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of the 1986 issues of THE JOURNAL OF PEDIATRICS are available to subscribers (only) from the Publisher, at a cost of $44.00 ($60.00 international) for Vol. 108 (January-June) and Vol. 109 (July-December), shipping charges included. Each bound volume contains subject and author indexes, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram, with the Journal name, volume number, and year stamped in gold on the spine. "Payment m u s t a c c o m p a n y all orders. Contact The C. V. Mosby Company, Circulation Department, 11830 Westline Industrial Dr., St. Louis, M O 63146, USA/800-325-4177, ext. 351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.