DIAGN MICROBIOL INFECT DIS 93 1993;17:93-96

BACTERIOLOGY

Comparison of Direct Specimen Testing Utilizing TestPack Strep A with Testing of Specimens Following a Two-Hour Broth Enrichment

Paul P. Bourbeau, Barbara J. Heifer, John P. Anhalt, and Donna W. Naumovitz

In a previous study, we demonstrated that a 2-h enrichment of throat swabs in broth followed by a direct fluorescent antibody test was significantly more sensitive than the Abbott TestPack Strep A for the detection of group-A streptococci. In this study, in an effort to simplify the technique, we utilized broth enrichment with an enzyme immunoassay detection method. A total of 1017 pharyngeal specimens were tested with TestPack Strep A (DTP), culture on a selective streptococcal agar, and

a 2-h broth enrichment in Todd-Hewitt broth (THB)followed by TestPack Strep A (ETP). Subculture of the THB was used to arbitrate discordant test results. The sensitivities and speci- ficities respectively were as follows: DTP (69% and 99%), ETP (90% and 98%), and SSA (97% and 100%). This en- richment method offers significantly greater sensitivity than direct testing for antigen yet still allows the potential for final- ized same-day reporting of results.

INTRODUCTION

Within the past decade, many laboratories have in- troduced rapid antigen tests for the direct detection of group-A streptococci (GAS) from throat swabs. Despite improvement in ease of performance and interpretation and a proliferation in the number and types of rapid antigen tests available, there has been no significant increase in test sensitivity. We share the opinion of some experts that the rapid antigen tests for GAS cannot be used yet as stand-alone tests; negative rapid antigen testing results should be con- firmed by culture (Bisno, 1991; Gerber et al., 1986; Kellogg et al., 1987; Vandepitte, 1991).

We have previously shown that a 2-h incubation

From the Department of Laboratory Medicine, Geisinger Medical Center, Danville, Pennsylvania, USA.

Address reprint requests to Dr. P.P. Bourbeau, Department of Laboratory Medicine, Geisinger Medical Center, Danville, PA 17822, USA.

Received 20 August 1992; revised and accepted 5 February 1993. © 1993 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/93/$6.00

in broth followed by a direct fluorescent antibody test produced sensitivity comparable to that of cul- ture (Anhalt et al., 1992). In this study, we continued to utilize the principle of broth enrichment but changed the detection method from a fluorescent antibody to a fixed-membrane enzyme immunoas- say, Abbott TestPack Strep A. The use of the TestPack Strep A abrogates the need for a fluorescence mi- croscope and a person experienced in interpretation.

MATERIALS A N D METHODS

All throat cultures from the emergency room and pediatric and family practice clinics of the Geisinger Medical Center in Danville, Pennsylvania, were col- lected on Culturette II swabs (Becton Dickinson Mi- crobiology Systems, Cockeysville, MD, USA). Upon receipt in the laboratory, the swabs were processed by a protocol that alternated weekly. For week 1, one swab was used for both culture and direct strep- tococcal antigen test. The second swab was used for a 2-h broth enrichment followed by direct strepto- coccal antigen test. For week 2, the protocol was

94 P.P. Bourbeau et al.

altered such that the swab used for culture was also used for the enrichment method, whereas the sec- ond swab was used for the direct antigen test only.

Culture was performed on group-A Selective Strep Agar (SSA) (Becton Dickinson). Incubation was at 35°C in 5%-10% CO2. Cultures were examined on the day following incubation, and those that were negative on day 1 were reincubated for another 24 h and then reexamined. All B-hemolytic streptococci were serotyped using the Streptex Latex A Reagent (Wellcome Reagents Division, Burroughs Wellcome, Research Triangle, NC, USA). Growth of f~-hemo- lytic streptococci on SSA plates was semiquantita- tively measured by growth in quadrants as follows: one or two quadrants, few organisms; three quad- rants, moderate number of organisms; and four quadrants, many organisms.

The direct streptococcal antigen test utilized was the Abbott TestPack Strep A (Abbott Laboratories, Abbott Park, IL, USA). The test was performed fol- lowing the manufacturer's instructions.

The enrichment technique was performed once daily in the laboratory. All swabs used for enrich- ment were refrigerated at 4°C until tested. Each swab was placed in I ml of Todd-Hewit t Broth (THB) and incubated at 35°C for 2 h. Following incubation, the swabs were wrung on the inside of the tubes and discarded. After mixing, 50 ~1 of the THB were plated onto a sheep blood agar (SBA) plate. (The inoculum on the SBA plate was streaked out, plates were in- cubated aerobically at 35°C for up to 2 days, and 13- hemolytic colonies were identified in the same man- ner as for the primary SSA plates.) The remainder of the THB was centrifuged for 5 rain at 700 g, and the supernatant was decanted and discarded. The pellet was resuspended in the remaining THB (50- 100 ~1). The TestPack Strep A extraction reagents were added directly to the test tube, and the antigen test was performed in the same manner as that for the direct specimen testing.

RESULTS

We tested 1017 specimens in this study. Of these, 230 specimens grew group-A streptococci on SSA and were defined as true positives. Using growth of GAS from the subculture of the THB to arbitrate differences between the DTP, ETP, and SSA results, an additional six specimens were defined as true positives, yielding a total of 236 true positives. These results are shown in Table 1. The sensitivity/specificity respectively of the three test results were as follows: DTP (69% and 99%), ETP (90% and 98%), and SSA (97% and 100%).

Table 2 presents a comparison of the sensitivity of DTP and ETP as a function of the amount of growth

of B-hemolytic streptococci on SSA plate. When many colonies were present on the SSA plate, both DTP and ETP had a sensitivity of >90%. With lesser amounts of GAS, however, differences became ev- ident. DTP had sensitivities of 71% and 42% for mod- erate numbers of and few colonies, respectively, whereas ETP had sensitivities of 95% and 71% for moderate numbers of and few colonies, respectively.

D I S C U S S I O N

The use of rapid antigen tests for the detection of GAS presents a dilemma for clinicians and labora- tories alike. On the one hand, there is a desire to diagnose streptococcal pharyngitis rapidly to hasten the initiation of appropriate therapy. On the other hand, there is a realization that, due to the inade- quate sensitivity of the rapid antigen detection meth- ods, all negative test results should be confirmed by culture, which requires up to 2 days for a definitive test result (Bisno, 1991; Gerber et al., 1986; Kellogg et al., 1987; Vandepitte, 1991).

We previously demonstrated that a 2-h incubation of throat swabs in THB followed by a direct flu- orescent antibody test for GAS was significantly more sensitive than the TestPack Strep A performed di- rectly from throat swabs (Anhalt et al., 1992). How- ever, the fluorescent antibody enrichment method (FA/EN) requires technical expertise and equipment not available in many laboratories. In addition, FA/EN must be batched to be performed cost- effectively, in part due to the controls that are utilized with each test run.

In this study, we wanted to know how the TestPack A--a product marketed as a direct antigen tes t - - would perform if utilized with a broth culture in- cubated for 2 h. The results of this study indicate 31% more true-positive tests with the ETP method compared with the DTP method. The sensitivity of ETP was 90%, whereas the sensitivity of SSA culture for 48 h was 97%. Although the difference is statis- tically significant (P < 0.01), we believe that the increased sensitivity of ETP compared with DTP may justify the use of ETP alone.

We realize that, although the ETP method offers the potential for same-day test results, it does not provide a truly rapid result, that is, a result while the patient is still in the office or clinic. Nonetheless, some laboratories, including ours, have found that the labor demands are prohibitive for providing dis- crete rapid antigen tests throughout the day (Lind- say et al., 1985). This results in batch testing the "rapid tests" several times each day. The results, consequently, are not usually available until after the patient has left the clinic.

Many physicians in our group practice are suffi-

Broth Enrichment for TestPack Strep A 95

TABLE 1 Comparison of Culture on Selective Streptococcal Agar (SSA) with DTP and ETP for Detection of Group-A Streptococci

No. of Result~

True False True False Sensitivity Specificity Sensitivity Test Positive Negative Negative Positive % % 95% CI b

SSA 230 6 781 97 94-98 DTP 162 74 771 10 69 99 62-74 ETP 213 23 763 18 90 ~ 98 85-93

aA total of 236 specimens were positive. bConfidence interval. cp < 0.01 compared with SSA (McNemar's test).

TABLE 2 Comparative Sensitivities of DTP and ETP as a Function of Amount of Growth on Selective Streptococcal Agar (SSA)

No. of Positive Results (% Sensitivity) at Indicated Growth Level a

Many Moderate Number Few Test Organisms b of Organisms Organisms d

SSA 60 115 55 DTP 56 (93) 82 (71) 23 (42) ETP 59 (98) 109 (95) 30 (71)

aFor definition of growth levels, see Materials and Methods. b26% of the total positive cultures. c50% of the total positive cultures. d24% of the total positive cultures.

ciently dissatisfied with the sensitivity of the rapid antigen tests that they do not order the test (un- published communications). These physicians have expressed significant interest in the concept of same- day reporting utilizing the enrichment method. We are planning to introduce a pilot project at one of our pediatric clinics for which the enrichment method will be utilized. For children meeting certain criteria, pharyngeal specimens will be collected each morn- ing during a "strep clinic." Specimens for the en- richment will be batched with results reported the same afternoon. During the pilot project, all ETP will be confirmed by a routine culture. Based upon the data in this study, we project a >97% accuracy rate for the enrichment method.

Use of this enrichment method precludes recov- ery or detection of bacteria other than GAS, and laboratories that presently culture for bacteria other than GAS from pharyngeal specimens may find this method unacceptable. Yet, for laboratories that now culture pharyngeal specimens only for GAS, this

method may be acceptable. It is also worth noting that the use of selective media such as SSA impedes recovery of much pharyngeal flora. In particular, SSA has been shown to inhibit the growth of both group-C and G ~-hemolytic streptococci (Evans and O'Neill, 1984).

We believe that the ETP method offers the po- tential for definitive same-day reporting for the di- agnosis of group-A streptococcal pharyngitis. The test as utilized in this study provides a sensitivity of 90% compared with a sensitivity of 97% for culture on SSA. Although that sensitivity may not be ac- ceptable for all microbiologists, it may be possible to increase the sensitivity of the technique by either a modification of the broth used for the enrichment or augmentation of the incubation period. However, the goal should be same-day reporting for the ma- jority of specimens tested in the laboratory. Lastly, this method should be adaptable to other direct an- tigen detection methods in addition to the TestPack Strep A.

REFERENCES

Anhalt JP, Heiter BJ, Naumovitz DW, Bourbeau PP (1992) Comparison of three methods for detection of group A

streptococci in throat swabs. J Clin Microbiol 30:2135- 2138.

96 P.P. Bourbeau et al.

Bisno AL (1991) Group A streptococcal infections and acute rheumatic fever. N Engl J Med 325:783-793.

Evans GL, O'Neill TE (1984) Development of an improved selective medium for the isolation of group A strep- tococci from throat cultures [abst C-136]. In Abstracts of the Annual Meeting of the American Society for Microbiol- ogy. Washington, DC: American Society for Microbi- ology, p 259.

Gerber MA, Randolph MF, Chanatry J, Wright LL, DeMeo KK, Anderson LR (1986) Antigen detection test for streptococcal pharyngitis: evaluation of sensitivity with respect to true infections. J Pediatr 108:654-658.

Kellogg JA, Landis RC, Nussbaum AS, Bankert DS (1987) Performance of an enzyme immunoassay test and an- aerobic culture for detection of group A streptococci in a pediatric practice versus a hospital practice. J Pediatr 111:18-21.

Lindsay AN, Swenson PH, Sanderson K (1985) Rapid di- agnosis of streptococcal pharyngitis [Letter]. J Pediatr 107:154.

Vandepitte J (1991) Streptococcal pharyngitis: a Belgian perspective. Pediatr Infect Dis 10:564-567.