eLetters to:

ELECTRONIC ARTICLE: Karen E. Gieseker, Martha H. Roe, Todd MacKenzie, and James K. Todd Evaluating the American Academy of Pediatrics Diagnostic Standard for Streptococcus pyogenes Pharyngitis: Backup Culture Versus Repeat Rapid Antigen Testing Pediatrics 2003; 111: e666-e670 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

Independence Day?

Lance A Chilton   (1 July 2003)

Independence of samples

James K. Todd   (3 July 2003)

Similar Study Published 15 Years Ago Questions Management of a Negative Rapid Strep Test

Christopher B. White   (13 November 2003)

Independence Day? 1 July 2003
Lance A Chilton, pediatrician Univ of New Mexico Send letter to journal: Re: Independence Day? lancekathy{at}yahoo.com Lance A Chilton	The authors of this article have generated very useful data regarding rapid strep tests and the culture gold standard. I wonder if they would be interested in commenting on the lack of independence between results on two separate rapid strep tests. If a single rapid strep test is 87.6% sensitive, it follows, I believe, that if the second test's results are independent of the first, there should be a roughly 98.2% sensitivity of two tests [1.00- (1.00-.876)^2]. That the sensitivity of the rapid-rapid strategy is only 91.4% suggests that these results are not independent one of another. Can the authors speculate why?
Independence of samples 3 July 2003
James K. Todd, Physician Children's Hospital Send letter to journal: Re: Independence of samples todd.james{at}tchden.org James K. Todd	Dr. Chilton raises an important issue; are the results of 2 swabs taken simultaneously really independent? The answer is, no! If one swab is positive, the other should be too, but this isn't always true. This sampling effect probably results from both an unequal amount of group A Strep on each of the two swabs and the limits of detection of the method used. This is why a backup culture appears to be more effective than a backup rapid test in our study - culture uses a different method of detection. At the same time, culture was also the basis for our gold standard, so its real value may be over-estimated. Our solution would be to do a single test on a single, larger volume, very thorough sample. Current swabs and rapid tests might require some modification, but it would be great to eliminate the need for backup while, at the same time, increasing sensitivity above 95%. In the past, tests that claim to have accomplished this have often relied on insensitive gold standards.
Similar Study Published 15 Years Ago Questions Management of a Negative Rapid Strep Test 13 November 2003
Christopher B. White, Pediatrician Medical College of Georgia Send letter to journal: Re: Similar Study Published 15 Years Ago Questions Management of a Negative Rapid Strep Test cwhite{at}mail.mcg.edu Christopher B. White	To the Editor:   I read with interest the recent article in Pediatrics by Karen Gieseker and her colleagues at the University of Colorado regarding the need for backup throat cultures for initially negative rapid antigen detection tests (RADT) for Streptococcus pyogenes.[1]  Although the number of patients in Dr. Gieseker’s study are far greater, her results are very similar to the results we published 15 years ago.[2]  Since our results were not referenced in her paper, I would like to summarize them here.   Our strategy was similar except that we used two “double swabs” for each patient, instead of two single swabs.  The RADT we used (Icon Strep A®, Hybritech, Inc., San Diego, CA) was an enzyme immunosorbent assay (EIA) test chosen based upon an in-vitro study we previously performed comparing eight different commercially available tests for the detection of GAS in broth.[3]  All swabs were obtained by one of the two investigators, to ensure consistency in technique.  One swab from the first double swab was tested using the RADT and the other underwent a conventional throat culture.  At the end of each half-day, the second double swab was similarly tested: one swab using the RADT, and the other cultured.  Those who read the second RADT and the two throat cultures were blinded to source of the swab and unaware of any of the results for the other swabs. Thus each patient had 4 different results:  two RADT’s and two throat cultures.  Table 1 shows our results.   Of the 264 patients tested, 21% had at least one throat culture positive for GAS (prevalence 21%).  The sensitivity of the either RADT compared with the throat culture was 87%.  What was most interesting, however, was that the sensitivity of the throat culture compared to the other throat culture was either 83% or 91% (depending on which throat culture was chosen to be the “gold standard” for comparison), which is essentially the same as the RADT compared to either throat culture (Table 2).  I would be interested in knowing what the sensitivity of the throat culture was in Gieseker’s study when one culture was compared to the other, since their study design was very similar to ours.    We concluded from our study that the throat culture was just as sensitive as the RADT (at least the EIA test that we studied) for the detection of GAS from a throat swab.  Our results were based upon a concordance rate of 95% between the two throat cultures, which is similar to those results obtained by other investigators.[4],[5],[6]  Given these results, not only should a negative RADT have a backup throat culture, but an initially negative throat culture should also be backed up with another throat culture as well.  In some cases, as Gieseker has concluded, a second RADT may also be an appropriate backup test.  The throat culture is a less than perfect gold standard, and it is unlikely that any RADT will have a sensitivity of >95% if the throat culture itself does not meet that standard.  Additionally, most of the currently available RADT’s are easier to perform and interpret than a throat culture.  Regardless of which test is used to make a diagnosis of streptococcal pharyngitis, however, the most critical step is proper swabbing of the throat to obtain adequate number of organisms for testing.  You can’t detect what isn’t there.   Table 1   Detection of Group A Streptococci by Rapid Antigen Detection Test and Culture     1st Double Swab 2nd Double Swab   RADT 1 T/C 1 RADT 2 T/C 2 No. + + + + 40 + + + - 1 + - + + 4 + - + - 9 + - - + 1 + - - - 7 - + + - 1 - + - + 3 - + - - 2 - - + - 16 - - - + 4 - - - - 176       Total 264           T/C:  Throat Culture RADT:  Rapid Antigen Detection Test     Table 2   Sensitivity and Specificity of the RADT and the T/C When Either T/C 1 or T/C 2 is Used as the “Gold Standard” for Comparison       Sensitivity (%)   Specificity (%)     T/C 1 T/C 2   T/C 1 T/C 2 RADT 1   87 86   90 92 RADT 2   89 85   87 87 T/C 1     83     98 T/C 2   91     96                 T/C:  Throat Culture RADT:  Rapid Antigen Detection Test     References: [1] 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:e666-e670. [2] Lewey S, White CB, Lieberman MM, Morales E.  Evaluation of the throat culture as a follow-up for an initially negative enzyme immunosorbent assay rapid streptococcal antigen detection test. Pediatr Infect Dis J 1988;7:765-69. [3] White CB, Lieberman MM, Morales E: An In Vitro Comparison of Eight Rapid Streptococcal Antigen Detection Kits. Journal of Pediatrics 1988;113:691-693. [4] Breese BB, Disney FA.  The accuracy of diagnosis of beta streptococcal infections on clinical grounds.  J Pediatr 1954;44:670-673. [5] Kaplan EL, Top FW, Dudding BA, et al.  Diagnosis of streptococcal pharyngitis:  differentiation of active infection from the carrier state in the symptomatic child. J Infect Dis 1971;123:490-501. [6] Battle CV, Glasgow LA.  Reliability of bacteriologic identification of beta-hemolytic streptococci in private offices.  Am J Dis Child 1971;122:134-6.