Letter

Trial of prescribing strategies in managing sore throat

BMJ 1997; 314 doi: 10.1136/bmj.314.7098.1904a (Published 28 June 1997)
Cite this as: BMJ 1997;314:1904.2

Access to the full text of this article requires a subscription or payment. Please log in or subscribe below.

Penicillin had no effect in patients negative for group A ß haemolytic streptococci

  1. M De Meyere, Professor of general practicea,
  2. M Bogaert, Professor of pharmacotherapya,
  3. G Verschraegen, Professor of microbiologya,
  4. I Mervielde, Professor of psychologyb
  1. a University of Ghent, University Hospital, 9000 Ghent, Belgium
  2. b University of Ghent, 9000 Ghent
  3. c Department of Family and Community Medicine, University of Toronto, Toronto, Canada
  4. d Primary Medical Care, Aldermoor Health Centre, Southampton SO16 5ST
  5. e Nightingale Surgery, Romsey SO51 7QN

    Editor—P Little and colleagues state that their study and the study by Whitfield and Hughes are the largest primary care trials in patients with sore throat.1 2 This is not so, and both of these studies can be criticised on methodological grounds.

    In 1987 American authors defined the 10 methodological criteria for valid randomised clinical trials in pharyngitis in primary health care: trials should be prospective, randomised, double blind, and placebo controlled; compliance should be checked by urine testing; there should be a sufficient number of subjects; throat swabs should be used; the patients included should be a representative sample; placebo and treatment groups should be comparable; and direct observation should be used.3 The study by Little and colleagues fulfils only seven of these 10 criteria. Compliance was not investigated, no throat swabs were used, and direct observation was made only at the first consultation. The prevalence of group A ß haemolytic streptococci is not stated: the representativeness of the sample is doubtful. Outcomes were documented in 582 subjects. Similarly, in the study of Whitfield and Hughes only seven criteria were fulfilled: compliance was not checked, no throat swabs were used, and the patients were not a representative sample (the prevalence of group A ß haemolytic streptococci was only 7.5%, which is much lower than the average population prevalence).2 The outcome was determined in 375 patients.

    As far as we could ascertain, only the study that we published in 1992 fulfils …

    Access to the full text of this article requires a subscription or payment

    Subscribe

    Article access

    Article access for 1 day

    Purchase this article for £20 $30 €32*

    The PDF version can be downloaded as your personal record

    * Prices do not include VAT

    Rapid responses

    THIS WEEK'S POLL

    BMJ most popular