Family Practice Vol. 20, No. 2, 108-111
© Oxford University Press 2003
Clinical Research |
Rapid test, throat culture and clinical assessment in the diagnosis of tonsillitis
Kirseberg Primary Health Care Centre and
a Department of Community Medicine, Lund University, Malmö University Hospital, Malmö, Sweden.
Correspondence to Dr N-O Månsson; E-mail: nils-ove.mansson{at}smi.mas.lu.se
Johansson L and Månsson N-O. Rapid test, throat culture and clinical assessment in the diagnosis of tonsillitis. Family Practice 2003; 20: 108–111.
Received 25 February 2002; Revised 9 August 2002; Accepted 4 November 2002.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objectives. Our aim was to identify a group of patients with clinical signs of tonsillitis to whom antibiotics could be prescribed without further diagnostic actions, and to compare the outcome of clinical assessment with the result of an antigen detection test using culture as the gold standard.
Methods. During two winter months, patients aged 
4 years attending for sore throat at three primary health care centres in Malmö, Sweden, were examined. Odds ratios, sensitivities, specificities and predictive values were calculated for clinical assessment and for an antigen detection test.
Results. Among the 169 participating patients, growth of group A ß-haemolytic streptococci (GAS) was found in 53 cultures, and 23 patients (14%) were clinically assessed as ‘absolutely positive’, representing positive clinical assessment. Nineteen had positive cultures for GAS. The sensitivity, specificity and predictive positive and negative value for the antigen detection test were 82, 96, 90 and 93%, respectively, and for positive clinical assessment 36, 97, 83 and 77%.
Conclusion. It is possible to identify a small group of patients with convincing signs of tonsillitis in which the specificity as well as the predictive positive value of the rapid test and the clinical assessment are close to equal. Antibiotics may be prescribed to these patients without further diagnostic actions.
Keywords. Antigen detection test, clinical assessment, rapid test, streptococci, throat culture, throat infection, tonsillitis.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
To distinguish between different aetiological agents merely by clinical assessment of an infection such as tonsillitis has long been considered almost impossible.1 Both recent and earlier studies have shown that early treatment of tonsillitis caused by group A ß-haemolytic streptococci (GAS) not only prevents complications but also shortens the duration of symptoms.2–4 Due to their good specificity, antigen detection tests are recommended as a screening method for GAS, before prescribing antibiotics.5 Consequently, clinical examination has been neglected, although, in daily clinical practice, the clinical findings in some patients with throat infections are so obvious that no further diagnostic methods are felt to be necessary.
The increasing problem with bacterial resistance to antibiotics6 and the costs involved make it important to evaluate common clinical problems such as sore throat continuously. The aim of this study, therefore, was to try to identify the group of patients with symptoms and signs that strongly influences the practitioner to start treatment with antibiotics without further diagnostic actions, and to compare the outcome of the clinical assessment with the result of an antigen detection test. In addition, the purpose was to distinguish the symptoms and signs with the highest specificity for infection with GAS to ensure a reliable clinical diagnosis.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Fifteen physicians, specialists in family medicine, from three health care centres in Malmö, Sweden, took part in the study. Data collection was performed from late January to the beginning of March.
Patients
Consecutive patients of both sexes presenting with a sore throat were included in the study. Children younger than 4 years were excluded because of anamnestic difficulties. In addition, the following were excluded: patients receiving current antibiotic treatment or treatment not completed in the past 72 h; patients with a complication demanding different management; and patients attending on Fridays after the last delivery to the bacteriological laboratory.
Examination
The physicians made a comprehensive assessment, based on the estimated probability of infection with GAS, using the grades ‘absolutely positive’, ‘positive’, ‘possibly positive’, ‘possibly negative’, ‘negative’, and ‘absolutely negative’ (Table 1
). The physicians also noted hypothetical management that would have been used in the normal clinical situation prior to receiving any test results (Table 1), and this decision was used to identify the most convincing clinically positive patients. The group in which a majority of the patients were given antibiotics without prior testing was considered the only clinically positive group in the following analysis.
|
The outcome of the clinical assessment was compared with an antigen detection test, using throat culture as the gold standard.
Tests and bacteriology
Throat swabs were obtained by staff nurses in the health centres’ laboratories. The rapid test Abbot Testpack Strep A Plus was used and the throat cultures were analysed at the Department of Clinical Microbiology, Malmö University Hospital. The specimens were inspected after 24 and 48 h, and growth of GAS was quantified as sparse, moderate or abundant.
Statistical methods
The chi-square test was used to compare differences between the diagnostic tests, and differences were considered statistically significant when P < 0.05. Logistic regression was performed for calculation of odds ratios and their 95% confidence intervals (CIs). All tests were two-sided.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In 1997, from January 27 to March 6, 169 patients were included, of whom 56% were women and 53% in the age interval 25–44 years.
Growth of GAS was found in 53 cultures, an incidence of 31%. Group C (GCS) and group G (GGS) ß-haemolytic streptococci were found in four cultures. Only cultures with growth of GAS were included in the following calculations.
Four groups dominated the comprehensive clinical assessment (Table 1
). Twenty-three patients (14%) were considered ‘absolutely positive’, the only group of patients of which a majority were given antibiotics without prior testing and, according to our criteria, the only clinically positive group. In this perspective, clinical assessment of infection with GAS turned out as follows: four were false positives, 19 true positives, 34 false negatives and 112 true negatives, giving a sensitivity of 36%, a specificity of 97%, a predictive positive value of 83% and a negative predictive value of 77%.
Fifty patients had a positive rapid test, four of which were false positives, and 46 were true positives, resulting in a sensitivity of 82%, a specificity of 96%, a predictive positive value of 90% and a predictive negative value of 93%. Thus, an overall comparison between a clinical assessment and the rapid test is in favour of the rapid test (P < 0.0001).
In the routine management of sore throat, the physicians in this investigation would have used a rapid test in 141 cases (83%). In 38 cases (22%), culture would have been used if the rapid test turned out to be negative. Twenty-five patients did not undergo any test other than clinical examination. Antibiotics were given to 14 (8% of all) of these patients, out of which four had co-existing infections that required antibiotics.
No single symptom, anamnestic statement or examination finding had both a high sensitivity and a high specificity (Table 2). However, a high specificity was found for known recurring tonsillitis, tonsil coating and patient’s recognition of the symptoms, which were also the three most common characteristics in the group considered ‘absolutely positive’.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In our study, which we believe is representative of general practice, clinical assessment is seldom considered as definitely positive or negative, possibly reflecting the awareness of the difficulty in separating streptococcal from non-streptococcal infection. This is probably also the reason why the rapid test was used to such a great extent.
The choice of throat culture as the gold standard was motivated by the tradition in Malmö at the time of the study. A better alternative might otherwise have been ASOT, i.e. the test of a significant increase in antistreptolysin O as an indication of streptococcal infection.
Clinical assessment often results in different levels of certainty, and we believe that one problem is finding the appropriate cut-off level, which in this study was chosen to be the most reliable out of five possible levels of clinical assessment, i.e. ‘absolutely positive’. This small group of patients presented a very convincing clinical pattern and had a high incidence of GAS. Thus, the high specificity (97%) for the clinical assessment is important; only a few patients were treated unnecessarily with antibiotics, which is important not only because of costs or fear of bacterial resistance, but also from the perspective of the debated7 necessity of treating streptococcal infections with antibiotics.
The group of patients considered ‘absolutely positive’ was also characterized by the statements, symptoms and signs with the highest specificity for this type of infection.
Little et al.7 studied re-attendance after different prescribing strategies in patients with sore throat, and concluded that to refrain from or delay prescription of antibiotics is an effective way to decrease the number of consultations with GPs without increasing the number of complications. However, the practitioners who took part in the present study hardly ever used this strategy. Instead, the rapid test was used to discriminate between patients that needed antibiotics and those who did not.
In summary, a small group of patients was distinguished, where the rapid test contributed very little to the diagnoses and where the clinical pattern was sufficient to motivate treatment with antibiotics. It is important that this group of patients is well defined to ensure appropriate use of antibiotics after solely a clinical diagnosis, and the rapid test must still be recommended as a valuable screening method in unclear cases.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Wannamaker, Lewis W. Perplexity and precision in the diagnosis of streptococcal pharyngitis. Am J Dis Child 1972; 124: 352–358.[Medline]
2 Dagnelie CF, Van Der Graaf Y, De Melker RA, Touw-Otten FWMM. Do patients with sore throats benefit from penicillin? A randomized double-blind placebo-controlled trial with penicillin V in general practice. Br J Gen Pract 1996; 46: 589–593.[ISI][Medline]
3 Bass JW. Treatment of streptococcal pharyngitis revisited. J Am Med Assoc 1986; 256: 740–743.[Abstract]
4 Hovelius B, Bygren P, Christensen P, Mårdh PA. Respiratory tract infections at a community care center—with emphasis on group A streptococci. Scand J Infect Dis, Suppl 1983; 39: 59–67.[Medline]
5 Schwann A, Ek E, Falck G et al. Aktuell rekommendation för decentralicerad diagnostik av hemolytiska streptokocker grupp A i svalgprov [Current recommendations for diagnostics of haemolytic streptococci group A in throat samples]. Läkartidningen 1991; 88: 1378.[Medline]
6 Melander E. Resistant pneumococci and use of antibiotics. Thesis. Malmö University Hospital, Lund University: Department of Community Medicine, 2000.
7 Little P, Gould C, Williamson I, Warner G, Gantley M, Kinmonth AL. Re-attendance and complications in a randomised trial of prescribing strategies for sore throat: the medicalising effect of prescribing antibiotics. Br Med J 1997; 315: 350–352.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Kelly Short Report: Can mouth swabs replace throat swabs?: Cross-sectional survey of the effectiveness of rapid streptococcal swabs of the buccal mucosa Can Fam Physician, September 1, 2007; 53(9): 1500 - 1501. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||