Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Can throat examination distinguish between bacterial and viral infective agents?

Three Part Question

In [patients with acute sore throat] is [clinical examination alone better than clinical examination and investigation] at [determining whether the infective agent is bacterial or viral]?

Clinical Scenario

An 8 year old presents to A&E with a sore throat and anterior cervical lymphadenopathy and a 3 day history of a cough. She is examined and told it is probably just a virus'. Is there any evidence for this statement or should she be investigated further?

Search Strategy

Medline OVID 1950—June week 3 2010, 75 papers, 6 papers relevant.

Cinahl, 12 papers found, no extra papers found.

Embase, 76 papers found, no extra papers relevant.

The Cochrane Database of Systematic Reviews (CDSR) June 2010.

(pharyngitis): ti, ab, kw 29 reviews 0 relevancies

(Exp PHARYNGITIS/OR exp LARYNGITIS/OR exp TONSILLITIS/OR exp Peritonsillar Abscess/OR OR OR OR (peritonsillar adj5 abscess).af. OR OR (throat adj5 infection$) OR (sore adj5 throat$) OR exp Infectious Mononucleosis/OR (infectious adj5 mononucleosis).af. OR (glandular adj5 fever).af.)AND {Streptococcus pyogenes/Or Streptococcal Infections/OR throat OR Bacteriological Techniques/OR (throat adj5 swab).af. OR nasopharyngeal OR ear nose throat OR pharyngeal swab OR bacterial OR Bacterial Infections/OR ‘Sensitivity and Specificity’/OR rapid antigen OR or BACTERIOLOGY/OR OR MICROBIOLOGY/} AND {physical OR Physical Examination/OR OR OTOLARYNGOLOGY/OR ear nose throat OR ENT exam$.mp. OR oropharyngeal exam$.mp. OR throat exam$.mp. OR centor score} limit to humans and English language.

Search Outcome

medline 65 papers found 6 relevant
cinahl 12 papers found no extra papers found
embase 76 papers found-no extra papers relevant
Cochrane-no relevant papers found

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Ebell et al,
9 studies which were large, blinded prospective studies containing over 300 patients with sore throat. The studies were required to note history and physical examination data and used throat culture as a reference standard. The aim was to identify patients with Group A B-haemolytic streptococcal infection.Systematic review, including 9 relevant randomised controlled trials. 1++ evidence.Likelihood ratio of one aspect of clinical examination detecting presence of GABHSTonsillar exudate (LR+=3.4) pharyngeal exudate (LR+=2.1) strep throat exposure in previous 2 weeks (LR+=1.9). The absence of symptoms was not efficient at ruling out the disease.The accuracy of clinical prediction rules in the different studies is presented in different statistical forms making it difficult to compare. Centor's score has not been validated in younger patients
Most valid clinical prediction rulesArea under ROC curve 0.79(good accuracy)
Schwartz et al,
A convenience sample of 192 patients aged 1 to 18 years who presented to a private peadiatric office with signs and symptoms of acute pharyngitis. A clinical examination was undertaken followed by a rapid antigen detection test and also a throat swab.Diagnostic studyPharyngeal findings of patient group with GABS compared to those withoutPharyngeal erythema 95% with vs 80% without. Palatal enanthem 44% v 10%. uvular erythema or petechiae 71% v 37%, uvular enanthem37% v 17%, Uvular oedema 69% v 50%, tonsillar erythema 77% v 55%, tonsillar exudate 75% v 38%( all p values less than 0.01 no specific figures given)We are not told how the patients were chosen for the study (if it was all suitable patients or just some) We are not specifically told if the examiner was blind to the results of the rapid antigen detection test. This study only includes children and is not validated in adults
Hjortdahl P, Melbye H,
174 patients who presented to the Emergency Department. Patients were required to be 18 years or older, attending between 16.00 and 21.00 with the chief complaint of a sore throat. Clinical examination was carried out and the doctor was required to complete a visual analogue scale about their degree of certainty of presence or absence of GABHS. Blood samples and throat swabs were taken. GABHS was diagnosed if a postive GABHS culture was found or a four fold or more increase or fall of AST or ADNase B serum during a 4 week observation period.Diagnostic studyNumber of incorrect diagnoses1 in 3Bias may be introduced because 40 different doctors of differing experience examined the patients. Only patients attending between 4 and 9 pm were seen. The prevalence of GABHS was quite high in this population (34%) this may be higher than in the UK and so the sbility to generalise this study may be affected. It is debatable whether the inclusion of patients with a negative culture but a four fold or greater change in AST or ADNase B serum titre ,into the GABHS diagnosed group is appropriate.
Fair degree of certainty of clinical diagnosis that the patient did not have GABSCorrect in 53/67 cases. True negative rate 79%
Fair degree of certainty that the patient does have GABHSCorrect in 21/33. True positive rate 64%
Correct identification of GABHS by clinical evaluation alone vs identification by immunological assay38/59 (sensitivity 0.64) 40/56 (sensitivity 0.71)
sSnsitivity of clinical examination vs immunological assaySensitivity 0.64 vs 0.71. Specificity 0.71 vs 0.93. Likelihood ratio 2.2 vs 8.7. postive predictive value 0.52 vs 0.82
Nawaz H, Smith DS, Mazhari R, et al
A convenience sample of 218 patients seen for sore throat during business hours on weekdays beween June 1993 and April 1996. No subjects excluded. Clinical data, including clinical predictors of GABHS was recorded on a standardized checklist data form and a throat swab was performed.Prospective cohort studyAspects of clinical examination with the highest probability of GABHSTender adenopathy and enlarged tonsils and pharyngeal erythema and tonsillar exudates. Postive predictor value 46. Sensitivity 71, specificity 77The sample studied was small and of convenience so there could have been selection bias. Some eligible patients were missed. Variation between physicians assessment may have existed.
Woods WA, Carter CT, Schlager TA
35 children between 2 and 3 years presenting to the Emergency department with symptoms of a URTI. Pharyngeal signs and symptoms were recorded and throat swabs taken.Prospective diagnostic studyClinical findings in GABHS and non-GABHS infection.FeverGABHS positive 90% and negative 76%There is no mention of how many of the eligible patients were included in the study. The sample size was too small for comparison of clinical findings in GABHS and non-GABHS so effectively no conclusions can be drawn.
Tonsillar exudate40% vs 28%
Cervical adenopathy60% vs 20%
Rhinorrhea40% vs 60%
McIsaac WJ, Kellner JD, Aufricht P et al
787 children and adults presenting with acute sore throat. Recommendations from 2 guidelines were compared with rapid testing alone, a clinical prediction rule (the modified Centor's score) and treatment for positive throat cultures only. Culture all was the gold standardDiagnostic studyCulture allSensitivity 100%, specificity 100%This study only included adults and children with a modified Centor score of 2 or more, however most guidelines support the use of clinical scores to to exclude patients. The sensitivity of the rapid test was only 83%. The study was not able to assess the impact of higher or lower GABHS prevalence.
Children-rapid test all, treat those with positive results, throat culture those with negative strep tests. Adults rapid test all, treat positive results, dont culture negative resultsChildren 100%, 99%, adults 76.7%, 99.2%
Children-treat as above. Adults rapid test all with a centor score of 2 or 3 and treat if positive. Treat all adults with a centor score over 4 empiricallyChildren as above, adults 78.1%, 95.8%
Treat children as above. test no adults and treat those with a centor score of 3 or 4 empiricallyChildren as above, adults 76.7%, 43.8%
Throat culture all with a centor score of 2 or 3.If score of 4 or more treat empirically.Children 100%, 90.3%. adults 100%, 96.5%
Rapid test on all children and adults, treat those with positive results without culture confirmation of negative resultsChildren 85.8%, 99% adults 56% 99.2%


The Ebell systematic review suggests that presence of a combination of tonsillar exudate, history of fever, anterior cervical lymphadenopathy and absence of cough is the best way of predicting GABHS from clinical examination. Schwartz also shows that significantly more GABHS pharyngitis cases have these clinical signs and symptoms. However Hjortdahl demonstrates the lower accuracy of clinical exam compared with immunological assay. These diagnostic studies aren't ideal and some may contain selection bias. The McIsaac study is the most useful study to decide the best investigation combinations.

Clinical Bottom Line

Clinical examination and investigation is better than clinical examination alone at dertermining whether the infective agent is bacterial or viral. The most sensitive and specific way to distinguish bacterial pharyngitis (ie GABHS) culture all adults with score 2 or 3 and treat those with score of 4 empirically, however this results in inappropriate antibiotic prescription in 4.8% (discussed in another BET). Therefore a compromise is to rapid test all and treat positive reults without culture confirmation of negative results. For children the best option was to rapid test all and treat those with positive results and culture negative results.


  1. Ebell MH, Smith MA, Henry BC et al. Does This Patient Have Strep Throat? JAMA 2000;284:2912–18.
  2. Schwartz RH, Gerber MA, McKay K. Pharyngeal findings of Group A Streptococcal Pharyngitis. Arch Pediatr Adolesc 1998;152:927–8.
  3. Hjortdahl P, Melbye H. Does near patient testing contribute to the diagnosis of streptococcal pharyngitis in adults? Scand J Prim Health Care 1994;12:70–6.
  4. Nawaz H, Smith DS, Mazhari R, et al Concordance of Clinical Findings and Clinical Judgement in the Diagnosis of Streptococcal Pharyngitis Academic Emergency Medicine Oct 2000 7(10):1104-9
  5. Woods WA, Carter CT, Schlager TA Detection of Group A streptococci in children under 3 years of age with pharyngitis Pediatric Emergency Care 1999 vol 15 No5
  6. McIsaac WJ, Kellner JD, Aufricht P et al Empirical Validation of Guidelines for the management of Pharyngitis in Children and Adults JAMA Apr 7 2004 vol291 No.13 pg 1587