Research article

Optimal management of adults with pharyngitis – a multi-criteria decision analysis

Sonal Singh^1†, James G Dolan^2† and Robert M Centor^3*

• * Corresponding author: Robert M Centor rcentor@uab.edu

• † Equal contributors

Author Affiliations

¹ Department of Medicine, Wake Forest University, Winston Salem, NC, USA.

² Department of Medicine, Unity Health System and the University of Rochester, Rochester, New York, USA.

³ Department of Medicine, University of Alabama-Birmingham, Birmingham, AL, USA.

For all author emails, please log on.

BMC Medical Informatics and Decision Making 2006, 6:14 doi:10.1186/1472-6947-6-14

Published: 13 March 2006

Abstract

Background

Current practice guidelines offer different management recommendations for adults presenting with a sore throat. The key issue is the extent to which the clinical likelihood of a Group A streptococcal infection should affect patient management decisions. To help resolve this issue, we conducted a multi-criteria decision analysis using the Analytic Hierarchy Process.

Methods

We defined optimal patient management using four criteria: 1) reduce symptom duration; 2) prevent infectious complications, local and systemic; 3) minimize antibiotic side effects, minor and anaphylaxis; and 4) achieve prudent use of antibiotics, avoiding both over-use and under-use. In our baseline analysis we assumed that all criteria and sub-criteria were equally important except minimizing anaphylactic side effects, which was judged very strongly more important than minimizing minor side effects. Management strategies included: a) No test, No treatment; b) Perform a rapid strep test and treat if positive; c) Perform a throat culture and treat if positive; d) Perform a rapid strep test and treat if positive; if negative obtain a throat culture and treat if positive; and e) treat without further tests. We defined four scenarios based on the likelihood of group A streptococcal infection using the Centor score, a well-validated clinical index. Published data were used to estimate the likelihoods of clinical outcomes and the test operating characteristics of the rapid strep test and throat culture for identifying group A streptococcal infections.

Results

Using the baseline assumptions, no testing and no treatment is preferred for patients with Centor scores of 1; two strategies – culture and treat if positive and rapid strep with culture of negative results – are equally preferable for patients with Centor scores of 2; and rapid strep with culture of negative results is the best management strategy for patients with Centor scores 3 or 4. These results are sensitive to the priorities assigned to the decision criteria, especially avoiding over-use versus under-use of antibiotics, and the population prevalence of Group A streptococcal pharyngitis.

Conclusion

The optimal clinical management of adults with sore throat depends on both the clinical probability of a group A streptococcal infection and clinical judgments that incorporate individual patient and practice circumstances.