Thwarting sore throats

The report by Warren McIsaac and colleagues that there is no difference in the sensitivity and specificity of a clinical sore throat score for patients seen in community-based family practices and those seen in an academic family medicine unit¹ is helpful for promoting the use of the sore throat score in the community at large. Nevertheless, one has to question the feasibility of implementing this tool on a broad scale, not because there are superior alternative approaches, but rather because of the limitations of the tool that front-line prescribers might perceive.

The medical literature suggests that antibiotics are used excessively to treat upper respiratory tract infections because physicians want to minimize the risk of failing to treat patients who would benefit from antibiotic therapy. Thus, the critical issue for the sore throat score is whether a sensitivity of 85% (or a false negative rate of 15%) will make practitioners sufficiently confident in the tool that they will abide by its recommendations.

It would be helpful if McIsaac and colleagues provided data on the percentage of patients who required an antibiotic prescription but did not get one on the basis of physician judgement. If physician judgement had a false negative rate of less than 15% this would imply that the physician threshold for committing an error of undertreatment is too high for physicians to follow the recommendations of the sore throat score.

Signature

Mitchell Levine
Centre for Evaluation of Medicines Department of Clinical Epidemiology & Biostatistics McMaster University Hamilton, Ont.