Inter-rater reliability of Centor score assessments between telemedicine and in-person examinations in patients with an acute sore throat

Background: It is uncertain whether the Centor criteria can be reliably assessed during telemedicine encounters with patients seeking care for a sore throat. Acquiring this knowledge is important as sore throat is a common reason for telemedicine consultations.

Objectives: primary objective: To compare the inter-rater reliability of Centor score assessments via telemedicine versus in-person examinations. Secondary objectives: To investigate whether the interrater reliability varies when assessing patients who are children versus adults, and whether the telemedicine physician considered conditions for assessment as adequate.

Methods: A cross-sectional study in which each patient initially underwent a telemedicine evaluation, followed by an in-person assessment conducted by an independent physician who was kept unaware of the outcome of the initial evaluation. Agreement between both assessments was measured using Cohen's kappa coefficient.

Results: During 2020-2023 189 patients with a mean age of 31 years (SD 18) were included. Among them, 114 were female and 148 adults. Agreement was low with kappa between 0.47(95% CI 0.38 - 0.56) to 0.58 (95% CI 0.43-0.72) when comparing assessments of lymph nodes, tonsils and the total Centor score. Kappa was potentially acceptable for history of fever and absence of cough. Subgrouping participants into children and adults did not affect kappa of the total Centor score.

Conclusion: Telemedicine examination in patients with an acute sore throat is not reliable for assessing Centor criteria.