Assessment of Centor criteria for a sore throat - telemedicine versus in-person physical examinations

Patrycja Woldan-Gradalska, Wojciech Gradalski, Sikandar Moradi, Martin Franzelius, Sara Folkerman, Eva-Maria Fuchs, Frida Liljegren, Therese Karlsson, Hálfdán Pétursson, Anette Larsson, Ingmarie Skoglund, Ronny Gunnarsson, Pär-Daniel Sundvall

Keywords: Centor score, Sore throat, Telemedicine, Pharyngotonsillitis, Primary health care

Background:

It is uncertain whether the Centor criteria can be reliably assessed during telemedicine encounters with patients seeking care for a sore throat.

Research questions:

Primary objective: To compare the inter-rater reliability of Centor score assessments via telemedicine versus in-person examinations.

Method:

Study Design: Cross-sectional blinded study.
Analysis: The interrater agreement between in-person physical examinations and telemedicine assessments were assessed using Cohen's kappa coefficient with a 95% confidence interval.
Setting: Six primary healthcare centres and three out-of-hours primary care centres in Region Västra Götaland, Sweden.
Participant selection: Patients ≥ 3 years seeking with a sore throat.
Intervention: Each patient received independent assessments from two physicians: telemedicine followed by face-to-face examination. Both physicians were blinded to each other's findings.
Outcome Measures: In-person physical examinations and telemedicine assessments of the four Centor criteria as well as for the total Centor score.

Results:

189 patients were included during 2020 to 2023. The participants mean age was 31 (SD 18, range 4-89), 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. Even if the participants were divided by children/adults or whether the conditions for telemedicine assessment of the Centor criteria were adequate or not, the level of agreement of the total Centor score did not change.

Conclusions:

The low agreement between in-person physical examination and telemedicine assessments of lymph nodes, tonsils, and the total Centor score shows that telemedicine do not have the same accuracy as in-person examinations.

Points for discussion:

Reliability of telemedicine in clinical assessments: implications for clinical decision-making in primary health care and antibiotic stewardship.

#10