Interexaminer reliability of pharyngeal injection and palatine tonsillar hypertrophy in a pediatric emergency department.

ABSTRACT

OBJECTIVES: To evaluate the interrater reliability of throat examinations in children according to the major and training stage. STUDY DESIGN: We performed a prospective observational study of interrater reliability. The participants included physicians with various amounts of experience and majors who were working in an urban, tertiary hospital. We collected 20 photos of the throats of children who presented to the pediatric emergency department (ED) and performed 2 surveys (with or without medical history). The primary outcome was the interrater agreement for pharyngeal injection (PI) and palatine tonsillar hypertrophy (PTH), and the secondary outcome was the interrater agreement for PI and PTH in subgroups of examiners divided by major and duration of clinical experience. RESULTS: Thirty-three examiners participated in this study. The overall percent agreement for PI was 0.669, and Fleiss' kappa was 0.296. The interrater reliability was similar before and after providing patients' medical history. The overall percent agreement for PTH was 0.408, and Kendall's W was 0.674. When the patients' medical history was provided, Kendall's W increased (0.692). In the subgroup analysis, Fleiss' kappa for PI ranged from 0.257 to 0.33, and Kendall's W for PTH ranged from 0.593 to 0.711. CONCLUSION: Examiners' agreement for PTH was more reliable than that for PI when evaluating children who visited the ED. The interrater reliability did not improve with increased clinical experience. These findings should be considered in the examination of pharyngeal pathology.