Abstract
BACKGROUND: The accuracy of students' heights and weights measured by school staff for body mass index (BMI) screening/reporting has not been established. This study examined school staffs' measurement accuracy, comparing accuracy by staff- and student-level characteristics. METHODS: School staff and researchers measured the height and weight of 1008 4th-8th grade students, within 1 month of each other. Bland-Altman plots, mean differences, and intraclass correlation coefficients (ICCs) were calculated to examine measurement accuracy. Linear mixed effects models assessed accuracy by staff- and student-level characteristics. RESULTS: Bland-Altman plots revealed no appreciable bias in differences between researcher and staff measurements. The mean absolute difference between researcher and school staff measurements were 1.0 ± 1.6 cm (height), 0.7 ± 1.8 kg (weight), and 0.4 ± 0.8 kg/m(2) (BMI). Inter-rater ICC values were ≥0.97, demonstrating "excellent" reliability. Categorical weight status was correctly classified for 94% of students (kappa 0.90), and for 96% with a BMI ≥95th% (kappa 0.94). Physical education (PE) teachers were slightly less accurate than school nurses in measuring height (0.4 cm less accurate; p = .045) and weight (0.4 kg; p = .015). CONCLUSIONS: School staff conducted height/weight measurements on 4th-8th grade students with high accuracy. Resultant school-based BMI reports using similar protocols should validly reflect weight status for almost all students.