Keyword: Routinely Collected Health Data
1 result found.
Original Article
Epidemiology and Health Data Insights, 2(5), 2026, ehdi061, https://doi.org/10.63946/ehdi/19690
ABSTRACT:
Introduction. Routine indicators of rehabilitation effectiveness are rarely checked against validated measures in children. Kazakhstan mandates the Rehabilitation Routing Scale, whose properties in children are unknown. We tested whether its change agrees with change on the Gross Motor Function Measure and the Barthel Index.
Methods. We analyzed 318 consecutive admissions to a national pediatric rehabilitation center (August to September 2025) from anonymized discharge records. Change scores were available for 306 children (Routing Scale), 215 (Gross Motor Function Measure) and 150 (Barthel Index). We estimated the magnitude of change, Spearman correlations of change scores (hypothesis: at least 0.30) with bootstrap 95% confidence intervals, tertile agreement and classification agreement.
Results. Most children improved on every scale (88%, 92% and 96%); the Routing Scale never recorded a decline. Admission scores agreed moderately (correlation 0.69 with the Gross Motor Function Measure and 0.52 with the Barthel Index). Change scores did not: 0.04 (95% confidence interval -0.10 to 0.18; 208 children) and -0.05 (-0.21 to 0.12; 143 children), while change on the two validated scales correlated weakly, at 0.29 (0.08 to 0.48; 90 children). Among 86 children with all three scales, 40 improved on all three and 38 improved on both validated scales but not by a ratio of at least 1.10 on the Routing Scale. Sensitivity analyses gave similar results, except that the Barthel Index correlation changed sign between age groups.
Conclusion. The national scale tracked baseline severity, but its change was not interchangeable with change on validated measures. Effectiveness reports should use more than one instrument. Limitations include a single center, a short stay, no test-retest data and no external anchor.
Methods. We analyzed 318 consecutive admissions to a national pediatric rehabilitation center (August to September 2025) from anonymized discharge records. Change scores were available for 306 children (Routing Scale), 215 (Gross Motor Function Measure) and 150 (Barthel Index). We estimated the magnitude of change, Spearman correlations of change scores (hypothesis: at least 0.30) with bootstrap 95% confidence intervals, tertile agreement and classification agreement.
Results. Most children improved on every scale (88%, 92% and 96%); the Routing Scale never recorded a decline. Admission scores agreed moderately (correlation 0.69 with the Gross Motor Function Measure and 0.52 with the Barthel Index). Change scores did not: 0.04 (95% confidence interval -0.10 to 0.18; 208 children) and -0.05 (-0.21 to 0.12; 143 children), while change on the two validated scales correlated weakly, at 0.29 (0.08 to 0.48; 90 children). Among 86 children with all three scales, 40 improved on all three and 38 improved on both validated scales but not by a ratio of at least 1.10 on the Routing Scale. Sensitivity analyses gave similar results, except that the Barthel Index correlation changed sign between age groups.
Conclusion. The national scale tracked baseline severity, but its change was not interchangeable with change on validated measures. Effectiveness reports should use more than one instrument. Limitations include a single center, a short stay, no test-retest data and no external anchor.
