Evaluating the use of parental reports to estimate health care resource utilization in children with suspected genetic disorders.

Abstract

A key step in evaluating the cost-effectiveness of diagnostic genome-wide sequencing (GWS) services is to measure the cost of prior and subsequent diagnosis-related health care resource utilization by patients. The majority of patients using diagnostic GWS services are children, and parental surveys are often used to complement utilization data abstracted from medical records. The objective of this study was to evaluate the validity of parental reports for children with very high levels of resource utilization. Primary caregivers of children enrolled in the CAUSES Research Clinic, a diagnostic GWS programme at B.C. Children's Hospital, completed an online survey. Parent-reported health care encounters for the 6-month period prior to survey completion were compared to utilization data abstracted from electronic medical records (EMR). The association between demographic characteristics and the probability of survey completion was tested using logistic regression. Agreement between parent-reported and EMR data was evaluated using Cohen's kappa, prevalence- and bias-adjusted kappa (PABAK), and the intraclass correlation coefficient (ICC). There were no major differences in demographic characteristics or resource utilization levels between families that completed the survey and those who did not. Agreement between parental reports and EMR data was high for hospitalizations (κ=0.71; PABAK=0.89; ICC=0.77) but lower for outpatient physician visits (κ=0.21; PABAK=0.48; ICC=0.27). Parental surveys are a valuable tool for estimating health care resource utilization during a 6-month recall period for children with suspected genetic disorders but are best used to complement utilization data collected from other sources.