Modifications to HEDIS and CSTE algorithms improve case recognition of pediatric asthma

Abstract

Our objective was to validate the Council of State and Territorial Epidemiologists (CSTE) definition of "probable" asthma and the Health Plan Employer Data and Information Set (HEDIS) definition of persistent asthma for diagnosis of pediatric asthma, and examine modifications that improve case recognition. CSTE and HEDIS criteria were applied to a cross-sectional study of 3,905 Medicaid children with physician-confirmed diagnosis of asthma/no asthma using a validated survey instrument based upon National Asthma Education and Prevention Program (NAEPP) Guidelines. Modified criteria were applied to another group of 1,458 non-Medicaid children from a managed care organization (MCO). Of 1,852 Medicaid children with physician-confirmed asthma, 906 had persistent asthma. CSTE identified 61% of children with "probable" asthma; HEDIS identified 44% of children with persistent asthma. Correct identification increased with greater disease severity. A modified CSTE increased sensitivity from 0.61 to 0.90, while maintaining high specificity. Three new HEDIS algorithms increased sensitivity from 0.44 to >0.84, with specificity >0.89. When applied prospectively to MCO children, these new algorithms demonstrated improved sensitivity. In conclusion, studies using current CSTE or HEDIS algorithms for case recognition underestimate asthma prevalence and overestimate asthma severity in children. Modified algorithms improve the identification of "probable" and persistent asthma.