Pediatric asthma surveillance using Medicaid claims

Area of Science:

• Pediatric Health
• Epidemiology
• Health Services Research

Background:

• Administrative claims data are frequently used to identify pediatric asthma cases.
• Varying case definitions can lead to significant differences in estimated asthma prevalence.
• Consistent classification of asthma cases over time is crucial for accurate outcome assessment.

Purpose of the Study:

• To determine pediatric asthma prevalence using multiple case definitions in administrative claims data.
• To evaluate the year-to-year consistency of asthma case classification.
• To compare the concordance of different asthma case definitions.

Main Methods:

• Retrospective cohort analysis of 357,729 children (0-18 years) using Michigan Medicaid claims (2001-2002).
• Six alternative asthma case definitions were applied, including Health Plan Employer Data Information Set (HEDIS) criteria.
• Definitions incorporated outpatient, emergency department, inpatient, and pharmacy claims.

Main Results:

• Asthma prevalence ranged from 3.7% (≥4 medication events) to 14.9% (any asthma utilization).
• 55.5% of cases meeting HEDIS criteria in 2001 also met them in 2002.
• The definition requiring ≥4 medication dispensing events showed the best year-to-year classification concordance.

Conclusions:

• Pediatric asthma case classification can be inconsistent year-to-year, irrespective of the definition used.
• Inconsistent classification may introduce bias in longitudinal studies assessing asthma outcomes.
• Classifying asthma cases by requiring four or more medication dispensing events appears to minimize misclassification bias.