Asthma's high prevalence and cost in urban populations has prompted Medicaid health plans to develop asthma disease management programs. Numerous ethical, regulatory and practical issues make it difficult for health plans to evaluate these programs using prospectively controlled studies. As a result, many programs have measured their success by comparing resource utilization data accrued prospectively with retrospective data. Asthma programs have been deemed successful by demonstrating a 50% reduction in resource use compared to retrospective data. This method for evaluating the success of asthma disease management programs is of questionable significance. Accordingly, we have created a template of the natural history of resource utilization by children with severe asthma on Medicaid. Method: We identified a cohort of children with asthma (ages 0-16) from an urban Medicaid claims data base who were continuously enrolled for 2 years. These children were divided into 3 groups based on their use of hospital resources: group 1 was comprised of children with “low” use (defined as no admissions and ≤ 1 ER visits); group 2 consisted of children with “high” use (defined as one admission or ≥ 2 ER visits); and group 3 was children with“extra-high” use (defined as ≥ 2 admissions or ≥ 5 ER visits). Of children identified as “high” use asthmatics in the 1st year, 209 of 306 (68%) became “low” users during the 2nd year. The “extra-high” use group showed similar rates of improvement; 249 of 437 (57%) became “low” users. Logistic regression analyses showed that African-American vs. Caucasian children (p≤0.005) and male vs. female children (p≤0.009) with “extra-high” use were 2 times more likely to remain in the “extra-high” use group. Conclusion: Children with severe asthma who receive Medicaid demonstrate large decreases in resource utilization without any interventions. The success of disease management programs for children with asthma in Medicaid populations should be measured against this template or within controlled clinical trials, rather than by comparison to utilization patterns in the previous year. Further work is needed to determine whether persistence in the “extra-high” use groups by African-American and male children is related to disease severity, ineffective services, or other causes.
