Asthma home teaching: two evaluation approaches

The aim of this research was to measure the impact of home-based teaching on reducing asthma admissions and emergency department (ED) visits for Medicaid-managed care patients utilizing two different study design methods. This was an historical-prospective study utilizing health plan administrative data, including membership files and medical claims. We identified 381 patients aged 2-56 with hospitalizations or ED visits for asthma. These high risk asthma members were recruited for a home-based teaching program to prevent future hospitalizations or ED visits. We evaluated program effectiveness using two quasi-experimental research designs: a "one-group pre/post-test design," where enrolled members served as their own control, and a more rigorous "untreated control group design with pre/post test," where results for enrolled members and a similar control group were compared pre/post test. Poisson regression models were used to investigate the dependence of member rates for asthma-related events on program enrollment, age, sex, race, and geographic region. Using the pre/post test design, members enrolled in the home-based teaching program demonstrated statistically significant reductions in hospital admissions and ED visits (p < 0.001). The untreated control group design, however, found no association between utilization and enrollment in the home-based teaching program (p = 0.510). Small differences were detected for subgroups. A marginally statistically significant impact of the program was found for Whites, but not for Blacks. The quasi-experimental design that utilized an external control group provided an approach that more accurately explained true disease management program impact. In addition, this approach allowed for subgroup analyses to detect opportunities for program improvement.