Comparison of hospitalizations, emergency department visits, and costs in a historical cohort of Texas Medicaid patients with chronic obstructive pulmonary disease, by initial medication regimen.

Background: Limited information is available on the relative outcomes and treatment costs of various pharmacotherapies for chronic obstructive pulmonary disease (COPD) in a Medicaid population.

Objective: This study compared the effects of initial medication regimens for COPD on COPD-related and all-cause events (hospitalizations and/or emergency department [ED] visits) and COPD-related and all-cause costs.

Methods: The study population was a historical cohort of Texas Medicaid beneficiaries aged 40 to 64 years with COPD-related medical costs (International Classification of Diseases, Ninth Revision, Clinical Modification codes 491.xx, 492.xx, 496.xx), 24 months of continuous Medicaid enrollment (12 months before and after the index prescription), and at least 1 prescription claim (index) for a combination product containing fluticasone propionate + salmeterol, an inhaled corticosteroid, salmeterol, or ipratropium between April 1, 2001, and March 31, 2003. The analyses of events employed Cox proportional hazards regression, controlling for baseline factors and preindex events. The analyses of costs used a 2-part model with logistic regression and generalized linear model to adjust for baseline characteristics and preindex utilization and costs.

Results: The study population included 6793 patients (1211 combination therapy, 968 inhaled corticosteroid, 401 salmeterol, and 4213 ipratropium). Only combination therapy was associated with a significantly lower risk for any COPD-related event (hazard ratio [HR] = 0.733; 95% CI, 0.650-0.826) and any all-cause event (HR = 0.906; 95% CI, 0.844-0.972) compared with ipratropium. COPD-related prescription costs were higher in all cohorts compared with the ipratropium cohort, but COPD-related medical costs were lower, offsetting the increase in prescription costs. For all-cause costs, prescription costs were higher in the combination-therapy cohort (+$415; P < 0.05) and the salmeterol cohort (+$247; P < 0.05) compared with the ipratropium cohort, but significant reductions in all-cause medical costs in the combination-therapy cohort (-$1735; P < 0.05) and salmeterol cohort (-$1547; P < 0.05) more than offset the increase in prescription costs.

Conclusions: In this historical population of Texas Medicaid beneficiaries, the combination-therapy cohort was 27% less likely to have a COPD-related event than the ipratropium cohort, 10% less likely to have any all-cause event, had similar COPD-related costs, and had reduced all-cause costs. Thus, compared with the ipratropium cohort, the combination-therapy cohort had an improvement in outcomes (based on the decreased time to a hospitalization or ED visit), with similar or decreased direct medical costs. Future research is needed in other patient groups.