Regression to the mean: a limited issue in disease management programs for chronic obstructive pulmonary disease.

Abstract

Our objective was to test for evidence of regression to the mean in chronic obstructive pulmonary disease (COPD)-related health care utilization in a Colorado Medicaid population that met the criteria for, but were not participating in, a COPD disease management (DM) program. National Jewish Medical and Research Center had enrolled individuals who (1) had a diagnosis of COPD for at least 1 year and (2) were active participants in Colorado Medicaid's 1-year DM program called breatheWise; the present study sought a comparator group for that population. In order to test for evidence of regression to the mean (ie, high utilization from the recruitment period reducing without active intervention) in this case management model, we conducted a case-controlled analysis of total spending for a comparator population that would have met the inclusion criteria for the DM program. The present study assessed health care utilization for fiscal years 2002 and 2003 in terms of total rates of emergency room (ER) visits and hospitalizations for all causes in the comparator group of COPD patients. In addition, total costs related to both ER visits and hospitalizations were compiled. In total, 354 individuals met the inclusion criteria and were identified as the comparator group. ER visits and hospitalizations were consistent for 2002 and 2003. ER visits totaled 314 and 315 in 2002 and 2003, respectively, indicating a 0.3% increase that was not significant. Hospitalizations decreased from 0.53 admissions per patient in 2002 to 0.48 in 2003-a 9.4% reduction that was not significant. With comparable rates of ER visits and hospitalizations, total costs for health care utilization remained virtually unchanged between 2002 and 2003. There is minimal evidence of regression to the mean over 2 consecutive years in the Colorado Medicaid patients with moderate to severe COPD.