Lappe, Jason M., Muhlestein, Joseph B., Lappe, Donald L., Badger, Rodney S., Bair, Tami L., Brockman, Ruth, French, Thomas K., Hofmann, Linda C., Horne, Benjamin D., Kralick-Goldberg, Susan, Nicponski, Nan, Orton, Janette A., Pearson, Robert R., Renlund, Dale G., Rimmasch, Holly, Roberts, Colleen, and Anderson, Jeffrey L.
Background: Despite recent advances in the treatment and prevention of cardiovascular disease, a treatment gap for secondary prevention medications still exists. Objective: To develop and implement a program ensuring appropriate prescription of aspirin, statins, [beta]-blockers, angiotensin-converting enzyme inhibitors, and warfarin at hospital discharge. Design: A nonrandomized before-after study comparing patients hospitalized before (1996-1998) and after (1999-2002) implementation of a discharge medication program (DMP). Patients were followed for up to 1 year. Setting: The 10 largest hospitals in the Utah-based Intermountain Health Care system. Patients: In the pre-DMP and DMP time periods, 26 000 and 31 465 patients, respectively, were admitted to cardiovascular services (n = 57 465). Measurements: Prescription of indicated medications at hospital discharge; postdischarge death or readmission. Results: By 1 year, the rate of prescription of each medication increased significantly to more than 90% (P < 0.001); this rate was sustained. At 1 year, unadjusted absolute event rates for readmission and death, respectively, were 210 per 1000 person-years and 96 per 1000 person-years before DMP implementation and 191 per 1000 person-years and 70 per 1000 person-years afterward. Relative risk for death and readmission at 30 days decreased after DMP implementation; hazard ratios (HRs) for death and readmission were 0.81 (95% CI, 0.73 to 0.89) and 0.92 (CI, 0.87 to 0.99) (P < 0.001 and P = 0.017, respectively). At 1 year, risk for death continued to decrease (hazard ratio, 0.79 [CI, 0.75 to 0.84]; P< 0.001) while risk for readmission stabilized (hazard ratio, 0.94 [CI, 0.90 to 0.98]; P = 0.002), probably because survivors had more opportunities to be readmitted. Limitations: The study design was observational and nonrandomized, and the authors could not control for potential con-founders or determine the extent to which secular trends accounted for the observed improvements. Conclusions: A relatively simple quality improvement program aimed at enhancing the prescription of appropriate discharge medications among cardiovascular patients is feasible and can be sustained within an integrated multihospital system. Such a program may be associated with improvements in cardiovascular readmission rates and mortality.