Improving Drug Prescribing for Oral Anti-diabetic Agents through Information Technology

BACKGROUND: In January 2004, the Veterans Affairs (VA) Maryland Health Care System implemented an e-prescribing tool to improve the appropriateness of rosiglitazone (Avandia) prescribing for diabetic patients through decision support, automated laboratory review and documentation at the point of prescription. This study evaluated the impact of the tool on rosiglitazone prescribing rates and patterns, and hemoglobin A1c (HbA1c) monitoring rates and levels. METHODS: A pre-post intervention study design was used to examine changes in diabetes prescribing, and the likelihood of receiving a rosiglitazone prescription or an HbA1c test, in a sample of medicated diabetic veterans (n=10171) from 2002-2005. A generalized estimating equation approach was used to account for repeated measures. The logistic regression model adjusted for age, marital status, mean HbA1c and number of drugs. RESULTS: In 2003, 18% of patients received a rosiglitazone prescription. In 2004, there was an 11% lower odds of receiving rosiglitazone after the tool was introduced (95% CI: 0.83 –0.95). For patients receiving rosiglitazone and insulin (a combination specifically discouraged because it may lead to congestive heart failure) there was a 43% decrease. These trends were sustained through 2005. There was no significant change in HbA1c mean levels or monitoring rates. CONCLUSIONS: Following the e-prescribing tool implementation, rosiglitazone use decreased and prescribing patterns improved, despite the fact that use at the national VA level increased substantially over the same time period. In addition, the tool did not worsen physiological (HbA1c) outcomes. This research suggests that such a tool can be effective, safe and sustainable.