This study will examine how a commercially available integrated health IT system impacts quality and resource use for patients with diabetes, using a natural experiment (2004-2012) within a large integrated delivery system (IDS). We will use a quasi- experimental pre-post design, which exploits the staggered implementation of integrated outpatient Health IT system between 2004-2008, and of the inpatient component between 2006-2010. The comprehensive Health IT system includes an electronic health record integrated with order-entry, decision-support, and a patient portal;the prior system was primarily based on a paper medical record. The IDS's existing automated databases will provide the data for this study and permit consistent capture of our outcomes and patient-level covariates both before and after the new Health IT system. We will focus on a cohort of 177,000 patients with diabetes across 1,028 primary care providers, 167 specialty clinicians (cardiologists, endocrinologists, pulmonologists), 21 Eds, and 18 medical centers within the IDS. The main predictor is use of Health IT by a patient's providers. The quality outcome measures include guideline-adherent prescription drug use and laboratory monitoring, drug adherence, and physiologic disease control (measured by laboratory tests) for diabetes and any of four other chronic diseases (Asthma, CAD, CHF, and HTN). The resource use outcome measures include outpatient visits, emergency department visits, non-elective hospitalizations. Using random effect models, we will test the hypotheses that Health IT use is associated with improved quality and lower visit rates. We will have excellent statistical power to detect small changes in our outcomes. We will make adjustments for patient, insurance, and organizational factors, including socio-economic status, case-mix, cost-sharing, and care delivery structure. Our research questions are broadly applicable across the nation, especially given high interest in Health IT, and the urgent need for improvements in the quality of patient care, and in patient healthcare resource use.
Health IT has compelling potential to transform clinical care, especially for the growing population of patients with diabetes, many of whom have complex healthcare needs. The implementation of an integrated Health IT system across a large delivery system creates a unique natural experiment and a valuable opportunity to investigate and quantify its effects on quality of care for patients with diabetes. Our research questions about Health IT effects on quality and resource use are broadly applicable across the nation, especially given high interest in Health IT, and the urgent need for improvements in the quality of patient care, and in patient healthcare resource use.
