Effects of team-based medication management on prescribing patterns, outcomes, and costs in veterans with diabetes
Lee, Kyung Min
MetadataShow full item record
In response to persistent shortages of primary care physicians, researchers have suggested increased use of non-physician providers to mitigate the consequences of the strained supply of physicians. As more health care systems implement a team-based approach in treating patients, better understanding the pattern of team care and its effects on outcomes and cost is critical in transitioning away from physician-centric care. This dissertation focuses on the prescribing component of care to explore the effects of team medication management on prescribing patterns, clinical outcomes, and care costs in Veterans with diabetes. In Chapter 2 I examine the outpatient prescribing patterns of oral antidiabetic drugs (OADs) over a 2-year period among patients who receive sole-provider prescribing or team-based prescribing. I find that team-prescribing patients receive significantly more sulfonylureas and metformin/sulfonyl urea combinations, suggesting that team prescribing may respond to patients’ needs quicker and intensify treatment earlier by switching or augmenting the initial medication. To determine the effect of prescribing modality on health outcomes, I compare mean changes in diabetes-related lab measures between the sole-provider models and team-prescribing models in Study 2. I perform 2-stage least squares regression to estimate the change in outcome measures from baseline to follow-up and find that while patients receiving NP prescribing achieve significantly greater reductions in glucose level compared to those receiving physician prescribing, the differences in HbA1c, glucose, and LDL reductions between team prescribing patients and sole-provider patients are insignificant. In Study 3 I explore the effect of prescribing model on health care costs. I use generalized linear modeling to estimate mean total outpatient cost among patients managed under different prescribing models. In addition, I perform logistic regression to estimate the likelihood of incurring any inpatient cost. I find no significant differences in mean outpatient cost or the likelihood of incurring any inpatient cost between sole-provider prescribing and team prescribing, suggesting that prescribing modality may not be associated with care costs. Evidence from this dissertation suggests that while team prescribing appears to provide more responsive medication management, it does not result in significant improvements in health outcomes or affect overall care costs.