3042 — A Simple, Four-Point Score for Evaluating the Likelihood of Medication, Diet, and Exercise Non-Adherence in Veterans with Hypertension
Crowley MJ, Grubber JM, and Bosworth HB, Center for Health Services Research in Primary Care, Durham VAMC;
Accurate identification of non-adherent Veterans with hypertension is essential, because interventions exist to improve blood pressure by targeting poor adherence. Clinicians lack convenient tools to help them predict patient adherence to key aspects of hypertension self-management: proper medication use, diet, and exercise. We developed a simple score to identify patients less likely to adhere to these aspects of self-management, and evaluated the use of this instrument in hypertensive Veterans.
We examined baseline data from 636 individuals enrolled in a hypertension trial, and created a categorical four-point score. Using adjusted logistic regression, we assessed the degree to which this score correlated with patient-reported medication adherence and patient-reported difficulty following diet and exercise recommendations. We confirmed the instrument’s performance using baseline data from a separate hypertension trial enrolling 591 Veterans.
Factors representing three domains, psychosocial (depressed mood, low accomplishment due to emotional issues), socioeconomic (lack of financial security), and health belief-related (belief myocardial infarction is likely with current lifestyle), correlated most strongly with patient-reported difficulty adhering to medication, diet, and exercise recommendations in the initial cohort. These factors comprised the four-point score. Odds of non-adherence appeared to increase with each tier on the four-point score. Relative to the lowest tier (score = 0), the adjusted OR of non-adherence among the highest tier (score >=3) was 2.61 [95% CI 1.43 to 4.77] for medications, 2.73 [95% CI 1.52 to 4.91] for diet, and 2.99 [95% CI 1.63 to 5.47] for exercise. Similar associations existed in the Veteran cohort. Relative to the lowest tier, the adjusted OR of non-adherence among the highest tier was 2.39 [95% CI 1.40 to 4.09] for medications, 1.93 [95% CI 1.11 to 3.33] for diet, and 2.27 [95% CI 1.30 to 3.97] for exercise. Predictive power ranged from fair to good (c = 0.62-0.73).
We created a four-point score designed to help clinicians identify hypertensive patients less likely to adhere to medication, diet, and exercise recommendations. The instrument performed well in Veteran and non-Veteran populations, and is simple enough to easily use clinically.
This simple instrument will help clinicians identify Veterans most likely to benefit from interventions targeting non-adherence.