Association of refill adherence and health care use among adults with hypertension in an urban health care system

Study Objectives. To determine the rates of undersupply, appropriate supply, and oversupply of antihypertensive drugs, as measured by refill adherence, among patients with complicated and uncomplicated hypertension (i.e., patients who have and have not, respectively, experienced hypertension-related target organ damage), and to examine the association of refill adherence with hospitalization and health care costs among these patients. Design. Retrospective analysis of electronic medical records. Setting. An urban, public health care system. Patients. A total of 15,206 patients aged 18 years or older whose electronic medical records indicated a clinical diagnosis of hypertension based on the International Classification of Diseases, Ninth Revision, Clinical Modification codes, and who had received at least one prescription of an antihypertensive drug from 1995-2001. Measurements and Main Results. We used multivariable analyses to investigate the association of refill adherence with hospitalization and costs. On average, 53% of patients had appropriate supplies (80-120% of supplies needed), 7% had undersupplies, and 40% had oversupplies of drug annually For patients with complicated hypertension, an undersupply of drug was associated with a 15% greater probability (p=0.009) and an oversupply with a 16% greater probability (p < 0.0001) of hospitalization. Among those with uncomplicated hypertension, oversupply was associated with an 11% greater probability (p=0.0002) of hospitalization; undersupply was not associated with greater probability of hospitalization. Total health care costs were lower for patients with undersupplies and higher for those with oversupplies of drug. Conclusion. Among adults in an urban health care system with complicated hypertension, both undersupply and oversupply of drug were associated with increased hospitalization rates. Monitoring refill adherence of patients, particularly those with low income, minority status, and complicated hypertension, may be useful for targeting patients with undersupplies of drug to encourage refill adherence and identifying patients with oversupplies, who are at high risk of hospitalization.