While medication therapy for control of chronic conditions such as hypertension, diabetes, and hyperlipidemia has improved dramatically over the past 20 years, control of these diseases and overall risk reduction for patients with chronic conditions remain uniformly poor. Many patient level characteristics, such as age, gender, degree of comorbidity, and financial hardship are predictive of adherence to medications, but they are extremely difficult to modify. The delivery of health care, on the other hand, is in the control of the health care organization. To date there has been limited research examining the characteristics of health care systems that support adherence to medications through better education, easier access, more subspecialty resources, or special quality improvement programs for specific diseases.
The aims of this study were to 1) identify variation in medication adherence at the clinic-level, 2) identify factors or combinations of factors that are associated with adherence and can also be modified by clinical or policy decisions to improve medication adherence and patient outcomes, and 3) identify patient perspectives of clinic-level support for medication adherence and patient self-management including facilitators and barriers to high adherence.
We identified a retrospective cohort of 444,418 VA primary care patients with diabetes treated in 559 clinics in FY 2006 to document variation in medication adherence across VA primary care clinics in FY 2007. Secondary data included patient-level diagnostic, clinical, and pharmacy data from VA national administrative databases, and organizational data from the VA Clinical Practice Organizational Survey. In addition, we designed and administered a new organizational survey to collect data from primary care providers about decision support and patient self-management support. We selected 82 different primary care clinics for the survey and approached multiple providers at each clinic. We used telephone interviews to help clarify discrepant responses from the same facility and to follow-up about specific programs and activities to improve adherence. Finally, we collected new organizational data from patient volunteers who attended focus group sessions about patient perspectives on factors that affect medication adherence. We used bivariate and multivariate statistical analyses to identify variation in adherence and factors associated with improved adherence and content analysis of focus group data.
Aim 1: We found a high level of variation in the percent of patients adherent to diabetes medications in primary care clinics, ranging from 56% to 81%. To compare adjusted adherence at the clinic level, we developed a patient-level risk adjustment model, which shows patient demographic characteristics available in administrative data explained 2.94% of the variation in adherence. The result of risk adjusted adherence at the clinic level found that the variation in the percent of patients adherent across clinics remained large (range: 58% - 81%). Differences in adherence between the top 25 clinics and bottom 25 clinics were statistically significant. The adjusted clinic adherence between affiliated clinics within the same health care system, i.e., VAMC, was statistically correlated, but adherence between unaffiliated clinics was uncorrelated. Aim 2: The results of the organizational surveys indicate that a number of organizational factors (provider-run group visits for diabetic patients, group patient education, individual patient education, pharmacist generated prescription renewal, and RN-generated interim refill) significantly predict adherence. However, inter-rater reliability among primary care staff at the same clinic was poor to fair. The overall survey response rate was 68% of the 82 clinics. Aim 3: The preliminary results of the patient focus group sessions indicate three emerging themes around facilitators of high adherence: the use of pill boxes, daily routines to remember to take medications, and a willingness to call the VA if/when there are problems with refills or renewals. Preliminary barriers include overall quality of pharmacy services (e.g. long wait times, staff with poor customer service skills, etc), confusion caused by having multiple prescription medications that need to be refilled at different times, and lifestyle barriers that cause individuals to miss or forget medications. Participants also offered a number of suggestions for how the VA could help them take their medications. The two most common suggestions were to align the refill dates of multiple prescription medications so patients do not constantly have to order refills and to increase the education provided to patients about diabetes self-management.
The discovery of significant variation in clinic-level adherence is important. The finding that clinic-level adherence was highly correlated among affiliated clinics within a health care system suggests that health care systems may play an important role in supporting patient adherence and self-management. Clinics or health care systems should consider specific programs to improve medication adherence, such as providing diabetic patient education using different approaches, improving processes for obtaining prescription renewals and refills, and improving quality of overall pharmacy services. Future work should cultivate and track organizational factors identified by patients and providers in this study, which improve patients' ability to self-manage diabetes and are not generally recorded in traditional data sets.
External Links for this Project
None at this time.