ATLANTA — More than one in five diabetic patients are nonadherent to their key medications—and the consequences soon show up in increased rates of hospitalization and mortality, Dr. P. Michael Ho said at the annual meeting of the American College of Cardiology.
Given this high rate of nonadherence and its marked negative impact, it seems appropriate to incorporate routine assessment of medication adherence into the clinical care of patients with diabetes, added Dr. Ho of the University of Colorado, Denver.
“Once medication nonadherence is identified, physicians should approach it similar to an elevated systolic blood pressure reading or a high LDL value; that is, it's a risk marker for adverse outcomes which requires treatment and follow-up,” he said.
Dr. Ho presented a retrospective cohort study of 11,532 diabetic patients at a large Denver-area nonprofit integrated managed care plan. The rate of nonadherence to oral hypoglycemic, statin, and antihypertensive therapies—all of which are recommended as core components of diabetes care in national guidelines based upon level I evidence—was 21% during the assessment year of 2003.
Nonadherence was defined as having filled prescriptions for these medications on less than 80% of days of follow-up. The lowest nonadherence rate was with antihypertensive medications (19%) while the highest (25%) was with statin therapy.
Outcomes were assessed during 2004 and the first one-third of 2005. Unadjusted all-cause mortality was 4% in adherent and 6% in nonadherent patients, a significant difference. The rate of all-cause hospitalization was 19% in the adherent group and 23% in nonadherent patients.
The mean LDL level in adherent patients was 85.5 mg/dL, compared with 98.2 mg/dL in the nonadherent subjects. Systolic and diastolic blood pressures and mean hemoglobin A1c values were also higher in the nonadherent cohort.
Nonadherent patients were significantly more likely to be female, younger, and had fewer comorbidities. Upon adjustment for these and other variables in a multivariate logistic regression analysis, medication nonadherence was associated with an 81% increase in the relative risk of all-cause mortality and a 58% increase in all-cause hospitalization. Each 25% increment in medication adherence was associated with a 25% relative risk reduction in mortality and a 17% decrease in hospitalization risk.
It's known from other studies that fewer than 50% of adults with diabetes achieve targets for LDL, blood pressure, and HbA1c. These data suggest medication nonadherence may be the explanation, Dr. Ho continued.
In terms of potential remedies for the poor rate of medication adherence, Dr. Ho said this problem deserves to be a high-priority research area. He noted there are encouraging reports out of Stanford (Calif.) University regarding the use of interactive voice-response technology when applied to groups of diabetic patients.
This system is similar to the automated telephone technology the airlines use when customers call in to make a reservation. The diabetic patient can phone in to the interactive system, or the system can phone the patient. Studies to date have shown this telephone-based strategy results in improved medication adherence and intermediate outcomes, according to Dr. Ho.