|
BACKGROUND/RATIONALE:
Assessment of medication adherence is an important tool in health services research and clinical care. One common measure of adherence obtainable from computerized records is the "medication possession ratio" (MPR), which reflects the reliability with which patients refill prescriptions. MPR is typically calculated by dividing the days' supply of medication dispensed by the period over which prescriptions are filled (as determined from refill intervals). Though the literature on MPR is extensive, no studies have examined the effect, in aggregate, of the different choices that are entailed in its computation. In addition, there are no studies that have compared the relationship of MPR to outcomes across a number of conditions, much less how this might be affected by different methods of calculation.
OBJECTIVE(S):
This study will evaluate refill-based adherence measures for three common and important medical conditions: diabetes, hypercholesterolemia, and hypertension. Objective I: To determine how the calculation of MPR is affected by the following decisions: 1) The number of refills over which it is measured; 2) The duration of time that is used to identify a therapeutic decision to interrupt treatment versus extreme lack of adherence; 3) The exclusion of "excess" drugs that are presumed to accumulate from the early refilling of prescriptions; and 4) the means by which adherence is estimated when more than one drug is used; Objective II: To optimize calculation of MPR for each selected medical condition; Objective III: To compare the relationship between the optimized measures of adherence and treatment success in patients with the selected medical conditions.
METHODS:
This will be an observational study using VA administrative data. The principal type of analysis will be outcomes assessment. Specific types of data will include outpatient prescriptions, laboratory data, demographics and relevant diagnoses. The population to be studied will be patients who are receiving outpatient medication therapy for diabetes mellitus, hypercholesterolemia or hypertension. For objective I, MPR will be assessed while varying the specified parameters. For objectives II and III, appropriate regression analyses will be performed using outcomes of hemoglobin A1c, low-density lipoprotein and blood pressure.
FINDINGS/RESULTS:
Choices made in calculation of MPR can substantially affect the percentage of patients who are deemed "non-adherent". The percentage of prescriptions for oral diabetes medications that had MPR<0.8 varied from approximately 5%-40% depending upon choices made in the calculation and sample stratifications. Separately varying the number of intervals between fills from 1-4 or the "clean periods" from 1-4 times the days' supply resulted in up to 2 to 3-fold differences in the percentage of prescriptions with MPR<0.8
In patients with diabetes, we found a significant association between MPR and hemoglobin A1c (HbA1c), a measure of diabetes control. We examined several subsets of patients, including those on monotherapy with oral medications who had at least five 90-day prescription fills. Patients on "high" or "low" dose monotherapy regimens with sub-optimal adherence, as reflected by MPR <0.8, were less likely to have HbA1c <7%. We found that those on higher medication doses had worse overall HbA1c control. The seemingly paradoxical observation likely reflects the presence of more severe diabetes, so that simply increasing the medication dose while maintaining monotherapy was insufficient to control blood sugars. This suggests that addressing both patient non-adherence as well as the need for adequate intensification of medications by providers is needed to improve overall diabetes control in the VA population.
IMPACT:
This study will assess the optimal way to calculate refill-based adherence measures for three common, important, and often inadequately controlled medical conditions: diabetes, hypertension and hyperlipidemia. This information will be useful in the care of individual patients and will also be of use in population-based assessments, including levels of adherence, the impact of non-adherence on outcomes and the effectiveness of system-wide efforts to improve adherence.
PUBLICATIONS:
None at this time.
DRA:
Health Services and Systems
DRE:
Resource Use and Cost, Quality of Care
Keywords:
Research measure, Adherence
MeSH Terms:
none
|
