Reliability of new measures of cost-related medication nonadherence

Although several national studies have attempted to measure medication nonadherence due to cost in cross-sectional studies of the elderly and disabled, little information exists on the psychometric properties of these measures over time.

Examine the test-retest reliability of several recently published measures of cost-related medication nonadherence, among elderly community.

We developed a questionnaire and tested the reliability of measures of cost-related medication nonadherence and general cost-reduction strategies in a sample of 185 elderly in eastern Massachusetts surveyed twice (1-2 months apart). General and medicine-specific cost-related nonadherence measures included: failure to fill or delayed refilling of a prescription due to its cost, skipping doses, or taking smaller doses to make a medicine last longer. We also tested the reliability of reported drug cost-reduction strategies, such as: using generic drugs; purchasing prescriptions via mail/internet or from outside the United States; receiving prescription samples from a doctor; and spending less on food, heat, or other basic needs to afford medicines. We used the McNemar test, a matched pair chi analysis, and Kappa statistics to examine the association of responses within patients between identical items asked at 2 points in time.

Kappa statistics for test-retest reliability ranged from 0.6 to 0.9 for all but one measure of cost-related nonadherence, and McNemar test statistics indicated no systematic change in the measures over time.

The estimated test-retest reliability of the measures of cost-related medication nonadherence were generally high. The measures have been integrated into the nationally representative Medicare Current Beneficiary Survey (MCBS), an ongoing national panel survey of Medicare beneficiaries, which will allow researchers and policymakers to identify changes in cost-related nonadherence among disabled and elderly Medicare beneficiaries.