Assessment of capacity to comply with medication regimens in older patients

To develop an instrument that will facilitate and focus the assessment of a patient's capacity to adhere to a medication regimen before its initiation.

This is a crossectional study that compares medical inpatients and outpatients to an age-matched, community-living, independent and relatively healthy group on their ability to adequately understand and implement hypothetical but realistic medication regimens.

Department of Veterans Affairs Medical Center, Sepulveda, California.

Fifty-five older subjects (65 years or older) were divided into three groups: (1) generally healthy comparisons (standard group) (n = 20); (2) medical outpatients (n = 15); and (3) medical inpatients ready for discharge (n = 20).

Older subjects were first tested on their capacity to comply with a difficult medication regimen presented in scenario form. If scores on the first scenario did not meet a standard group-derived cutoff point, further testing was conducted with a simpler scenario to identify greater levels of impairment.

The outpatient group had significantly lower scenario scores than did the healthy comparison group (P < .03). The simpler scenario also showed a trend toward outpatient impairment (P = .06). In the comparison group, only 5% failed Scenario 1, and none failed Scenario 2. The outpatient group had the most difficulty, with 40% failing Scenario 1 and one-third of those failing Scenario 2. This differed significantly from the comparison groups (Fisher's Exact P < .03). In the inpatient group 20% failed Scenario 1 and 75% of those failing Scenario 2. The sensitivity and specificity of the Folstein Mini-Mental State Examination in identifying scenario-impaired subjects were 73% and 80%, respectively. Question types were analyzed to determine which questions were most frequently missed. Memory and judgment questions were found overall to be the most frequently missed. Healthy controls missed some judgment questions; however, the outpatient group was significantly worse in this category (chi 2 = 5.08; P = .01). All three groups improved their scenario performance significantly with question cueing.

A significant number of medically ill outpatients encountered difficulty in understanding or remembering correctly hypothetical but realistic medication regimens. This suggests that an older medical patient's cognitive and functional capacity to comply with medication regimens of differing complexity can be specifically assessed before the start of the regimen and probably should be assessed in patients whose compliance capacity is in question. The assessment instrument under development in this study may be helpful in detecting those who need assistance with medications, thus identifying the need for intervention before poor compliance can lead to increased morbidity, rehospitalization, and increased medical costs.