Potential for drug interactions in seniors with osteoarthritis.

OBJECTIVE To document the potential for drug interactions in seniors with osteoarthritis and to consider the usefulness of computerized support for detecting clinically important interactions.

DESIGN Self-administered mailed survey. One question requested a list of all medications (prescribed drugs and self-care products, including herbal and "natural" health products) taken in the last 7 days. Interactions among all medications were assessed using an on-line software package.

SETTING Three urban primary care practices in Nova Scotia.

PARTICIPANTS Questionnaires were sent to 244 patients aged 65 years and older with physician-confirmed osteoarthritis.

MAIN OUTCOME MEASURES Number of potential interactions and level of clinical significance associated with each.

RESULTS Response rate was 78% (n = 191); 174 respondents (92%) supplied information on medications. Respondents took an average of 4.7 products of which 2.8 were prescription medications and 1.9 were self-care products. A total of 214 potential interactions were identified; 30 (14%) of these were clinically significant. Most interactions involved nonprescription products, most frequently acetylsalicylic acid. Recommendations in 29 of these 30 clinically significant interactions were cautionary, advising such measures as closer monitoring of blood tests, observation for toxic effects, or making patients aware of side effects. Only 1 interaction prompted a recommendation for avoidance. Respondents reported use of 7 different herbal and natural health products; these products were associated with 5 clinically insignificant interactions.

CONCLUSION Risk of drug interactions in seniors might be high, but few interactions are clinically significant. Only 1 found in our study carried a recommendation for avoidance. The on-line program reported all significant interactions, but the high proportion of insignificant interactions (6 : 1) also reported could lead physicians to override computer-generated alerts.