Effect of written instructions on accuracy of self-reporting medication regimen in glaucoma patients

Our purpose was to evaluate the accuracy of self-reporting of the prescribed medication regimen in a glaucoma population, identify contributing factors, and assess the effect of written instructions.

All patients at an urban resident glaucoma clinic were offered participation in this prospective, case-controlled study. Two trained interviewers administered a confidential questionnaire consisting of six questions regarding the name and dosage of ophthalmic medications, education level, ability to read, and age. A verbatim response for each question was recorded on the questionnaire. At the end of the visit, patients were given a written chart describing their ophthalmic medications, frequency, and dosage. At their next scheduled visit, the same questionnaire was repeated. The patients' responses from both visits were compared with the regimen they were prescribed and with each other. This information was used to determine the accuracy of reporting medications.

A total of 193 patients were enrolled in the study over a 10-month period; 164 patients completed both phases (85% completion rate). The study population consisted of 85 women and 79 men with a mean age of 68.40+/-11.6 years. Eighty-four patients had attained less than 12th grade education and 80 had completed high school. At the first visit, 66 patients (40%) showed less than 100% accuracy. Forty-nine of 84 (58%) patients who had not completed high school education showed less than 100% accuracy with a mean score of 65% (P=0.001), whereas 17 of 80 (21%) of patients who had completed high school showed less than 100% accuracy with a mean score of 87% (P=0.001). After written instruction, the accuracy of reporting improved by 23.36+/-30.8 percentile points in patients without completion of high school education and improved by 8.46+/-21.7 percentile points in patients who had completed high school (P<0.001). The mean number of ophthalmic medications prescribed was 2.10+/-0.93 (range 1 to 4). Patients on one medication had 100% accuracy in reporting 82% of the time, whereas those with four medications had 100% accuracy 21% of the time. Gender, age, and race of the patient were not correlated with the accuracy of self-reporting (P>0.05).

The education level of the patient and the number of medications showed direct correlation with patients' ability to report medications accurately. Patients showed improvement in accuracy of reporting medications when given written instructions about their regimen, regardless of their level of education or number of medications.