Nonadherence to insulin therapy in low-income, type 2 diabetic patients

To evaluate the psychosocial barriers to insulin use in low-income, type 2 diabetic patients; the clinical characteristics of these patients; and the possible causes of nonadherence to insulin regimens months after prescription.

We studied a prospective cohort of low-income patients with type 2 diabetes mellitus, aged 45 to 75 years, attending a tertiary health care center in Mexico City, Mexico. Patients were eligible if their diabetes was not controlled with oral agents, and they were excluded if they had type 1 diabetes mellitus, a secondary cause of diabetes, had been admitted to the hospital within the month before study commencement, had been previously treated with insulin, had severe diabetic complications, or had a chronic or disabling medical illness. All patients were prescribed 6 to 10 units of neutral protamine Hagedorn insulin before bedtime and received a referral to visit with a diabetes nurse educator. The main outcome was adherence or nonadherence to insulin therapy, and it was correlated with several variables including attitudes toward insulin, diabetes self-management, diabetes-related knowledge, depression, and diabetes-related distress. Brief medical history, complete battery of questionnaires, and laboratory workup were obtained at baseline and 1 to 3 months after insulin prescription.

Twenty-nine consecutive patients were included. Mean age (+/- standard deviation) was 59 +/- 8 years, 18 (62%) were women, mean diabetes duration was 14 +/- 9 years, and mean hemoglobin A1c level was 10.8 +/- 1.4%. Negative attitudes toward insulin were very common, particularly in patients with less education and poorer diabetes-related knowledge (odds ratio, 6.2; 95% confidence interval, 1.04-47.3; P = .02). Even when they received precise recommendations, 12 patients (41%) did not adhere to insulin treatment. Patients who did not adhere to therapy were most commonly women and were depressed (P = .05). Improved adherence was significantly associated with the additional support of a diabetes nurse educator (odds ratio, 6.6; 95% confidence interval, 1.0-55.7; P = .02).

Improving patient perception and acceptance of insulin with the help of diabetes educators can facilitate earlier and more aggressive intervention and thus optimize glycemic control.