Patients with low literacy have problems understanding medical instructions, reading prescriptions, and managing chronic diseases. Many patients with diabetes have low literacy. It is not known whether interventions targeting literacy are associated with improvements in diabetes self-management. Rothman and colleagues conducted a randomized controlled trial to determine whether a comprehensive strategy aimed at low-literacy patients could improve diabetes-related outcomes.

The study included English-speaking adults with poorly controlled type 2 diabetes in a general internal medicine practice. Before randomization, all patients attended a one-hour educational session about diabetes treatment recommendations. Control patients received usual care from their primary care physicians. Intervention patients had intensive diabetes management with one-to-one educational sessions, application of evidence-based guidelines, and strategies aimed at eliminating logistical and practical barriers that could prevent optimal care, such as transportation, communication, and insurance issues. Lead personnel included pharmacists and a diabetes care coordinator. This staff had information about patients’ literacy status and tailored their approach accordingly. For low-literacy patients, pictures and simple, verbally-based explanations were used.

Hemoglobin A1C levels and systolic blood pressure were measured at baseline, six, and 12 months. Literacy was assessed at enrollment. Improvement of goal A1C and blood pressure levels was analyzed, looking at differences between control and intervention groups stratified by literacy level. Literacy less than or equal to a sixth-grade reading level was considered low literacy. The study was powered to detect a 1 percent difference in A1C between groups and a 10-mm Hg difference in blood pressure, but not to detect differences by literacy status.

In general, the study population had poor glycemic control and low socioeconomic status. More than one third had low literacy. A1C levels improved significantly more in patients in the intervention group than in the control group (2.1 percent improvement compared with 1.2 percent). Only small improvements were found when high-literacy patients were compared in the two groups, whereas low-literacy patients had greater improvement than their control counterparts. Low-literacy patients in the intervention group were more likely to reach target A1C levels than control patients at 12 months. This difference was not maintained in high-literacy patients. Regarding systolic blood pressure, intervention patients did better than control patients (with a 7.6-mm Hg difference), but in this instance, the benefit was similar for low- and high-literacy groups, and this applied to patients who reached goal systolic blood pressure at 12 months.

The authors conclude that interventions to help patients with diabetes lower A1C levels preferentially benefit patients with low literacy, even after adjusting for baseline characteristics. The authors attribute this benefit to specific techniques, including simplification of instructions and asking patients to relay back what they learned to ensure the information was understood. However, this benefit did not carry over to blood pressure management. It is unclear why systolic blood pressure improvement did not vary by literacy status, but it may be that the reason for success in lowering blood pressure was more physician- than patient-dependent. This study was small and insufficiently powered, which limits its generalizability.