Development and Evaluation of a Low‐Literacy Discharge Medication Education Tool

Background:

Medication nonadherence is a risk factor for poor clinical outcomes among cardiac patients. Using a randomized, controlled clinical trial, we developed and evaluated a low‐literacy picture‐ and icon‐based discharge medication education tool for teaching patients about their medications at the time of hospital discharge.

Methods:

We developed a prototype of a printable outpatient medication schedule, customizable for each patient, featuring instruction‐specific icons and color pictures of the pills a patient receives at hospital discharge. This prototype was shown to approximately 30 nurses in shift change meetings, with feedback used to make rapid‐cycle serial improvements until saturation was reached. Additionally, 10 nurse‐patient pairs pilot‐tested the tool and provided feedback to the principal investigator. To evaluate the tool, we conducted a randomized, controlled clinical trial testing its effectiveness on postdischarge medication adherence. At 2 weeks postdis‐charge, via telephone interview, we collected self‐reported medication adherence. To validate this self‐report, we again interviewed patients 4 weeks postdischarge and asked each participant to count the number of pills she or he had remaining for each medication. We also used automated pharmacy data to assess for delay in refilling each medication from the 30‐day supply given at discharge. In a self‐administered survey, we asked each discharging nurse to rate the acceptability of the tool and report time spent teaching the patient.

Results:

Average age of our participants was 56 (SD 12). Most participants were Latino (75%); 13% were African American. Most (62%) were uninsured, 46% had limited English proficiency, and 47% had inadequate health literacy. These characteristics did not vary significantly by randomization arm. Of the 210 participants randomized and eligible for the week 2 interview, 79% completed this assessment. Mean self‐reported medication adherence at week 2 was 70% (95% CI: 62%‐79%) among intervention participants compared with 78% (95% CI: 72%‐84%) among controls (P = .13). Week 2 self‐reported adherence was significantly correlated with pill counts (P = .008) and refill delays (P = .048). Nurses implementing the intervention arm returned 66 of 135 distributed surveys. Of these, 74% indicated that if the tool were available routinely, they would use it for “every patient.” There was a nonsignificant trend toward nurses spending less time teaching when the tool was utilized.

Conclusions:

We developed a low‐literacy icon and picture‐based tool for teaching cardiac patients about their medications at the time of hospital discharge. The nursing staff found the tool to be very useful, and there was a trend toward less time spent when using the tool. However, secondary to unexpectedly wide confidence intervals for self‐reported adherence and a high rate of adherence in the control group, we were unable to demonstrate the tool's effect on postdischarge medication adherence.

Author Disclosure:

K. Cordasco, none; S. Asch, none; J. Guterman, none; S. Gross‐Schulman, none; L. Ramer, none; D. Bell, none; I. Franco, none; C. Mangione, none.