Evaluation of nurse practitioners' assessment of patient reading literacy levels

Abstract

Based on results of the National Adult Literacy Survey of 1992, it was possible one fourth of patients who received health care from nurse practitioners may not have been able to understand instructions on prescription bottles, or patients’ education materials, use written directions for tests, or give true informed consent. Because of shame and fear, patients do not tend to self-report low literacy. This study compared nurse practitioners’ utilization of indirect literacy assessment, that is, assessment strictly based on clinic interaction, and direct literacy assessment, by a word recognition test. The Rapid Estimate of Adult Literacy in Medicine-Revised (REALM-R) was administered to 35 patients to screen for low literacy. Two nurse practitioners were asked “Do you think this patients reads at or above the 6th grade level, or below the 6th grade level?” after clinical interaction with the patients. Nurse practitioners accuracy, utilizing the indirect method of literacy assessment, the clinic interaction, was accurate 83% of the time, (n=29). Based on direct assessment by REALM-R testing, nurse practitioners inaccurately assessed patient reading literacy 17% (n=6) of the time. Of the patients who were inaccurately assessed for literacy levels, 67% of the time the reading skill level was overestimated, (n=4), and 33% of the time (n=2) the literacy level was underestimated. Chi-square results indicated no significant relationship between nurse practitioners’ indirect assessment methodology and REALM-R instrumentation. Nurse practitioners correctly assessed patient reading literacy skill levels in approximately 8 of 10 patients based solely on clinical interaction. When inaccuracy occurred, the tendency was to overestimate patients’ literacy ability; in the study, 12% of patients were unable to read six of eight common medical words, indicating a reading skill level below the 6th grade. Written materials, such as most prescriptions and package inserts, consent and registration forms, and many patient education materials would be intelligible for those four patients placing them at greater risk for poor health outcomes associated with poor literacy skills. Nurse practitioners could have utilized the REALM-R word recognition tool to detect patient illiteracy not recognized during the clinical interaction.