Background: Low Health Literacy (HL) is associated with poor health outcomes (1), higher utilization of care, increased risk for hospitalization and increased risk for 30-day rehospitalization (2). HL is composed of three components which includes obtaining the information (reading), comprehending the information and applying it to one’s daily life (3). Patients may master one component, but struggle in other areas. To date, there is little data to identify which groups of patients are at higher risk for lower HL. The aim of this study was to evaluate if there are differences between reading level and comprehension scores across various demographics for patients admitted to the hospital.
Methods: This cross-sectional study was conducted from September 2021 to April 2023 on adult patients admitted to John Dempsey Hospital. All adult patients were eligible for screening, except those with altered mental status, those with airborne precautions, patients from the Department of Corrections, or those in the ICU. Screening was conducted using iPads with questions from the Rapid Estimate of Adult Literacy in Medicine-Short Form (REALM-SF). Information was also collected on age, gender, race, and ethnicity. Two HL scores were tabulated. Patients were first asked to read 7 words within 5 seconds from the REALM-SF. 1 point was given for each word read correctly for a maximum total of 7 points. An additional step to the REALM-SF was added by asking the patient to define the same word. A score of 1 point was given for correct definitions based on the Oxford Dictionary, and 0 points for partial, incorrect definitions, or if the patient did not know the definition. Binary logistic regression was calculated to determine if REALM-SF reading scores were dependent on age, sex, race, ethnicity, gender or comprehension scores. Paired t-tests were used to examine the mean reading and comprehension scores separated based on race, ethnicity, age and gender. All statistics were calculated using SPSS v29.
Results: There were 466 patients screened by 85 volunteers for HL through REALM-SF with a mean reading score of 6.6 and mean comprehension score of 3.9. Binary logistic regression found that those with perfect reading scores were 574% more likely (p=< 0.001; 95% CI 4.0-12.3) to score higher on their comprehension score. Hispanic patients were 71% (p-0.019 CI: 0.10-0.82) less likely to do well compared to non-Hispanics. Black patients had lower reading (6.1 vs. 6.7, p=0.002) and comprehension scores (3.4 vs. 4.0, p=0.021) compared to non-black patients. Female patients had higher comprehension scores compared to men (4.2 vs. 3.7, p=0.003). In terms of age, patients in the highest age cohort (83-97 years of age) had a lower mean comprehension score compared to the mean of other age groups (3.41 vs. 4.04 p=0.003).
Conclusions: There are significant differences in both reading and comprehension scores across various demographic variables. Male, black, and older patients have shown to have lower levels of comprehension for medical terminology, meaning that simply asking a patient to repeat terms or handing them written literature upon discharge may put them at higher risk to not treat their disease appropriately. Furthermore, our data found a strong likelihood between predicting one’s performance on the REALM-SF and ability to define that word correctly. This further validates that our method accurately assessed both types of HL and that future research needs to be completed to determine why these differences are present.
