Reliability and Validity of the Spoken Knowledge in Low Literacy in Diabetes in Measuring Diabetes Knowledge Among Hispanics With Type 2 Diabetes

Abstract

Purpose The purpose of the study was to examine the reliability and validity of the Spoken Knowledge in Low Literacy in Diabetes (SKILLD) scale in measuring diabetes knowledge among Hispanics with type 2 diabetes (T2DM). Methods A total of 111 Hispanic participants with T2DM were recruited from clinics and churches in the Southeastern US. The Spanish version of the SKILLD was used. Internal consistency reliability and exploratory factor analysis (EFA) were examined for the Spanish SKILLD. Spearman rank correlations of SKILLD scores with the Summary of Diabetes Self-Care Activities (SDSCA) scores were inspected for evidence of convergent validity along with A1C level and duration of diabetes. Results The average SKILLD score was 35.1% (SD = 23.4%). The correct response for each SKILLD item ranged from 4.5% to 56.8%. The estimated reliability via internal consistency was adequate (Kuder-Richardson 20 = 0.706). EFA of the SKILLD items modestly indicated 1 factor could be retained. Spearman rank correlations of SKILLD scores with diabetes self-care activities were significant and positive for foot self-care. Higher SKILLD scores were significantly correlated with duration of diabetes, using insulin, having a high education level, ever attending a diabetes class, or having a parent with diabetes. Conclusion The Spanish SKILLD showed acceptable reliability and adequate validity in this sample of Hispanics with T2DM. SKILLD scores indicated low diabetes knowledge in this sample. Patient-centered diabetes education tailored to low literacy needs of Hispanics is needed.