Development of a foot care knowledge test for elderly people with diabetes

Abstract

<p>Purpose of the study was to develop a valid and reliable instrument to measure foot care knowledge in elderly people with diabetes. Utilizing a three phase design, Phase I focused on survey of foot care content deemed important by a geographically random, stratified sample of 90 diabetes nurse educators. With 71% of subjects responding, results indicated that Foot/Nail Care, Footwear/Shoes, and Footwear/Socks, Stockings, Hosiery were considered the three most important aspects of foot care related to elderly people with diabetes. Data were used in Phase II to design a 25 item, multiple choice foot care knowledge test corresponding to a 3rd grade reading level with face and content validity as judged by diabetes experts. Pilot testing of the instrument was conducted in Phase III of the study using 102 elderly people with diabetes and 103 without diabetes. Results yielded an index of internal consistency of.80 utilizing KR20 formula with a modest test-retest correlation of.61 as derived from a subset of respondents with diabetes. Construct validity of the instrument was established via principal components factor analysis indicating the presence of a one factor model as denoted by negligible to low partial correlations between test items; scree plot of eigenvalues; roots and proportion variance criteria; and, measure of sampling adequacy. Further construct validity of the instrument was established employing criterion group analyses in which subjects with diabetes, subjects with diabetes who received concentrated foot care education, and subjects having received any formal diabetes education all scored significantly higher on the foot care knowledge test than elderly subjects without diabetes. As a composite, the foot care knowledge test yielded an item difficulty index of 64% with values ranging from 22% to 92%. All items exhibited positive discriminating power to distinguish high from low test scorers with the test, as a whole, yielding an average discriminating power of.68 with values ranging from.57 to.82. Test item distracters (i.e., incorrect responses) were effective in attracting selection by a greater proportion of low versus high test scorers. Psychometric evaluation of the instrument and indices of item analyses indicate the Diabetes Foot Care Knowledge Test represents a valid and reliable instrument for use with elderly patients with diabetes.</p>