Reliability and validity of an adherence score sheet to monitor adherence of patients with diabetes to personalised nutrition education

Abstract

BACKGROUND : Dietary management of diabetes relies on the patient's ability to adhere to the nutrition guidelines presented to them by a health care professional. Traditional analysis of dietary intake to monitor adherence can be tedious, and there is a need for a short, easy-to-use tool to measure dietary adherence from food records. This study aimed to determine the reliability and validity of an Adherence Score Sheet (ASS) developed to quantify dietary adherence to personalised nutrition education. METHODOLOGY : 67 three-day food records of patients with diabetes who received nutrition education were scored using the ASS Adherence scores were given for adhering to glycemic index (GI), glycemic load (GL), protein and fat guidelines, and a total adherence score. Intraclass correlation (ICC) and Bland–Altman plots were used to demonstrate the amount of agreement between ASS from two independentraters (inter-rater reliability) and ASS score from Rater 1 [test method] vs ASS obtained from dietary analysis (DA) program [reference method] for the assessment of concurrent validity. RESULTS : ICC values for the total adherence score showed good agreement (ICC = 0.74) between raters, with subcategories agreement being fair to excellent (ICC ranging from 0.56 to 0.81). The Bland–Altman plots for all 5 categories (GI, GL, protein, fat, and total score) indicated acceptable agreement between Rater 1 and Rater 2. ICC values for all categories indicated excellent validity between Rater 1 and the DA scores. The Bland–Altman plots for validity indicated overall acceptable agreement between Rater 1 and DA for all 5 categories. CONCLUSION : Our findings reveal that the ASS is a reliable and valid tool to determine and quantify adherence to personalised nutrition education among patients with diabetes.