Engaging patients and families in nutritional monitoring: A validation study of food record charts to quantify food intake

BACKGROUND & AIMS: Malnutrition is common in hospitalised patients and contributes to poor clinical outcomes. To support adequate nutritional intake in patients, accurate assessment of dietary food intake is critical, but it remains challenging and time-consuming. The present study aims to assess how accurate patients and family members estimate food intake using food record charts (FRCs) compared with weighed food records (WFRs). METHODS: In a cross-sectional study, 30 patients (≥18 y, Dutch-speaking, no delirium and no isolation restrictions) and 30 family members (≥18 y, Dutch-speaking, non-healthcare professionals) estimated simulated food consumption of nine different hospital meals (three breakfasts, three lunches, and three dinners) consisting of 79 different food items with FRCs, and these estimates were compared to WFRs. Subgroup analyses were performed for food consumption estimations by food item, including energy and protein content, food consistency, consumed amount, and food groups. Bland-Altman plots and inter-rater agreement were used to identify the accuracy of food intake estimation. Values are presented as mean ± SD. RESULTS: Food consumption estimated by patients using FRCs was comparable to food consumption measured by WFRs with a mean overestimation of 1.2 ± 8.1 % (p = 0.178), whereas family members overestimated intake by 2.2 ± 7.5 % with FRCs compared to WFRs (p = 0.012). Protein-dense products (>10 g/100 g) were underestimated by ∼2 %, while products with lower consumption (