Comparison of the Interobserver Variability of 2 Different Methods of Dietary Assessment in a Geriatric Ward: A Pilot Study

Can hospitalized patients adequately estimate their own food intake? A cross-sectional pilot study

ABSTRACT Objective To assess and identify factors linked to the accuracy of patients’ food intake estimations through a self-monitoring instrument filled in by the patient. Methods This cross-sectional study approached adult hospital patients subjected to regular or therapeutic diets. The actual food intake percentage was obtained by the ratio between the actual food intake and the amount of food served x 100. Patients had to complete a food chart including 9 circles representing plates of food in percentages (increasing scale of 12.5%) to represent their food intake at lunch and/or dinner. The Bland-Altman method assessed the agreement between the actual and the estimated values. The associations between variables (age, sex, hospitalization day, diet prescription, amount of food served and actual food intake percentage) and the accuracy of the food intake estimation (adequate ±10%, overestimated and underestimated) were evaluated through univariate multinomial logistic regression. Results Ninety-six patients were evaluated (51.0% male; 44.0±15.8 years of age). The Bland-Altman analysis showed good agreement between the actual and the estimated food intake. The actual food intake percentage was the only variable associated with the accuracy of the food intake estimation. Conclusion Most patients (~70%) adequately estimated their food intake using the 9-point food chart tested. Furthermore, the only factor linked to the accuracy of the food intake estimation was the actual food-intake percentage. These findings provide preliminary support for the usefulness of this instrument. However, it must be tested in a representative sample of hospitalized patients.

Using digital photography in a clinical setting: a valid, accurate, and applicable method to assess food intake

BACKGROUND/OBJECTIVES

Regular monitoring of food intake is hardly integrated in clinical routine. Therefore, the aim was to examine the validity, accuracy, and applicability of an appropriate and also quick and easy-to-use tool for recording food intake in a clinical setting.

SUBJECTS/METHODS

Two digital photography methods, the postMeal method with a picture after the meal, the pre-postMeal method with a picture before and after the meal, and the visual estimation method (plate diagram; PD) were compared against the reference method (weighed food records; WFR). A total of 420 dishes from lunch (7 weeks) were estimated with both photography methods and the visual method. Validity, applicability, accuracy, and precision of the estimation methods, and additionally food waste, macronutrient composition, and energy content were examined.

RESULTS

Tests of validity revealed stronger correlations for photography methods (postMeal: r = 0.971, p < 0.001; pre-postMeal: r = 0.995, p < 0.001) compared to the visual estimation method (r = 0.810; p < 0.001). The pre-postMeal method showed smaller variability (bias < 1 g) and also smaller overestimation and underestimation. This method accurately and precisely estimated portion sizes in all food items. Furthermore, the total food waste was 22% for lunch over the study period. The highest food waste was observed in salads and the lowest in desserts.

CONCLUSIONS

The pre-postMeal digital photography method is valid, accurate, and applicable in monitoring food intake in clinical setting, which enables a quantitative and qualitative dietary assessment. Thus, nutritional care might be initiated earlier. This method might be also advantageous for quantitative and qualitative evaluation of food waste, with a resultantly reduction in costs.