Estimating energy requirements in obese and non-obese patients: a survey of dietetic practice in the United Kingdom

Abstract

Background: Many methods are available to determine energy requirements; however, all have limitations and their use in clinical practice is variable and not universally understood (Green et al., 2008). Estimating energy requirements of obese patients is particularly problematic (Breen et al., 2004). The aim of the current survey was to investigate current practice in the estimation of energy requirements in an obese and non‐obese patient in a large cohort of UK dietitians. Methods: A cross‐sectional web‐based survey of UK Registered Dietitians was performed. An opportunistic sample was recruited via e‐mail providing a link to the survey (contact details openly available on NHS Trust websites) and through the online newsletter of the British Dietetic Association. The anonymous online questionnaire was developed specifically for this project by dietitians working in nutrition support and was based on the structure of two previous surveys (Reeves et al., 2003; Green et al., 2008). Respondents were asked to estimate energy requirements using two theoretical case scenarios: one patient was obese and one was not. Demographic information including training, experience and job role were also collected. Data were analysed using SPSS, version 18 (SPSS Inc., Chicago, IL, USA); chi‐squared tests for independence, Kruskal–Wallis and Mann–Whitney U‐tests were performed. Results: Six hundred and seventy‐two responses were received from all areas of the UK. For the non‐obese patient, prediction equations and adjustment for metabolic stress and physical activity was used by 90.3% of respondents. The median estimated energy requirement was 8704 kJ (2079 kcal) [interquartile range (IQR): 8122–9295 kJ (1940–2220 kcal)] day−1. The median target volume of feed prescribed was 2000 (IQR: 2000–2000) mL day−1; significantly less than estimated requirements (P < 0.001). Estimated energy requirement using kcal/kg method was significantly lower compared to the equations method: 7536 kJ (1800 kcal) [range: 2428–10798 kJ (580–2579 kcal)] versus 8704 kJ (2079 kcal) [range: 2428–12385 kJ (580–2958 kcal)] day−1 (P < 0.001). For the obese patient, prediction equations to estimate basal metabolic rate (BMR) alone was used by 50.7% of respondents. Nutrition support dietitians used a lower stress factor compared to non‐nutrition support dietitians (10.3 ± 6.3 versus 13.6 ± 6.05 %; P = 0.016). The method used to estimate energy requirements was associated with years in clinical practice (P < 0.001); those in practice <5 years were more likely to use BMR alone when estimating energy requirements for the obese patient compared to those in practice >10 years. Respondents used a significantly lower kcal kg−1 for the obese patient (25 (IQR: 20–30) kcal kg−1) compared to the non‐obese patient (30 (IQR: 25–35) kcal kg−1) (P = 0.014). Discussion: This survey found the majority of respondents used prediction equations to estimate energy requirements which is similar to the results of previous surveys (Green et al., 2008). Many more respondents used BMR al...