How do guideline recommended energy targets compare with measured energy expenditure in critically ill adults with obesity: A systematic literature review

Background: Critically ill patients with obesity have unique and complex nutritional needs, with clinical practice guidelines conflicting regarding recommended energy targets. The aim of this systematic review was to 1) describe measured resting energy expenditure (mREE) reported in the literature and; 2) compare mREE to predicted energy targets using the European (ESPEN) and American (ASPEN) guideline recommendations when indirect calorimetry is not available in critically ill patients with obesity. Methods: The protocol was registered apriori and literature was searched until 17th March, 2022. Original studies were included if they reported mREE using indirect calorimetry in critically ill patients with obesity (BMI≥30 kg/m²). Group-level mREE data was reported as per the primary publication using mean ± standard deviation or median [interquartile range]. Where individual patient data was available, Bland–Altman analysis was used to assess mean bias (95% limits of agreement) between guideline recommendations and mREE targets (i.e. ASPEN for BMI 30–50, 11–14 kcal/kg actual weight compared to 70% mREE and ESPEN 20–25 kcal/kg adjusted weight compared to 100% mREE). Accuracy was assessed by the percentage (%) of estimates within ±10% of mREE targets. Results: After searching 8019 articles, 24 studies were included. mREE ranged from 1607 ± 385 to 2919 [2318–3362]kcal and 12–32kcal/actual body weight. For the ASPEN recommendations of 11–14 kcal/kg, a mean bias of −18% (−50% to +13%) and 4% (−36% to +44%) was observed, respectively (n = 104). For the ESPEN recommendations 20–25 kcal/kg, a bias of −22% (−51% to +7%) and −4% (−43% to +34%), was observed, respectively (n = 114). The guideline recommendations were able to accurately predict mREE targets on 30%–39% occasions (11–14 kcal/kg actual) and 15%–45% occasions (20–25 kcal/kg adjusted), for ASPEN and ESPEN recommendations, respectively. Conclusions: Measured energy expenditure in critically ill patients with obesity is variable. Energy targets generated using predictive equations recommended in both the ASPEN and ESPEN clinical guidelines have poor agreement with mREE and are frequently not able to accurately predict within ±10% of mREE, most commonly underestimating energy needs.