Nutrition-related outcomes and dietary intake in non–mechanically ventilated critically ill adult patients: A pilot observational descriptive study

Background: Critically ill patients who do not receive invasive mechanical ventilation (IMV) are a growing population, experiencing complex interventions that may impair dietary intake and nutrition-related outcomes. Objectives: The objectives of this study were to quantify intake and nutrition-related outcomes of non-IMV critically ill patients and to establish feasibility of methods to measure nutrition-related outcomes in this population. Methods: Non-IMV adult patients expected to remain in the intensive care unit (ICU) for ≥24 h were eligible. Nutrition-related outcomes were assessed at baseline by subjective global assessment (SGA); on alternate study days by mid-upper arm circumference (MUAC), calf circumference (CC), and ultrasound of quadriceps muscle layer thickness (QMLT); and daily by body weight and bioelectrical impedance analysis (BIA). Data were censored at day 5 or ICU discharge. Dietary intake from all sources, including oral intake via investigator-led weighed food records, was quantified on days 1–3. Feasibility was defined as data completion rate ≥70%. Data are expressed as mean (standard deviation) or median [interquartile range (IQR)]. Results: Twenty-three patients consented (50% male; 53 [42–64] y; ICU stay: 2.8 [1.9–4.0] d). Nutrition-related outcomes at baseline and ICU discharge were as follows: MUAC: 33.2 (8.6) cm (n = 18) and 29.3 (5.4) cm (n = 6); CC: 39.5 (7.4) cm (n = 16) and 37.5 (6.2) cm (n = 6); body weight: 95.3 (34.8) kg (n = 19) and 95.6 (41.0) kg (n = 10); and QMLT: 2.6 (0.8) cm (n = 15) and 2.5 (0.3) cm (n = 5), respectively. Oral intake provided 3155 [1942–5580] kJ and 32 [20–53] g protein, with poor appetite identified as a major barrier. MUAC, CC, QMLT, and SGA were feasible, while BIA and body weight were not. Conclusions: Oral intake in critically ill patients not requiring IMV is below estimated requirements, largely because of poor appetite. The small sample and short study duration were not sufficient to quantify changes in nutrition-related outcomes. MUAC, CC, QMLT, and SGA are feasible methods to assess nutrition-related outcomes at a single time point in this population.