| Abstract [eng] |
Background. Malnutrition prevalence in the intensive care unit (ICU) ranges from 38–78% and is independently associated with increased mortality, prolonged hospitalisation, and infectious complications. Accurate nutritional risk assessment and energy expenditure estimation represent key clinical challenges, addressed through methods ranging from clinical screening tools to instrumental measurements. The interrelationships and practical diagnostic value of these methods in critically ill patients remain insufficiently studied. Aim. To evaluate the association of biochemical markers and clinical nutritional assessment tools – nutritional risk scores and instrumental methods – with metabolic status changes and clinical outcomes in an intensive care patient population. Objectives. 1. To assess the association between NRS 2002 and mNUTRIC nutritional risk scores and clinical outcomes in ICU patients. 2. To determine the association of urea – creatinine ratio with metabolic phase transitions and to assess the significance of hypophosphataemia episodes as a metabolic marker in ICU patients. 3. To evaluate the informativeness of indirect calorimetry and bioelectrical impedance analysis compared to standard ESPEN formulae for energy and protein requirement assessment. Methods. Data from an ongoing prospective observational study (2023–2027) at Vilnius University Hospital Santaros Klinikos ICU was analysed. Seventy-six mechanically ventilated adult patients were included. Nutritional risk was evaluated using NRS 2002 and mNUTRIC scales. Hypophosphatemia episodes and urea-to-creatinine ratio dynamics were analysed as metabolic markers. Energy expenditure was measured using the COSMED Q-NRG indirect calorimeter (IC) and compared to the ESPEN 25 kcal/kg/day formula. Body composition was assessed by bioelectrical impedance analysis (BIA). Results. Mean age was 61,2 ± 14,1 years; males constituted 63,2%. ICU survival was 57,9% and hospital survival 52,6%. The mNUTRIC score was a significant predictor of ICU mortality (OR 1,47; p=0.007) and hospital mortality (OR 1,60; p=0.004). Agreement between NRS 2002 and mNUTRIC was poor (κ=0,10). The mean difference between ESPEN formula and IC was 90,6 kcal/day (p=0.014) with wide limits of agreement (-706.4 to +525.2 kcal/day). Only 38,9% of IC measurements fell within ±10% of the ESPEN target. Cumulative caloric discrepancy was significantly greater in patients with ICU stay exceeding 10 days (4354 vs 1933 kcal; p=0,0005). Protein requirements calculated from lean body mass and ESPEN formula did not differ significantly (p=0,07), but individual variation was substantial (-24,5 to +19,8 g/day). Urea-to-creatinine ratio analysis was limited due to insufficient measurement frequency in routine clinical practice. Conclusions. The mNUTRIC score is a reliable predictor of ICU mortality and outperforms NRS 2002 in the critically ill population. The ESPEN 25 kcal/kg/day formula systematically overestimates energy requirements and cannot capture individual metabolic variability, particularly during prolonged ICU stays. Indirect calorimetry remains the most accurate method for energy expenditure assessment. The potential of biochemical markers – urea-to-creatinine ratio and hypophosphatemia – as indicators of metabolic phase transitions warrants further investigation. |
