82 CHAPTER 5 5.2.2 Data collection Data were extracted from the electronic hospital information system (EPIC Systems, Wisconsin, USA). Collected parameters included demographic data, reason for admission (acute or elective, surgical or medical), body mass index (BMI), comorbidities (presence of sepsis and active malignant disease), baseline laboratory measurements (C-reactive protein, albumin, and creatinine), the APACHE III score over the first 24 h of ICU-admission, length of stay at ICU and hospital, ICU and hospital survival, and survival after 3, 6, and 12-months. 5.2.3 Bio-impedance analysis In line with standard care protocol, BIA-measurements were performed preferably within hours after ICU-admission and no later than 24 h after ICU-admission. To perform BIA, two electrodes were placed on the wrist and dorsal site of the hand and on the ipsilateral ankle and forefoot with a distance of at least 5 cmbetween the electrodes (Biatrodes, Akern Srl, Italy). An alternating current (400 mA and 50-kHz) was sent between the electrodes and the resistance (R), reactance (Xc), and phase angle (PhA) were measured (BIA 101 Anniversary Sport Edition analyser, Akern Srl). BIA is based on the electrical principle that the body is a circuit with a given R and Xc. R reflects the opposition of current flow through intracellular and extracellular solution and Xc reflects the capacitance of the cells to store energy.14 PhA is a ratio of whole-body cellular health and integrity that can be derived fromR and Xc as the arc tangent of Xc/R and represents the difference between voltage and current. BIA-derived measurements can vary based on factors like sex, age, and body composition. In addition, changes in body hydration, among other alterations in physical status during ICU-admission, can alter cellular resistance and therefore PhA.15 In this study, baseline PhA was used to assess whole body cellular health as a marker for malnutrition and physical frailty. PhA values were compared to previously established cut-off values based on a healthy and well-fed control population (PhA ≥5 and ≤7°).16,17 For men, a PhA beneath 5° and for women a PhAbeneath 4.6° is considered to reflect an impaired health ormalnourishment. 5.2.4 Statistical analysis The primary aim of this study was to assess all-cause 1-year mortality after critical illness. Data are presented as median and interquartile range (IQR). For comparison between groups, an independent sample t-test was used in case of normal distribution or a Mann-Whitney test in case of non-normal distribution. A chi-square test was performed in case of categorical variables. A multivariate logistic regression analysis was performed (backwardWald). In addition, a receiver-
RkJQdWJsaXNoZXIy MjY0ODMw