Presented at ESPEN in Florence, Italy September 2008
OBJECTIVE: Chronically ventilated patients have nutritional modifications during their hospital stay. These parameters may influence the weaning from the ventilator process. Bio impedance technique could assess in a repetitive way the body composition of these patients and their ability to be weaned.
SUBJECTS / METHODS: All the patients admitted to a rehabilitation geriatric center for weaning from the ventilator were included to the study. In addition to anthropometric, functional and laboratory parameters, body composition was evaluated using multi-frequency bioelectrical impedance analysis (BIA) (Bodystat, QuadScan 4000, British Isles) in the attempt to detect fluid, fat and lean mass variations. Weaned and non weaned patients were compared using chi-square test analysis and predictors for successful weaning were detected using one way ANOVA , independent samples test and test for equality of means.
RESULTS: Four hundred eleven patients were admitted between 2004 and 2007. Undernutrition was detected in 11.9%, normal BMI was present in 25.1%, 30.2% were overweight and 32.8% were obese. They were 75.6 ± 11.2 years old, weighted 71.8± 18.2 kg. Body composition was achievable and body composition was: Fat mass 24.6 ± 13.6 kg, lean body mass 47.2 ± 16.2 and extracellular water 40.6 ± 12.9 kg. 214 were weaned while 197 not. Predictor factors were number of ventilation days before admission (p=0.05), functional index before admission, presence of decubitus ulcers (p=0.001), APACHE II score on admission (p=0.001), fat free mass index (0.02) at admission and at discharge, ECW and lean body mass at discharge (p<0.05). Loss of extracellular water was a strong predictive index (p<0.04). Calorie intake, level of albumin or glucose were not predictive factors.
CONCLUSION: In addition to factors related to admission, modifications in body composition and mainly decrease in extracellular water are associated with successful weaning and can be detected by body composition analysis.