Measuring body composition in critically ill patients.

Background: Severe illness (SI), often requiring intensive care (IC), is commonly associated with dynamic changes in body composition (BC). Most currently available BC measurement techniques are static, disturbed by the illness, or require risk/ benefit compromise before moving the patient to the BC measurement site. Few BC studies of such ill people have been reported. Previously, we studied methods for assessing height and weight in IC patients. Objective(s): We aim to study abdominal CT images, performed as part of illness diagnosis and progress, in several groups of IC and SI patients. Method(s): We measured L3 muscle area (MA), L4-5 visceral fat (VF) and subcutaneous fat (ScF) areas in several patient groups: (a) all IC patients requiring diagnostic abdominal CT scanning over a 12 month period, with matched general ward controls (b) all IC patients with acute pancreatitis (c) one SI patient with an inherited metabolic disorder (glycogen storage disease GSD 1b), and (d) one IC patient with an acquired metabolic disorder (organic aciduria). Result(s): 67 IC patients were matched to a cohort of 68 ward inpatients. No differences in BC were noted. There was no statistical significance in survival to discharge, based on BC. In the IC acute pancreatitis group, 3 women and 18 men had CT scans analysed on two separate occasions (mean betweenscan time =9.4 days) within their hospital admission. ICU mortality was 9%. There was a decrease in median VF from 229.2 cm2 to 202.1 cm2 (p < 0.01) and a decrease in median VF:ScF ratio from 1.20 to 1.05 (p < 0.01) during the acute illness. MA did not change significantly. The single IC and SI case studies showed reduced MA. Conclusion(s): Precise BC can be measured and followed in IC/ SI patients from incident and sequential abdominal CT scans. Correction for measured or estimated height and weight from knee height and arm circumference is needed. BC measurements in IC/SI patients may add to diagnosis and prognosis.