WEIGHT CHANGES IN CRITICALLY ILL PATIENTS EVALUATED BY FLUID BALANCESAND IMPEDANCE MEASUREMENTS

Objective. To study simple, rapid, and predictive methods to determine body weight changes in critically ill patients. Design: Prospective, consecutive sample. Setting: Medical intensive care unit of a universi ty hospital. Patients: Thirty-one consecutive patients. Interventions: Calculated weight changes, using day-to-day and cumulative fluid bala nces corrected (in two ways) for insensible losses, were compared with the actual weight changes (mattress bascule). A tetrapolar impedance technique measuring resistance was evaluated for estimating weight cha nges. Measurements and Main Results: No reliable relationship was foun d between calculated weight changes using fluid balances corrected for insensible loss and the observed weight changes. An intraindividual r elationship was found between actual weight changes and changes in res istance measured with the tetrapolar impedance technique in a group of 24 critically ill patients with large weight changes (11.1 +/- 6.7 kg ). No such intraindividual relationship was found in seven patients wi th small weight changes (3.1 +/- 2.2 kg). In each patient, the slope c oefficient of the change in weight and resistance relationship differe d; this individual slope coefficient could be an indication for hydrat ion. Conclusions: Calculated fluid balances are not predictive for act ual weight changes in critically ill patients. Absolute weight measure ments are indispensable. Changes in resistance correlated with weight changes in individual patients if weight changes were >3 kg.