ABSTRACT
The bromide-82 dilution space (extracellular space, ECS) and blood volume (BV) were measured in 21 patients with esophageal and gastric cancer and in 27 patients 18-96 months after total gastrectomy. Resistance (R) and reactance (Xc) from bioelectrical impedance measurements were used to obtain multiple regression equations for ECS and BV. The variables weight, gender, and height 2/Xc were independent predictors of ECS (r = 0.767; P less than 0.0001). Height 2/R and gender were predictors of blood volume (r = 0.856; P less than 0.0001). The mean difference between the Br space and the ECS predicted from impedance measurements was 0 +/- 1.54 (mean +/- SD). The limits of agreement (+/- 2 SD) were therefore +/- 3.08 l or 19.6% of the mean Br space of 15.7 l. The limits of agreement for BV were +/- 789 ml or +/- 19.7% of the average BV of 4008 ml. It is concluded that bioelectrical impedance plethysmography using a single frequency can be used for the estimation of ECS and BV. The wide limits of agreement, however, may limit its used in clinical practice.
Subject(s)
Blood Volume Determination/methods , Body Composition , Extracellular Space , Plethysmography, Impedance , Adult , Aged , Anthropometry , Body Mass Index , Body Water , Bromine Radioisotopes , Esophageal Neoplasms/pathology , Evaluation Studies as Topic , Female , Gastrectomy , Humans , Indicator Dilution Techniques , Male , Middle Aged , Postoperative Period , Stomach Neoplasms/pathology , Stomach Neoplasms/surgeryABSTRACT
The incidence of postoperative delirium was assessed in 92 patients on a general surgical intensive care unit. Postoperative delirium was diagnosed in 39 patients (42%). Most of the cases were diagnosed on the second postoperative day and the median duration was seven days. Ten patients with delirium (25%) had a lethal outcome, compared to a 13% mortality of the whole population. In univariate analysis the variables age, preoperative therapy for heart failure, respirator therapy, dobutamin therapy and lowest capillary pO2 on day one were significantly associated with later development of delirium, whereas preoperative peritonitis, and history of stroke or hypertension were only borderline significant. A predictive model with three parameters emerged from multiple logistic regression analysis: after correction for age (p = 0.001), respirator therapy (p = 0.020), and capillary pO2 on day one (p = 0.049) none of the remaining variables proved of additional significance. The statistical model yielded a predictive accuracy of 78%.
