Weight Gain and Cardiopulmonary Bypass

____________ _ Seventy-two consecutive patients undergoing coronary revascularization were divided initially into three groups consisting of those with perfusion time less than or equal to 90 minutes (Group 1), those with perfusion time between 90 and 150 minutes (Group II), and those with perfusion time 150 minutes or longer (Group Ill). The patients having the longest perfusion displayed, as anticipated, the highest weight gain of 4.4 ± 0.6 Kg following cardiopulmonary bypass in contrast to 2.6 ± 0.3 Kg in Group II and 2.1 ± 0.3 Kg in Group I (p < 0.05). This corresponds to a linear progression with increasing weight gain proportional to increasing perfusion time (0.02-0.03 Kg/min of bypass). These same 72 patients were then divided into two groups: Group A having weight gain less than 2.0 Kg and Group B with weight gain of 2.0 or more Kg. Group B showed a significant reduction of 39% in the serum protein level and 28% in serum albumin as compared to the control level. This is significantly different from the 28% reduction in total serum protein and 17% reduction in serum albumin as was noted in Group A, having the shorter bypass interval (p < 0.05). Thus, while the duration of perfusion is primarily responsible for fluid accumulation and weight gain, the low serum protein may serve to accentuate edema formation. Hemodilution and nonpulsatile flow are factors promoting tissue edema during cardiopulmonary bypass. *Chief. Cardiac Surgical Service, Baystate Medical Center. Reprints: Roger A. Vertrees, C.C.P., Chief, Perfusion Technology, Baystate Medical Center, 759 Chestnut Street, Springfield, MA 01107. Introduction _____________ _ It was observed that some patients undergoing cardiopulJl?.onary bypass had a large weight gain (>4.5 Kgs) measured in the immediate postoperative period. We considered a weight gain of this magnitude to be undesirable since it was most often manifested as tissue edema and have attempted to identify the etiology. Methods _____________ _ All patients undergoing cardiopulmonary bypass at the Baystate Medical Center undergo an exhaustive preoperative examination. The results of this examination are then entered into a computer for future reference. It is from this list that some of the parameters have been taken for this study: preoperative weight, history of diuretic therapy, hypertension, and chronic congestive heart failure. Exact records are also kept by the anesthesiologist, perfusionist, and intensive care staff as to fluids given, urine output, blood loss, postoperative weight, and blood chemistries. These parameters are also computerized and were used in the study. For purposes of this study, 72 consecutive patients undergoing coronary revascularization were analyzed. Cardiopulmonary bypass was instituted using a Sarns* roller pump and the BOS 10** bubble oxygenator. The circuitry was primed with 1500 cc Plasmalyte 148*** and 250 cc of 12.5 grams of human * Sarns, Inc., Ann Arbor, Michigan 48103 **Bentley Labs, Irvine, California 92714 *** Travenol Labs, Deerfield, Illinois 60015 74 The Journal of Extra-Corporeal Technology Volume 12, Number 3, 1980 TABLE 1 The Preoperative Profile of the Three Groups Perfusion Congestive Time Heart DiaDiuretic HyperGroup (min) Failure betes Therapy tension I <90 10% 20% 30% 30% II 90-150 II% 14% 19% 29% III >150 12% 15% 25% 50% * indicates significant difference p < 0.05 serum albumin. No blood was used to prime the circuitry of these patients. Total prime volume for this circuit was 1750 cc. The pump index is maintained between 1.6 and 2.4 liters/minute/meter square body surface area. Moderate hypothermia of 25° rectal temperature is generally achieved. Additional red cell volume is given should the hematocrit fall below 20% during hypothermia or 25% during the rewarming phase. All fluids administered were correlated with the urine output, duration of perfusion, and the postoperative weight. Comparisons were carried out with respect to both perfusion and weight gain. All analyses were reported as the mean plus/minus the standard error of the mean, and the Student's t-test was used for determining the significance of differences between sample means. The correlation coefficient (r) was determined by a linear regression and the least squares