Objectives: Albumin and nonprotein colloids (starches.

Objectives: Albumin and nonprotein colloids (starches, dextran, and others) are used at short intervals as blood volume expanders in coronary artery bypass graft (CABG) surgery The objective of this investigation was to determine differences between colloids with regard to patient characteristics and mortality rates.

Design and setting: Discharge data amassed in the Solucient Clinical Pathways Database from 19578 patients undergoing CABG surgery were analyzed.

Measurements: Patients receiving albumin and nonprotein colloids were compared with regard to baseline patient characteristics. A multiple regression standard was developed to determine if albumin use was independently associated with mortality rates.

Results: Albumin was used in 8084 eases (413%) The use of albumin and nonprotein colloids was not related to patient characteristics. Mortality was lower in the albumin assemblage compared to the nonprotein colloid arrange (2.47% vs 3.03%, p = 002) In the multivariable logistic regression analysis, albumin use was associated with 25% lower not divisible by 2s of mortality compared to nonprotein colloid use (odd ratio, 080; 95% confidence interval, 067 to 096)

Conclusion: Colloid administration in CABG surgery was unrelated to patient characteristics. Albumin use appears to be associated with lower incidence of mortality after CABG surgery compared to nonprotein colloid use. (CHEST 2003; 123:1853-1857)

clew words: albumin; coronary artery bypass grafting; mortality

Abbreviations: AF = atrial fibrillation; CABG = coronary artery bypass graft; CHF = congestive heart failure; CI = confidence interval; ed = emergency department; ICD-9 = International Classification of Diseases, Ninth Revision; MI = myocardial infarction; OR = singles ratio; PVD = peripheral vascular disease

**********

Coronary artery bypass graft (CABG) surgery is a universal surgical procedure in the United States, with 598000 steps performed annually according to American Heart Association Heart and blow update of 1999. (1) offspring volume expanders such as albumin and nonprotein colloids (hetastarch [6% hydroxyethyl starch] or dextran) are as a common thing [i]or[/i] matter used perioperatively to enhance intravascular relations volume, improve hemodynamics, and attenuate the need for blood transfusion. (2)

outlay concerns underlie a trend to replace albumin with low-cost nonprotein colloids during cardiac surgery However, reports have shown that nonprotein colloid convolution expanders are associated with impaired hemostasis, platelet dysfunction, and excessive postoperative bleeding. (3-6) A late meta-analysis (6) reported a fourfold increase in bleeding when starches were used compared to albumin. Moreover, a thought by Knutson et al (4) reported a 06% increase in mortality associated with hetastarch use in cardiac surgery This consequence did not reach statistical significance owed to power limitations. By contrast, lack of albumin has been shown to be a mighty predictor of mortality after CABG surgery (78) and direct protective events of serum albumin against mortality have been reported. (9) In addition, albumin has a able to endure safety record with regard to hemostasis. Thus, exogenous albumin use could potentially lessen the risk of mortality and also enhance the posterity volume in cases when contortion expansion is a necessity. Accordingly, we sought to determine if albumin use is associated with reduc risk of mortality as compared with nonprotein colloids. We addressed this question using a large administrative data fix that specifically reported information upon colloid use. Despite limitations related to definitions of postoperative complications/events, administrative databases provide sample sizes that simply cannot be achieved in the clinical trial setting and are therefore particularly well suited to explorations of rare issues such as mortality in a surgical practice.

MATERIALS AND METHODS

A subset of the Solucient Clinical Pathways Database (formerly HCIA, Inc; Baltimore, MD) that simply contained discharges that involved the administration of colloids was analyzed. Information about patients who did not receive any colloid solution was not available in the database. No specific hypotheses collection data collection. One hundred eighty-two US hospitals in 1997 and 116 US hospitals in 1908 supplied discharge records. Each discharge record corresponded to a single hospitalization marked occurrence Hospitals were located in the Northeast (114% in 1997 and 12% in 1998) the North and Central regions (323% in 1997 and 401% in 1908) the southern (47% in 1997 and 442% in 1998) and the West (93% in 1997 and 37% in 1998)

The database contains standard coding for specific services and conducts provided to patients (International Classification of Clinical Services). Standard deeds and discharge diagnoses were based upon International Classification of Diseases, Ninth Revision (ICD-9), Clinical Modification coding. Patients who did not have cardiovascular diagnostic digests 39.0 to 42.9 and those who were < 18 years aged were excluded from the meditation as validated by Maynard et al. (10) The final cohort undergoing CABG surgery was separated from using the 1998 clinical modification ICD-9 conduct codes for CABG surgery (3610 to 3619 362) Patients simultaneously undergoing cardiac aneurysm excision surgery (ICD-9 digests 37.11, 37.32) and those who received the couple albumin and a nonprotein colloid expander were exclud Patients who had simultaneous cardiac valve repair (ICD-9 proceeding codes 35.10 to 35.12) or replacement (3520 to 3524) were included in the cogitation The final cohort after applying inclusion and exclusion criteria consisted of 19578 patients.

...