Portal de Periódicos da CAPES

EFFECT OF ACUTE NORMOVOLEMIC HEMODILUTION ON ALLOGENEIC BLOOD TRANSFUSION DURING CARDIAC SURGERY; A META-ANALYSIS

1998; Lippincott Williams & Wilkins; Volume: 86; Issue: Supplement Linguagem: Inglês

10.1097/00000539-199804001-00021

1526-7598

Paul G. Loubser, Ileana A. Dejuan,

Cardiac and Coronary Surgery Techniques

Introduction. Acute normovolemic hemodilution (ANH) is used in conjunction with whole blood sequestration (WBS) during surgery as a means of avoiding transfusion of allogeneic blood. [1] However, randomized studies of ANH have produced conflicting results. [2] Meta-analysis is a quantitative review procedure in which a large collection of results from individual studies are statistically analysed. [3] This study was performed to examine the efficacy of ANH as a blood conservation technique in cardiac surgery using meta-analysis. Methods. A bibliography was assembled by performing a detailed search of the Medline database, coupled with cross-referencing from 1966 - 1997. Publications were sorted into separate categories including: human studies, animal studies, type of surgery by specialty, type of study (mathematical, cost-analysis), etc. Human studies in patients undergoing cardiovascular surgery were then selected according to study N, control group, reported data, significance, etc. The effect of ANH on transfusion of allogeneic blood products (Sub-study A) and on surgical blood loss (Substudy B) was then examined using Meta-analysis Programs, Version 4.3. For each sub-study, the treatment (ANH) effect size was calculated, i.e. Hedges d. Effect sizes were then combined and analyzed using the Weighted Integration Model. Homogeneity Q of the patient population was examined using the Random Effects Model. p<0.05 was used for statistical significance. Results. Of the 122 publications retrieved, 22 focused on cardiovascular surgery in humans, viz. cardiac surgery; n = 14, vascular surgery; n = 8. Seven of the cardiac surgery trials were eligible for Sub-study A, while 5 were eligible for Sub-study B. Volume of sequestered whole blood ranged from 450 - 2100 mLs and was either heparinized or collected in citrate anticoagulant. One study compared the effects of sequestered heparinized whole blood with citrated whole blood, incorporating 2 control groups - these results were analyzed separately. Another study pooled data for patients undergoing sequestration of both heparinized and citrated whole blood. For Sub-study A, all 7 trials (sample size - 723) reported that allogeneic blood transfusion requirements were decreased in patients undergoing ANH (mean effect size = 2.0 +/- 0.09, p<0.001; homogeneity Q = 676.1, p<0.001; sampling error variance = 1.13%. For Sub-study B, all 5 trials (sample size - 351) reported that blood loss decreased in patients undergoing ANH (mean effect size = 0.8 +/- 0.11, p<0.001; homogeneity Q = 415.8, p<0.001; sampling error variance = 1%). A Fail-Safe N was estimated to exceed 25 for both sub-studies. Discussion. These results suggest that ANH is an effective blood conservation technique. Minimizing surgical blood loss decreases the need for transfusion of allogeneic blood. However, the population sample reported above does not meet a test of homogeneity. Over 75% of the variance was due to systematic factors, i.e. study characteristics and variables not identified or reported in the studies. Therefore, additional trials that tightly control patient sampling and study design are required.