Predictive factors for perioperative blood transfusions in rectal resection for cancer: A multivariate analysis of a group of 212 patients

Citation
Sp. Benoist et al., Predictive factors for perioperative blood transfusions in rectal resection for cancer: A multivariate analysis of a group of 212 patients, SURGERY, 129(4), 2001, pp. 433-439
Citations number
35
Categorie Soggetti
Surgery,"Medical Research Diagnosis & Treatment
Journal title
SURGERY
ISSN journal
00396060 → ACNP
Volume
129
Issue
4
Year of publication
2001
Pages
433 - 439
Database
ISI
SICI code
0039-6060(200104)129:4<433:PFFPBT>2.0.ZU;2-S
Abstract
Background. In colorectal cancer surgery, allogeneic blood transfusions hav e reportedly been associated with higher rates of postoperative complicatio ns and tumor recurrence. However, because of the increased cost of alternat ive types of blood transfusions (eg, the use of autologous blood or erythro poietin administration), their routine use cannot be recommended. This stud y evaluated the risk factors for perioperative blood transfusions in resect ion for rectal cancer in order to identify patients who could benefit from these methods. Methods. From 1990 to 1997, 212 consecutive patients who underwent elective rectal resection for cancer were reviewed. The associations between periop erative heterologous blood transfusion and 18 patient-, tumor-, surgical-, and treatment-related variables were assessed by univariate and multivariat e analysis. Results. Of the 212 patients, 72 (34%) received transfusions. Multivariate analysis revealed that 5 preoperative variables were significant risk facto rs for preoperative blood transfusion: age > 65 years (P = .03), body mass index > 27 kg/m(2) (P = .04), preoperative hemoglobin less than or equal to 12.5 g/dL (P < .0001), American Society of Anesthesiologists status > 2 (P = .024), and additional surgical procedures (P = .018). In patients with a nemia, the risk of transfusion was at least 47% in patients with 1 other ri sk factor or more. In nonanemic patients, the risk of transfusion was under 11 % in patients with 1 risk factor or none, but increased to 47 % in thos e with 2 or more risk factors. Conclusions. Our analysis of risk factors for perioperative blood transfusi on in rectal resection for cancer must be considered to constitute guidelin es for a more responsible use of the expensive alternatives to allogeneic b lood transfusion.