Evaluating variability in use of intravenous albumin in patients undergoing surgery for cancer.

BACKGROUND

Despite numerous randomized controlled trials finding that albumin is not associated with improved patient outcomes, transfusion practice is highly variable. We examined the variability and impact of albumin transfusion on outcomes in cancer surgery.

METHODS

We included consecutive adults undergoing cancer surgery between 2018 and 2021 in Ontario, Canada. The primary exposure was the proportion of patients who received perioperative albumin. The secondary outcomes were hospital length of stay and the incidence of infection, anemia, venous thromboembolism, and mortality in albumin-treated versus non-albumin-treated patients in a case-control analysis.

RESULTS

Of 155 166 cancer surgeries (66.8% female patients, median age 62.9 yr), 2.5% received perioperative albumin. The cancer surgery types with the highest proportion of patients receiving albumin were hepato-pancreato-biliary (24.8%) and colorectal (18.6%). Of 104 facilities, 12.5% had nonrandom outliers for albumin use in at least 1 cancer type (p = 0.0004). Patient outcomes were different in case-control matched cohorts for colorectal and hepato-pancreato-biliary surgeries, including a higher rate of infection, venous thromboembolism, and mortality in patients treated with albumin (cases) than those who were not (controls).

CONCLUSION

Albumin transfusion rates were highly variable among hospitals for the same cancer type. Quality improvement initiatives are warranted to curtail unnecessary albumin transfusions in the perioperative period.