Translating Surgical Transfusion Evidence into Practice: A Multicenter Prospective Observational Study

Abstract

Background: Randomized controlled trials have confirmed the safety of a restrictive red-cell tranfusion threshold (8 g/dL) versus a liberal threshold in cardiac and orthopedic surgeries. We aim to examine whether these findings can be replicated with observational data and generalized to broader specialties and a lower threshold. Methods: This is a prospective observational study conducted in three region-representative hospitals in China. Inpatients aged ≥18 years undergoing surgeries from six red cell-consumptive specialties (2015–2016) were defined as the base population. Three criteria were used to derive an analytic population: no significant intraoperative bleeding (<500 mL); hemoglobin at 7–10g/dL; and stratified by specialties (CCV and nonCCV group). Propensity-score matching was used to minimize confounding effect when necessary. Findings: Of 36 607 patients, 3930 (10·7%) received transfusion. The transfusion rate significantly increased with the intraoperative bleeding and decreased with the hemoglobin (p trend <0·0001). After applying the three criteria and propensity-score matching, transfusion was unrelated with death (OR [95% CI]: CCV, 1·51 [0·25–9·23]; nonCCV, 1·41 [0·44–4·46]), composite complication (CCV, 1·36 [0·63–2·95]; nonCCV, 1·61 [0·98–2·66]), but was related with infection in nonCCV (2·15 [1·12–4·12]). Patient outcomes showed no significant differences after transfusions at hemoglobin of 7–8g/dL vs. 8–10g/dL. Interpretation: We have replicated the findings of transfusion RCTs in six specialties and a lower threshold. Tailoring both the design and analysis is crucial for using observational data to generalize RCT evidence into wide surgical transfusion practice. Funding Statement: The National Health and Family Planning Commission of China (no. 201402017). Declaration of Interests: Authors have nothing to disclose. Ethics Approval Statement: The need for written informed consent was waived by the institutional review board of Peking Union Medical College Hospital (the leading research site of MSCP; date: July 26th, 2013; no. S-574) considering that the project was a quality improvement endeavor and individual information was analyzed anonymously.