Feasibility of a Restrictive Transfusion Strategy for Patients Undergoing Malignant Tumour Resection: A Multicentre Prospective Observational Study

Abstract

Background The recommended transfusion threshold for surgical oncology patients remains unclear owing to insufficient evidence from randomized controlled trials. Evidence from observational studies has been questioned because of huge patient heterogeneity. We aimed to assess the safety of restrictive transfusion at a minimum tolerable haemoglobin threshold for surgical oncology patients with observational data using a tailored design and analysis. Methods Data were collected from four representative regional hospitals in China during 2015 − 2016. Surgical inpatients with seven types of cancer (total cancers) comprised 6055 participants as the base population. In our design, we (1) derived a primary analytic sample by the interested zone of transfusion decision: 6–10 g/dL; (2) selected the control group according to trigger haemoglobin threshold; (3) evaluated the patient heterogeneity between the transfused and non-transfused groups with key covariates according to standardized mean difference (SMD) values. Analysis is mainly based on two datasets: The base population to obtain a natural pattern of transfusion practice; and the primary analytic sample to evaluate the safety of a restrictive transfusion strategy. Results In the base population, 836 (13.81%) were transfused and showed high heterogeneity compared with non-transfused patients (SMD: 7.1–71.31%); there was a significant positive association between transfusion and the composite outcome (in-hospital complications and death) (P < 0.05). In the primary analytic sample, patient heterogeneity was greatly reduced (all SMD ≤ 10%). Compared with non-transfusion, transfusion no longer increased the risk of the composite outcome for total cancers at haemoglobin 6–10 g/dL (odds ratio [OR]: 1.18, 95% confidence interval [CI]: 0.71–1.98), especially for colorectal cancer at 6 − 8 g/dL (OR: 0.54, 95% CI: 0.17–1.68). Compared with those transfused at haemoglobin 8–10 g/dL, patients transfused at 6–8 g/dL did not increase the risk of the composite outcome for total cancers (OR: 1.08, 95% CI: 0.44–2.65), especially for colorectal cancer (OR: 0.46, 95% CI: 0.12–1.82). Conclusions A restrictive transfusion threshold of 8 g/dL may be feasible for total cancers, with a threshold as low as 6 g/dL for colorectal cancer. Restrictive transfusion evidence can be complemented with observational data using a tailored design and analysis.