Association between cell saver transfusion to estimated blood loss ratio and risk of pulmonary embolism after adult spinal deformity surgery.

Abstract

The objective of this study was to determine whether increased cell saver (CS) salvage transfusion to estimated blood loss (EBL) ratio (CS:EBL) is a driver in the development of cardiopulmonary (CP) and/or renal complications. Patients with adult spinal deformity (ASD) enrolled in a multicenter, observational prospective study from 13 spinal deformity centers in North America were queried. Eligible participants met at least one of the following radiographic and/or procedural inclusion criteria: pelvic incidence-lumbar lordosis mismatch ≥ 25°, T1 pelvic angle ≥ 30°, sagittal vertical axis ≥ 15 cm, thoracic scoliosis ≥ 70°, thoracolumbar/lumbar scoliosis ≥ 50°, global coronal malalignment ≥ 7 cm, having undergone a 3-column osteotomy, spinal fusion ≥ 12 levels, and/or age ≥ 65 years with ≥ 7 levels of instrumentation. Patients were dichotomized based on whether their CS:EBL was ≥ or < 0.33. Patients who did not have CS transfused intraoperatively were excluded. Key outcomes included renal and CP-related medical complications. Patient characteristics, preoperative laboratory results, operative data, and radiographic parameters were compared using appropriate statistical tests. A conceptual multivariable logistic regression model was built to assess risk factors associated with the primary outcome. Four hundred six patients were included in this analysis, including 10.6% (n = 43) with a CS:EBL ≥ 0.33 and 89.4% (n = 363) with a CS:EBL < 0.33. The patients with a CS:EBL ≥ 0.33 were significantly older (mean 66.2 [SD 12.2] vs 58.9 [SD 16.4] years, p = 0.0007), experienced less EBL intraoperatively (mean 1048.3 [SD 852.2] vs 1695.6 [SD 1295.3] ml, p < 0.0001), had fewer total instrumented levels (mean 12.2 [SD 3.3] vs 14.1 [SD 3.6] levels, p = 0.0001), had fewer posterior column osteotomies performed (72.1% vs 86.8%, p = 0.0103), and had less major coronal Cobb angle correction (mean -17.0° [SD 14.6°] vs -22.7° [SD 16.7°], p = 0.0373). Despite comparable transfusion rates, patients with a CS:EBL ≥ 0.33 had fewer packed red blood cell (RBC), fresh frozen plasma, and platelet units transfused intraoperatively (p < 0.05). No significant differences were observed among overall CP and renal complications. However, when stratifying CP complications by type, patients with a CS:EBL ≥ 0.33 experienced a significantly greater rate of pulmonary embolisms (PEs; 9.3% vs 1.4%, p = 0.0093) within 30 days of surgery. A multivariable logistic regression model adjusted for the significant differences between the two groups found a CS:EBL ≥ 0.33 to be an independent risk factor for the development of a PE, conferring an odds ratio of 6.57 (95% CI 1.75-24.66) with excellent model diagnostics (model p value = 0.0031, area under the receiver operating characteristic curve = 0.92). Patients with a high CS:EBL were at a significantly greater risk of developing a PE within the early postoperative period independent of blood loss and operative complexity. The findings support the re-evaluation of salvaged RBC use in this patient population based on perceived benefits.