Study design: Retrospective analysis of prospectively-collected data.
Objective: This study aims to define clinically relevant blood loss in adult spinal deformity (ASD) surgery.
Background: Current definitions of excessive blood loss following spine surgery are highly variable and may be suboptimal in predicting adverse events (AE).
Methods: Adults undergoing complex ASD surgery were included. Estimated blood loss (EBL) was extracted for investigation, and estimated blood volume loss (EBVL) was calculated by dividing EBL by the preoperative blood volume utilizing Nadler's formula. LASSO regression was performed to identify five variables from demographic and peri-operative parameters. Logistic regression was subsequently performed to generate a receiver operating characteristics (ROC) curve and estimate an optimal threshold for EBL and EBVL. Finally, the proportion of patients with AE plotted against EBL and EBVL to confirm the identified thresholds.
Results: In total 552 patients were included with a mean age of 60.7±15.1 years, 68% females, mean CCI was 1.0±1.6, and 22% experienced AEs. LASSO regression identified ASA score, baseline hypertension, preoperative albumin, and use of intra-operative crystalloids as the top predictors of an AE, in addition to EBL/EBVL. Logistic regression resulted in ROC curve which was used to identify a cut-off of 2.3 liters of EBL and 42% for EBVL. Patients exceeding these thresholds had AE rates of 36% (odds-ratio: 2.1, 95% CI [1.2-3.6]) and 31% (odds-ratio: 1.7, 95% CI [1.1-2.8]), compared to 21% for those below the thresholds of EBL and EBVL, respectively.
Conclusion: In complex ASD surgery, intraoperative EBL of 2.3 liters and an EBVL of 42% are associated with clinically-significant AEs. These thresholds may be useful in guiding preoperative-patient-counseling, healthcare system quality initiatives, and clinical perioperative bloodloss management strategies in patients undergoing complex spine surgery. Additionally, similar methodology could be performed in other specialties to establish procedure-specific clinically-relevant bloodloss thresholds.
Copyright © 2024 Wolters Kluwer Health, Inc. All rights reserved.