Cell Saver Use is Not Associated With Increased Morbidity But Reflects Transfusion Burden in Pediatric Spinal Deformity Surgery: A Retrospective Cohort Study.

INTRODUCTION: Cell saver (CS) is widely used in spinal fusion to reduce exposure to allogeneic transfusion, which carries risks such as infection and prolonged hospitalization. Pediatric spinal fusion is frequently associated with significant intraoperative blood loss. Concerns persist that reinfused blood could increase morbidity. We therefore evaluated the association between CS exposure and postoperative outcomes after pediatric spinal fusion using the American College of Surgeons National Surgical Quality Improvement Program Pediatric database.

METHODS: Among patients who underwent spinal fusion surgery for pediatric spinal deformity between 2021 and 2023, CS exposure was categorized as none, low (<3.5 mL/kg), or high (≥3.5 mL/kg). Inverse probability of treatment weighting (IPTW) was used to balance demographics, comorbidities, and surgical complexity across groups. Logistic regression assessed outcomes, with global tests followed by Tukey-adjusted pairwise contrasts. The primary outcome was all-cause morbidity; secondary outcomes included reoperation rates, discharge to home, allogeneic transfusion, and other system-based complications.

RESULTS: There were 20,376 patients included in this study (no CS: 6967; low CS: 6811; high CS: 6598). High CS patients had more complex baseline profiles, including higher ASA III-IV status (42.4% vs. 28.6% no CS, 26.3% low CS), neuromuscular scoliosis (27.4% vs. 15.9% and 11.6%), and ≥13 levels fused (52.8% vs. 26.3% and 33.6%; all P<0.001). After IPTW, all-cause morbidity did not differ significantly between groups (P=0.75). Respiratory, renal, and infectious complications, discharge to home, and reoperation were also not significantly different (all P>0.05). Neurological complications varied significantly (P=0.013), but only the high CS versus no CS comparison was significant (OR=1.60; P=0.015). Intraoperative allogeneic transfusion rates differed (P<0.001), with low CS associated with reduced rates (OR=0.62; P<0.001) and high CS with increased rates (OR=1.83; P<0.001).

CONCLUSIONS: CS use was not associated with increased overall morbidity within 30 days. High CS volume identified a subgroup with higher odds of neurological complications and greater transfusion rates, likely reflecting residual confounding or bleeding severity rather than a direct CS effect. Low CS exposure was associated with reduced allogeneic transfusions.

STUDY DESIGN: Retrospective cohort study.

LEVEL OF EVIDENCE: Level III.