Abstract Introduction: Acute Chest Syndrome (ACS), a severe complication of Sickle Cell Disease (SCD) driven by a cascade of vaso-occlusion, inflammation, and hemolysis, is a leading cause of mortality in adults and frequently progresses to Acute Respiratory Distress Syndrome (ARDS). While Extracoreal Membrane Oxygenation (ECMO) is a potential salvage therapy, its effectiveness is difficult to evaluate due to severe confounding by indication; as a last-resort therapy, it is reserved for patients with the highest risk of mortality, making it challenging to separate the treatment's effect from the underlying severity of illness. This study aimed to estimate the effect of ECMO on in-hospital outcomes in a rigorously selected, comparable cohort of critically ill SCD-ACS patients. Methods: We performed a retrospective cohort study using the Healthcare Cost and Utilization Project (HCUP) National Readmissions Database (NRD) from 2018 to 2022. We identified adult (≥18 years), non-elective hospitalizations for SCD-ACS (ICD-10-CM: D57. 01, D57. 211, D57. 411, D57. 811). To create a clinically comparable cohort, we restricted our analysis to patients with severe organ failure, defined by the concurrent receipt of both invasive mechanical ventilation (ICD-10-PCS: 5A19*) and transfusion (ICD-10-PCS: 302*). The primary exposure was ECMO use (ICD-10-PCS: 5A1522). To account for baseline differences, we employed Inverse Probability of Treatment Weighting (IPTW), which creates a pseudo-population where measured confounders are balanced between treatment groups. Propensity scores were derived from a multivariable logistic regression model including patient demographics, hospital characteristics, APR-DRG severity scores (as categorical variables), and comorbidities identified using HCUP's Clinical Classifications Software Refined (CCSR). The primary outcome was modeled using survey-weighted logistic regression, with a two-sided p-value 0. 05 considered statistically significant. All analyses were performed in STATA 19. 5. Results: From 19, 065 SCD-ACS hospitalizations, our criteria isolated a final analytic cohort of 564 critically ill patients (national estimate: 1, 047). Within this sample, 17 patients received ECMO (national estimate: 33 cases) ; of these, 13 (76. 5%) received veno-arterial (VA) ECMO, 1 (5. 9%) received veno-venous (VV) ECMO, and 3 (17. 6%) were unspecified. Unadjusted in-hospital mortality was nearly identical between groups (29. 4% for ECMO vs. 29. 7% for non-ECMO), despite significant baseline imbalances. The ECMO group was younger (mean age 30. 4 vs. 36. 4 years, p=0. 014) but had a uniformly higher APR-DRG risk of mortality (p0. 001). After IPTW adjustment, most baseline differences were resolved; however, an imbalance in hospital bed size persisted (p=0. 004). In the primary IPTW-adjusted analysis, ECMO use was associated with a statistically non-significant 55% reduction in the odds of in-hospital mortality (Adjusted Odds Ratio 0. 45; 95% Confidence Interval 0. 11-1. 80; p=0. 256). Among survivors, ECMO was associated with profound increases in resource utilization, including significantly longer hospital stays (adjusted mean: 44. 8 vs. 20. 3 days; Adjusted Rate Ratio aRR 2. 21; 95% CI 1. 52-3. 22) and substantially higher total charges (aRR 2. 65; 95% CI 1. 75-3. 99), corresponding to an adjusted mean cost of 844, 731 versus 319, 357 for non-ECMO patients. Conclusion: In this nationally representative analysis, ECMO use was associated with a strong trend towards improved survival among a carefully selected cohort of critically ill adults with SCD-ACS. The substantial resource utilization observed is consistent with the high acuity of this population, a majority of whom required veno-arterial (VA) ECMO for profound cardiorespiratory collapse. While the rarity of ECMO use limited statistical power, our findings provide crucial evidence to support its selective use in expert centers rather than being viewed as a futile intervention. There is a need for larger, prospective studies to validate these findings, better define which patients are most likely to benefit, and explore ways to reduce complications in this high-risk group.
Amalraj et al. (2025) studied this question.