AUTHORS
Jennifer Hsu, Charles Abraham Joseph Larson, BS2, Rishi Kumar Nanda, DO, MS3, Daniel Thomas Thomas Jones, DO, MBA, MHS4, Kyaw Zin Thein, MD5
AFFILIATIONS
1 Brown University, Providence, Rhode Island, United States
2 Trinity School of Medicine Office of Admissions & Administration, Alpharetta, Georgia, United States
3 Kirk Kerkorian School of Medicine at University of Nevada, Las Vegas, Las Vegas, Nevada, United States
4 Sunrise Health GME Consortium, Las Vegas, Nevada, United States
5 Comprehensive Cancer Centers of Nevada - Central Valley/Twain, Las Vegas, Nevada, United States
ABSTRACT
Background:
A substantial proportion of patients with acute leukemias require emergency medical intervention within the first 24-48 hours to prevent early mortality. While inter-hospital transfer from non-tertiary facilities to specialized academic medical centers is common to escalate care, logistical barriers may delay time-sensitive interventions.
Methods:
This retrospective, cross-sectional study utilized the 2016-2022 National Inpatient Sample (NIS).
Adult hospitalizations (≥18 years) with a principal or secondary diagnosis of acute leukemia were identified using ICD-10-CM code families for acute myeloid leukemia, acute lymphoblastic leukemia, acute promyelocytic leukemia, or other specified acute leukemias. To isolate acute oncological emergencies, the cohort was further refined by ≥1 secondary code for: hyperleukocytosis/leukostasis, tumor lysis syndrome, disseminated intravascular coagulation, severe neutropenic fever, acute respiratory failure, or sepsis.
The primary exposure variable was admission source status, categorizing Inter-Hospital Transfer (transferred in from another short-term acute care hospital) versus Direct Admission. Discharges transferred out to another acute care facility were excluded.
The primary endpoint was all-cause in-hospital mortality. Secondary outcomes included emergency procedural interventions (mechanical ventilation, therapeutic apheresis, and central venous catheter placement), length of stay (LOS), and adjusted total hospital charges.
Multivariable logistic regression and survey-weighted linear models evaluated the independent association between transfer status and outcomes. Models adjusted for patient demographics, hospital level characteristics, baseline Elixhauser Comorbidity index, and acute illness severity. Survey weights (DISCWT) and relevant sampling design parameters were applied.
Results:
A total of 24,918 unweighted adult hospitalizations with acute leukemia presenting with concurrent oncologic emergencies between 2016-2022 were evaluated.
After full multivariable adjustment for sociodemographics, baseline comorbidities, hospital characteristics, and concurrent acute clinical crises (such as tumor lysis syndrome, DIC, neutropenic fever, respiratory failure, and sepsis), inter-hospital transfer (IHT) was an independent predictor of adverse inpatient outcomes.
Patients admitted through IHT faced a significantly higher risk of inpatient mortality compared to direct admissions, a 17.9% increase in adjusted odds of all-cause in-hospital deaths (aOR = 1.179; 95% CI: 1.084–1.282). Transfer status also independently predicted elevated odds across urgent procedural interventions. Transferred patients experienced around a 46.5% increase in the odds of requiring mechanical ventilation (aOR = 1.465; 95% CI: 1.336–1.607) and a 67.8% increase in undergoing therapeutic apheresis (aOR = 1.678; 95% CI: 1.419–1.983) relative to direct admissions. However, the odds of central venous catheter placement did not differ significantly between groups (aOR = 1.004; 95% CI: 0.891–1.131). Additionally, IHT was a key driver of increased healthcare system utilization, leading to a 15.7% increase in total length of stay (ratio = 1.157; 95% CI: 1.123–1.191), and a 12.1% increase in inflation-adjusted total hospital charges (ratio = 1.121; 95% CI: 1.079–1.164).
Acute organ system decompensation was a primary clinical driver of inpatient mortality. Concomitant acute respiratory failure marked the single strongest indicator of death, increasing mortality odds over six-fold (aOR = 6.185; 95% CI: 5.684–6.731), whereas concurrent sepsis (aOR = 2.496; 95% CI: 2.316–2.690) and disseminated intravascular coagulation (aOR = 2.239; 95% CI: 2.014–2.489) more than doubled the chance of in-hospital death. Sociodemographic disparities were also evident: male sex lowered mortality (aOR=0.856), self-pay raised mortality (aOR=1.276) and apheresis odds (aOR=2.062) vs. Medicare, and large hospitals raised mortality vs. small (aOR=1.149; all p<0.05).
Conclusion:
Inter-hospital transfer for acute leukemia emergencies is independently associated with heightened inpatient mortality, procedural escalation, and substantial healthcare resources. This analysis also underlines significant sociodemographic and systemic disparities. Proactive vigilance and streamlined inpatient protocols are crucial to navigating critical complications.
DISCLOSURES
Jennifer Hsu: No relevant conflicts of interest to declare.