Blood Clots Are Rising Sharply in Leukemia Patients, National Study Warns

Venous thromboembolism, the umbrella term for dangerous blood clots that form in the deep veins of the legs or travel to the lungs, has long been recognized as an underappreciated complication of acute leukemia. A new nationally representative analysis of hospital admissions across the United States now puts hard numbers on the problem, and the findings are sobering: roughly one in every twenty-two hospitalized adults with acute leukemia carries a diagnosis of venous thromboembolism, the burden has risen significantly over recent years, and the presence of clots is tied to a nearly fifty percent increase in the risk of dying during the hospital stay.

The study, published in the Annals of Hematology, drew on the National Inpatient Sample, the largest publicly available all-payer hospital database in the United States. The researchers, led by co-first authors May AlMoshary of Princess Nourah bint Abdulrahman University in Riyadh and Ahmed Elbataa of Al-Azhar University in Cairo, examined seven years of admissions data spanning 2016 through 2022. They identified 191,640 weighted hospitalizations of adults admitted with a principal diagnosis of acute leukemia, using standardized ICD-10-CM diagnostic codes to ensure consistent case identification across hundreds of hospitals nationwide.

Within that enormous cohort, 8,880 hospitalizations, or 4.6 percent, involved venous thromboembolism. That figure may sound modest, but in the context of a disease whose treatment already involves profound disruptions of the blood system, it represents a substantial clinical burden. More troubling is the trajectory: after adjusting for patient and hospital characteristics, the prevalence of venous thromboembolism climbed significantly between 2016 and 2022, a trend the authors describe as statistically robust with a p-value of 0.032. The burden peaked at 5.3 percent in 2021, suggesting that the pressures of the pandemic era, or evolving treatment patterns, may have intensified the thrombotic risk faced by leukemia patients.

Acute leukemia occupies a peculiar and dangerous position in hematology. Unlike solid tumors, which promote clotting through local inflammation and vessel compression, leukemia floods the bone marrow and bloodstream with malignant white blood cells that actively subvert the coagulation system. Patients simultaneously face thrombocytopenia, a dangerous shortage of platelets that predisposes them to catastrophic bleeding, and a hypercoagulable state driven by tumor cell procoagulants, inflammatory cytokines, and chemotherapy itself. The result is a clinical tightrope: the same patient may need transfusions to survive a hemorrhage and anticoagulants to survive a clot, often within the same hospital week.

This competing thrombohemorrhagic dynamic is precisely why the research team chose blood product transfusion as the primary outcome of their analysis. Transfusion requirements, they argue, serve as a clinically meaningful proxy for the overall balance between clotting and bleeding that defines acute leukemia care. What they found was unexpected. After multivariable adjustment, patients with venous thromboembolism actually had lower odds of receiving a blood product transfusion, with an adjusted odds ratio of 0.842. The authors are careful about interpretation, noting that this paradoxical inverse association may reflect unmeasured disease severity rather than a genuine protective effect, and that clinicians may withhold transfusions or anticoagulation in ways that confound simple associations.

The other outcomes were far less ambiguous. Venous thromboembolism was independently associated with a 47.8 percent increase in the odds of dying in the hospital, an adjusted odds ratio of 1.478 with a p-value below 0.001. Patients with clots stayed in the hospital dramatically longer, an average of 8.41 additional days, and generated substantially higher costs, with inflation-adjusted hospital charges exceeding those of clot-free admissions by an average of $127,633. For a disease already among the most expensive to treat in medicine, the economic toll of thrombotic complications alone is staggering.

Methodological rigor is a strength of the analysis. Because the National Inpatient Sample uses a complex survey design with weighting to represent the full national population of hospitalizations, the researchers employed multivariable regression models that accounted for that design, ensuring their estimates reflect national patterns rather than the quirks of any hospital subset. To test the durability of their findings, they ran propensity-score weighted sensitivity analyses using average treatment effect on the treated weights, a technique that statistically balances the clot and no-clot groups on observed characteristics. Every major association, from the mortality signal to the transfusion paradox, remained stable under this scrutiny.

The rising trend in clot prevalence deserves particular attention. Several forces could be driving it. Treatment landscapes for acute leukemia have shifted over the study window, with wider adoption of intensive induction regimens, novel targeted agents, and supportive care advances that keep patients alive longer and thus exposed to thrombotic risk for greater spans of time. Improved imaging, particularly more routine use of computed tomography pulmonary angiography, may also be detecting clots that previously went undiagnosed. And the 2021 peak coincides with the COVID-19 pandemic, a period when viral infection itself became a potent prothrombotic force and hospital systems were strained to their limits.

For clinicians, the message is that venous thromboembolism should be treated as a first-order complication of acute leukemia rather than an afterthought. Prophylaxis decisions in this population are genuinely difficult, because standard anticoagulation carries hemorrhagic risk in patients whose platelet counts can plummet without warning. Yet the mortality association documented here suggests that under-recognition or undertreatment of thrombosis carries its own lethal price. The authors suggest that their nationally representative figures can help calibrate risk-benefit calculations, identifying which hospitalized leukemia patients might merit closer surveillance, earlier imaging, or tailored prophylactic strategies.

The study also carries implications for health systems and researchers. The added length of stay and cost burden quantified here, more than eight extra bed-days and over $127,000 in additional charges per complicated admission, translate into enormous aggregate expenditure when multiplied across the tens of thousands of leukemia admissions each year. Future work, the team notes, should probe the mechanisms behind the transfusion paradox and the drivers of the rising trend, ideally with granular clinical data that the administrative coding of the National Inpatient Sample cannot provide. For now, the analysis stands as one of the clearest national portraits to date of a quiet killer complicating one of medicine’s most aggressive cancers, and a reminder that in acute leukemia, the battle between clotting and bleeding is fought on every hospital ward, every day.

Subject of Research: Venous thromboembolism prevalence and inpatient outcomes in hospitalized adults with acute leukemia in the United States, 2016–2022

Article Title: National trends and inpatient outcomes of venous thromboembolism in hospitalized acute leukemia: a propensity-score weighted analysis, 2016–2022

Article References: AlMoshary, M., Elbataa, A., Mansour, A., Murad, M. R., Alabed, A., ALSubhi, S. A., Alzarea, E. A., Halawani, S. H., Bajuaifer, N. A., Basonbul, A. A., Abdelbary, L., & Bakhsh, E. (2026). National trends and inpatient outcomes of venous thromboembolism in hospitalized acute leukemia: a propensity-score weighted analysis, 2016–2022. Annals of Hematology. https://doi.org/10.1007/s00277-026-07312-6

Image Credits: AI Generated

DOI: 10.1007/s00277-026-07312-6

Keywords: venous thromboembolism, acute leukemia, National Inpatient Sample, in-hospital mortality, blood transfusion, hospital length of stay, propensity-score weighting, hematology, thrombosis, hospitalization costs, ICD-10-CM, retrospective cohort