Hospital at Home Lets Leukemia Patients Recover Safely Outside the Ward

For decades, the standard of care for acute myeloid leukemia has been brutal in its logistics as much as in its biology. Fit patients diagnosed with this aggressive cancer of the blood and bone marrow typically receive intensive chemotherapy, which destroys not only the leukemic cells but also the healthy bone marrow that produces infection-fighting neutrophils. The result is a prolonged period of profound neutropenia, during which even the bacteria that normally live harmlessly on skin and in the gut can become life-threatening threats. To manage that risk, patients have traditionally been confined to hospital wards for roughly four weeks during the first induction cycle and two to three weeks during each subsequent consolidation cycle. A new retrospective study from Sweden, published in Supportive Care in Cancer, now challenges that assumption, showing that with a well-organized hospital-at-home program, most of that time can be spent safely in the patient’s own bed rather than a hospital one.

The research, led by Kristina Warolén and Martin Jädersten of Karolinska University Hospital together with colleagues at Sahlgrenska University Hospital in Gothenburg, compared two cohorts of intensively treated AML patients diagnosed between January 2020 and October 2023. The Stockholm cohort, comprising 116 patients, had access to the Stockholm Region’s hospital-at-home (HaH) service, an infrastructure of 11 healthcare companies operating 38 units across 8 geographic areas. The Gothenburg cohort, 49 patients, received conventional inpatient care without such support. Because both hospitals sit in major urban centers and care is delivered free of charge through Sweden’s universal healthcare system, the researchers hypothesized that the two populations would be broadly comparable, and indeed the cohorts were well matched for age, comorbidity as measured by the Charlson Comorbidity Index, and disease risk according to the European Leukemia Net 2022 classification.

The Stockholm HaH model is remarkable for its scope. At home, patients can receive advanced supportive care that most health systems would consider strictly inpatient territory: intravenous fluids, parenteral nutrition, intravenous medications, and blood and platelet transfusions, the latter sometimes administered at local HaH hubs. Each unit is staffed by physicians, nurses, and other healthcare professionals capable of telephone support and emergency house calls around the clock. Although HaH staff receive no formal additional training in hematology, they manage a wide range of advanced medical conditions in oncology and beyond. Laboratory results are reviewed simultaneously by HaH teams and by junior physicians on the hematology ward through a shared electronic health record, and a joint management plan for transfusions and neutropenic complications is documented for every patient. The maximum distance from a patient’s home to Karolinska University Hospital in Huddinge is around 150 kilometers, and in the Stockholm archipelago, helicopter service is available for emergencies.

The primary endpoint of the study was a composite of death or requirement of a higher level of care, defined as the need for vasopressors, high-flow oxygen or non-invasive ventilation, or intensive care admission, during the first two treatment cycles. The results were striking in their equivalence. After the first cycle, 12 of 116 Stockholm patients (10 percent) met the primary endpoint, compared with 4 of 49 Gothenburg patients (8 percent), a difference that was statistically indistinguishable. Crucially, ten of the twelve Stockholm events occurred in patients who had never been discharged to HaH at all, and five of those patients died during the first cycle. Only two patients discharged to home care were later readmitted requiring a higher level of care, one needing vasopressors for a single day and one needing high-flow oxygen for a single day. No deaths occurred after discharge to HaH.

The efficiency gains, by contrast, were dramatic. During the first cycle, 94 percent of Stockholm patients were discharged, versus just 2 percent in Gothenburg, where 48 of 49 patients remained hospitalized for the entirety of the neutropenic phase. Median inpatient stay after the first cycle was 16 days in Stockholm versus 28 days in Gothenburg, and after the second cycle 9 versus 18 days. Over the first two cycles, Stockholm patients spent on average three additional weeks at home, roughly a third of the entire treatment period. Readmissions were common, with 50 of 109 discharged patients returning to hospital during the first cycle, but the reasons were revealing: clinical deterioration with fever, other symptoms, adherence to older protocols, patients not yet enrolled in the HaH program, persistent fever without deterioration, and in only two cases, patient preference.

Perhaps the most technically interesting finding concerns febrile neutropenia, the fever that signals possible infection during the neutropenic nadir. Among 150 neutropenic episodes in discharged Stockholm patients across the first two cycles, fever developed in 130 cases, or 86.7 percent. Of these, 38 episodes, 29 percent, were resolved entirely with intravenous antibiotics administered at home, while the majority of admitted patients received their first antibiotic dose at home before transfer, minimizing the time from fever onset to treatment. Early administration of antibiotics in febrile neutropenia is known to reduce morbidity and mortality, and the HaH model’s ability to deliver piperacillin-tazobactam, increasingly as a continuous infusion, often within an hour of fever onset may have favorably influenced outcomes even for patients ultimately admitted. Notably, patients treated entirely at home had significantly fewer positive blood cultures, 11 percent versus 39 percent among admitted patients, likely reflecting less severe illness.

The study also generated data relevant to evolving international guidelines. In 42 cases, Stockholm clinicians discontinued intravenous antibiotics while patients were still neutropenic, a practice that the older ECIL-4 guidelines from 2011 discouraged, recommending an additional 24 to 48 hours of hospitalization. In practice, 76 percent of these patients did not need to restart antibiotics within 72 hours, and although ten did restart, none required a higher level of care or died. The updated ECIL-10 guidelines from 2024 now permit close monitoring of neutropenic patients discontinuing antibiotics in either inpatient or outpatient settings, bringing official recommendations into line with what the Stockholm team had been doing for years.

There were also unexpected signals suggesting that home care may do more than merely relocate treatment. Stockholm patients required 40 percent fewer days of intravenous antibiotics than their Gothenburg counterparts across the first two cycles, a difference the authors attribute partly to more liberal use of G-CSF, which can shorten neutropenia by up to five days, but possibly also to faster recovery driven by better nutrition, greater physical activity, and a lower risk of acquiring multidrug-resistant bacteria outside the hospital environment. Intriguingly, while three-month and twelve-month survival were comparable between cohorts, there was a trend toward better long-term survival in Stockholm beyond the first year, with a hazard ratio of 0.62 that did not reach statistical significance but raises the possibility that home-based care supports immune function and tolerance of subsequent therapies, including stem cell transplantation.

The authors are careful about limitations. The retrospective design, modest sample sizes, and reduced statistical power to detect rare but severe adverse events all temper the conclusions, and differences in local practice between the two centers, particularly G-CSF use, complicate direct comparison. Factors such as having a relative at home, distance from hospital, compliance, and language skills likely influenced who could be managed at home but were not systematically evaluated, and there was no structured assessment of patient experience, though routine questionnaires suggested near-universal gratitude for the program, with patients valuing home-cooked meals, emotional support, and time with family. Still, out of nearly 200 episodes of outpatient care of neutropenic AML patients, only 2 percent of readmissions required a higher level of care, no worse than expected for this heavily treated population.

The implications extend well beyond Sweden. As hospital-at-home services expand internationally, driven in part by lessons from the COVID-19 pandemic, which originally pushed the Stockholm program to keep febrile patients home to reduce SARS-CoV-2 exposure and relieve bed shortages, this study offers a practical blueprint for extending home care to one of the most vulnerable populations in oncology. It required no additional on-call staffing or dedicated resources, leveraging existing HaH infrastructure alongside AML-specific protocols. Karolinska has already moved further, offering selected patients consolidation chemotherapy at home via electronic infusion pumps since November 2022, and a prospective study of quality of life, depression, nutrition, and muscle mass is underway. If confirmed in prospective trials, the message is simple and potentially transformative: for many patients with acute myeloid leukemia, the safest place to endure the hardest weeks of treatment may be home.

Subject of Research: Outpatient management of intensively treated acute myeloid leukemia through a hospital-at-home program

Article Title: Hospital at home enables safe and efficient outpatient management of intensively treated acute myeloid leukemia

Article References: Warolén, K., Blennow, O., Ravn Landtblom, A., Nilsson, C., Kullänger Schalling, L., Andersson, C., Lind, C., Hållander, S., Westerberg, S., Stjernfeldt, A.-S., Sporre, E., Leskinen, H., Söderholm, L., Wennström, L., Broliden, P. A., Deneberg, S., Espman, E., von Bahr, L., & Jädersten, M. (2026). Hospital at home enables safe and efficient outpatient management of intensively treated acute myeloid leukemia. Supportive Care in Cancer, 34(10), Article 954. https://doi.org/10.1007/s00520-026-11171-2

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11171-2

Keywords: acute myeloid leukemia, hospital at home, febrile neutropenia, intensive chemotherapy, outpatient care, Karolinska University Hospital, supportive care, intravenous antibiotics, neutropenia, home care, hematology, patient safety