Nearly One in Five Cancer Therapy Hospitalizations Could Have Been Prevented, Study Finds

Every year, millions of patients around the world receive systemic anticancer therapy—chemotherapy, molecularly targeted agents, and immune checkpoint inhibitors that have transformed cancer from a frequently fatal disease into a chronic, often manageable condition. Yet the same drugs that extend lives can also land patients in the hospital. A new retrospective study from Gifu University Hospital in Japan now offers one of the most detailed looks to date at why cancer patients on drug therapy are admitted to hospital unexpectedly, and how many of those admissions might never have happened. The answer, published in Supportive Care in Cancer, is striking: roughly one in five unplanned hospitalizations tied to treatment side effects was judged potentially preventable, most often because guideline-recommended supportive care was never given.

The research team, led by pharmacist Momoka Endo and colleagues, combed through every unplanned hospital admission recorded at their institution over a thirteen-month window, from March 2023 to March 2024. Out of 3,135 unplanned admissions in total, 343 involved patients who had recently received systemic anticancer therapy, and 161 of those—about 5.1 percent of all unplanned hospitalizations—were directly attributable to treatment-related adverse events. The median patient was 68 years old, and the cases spanned ages 5 to 92, underscoring how broadly cancer drug toxicity cuts across demographics. Laboratory values at admission told a story of physiologic stress: median serum albumin of 3.2 grams per deciliter, a marker of nutritional and inflammatory status, and a median C-reactive protein of 4.79 milligrams per deciliter, indicating substantial systemic inflammation.

What actually sent these patients to the hospital? Infections dominated the list. Non-pneumonic infections suggestive of treatment-related immunosuppression—enteritis, cholangitis, herpes zoster, and COVID-19 among them—accounted for 40 cases, or 24.8 percent of the total. Febrile neutropenia, the dangerous combination of fever and critically low white blood cell counts that oncologists dread, drove 23 admissions, or 14.3 percent. Fever and fatigue accounted for another 22 cases, pneumonia for 20, and gastrointestinal symptoms such as nausea, vomiting, and diarrhea for 14. Nine admissions, 5.6 percent, stemmed from immune-related adverse events, the distinctive toxicities that arise when checkpoint inhibitors unleash the immune system against healthy tissue. Severity was graded using the Common Terminology Criteria for Adverse Events, version 5.0, the standard lexicon for oncology toxicity.

To determine how many of these admissions were avoidable, the investigators did something unusually rigorous for a retrospective chart review. Two pharmacists independently confirmed the causal link between each hospitalization and anticancer therapy, resolving disagreements by consensus. A multidisciplinary panel—two pharmacists, a physician, and a nurse—then re-examined each case, asking a deliberately constrained question: could the admission reasonably have been avoided through guideline-concordant interventions implemented before the adverse event began, using only the clinical information available at that time? The framework drew on domestic and international guidelines covering prophylactic granulocyte colony-stimulating factor, antiemetic therapy, antimicrobial prophylaxis, dose adjustment per package inserts, vaccination, and monitoring systems. A case was labeled potentially preventable if a recommended prophylactic intervention had been omitted, if treatment intensity had not been modified despite known risk factors, or if assessment was delayed after early symptoms appeared.

The verdict: 30 of the 161 cases, or 18.6 percent, were potentially preventable, with a 95 percent confidence interval of 13.4 to 25.4 percent. The largest share of missed opportunities—15 cases, half of all preventable admissions—involved optimization of supportive care. Seven patients needed better granulocyte colony-stimulating factor prophylaxis to ward off febrile neutropenia, three needed prophylactic antimicrobials, four needed intensified antiemetic regimens, and one needed better management of comorbidities. Six cases pointed to suboptimal anticancer therapy itself, where doses should have been reduced or modified based on patient characteristics. Another six involved failures of early detection—insufficient monitoring and patient education that might have caught deterioration before it required an ambulance. Five cases involved preventive measures that should have preceded treatment altogether, chiefly varicella-zoster vaccination.

Febrile neutropenia was the single most common event among the preventable admissions, appearing in 10 of the 30 cases, followed by gastrointestinal symptoms in 6. That pattern is clinically meaningful because febrile neutropenia is precisely the complication that prophylactic G-CSF is designed to prevent. International guidelines, including the American Society of Clinical Oncology recommendations cited by the authors, specify risk thresholds for primary and secondary prophylaxis, and real-world evidence from Japan has confirmed the effectiveness of pegfilgrastim in routine practice. When a hospitalization occurs because that well-established prophylaxis was never given, the gap is not scientific uncertainty—it is a systems failure in translating evidence into practice.

The antiemetic findings reveal a subtler problem. Standard prophylaxis was administered according to Japan Society of Clinical Oncology guidelines, yet in some patients, escalation based on nausea and vomiting experienced in prior treatment cycles was never implemented. Chemotherapy-induced nausea and vomiting is famously individual: some patients are inherently vulnerable, and a patient who vomited through cycle one is highly likely to vomit through cycle two unless the regimen is intensified. The authors point to emerging risk-stratification tools, such as those validated in the MyRisk trial, that could support individualized antiemetic strategies even in the first cycle. Similarly, while guidelines do not endorse universal antimicrobial prophylaxis, the study suggests that selected high-risk patients—those who are elderly, have a history of severe neutropenia, or carry multiple overlapping immunosuppressive factors—may merit risk-adapted prophylaxis that was not delivered.

Context matters for interpreting the 18.6 percent figure. Large administrative database studies in the United States have estimated that only about 5 to 7 percent of systemic anticancer therapy-related hospitalizations are potentially preventable, while detailed chart-review studies have suggested figures as high as 21 percent among patients with gastrointestinal malignancies. The Japanese study sits toward the higher end of that spectrum, likely reflecting the granularity of its method: administrative codes cannot capture a missing dose of prophylaxis or a delayed phone call, but a multidisciplinary team reading the full medical record can. The authors also note that most prior evidence comes from Western countries, and healthcare delivery systems in Asia may differ in ways that affect both the risk of unplanned admission and the opportunities to prevent it.

The study is candid about its limitations. It was single-center and retrospective, with a modest sample size, so institutional and regional factors may shape the results. Immune-related adverse events occurring more than 30 days after checkpoint inhibitor administration fell outside the observation window and were excluded. Preventability judgments, though structured and consensus-based, ultimately rest on retrospective clinical opinion; the team used a study-specific framework rather than a validated instrument, could not quantify inter-rater reliability, and did not prospectively test whether the identified interventions would actually prevent admissions. The remaining 131 cases were judged difficult to prevent, driven by disease progression, unpredictable serious toxicities, or adverse events that occurred despite appropriate supportive care—a reminder that cancer drug toxicity cannot be engineered away entirely.

Even with those caveats, the message is actionable and, in its way, optimistic. Nearly one in five treatment-related hospital admissions at this institution traced back to identifiable, correctable gaps: a prophylactic drug not given, a dose not adjusted for renal function or age, a worsening symptom not escalated, a vaccine not administered before immunosuppression began. The authors argue that standardized, regimen-specific supportive care protocols, systematic risk stratification, and team-based care models—incorporating electronic medical record alerts, pharmacist-led medication review, and nurse-led monitoring—could close many of these gaps. Structured follow-up for high-risk patients, including scheduled outpatient visits, telephone monitoring, and education that teaches patients when to call, may enable earlier intervention as symptoms worsen. As cancer care continues its migration into outpatient settings, where patients spend most of their treatment journey away from clinical oversight, the study suggests that the cheapest hospital bed is the one a patient never needs—and that a meaningful fraction of those beds can be freed with tools medicine already has.

Subject of Research: Preventability of unplanned hospitalizations due to adverse events in patients receiving systemic anticancer therapy

Article Title: Potentially preventable unplanned hospitalizations due to adverse events in patients receiving systemic anticancer therapy: a retrospective single-center study

Article References: Potentially preventable unplanned hospitalizations due to adverse events in patients receiving systemic anticancer therapy: a retrospective single-center study. (n.d.). https://doi.org/10.1007/s00520-026-11351-0

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11351-0

Keywords: systemic anticancer therapy, unplanned hospitalization, adverse events, febrile neutropenia, supportive care, G-CSF prophylaxis, antiemetic therapy, antimicrobial prophylaxis, vaccination, patient safety, oncology, pharmacists