Hospice Medications Associated with Increased Mortality Risk in Dementia Patients

Hospice care is fundamentally designed to provide comfort, peace, and dignity to patients navigating the end-of-life journey. Yet, when focusing on the increasingly significant demographic of Americans living with dementia who are enrolled in hospice programs, the trajectory of care becomes notably more complex. Unlike many terminal conditions with relatively predictable courses, dementia presents a prolonged and unpredictable decline, underscoring the critical need for treatment plans that are carefully aligned with each patient’s individual goals and stage of illness. This nuanced approach is essential to ensuring that therapeutic interventions genuinely enhance quality of life rather than inadvertently inflicting harm.

A groundbreaking study conducted by researchers at the University of Michigan sheds light on a pressing concern within this domain: the use of benzodiazepines and antipsychotics in hospice patients with dementia. Published in the reputable journal JAMA Network Open, the analysis reveals a stark association between these commonly prescribed psychoactive medications and increased mortality risk among this vulnerable population. Specifically, the study evaluated data from over 139,000 nursing home residents with Alzheimer’s disease and related dementias who were enrolled in hospice between 2014 and 2018, identifying that initiation of benzodiazepines after hospice enrollment was linked to a 41% increased likelihood of death within six months. Similarly, patients starting on antipsychotics faced a 16% higher chance of mortality within the same time frame compared to matched counterparts who did not receive these drugs.

The findings emerge against a backdrop where nearly half of these patients were prescribed benzodiazepines and approximately 13% received antipsychotics shortly after hospice admission. This prescribing pattern is particularly alarming given the average hospice stay exceeded 130 days, a period much lengthier than the terminal weeks commonly associated with hospice care. This extended duration highlights the critical fact that many hospice enrollees with dementia are not imminently dying but are instead experiencing a protracted decline, which demands a reassessment of clinical approaches tailored to their unique disease course.

Benzodiazepines, such as Ativan and Valium, and antipsychotics, including medications like Haldol and Zyprexa, are often employed to manage complex neuropsychiatric symptoms in dementia. These symptoms—ranging from agitation and anxiety to episodes of delirium—can be profoundly distressing for patients and their families. While these drugs may offer symptomatic relief in some cases, their use in frail, older adults carries well-documented risks. These include sedation, increased confusion, and heightened susceptibility to falls, all of which can significantly exacerbate morbidity. Importantly, the U.S. Food and Drug Administration has issued boxed warnings on antipsychotics warning of increased mortality risk in dementia patients, a cautionary note echoed in emerging concerns about benzodiazepines.

The study’s lead author, Dr. Lauren B. Gerlach, a geriatric psychiatrist, emphasizes the pivotal challenge facing hospice care providers: prognostication in dementia is notoriously difficult. The six-month life expectancy criterion that determines Medicare hospice eligibility often fails to capture the prolonged and erratic progression of dementia. Consequently, nearly one in five patients exceeds this timeframe, risking discharge from hospice services even while continuing to experience significant needs for symptom management and support. This structural mismatch calls for urgent re-examination of both policy frameworks and clinical guidelines.

Moreover, Gerlach highlights systemic gaps in monitoring and oversight of medication prescribing within hospice settings. A critical source of the current data stems from a brief period between 2014 and 2018 when hospices were required to report prescribing information to federal authorities. Since then, this transparency has been lost, rendering hospice medication prescribing effectively a “black box” to regulators and researchers alike. Unlike nursing homes, where benzodiazepine and antipsychotic use is rigorously tracked and factored into quality ratings, no such accountability mechanisms exist within hospice care, leaving a glaring blind spot in quality assessment and patient safety monitoring.

The variability in prescribing practices is another dimension underscored by the University of Michigan team’s prior research, which revealed drastic disparities among hospice agencies. Their data indicated a staggering range in benzodiazepine prescribing rates—from 12% to as high as 80%—and antipsychotic use varying from 6% to 62%, after controlling for patient characteristics. Notably, larger and for-profit hospice agencies tended to prescribe these medications more frequently. Such heterogeneity suggests that beyond clinical indicators, institutional culture, agency policies, and perhaps economic incentives may be influencing prescribing behavior, introducing a layer of inconsistency that may imperil patient outcomes.

The implications of these findings extend deeply into clinical practice and healthcare policy. As the proportion of hospice patients diagnosed with dementia continues to climb—currently representing 60% of all hospice enrollees—the need for more dementia-specific hospice care models becomes increasingly critical. The standard Medicare hospice benefit, staples of which were established when hospice patients predominantly had terminal cancer with predictable decline trajectories, inadequately addresses the unique course of neurodegenerative illnesses such as dementia. Tailored care models that accommodate prolonged, variable progression and optimize individualized symptom management are essential.

Educational initiatives and clinical guidelines must further evolve to support hospice clinicians in making nuanced pharmacological decisions balancing symptom relief and potential harm. The study’s authors call for enhanced prescribing transparency and robust quality measures focused specifically on dementia care in hospice. These would facilitate ongoing assessment and recalibration of care strategies to ensure that medications serve to improve patient well-being rather than contribute to premature mortality.

Ultimately, the research spearheaded by Dr. Gerlach and her colleagues catalyzes a critical conversation around appropriate end-of-life care for people with dementia. It challenges healthcare providers, policymakers, and hospice organizations to confront the tension between symptom management and associated risks in a population characterized by complex, fluctuating needs. This work invites a reassessment of existing hospice structures, urging the development of more sophisticated, evidence-informed, and patient-centered frameworks that honor the dignity and life quality of those living with dementia until their final days.

In conclusion, this seminal study amplifies an urgent call to action for the healthcare community. By illuminating the risks linked to benzodiazepine and antipsychotic use in dementia hospice patients—risks that are both substantial and potentially modifiable—it lays the groundwork for reimagining hospice care paradigms aligned with the realities of neurodegenerative disease trajectories. The ultimate goal is clearer: providing care that truly enhances comfort, autonomy, and respect in the twilight of life.

• Gerlach, L.B., Zhang, L., Kim, H.M., Teno, J., Maust, D.T. (2025). Benzodiazepine or Antipsychotic Use and Mortality Risk Among Patients With Dementia in Hospice Care. JAMA Network Open. DOI: 10.1001/jamanetworkopen.2025.37551
  Keywords: Hospice care, Dementia, Benzodiazepines, Antipsychotics, Mortality risk, Alzheimer’s disease, Medicare, Prescribing patterns, End-of-life care, Geriatrics, Psychiatry, Health policy