Background: Most Americans die in the acute care hospital, where aggressive, life-prolonging interventions are readily performed. Although patients with incurable illness might prefer palliative care, perceived differences in prognosis by physicians may influence the type of care provided. Patients with advanced cancer and advanced dementia represent 2 extremes in the use of hospice services and may also be treated differently in the acute care hospital. We tested this hypothesis and quantitated the use of nonpalliative interventions in hospitalized, incurably ill patients.
Methods: Charts of elderly patients with advanced dementia or metastatic solid tumor malignancy who died during a 13-month period in a tertiary care acute teaching hospital were reviewed. Main outcome measures included the number of patients receiving invasive of noninvasive (but complex) diagnostic tests, invasive nonpalliative treatments, cardiopulmonary resuscitation, systemic antibiotics, and do-not-resuscitate orders.
Results: Charts of 164 patients (80 with dementia and 84 with cancer) were reviewed. Overall, 47% received invasive nonpalliative treatments. Controlling for age, sex, length of stay, and insurance status, the groups were equally likely to receive nonpalliative treatments (P = .75), but patients with dementia were more likely to receive new feeding tubes (P = .02). Cardiopulmonary resuscitation was attempted for 24% of each group. Patients with cancer more often received invasive (41% vs 13%; P = .002) and complex noninvasive diagnostic tests (49% vs 23%; P = .02). Overall, 88% received antibiotics, often empirically, but, controlling for neutropenia and invasive tests and treatments, patients with dementia were significantly more likely to receive antibiotics for an identifiable infection (P = .004).
Conclusions: Incurably ill patients often receive nonpalliative interventions at the end of life. Patients with cancer receive more diagnostic tests, but patients with dementia receive more enteral tube feeding. Patients commonly receive systemic antibiotics, often empirically. Cardiopulmonary resuscitation is equally applied, but is out of proportion to expected survival.