Concentrating hospital-wide deaths in a palliative care unit: the effect on place of death and system-wide mortality.

INTRODUCTION: We studied the impact of an 11-bed inpatient palliative care unit (PCU) on site of death and observed mortality in the health system, oncology, and palliative care units. Observers were concerned that an active PCU would attract dying patients and worsen comparative mortality rates for Medicare and U.S. News & World Report comparisons. METHODS: We reviewed 10 years of experience with all patients who died in the hospital before and after we opened our PCU in 2000. RESULTS: The PCU concentrated dying patients on the PCU but total deaths did not change over 10 years and remained approximately 3% of admissions. Within 2 years, one quarter of all health system decedents died on the PCU. The proportion who died on the oncology floor and general units declined, but the number of intensive care unit deaths did not change. CONCLUSIONS: An inpatient PCU did not increase the hospital-wide death rate. The PCU did change the site of death to a more appropriate venue for one quarter of patients.

A high-volume specialist palliative care unit and team may reduce in-hospital end-of-life care costs.

BACKGROUND: Current end-of-life hospital care can be of poor quality and high cost. High volume and/or specialist care, and standardized care with clinical practice guidelines, has improved outcomes and costs in other areas of cancer care. METHODS: The objective of this study was to measure the impact of the palliative care unit (PCU) on the cost of care. The PCU is a dedicated 11-bed inpatient (PCU) staffed by a high-volume specialist team using standardized care. We compared daily charges and costs of the days prior to PCU transfer to the stay in the PCU, for patients who died in the first 6 months after the PCU opened May 2000. We performed a case-control study by matching 38 PCU patients by diagnosis and age to contemporary patients who died outside the PCU cared for by other medical or surgical teams, to adjust for potential differences in the patients or goals of care. RESULTS: The unit admitted 237 patients from May to December 2000. Fifty-two percent had cancer followed by vascular events, immunodeficiency, or organ failure. For the 123 patients with both non-PCU and PCU days, daily charges and costs were reduced by 66% overall and 74% in "other" (medications, diagnostics, etc.) after transfer to the PCU (p < 0.0001 for all). Comparing the 38 contemporary control patients who died outside the PCU to similar patients who died in the PCU, daily charges were 59% lower (US dollars 5304 +/- 5850 to US dollars 2172 +/- 2250, p = 0.005), direct costs 56% lower (US dollars 1441 +/- 1438 to US dollars 632 +/- 690, p = 0.004), and total costs 57% lower (US dollars 2538 +/- 2918 to US dollars 1095 +/- 1153, p = 0.009). CONCLUSIONS: Appropriate standardized care of medically complex terminally ill patients in a high-volume, specialized unit may significantly lower cost. These results should be confirmed in a randomized study but such studies are difficult to perform.