BACKGROUND/RATIONALE:
The predominant non-hospice palliative care (PC) delivery model in the U.S. is the interdisciplinary in-patient palliative care consultation team (PCCT). The Veterans Health Administration (VHA) recognizes the need for high quality hospice and palliative care (HPC) for hospitalized veterans: all VHA facilities are mandated to have an interdisciplinary PCCT to respond to inpatient consultations, among other responsibilities. In light of the growing use of such teams, appropriately designed and powered studies are necessary to demonstrate their effect on resource use and costs of care.
OBJECTIVE(S):
The research objectives of the study are:
1. To compare total direct costs and pharmacy and ancillary (laboratory and radiology) costs of hospitalized veterans with advanced illness who receive PC versus UC, adjusting for the treatment selection effect. To compare total direct costs and pharmacy and ancillary (laboratory and radiology) costs of hospitalized veterans who received PC versus UC during their terminal hospitalization in the study period, adjusting for the treatment selection effect
2. To compare the probability of ICU admission and length of stay (LOS) for hospitalized veterans with advanced illness who receive PC versus UC, adjusting for the treatment selection effect. To compare the probability of ICU admission and length of stay (LOS) for veterans who received PC versus UC during their terminal hospitalization in the study period, adjusting for the treatment selection effect
METHODS:
This is an observational, retrospective cost study using a VA payer perspective of all patients who died at 5 acute care facilities in VISN 3 during FY 03 and FY 04. Dependent variables include ICU use and per diem pharmacy, laboratory, radiology, and total direct costs during the patient’s terminal hospitalization. The independent variable of interest is whether the patient was seen by the PCCT during the terminal hospitalization. Other independent variables include principal diagnosis, comorbidty, surgical or medical stay, age, length of stay (LOS) and ICU LOS. Data sources will include DSS National Clinical Data Extracts (NCDE), patients’ medical records and VISN 3 palliative care program data. Three generalized linear models (GLM) will be estimated for total direct, ancillary (laboratory and radiology costs) and pharmacy costs. A Probit regression will be used to examine the relationship between ICU admission and PCCT care. All models will be adjusted for the treatment selection effect.
FINDINGS/RESULTS:
There are no results at this time.
IMPACT:
Our long-term goal is to inform current and future efforts of VHA to expand PC programs and ultimately, to improve the care received by veterans with life-limiting illness.
PUBLICATIONS:
Journal Articles
- Penrod JD, Deb P, Luhrs C, Dellenbaugh C, Zhu CW, Hochman T, Maciejewski ML, Granieri E, Morrison RS. Cost and utilization outcomes of patients receiving hospital-based palliative care consultation. Journal of Palliative Medicine. 2006; 9(4): 855-60.
