BACKGROUND/RATIONALE:
In June 2004, VA appointed its first director of Hospice and Palliative Care (HPC) as part of its efforts to strengthen the provision of hospice and palliative care within VA. Initiatives have included development of inpatient hospice and palliative care units within each facility, educating inpatient staff in providing quality end-of-life care, and collaboration with community hospice agencies through the national hospice-veteran partnership program. FY2006 strategic initiatives for the hospice and palliative care program include: (1) improving access to hospice and palliative care in both inpatient and outpatient settings, 2) promoting quality improvement through program development and outcome measurement and 3) enhancing staff expertise in the delivery of care at the end of life. The action plan for improving access to hospice and palliative care in inpatient and outpatient settings includes the exploration of automated case finding techniques. This specific task was articulated because a FY2005 survey found that of 81% of facilities had no automated case finding method to identify veterans appropriate for HPC (FY2005 Status Report).

OBJECTIVE(S):
The objectives of the project were:

(1) to work with an expert panel to identify diagnoses and/or events in inpatient, outpatient and long term care settings that could indicate veterans at risk for needing specialized end-of-life care services;
(2) to create computer algorithms for these indicators using data elements available in the various national VA databases;
(3) to determine the prevalence of these indicators by applying them to various national VA databases; and
(4) to test the final indicators agreed upon by the expert panel by merging patients identified by the indicators with mortality data to see how predictive the indicators are.

METHODS:
In collaboration with the expert panel, we identified potential indicators of mortality. Algorithms for these indicators used ICD-9 codes, bedsections, and clinic stop, and an index date proxying initial diagnosis of a condition/event. Data sources were VA inpatient and outpatient databases (FY2001-FY2005). Mortality data was from VIReC (deaths through March 2006). Survival analysis (SAS LifeReg) was used to predict months of survival. Median months survival was calculated by inpatient/outpatient status, condition, and age

FINDINGS/RESULTS:
Patients in groups with median months of survival <=12 were considered appropriate candidates for referral to HPC. Those identified included inpatients with head/neck, trachea/bronchus/lung, prostate, colon, liver, lymphomas, pancreatic, acute leukemias, melanoma, and central nervous system cancers. Additionally, based on supplementary criteria, patients with 2 hospitalizations for either CHF or COPD within a year, and ICU patients with either cancer diagnoses or lengths of stay greater than 10 days were identified as potentially appropriate candidates.

IMPACT:
The director of HPC presented the findings of this report to GEC staff and program managers for feedback. At the national level, the question is whether or not this work might serve as the basis for a quality performance measure. At the individual facility level, the question is whether or not this work might be helpful in identifying individuals appropriate for referral to HPC. HPC is often seen as contributing significantly to quality of care as perceived by end-of-life patients and their families. It is often essential in allowing patients to remain in their homes at the end-of-life.

---

External Links for this Project

---

PUBLICATIONS:

None at this time.

DRA: Aging, Older Veterans' Health and Care
DRE: none
Keywords: Cost, End-of-life, Hospice
MeSH Terms: none