The power of a plan: The impact of palliative care decisions at hospital discharge on readmission rates
Joshua Fabie, MS2; Kathryn Walker, PharmD, BCPS, CPE; Christopher Kearney, MD
Georgetown University School of Medicine; MedStar Union Memorial Hospital
Georgetown University
Results
Introduction
important targets for contemporary health care. Hospital or intensive care unit stays and life-
1.
Consulted
in 3 days
(n=95)
sustaining procedures performed near the end of life represent a substantial portion of health
care spending and are frequently associated with a worse quality of death1. Improving the
Initial consults
(n=236)
quality of health care for patients at or nearing the end of life and yielding cost savings are joint
benefits of an effective palliative medicine program. Referral to hospice, one important element
of palliative care, is associated with reductions in high-intensity care, particularly ICU
admissions, hospital admissions, and emergency department
visits2.
Verified date of
death
(n=180)
predictive for lower costs, and earlier palliative care consultation during hospital admission is
Palliative
(n=64)
67% agreed
to PC plan
Palliative
(n=51)
60% agreed
to PC plan
associated with lower cost of hospital stay for patients admitted with an advanced cancer
diagnosis3. This project aims to further our understanding of the role that palliative medicine
plays in optimizing end-of-life care for patients with serious illnesses.
2.
80
Plan of care distribution among
estimated prognosis
3.
3)
To compare hospital readmission rates and costs between patients who opted for
a palliative plan of care at discharge versus a similar group of patients who did
not agree to a palliative plan
To compare index hospitalization charges based on the timing of palliative
consultation and care plan decisions
Number of patients
2)
Index hospitalizations occurred over three months within three inner-city teaching hospitals
reviewed retrospectively to collect data for all patients surviving to discharge after receiving
an initial palliative consultation from December 2012 to April 2013.
The Social Security Death Index, obituary archives, and the EMR were queried in an attempt
to locate the dates of death of study participants. Death dates were assumed to belong to
participants if the name, date of birth, and state of residence listed in the external source
matched existing hospital records. Only patients with confirmed dates of death were
included in this study.
40
Readmission (RA) data included only MedStar hospital readmissions.
25%
21%
26%
7%
8%
0%
10
20
30
40
Days after discharge
Palliative adjusted
Non-palliative
50
60
Non-palliative adjusted
17%
20
Palliative (n=115)
56%
29%
0
Hours <1 week
<1
1-6
7-12 >1 year
month months months
<1 week
7-12 months
<1 month
>1 year
Figure 2. Discharge care plans organized by clinician estimated prognosis show increased
preference for a PC plan when prognosis is less than six months.
Figure 3. Clinicians tend to overestimate life expectancy in the last year of life, particularly in
patients surviving one month or less.
4.
30
60
30
60
Readmission rate
7%
8%
42%
55%
Mortality
80%
86%
29%
48%
Adjusted RA rate
21%
26%
45%
70%
$608.28
$662.66
$7,639.24
$12,441.42
Charges per RA
$8,744.04
$8,467.30
$18,390.76
$22,463.67
Total charges
$69,952.35
$76,205.69
$496,550.56
$808,692.12
Figure 5 and associated table. A palliative plan of care is associated with decreased all-cause 30 day
readmission rates (p<.0001), (adjusted p=.0003). Charges per readmission are less in the palliative
group at 30 days (p=.0119) and 60 days (p=.0028).
Discussion
Index admission charges by days to consult and plan of care
$60k
$54,852
$40k
A palliative plan of care is associated with decreased hospital
readmission rates and related charges. Early palliative care
consultation and agreement for a palliative plan of care following
discharge are linked to decreased charges for the index admission.
$39,741
$27,017
Days
Charges per patient
Hours
1-6 months
Non-palliative (n=65)
$29,667
The timing of consultation did not impact the likelihood of patients and
families agreeing to a palliative plan of care at discharge.
$20k
$12,408 $17,175
RA data will be
reported as all-cause overall RA rates as well as adjusted RA rates to account for
confounding due to expected high mortality rates. Adjusted RA rates represent the percent
of patients readmitted out of those still alive at each time point.
•
55%
42%
Palliative
Estimated
Prognosis 16%
31%
Patients were included in the palliative plan of care group if there was agreement at
discharge that care would be focused on comfort (e.g., hospice or MOLST form limiting
aggressive measures).
•
7%
Estimated prognosis
Palliative decision
Non-palliative decision
Hospital), each with an established interdisciplinary palliative consultation team. Charts were
•
45%
29%
To compare the palliative care (PC) group versus the non-palliative group in terms
of demographics, primary diagnoses, estimated prognosis, and duration of
survival after discharge
(MedStar Union Memorial Hospital, MedStar Good Samaritan Hospital, and MedStar Harbor
•
50%
6%
Survival
60
Methods
•
Estimated prognosis vs. actual
survival
70%
0
Objectives
1)
Readmission rates by plan of care and adjusted for mortality
75%
Consulted
after 3 days
Non-palliative
(n=85)
(n=34)
Unverified
date of death
(n=56)
Access to palliation is
5.
Non-palliative
(n=31)
Percent of patients readmitted
Achieving overall reductions in health care costs and reducing hospital readmission rates are
Financial data were reported from MedStar InfoMart for the study patients to compare
charges associated with index hospital admissions and readmissions.
N=6
N=3
1
Consult within 3 days (n=95)
Palliative decision
N=5
1
N=3
4
References
1.
Zhang B, et al. 2009. Health care costs in the last week of life: associations with end-of-life conversations. Arch Intern Med.
2009 Mar 9;169(5):480-8.
2.
Bergman J, et al. Hospice use and high-intensity care in men dying of prostate cancer. Archives of internal medicine.
2011;171(3):204-10.
3.
May P, et al. 2015. Prospective Cohort Study of Hospital Palliative Care Teams for Inpatients With Advanced Cancer: Earlier
Consultation Is Associated With Larger Cost-Saving Effect. J Clin Oncol. 2015 Jun 8.
Consult after 3 days (n=85)
Non-palliative decision
Mean Charges
Figure 4. Factors associated with increased index hospital charges include late
consultation (p<.0001) and non-palliative plan of care at discharge (p<.05).
Poster produced by Faculty & Curriculum Support (FACS), Georgetown University School of Medicine