Skip to main content
. Author manuscript; available in PMC: 2017 Jun 14.
Published in final edited form as: JAMA Intern Med. 2016 Aug 1;176(8):1095–1102. doi: 10.1001/jamainternmed.2016.1200

Table 2.

Adjusteda Proportions for Measures of Care at the End of Life and Family Perceptions of Quality Outcomes by Diagnosis

Outcomes Cancer Dementia End-stage Renal Disease Cardiopulmonary Failure Frailty Other P Valueb
All Veteran Decedents (N=57,728) (N ((%))c 23,523 (40.8) 3675 (6.4) 2265 (3.9) 13,854 (24.0) 9931 (17.2) 4480 (7.8)

Measures of Care at the End of Life
Palliative care consultation 73.5 61.4 50.4 46.7 43.7 41.5 <.001
Do-not-resuscitate order 95.3 93.5 87.0 86.3 88.6 83.9 <.001
Died in inpatient hospice 42.9 32.3 24.3 22.9 20.3 20.6 <.001
Died in the intensive care unit 13.4 8.9 32.3 34.1 35.2 37.4 <.001

Bereaved Family Survey Participants (n=34,005) (%)d,e 40.3 6.6 3.7 24.4 17.4 7.6

Overall rating of patient’s care was excellent 59.2 59.3 54.8h 54.8f 53.7f 55.0f <.001
Providers always listened to concerns 73.8 75.7 68.6f 71.5f 70.5f 73.0 <.001
Providers always provided the medical treatment that patient and family wanted 79.1 80.4 73.4f 76.8f 76.5f 77.4 <.001
Providers always kept family informed about patient’s condition and treatment 68.2 71.1h 63.8g 65.9f 66.6 67.5 .0011
Providers always gave enough emotional support prior to the patient’s death 64.6 67.5h 61.5 62.1f 62.0h 63.3 <.001
Patient had frequent uncontrolled paini 55.0 49.4f 54.3 55.9 53.3 55.3 .003
a

Analyses adjusted for age, race, gender, next-of-kin relationship, and comorbidity, and clustered by facility.

b

P value for the overall association of diagnosis with each outcome

c

Table excludes 25 decedents from Table 1 who were missing covariate data. P < .003 for the comparison of each diagnosis versus cancer for all “measures of care at the end of life” outcomes

d

Analyses weighted for survey non-response. Table excludes 10 Bereaved Family Survey participants for whom no survey weight could be calculated because of missing covariate data.

e

Among the 34,005 Bereaved Family Survey participants for whom survey weight could be calculated. Sample size varies to reflect % missing data for individual survey items: overall rating (1.4), providers listened (2.6), provision of desired medical treatment (3.5), provider informed family (1.9), frequent uncontrolled pain (13.7), and providers gave enough emotional support (3.0)

f

P < .001 for the comparison of each diagnosis versus cancer

g

P < .01 for the comparison of each diagnosis versus cancer

h

P < .05 for the comparison of each diagnosis versus cancer

i

Includes only the 81.7% of patients who, per family report, experienced pain