Table 4. Medical Care, Location of Death, Quality of Death and Survival by EOL Discussion among Deceased Propensity-Score Matched Patients (N=145) a.
| Patient discussed EOL care preferences with physician | ||||||
|---|---|---|---|---|---|---|
| Yes | No | AOR b | (95%CI) | p-value | ||
| No. of cases; N (%) | 75 (51.7) | 70 (48.3) | ||||
| Medical care received in last week; N (%) | ||||||
| ICU stay | 2 (2.7) | 10 (14.3) | 0.01 | (0.02-0.6) | 0.01* | |
| Ventilator use | 1 (1.3) | 10 (14.3) | 0.03 | (0.002-0.3) | 0.005** | |
| Resuscitation | 1 (1.4) | 6 (8.7) | 0.1 | (0.02-1.3) | 0.09§ | |
| Chemotherapy | 4 (5.3) | 7 (10.0) | 0.5 | (0.1-1.8) | 0.30 | |
| Inpatient hospice utilized | 8 (10.7) | 5 (7.1) | 1.8 | (0.5-6.5) | 0.34 | |
| Inpatient hospice stay ≥ 1 week | 4 (5.4) | 2 (3.0) | 3.7 | (0.4-38.2) | 0.27 | |
| Outpatient hospice utilized | 58 (77.3) | 40 (57.1) | 3.2 | (1.5-6.9) | 0.004** | |
| Outpatient hospice stay ≥ 1 week | 52 (69.3) | 34 (50.8) | 2.5 | (1.2-5.0) | 0.01* | |
| Place of Death; N (%) | ||||||
| ICU | 2 (2.7) | 9 (12.9) | 0.1 | (0.03-0.7) | 0.02* | |
| Hospital | 15 (20.0) | 18 (25.7) | 0.7 | (0.3-1.6) | 0.45 | |
| Inpatient Hospice | 5 (6.7) | 3 (4.3) | 1.9 | (0.4-8.8) | 0.44 | |
| Home | 47 (62.7) | 38 (54.3) | 1.3 | (0.6-2.6) | 0.49 | |
| Quality of Life in the Last Week of Life; cmean (S.D.) | β | SE | p-value | |||
| Psychological Distress | 3.7 (3.0) | 3.2 (3.3) | 0.5 | 0.6 | 0.37 | |
| Physical Distress | 3.6 (3.2) | 4.5 (3.7) | -1.2 | 0.6 | 0.04* | |
| Quality of Death | 6.3 (2.7) | 5.7 (3.3) | 0.5 | 0.5 | 0.39 | |
| HR | (95%CI) | p-value | ||||
| Survival; median (quartiles) | 88 [54-218] | 85 [30-253] | 0.8 | (0.6-1.1) | 0.22 | |
§
Borderline significant 0.05≤p-value<0.10
*
p-value<0.05
**
p-value<0.01
***
p-value<0.001
a
Among the deceased propensity-matched patients, all variables shown in Table 2 did not differ by EOL discussions.
b
The OR here is conditional on quintiles of predicted propensity scores and adjusted for confounders of socio-demographic characteristics, health status measures, recruitment sites, terminal illness acknowledgement, treatment preferences and survival if they remain significant in the multivariate model.
c
Rated at postmortem. Higher score indicates more distress for psychological and physical scales whereas higher score indicates better quality of life.