Table 1. Demographic Characteristics of the Cohorts, Including Those Discharged From Inpatient Hospital to Death.
| Characteristic | Full sample, patients, No. (%) (n = 405) | Matched sample, patients, % (n = 395)a | Unmatched HIH patients, No. (%) (n = 10) | ||||
|---|---|---|---|---|---|---|---|
| T-HIH (n = 108) | Controls (n = 297) | Bias, % | T-HIH (n = 98) | Controls (n = 297) | Bias, % | ||
| Age, mean (SD), y | 67.7 (1.14) | 66.6 (0.72) | 9.4 | 68.2 | 68.4 | −1.9 | 63.2 (0.16) |
| Married | 28 (25 .9) |
120 (40.4) | −31.0 | 28.6 | 31.9 | −7.2 | NA |
| VHA priority status P1A | 42 (38.9) | 117 (39.4) | −1.0 | 38.8 | 35.2 | 7.4 | 4 (40.0) |
| Hospitalizations in the prior 30 d | 13 (12.0) | 39 (13.1) | −3.3 | 13.3 | 11.9 | 4.1 | NA |
| Intensive care unit use during index hospitalization | 13 (12.0) | 62 (20.9) | −23.9 | 13.3 | 15.7 | −6.5 | NA |
| JEN Frailty Index score (reference, 0-5) | |||||||
| 6 | 21 (19.4) | 57 (19.2) | 0.6 | 19.4 | 18.8 | 1.6 | 2 (20.0) |
| ≥7 | 52 (48.1) | 124 (41.8) | 12.8 | 46.9 | 44.4 | 5.2 | 6 (60.0) |
| VHA Care Assessment of Need score, probability of an event at 90 d (reference, 90-d event) | |||||||
| ≤0.05 | 20 (18.5) | 84 (28.3) | −23.1 | 20.4 | 19.8 | 1.4 | NA |
| >0.05 to ≤0.29 | 40 (37.0) | 143 (48.1) | −22.5 | 40.8 | 41.5 | −1.3 | NA |
| >0.29 | 35 (32.4) | 33 (11.1) | 53.2 | 26.5 | 22.5 | 10.0 | 9 (90.0) |
| Year (reference, 2012-2014) | |||||||
| 2015-2016 | 43 (42.6) | 95 (32.0) | 22.0 | 43.9 | 41.4 | 5.1 | 3 (30.0) |
| 2017-2018 | 38 (35.2) | 60 (20.2) | 33.8 | 31.6 | 34.6 | −6.8 | 7 (70.0) |
| Major diagnostic group of the index hospitalization (reference, all other types) | |||||||
| Respiratory | 14 (13.0) | 72 (24.2) | −29.2 | 14.3 | 15.4 | −3.0 | NA |
| Circulatory | 80 (74.1) | 133 (44.8) | 62.3 | 72.4 | 73.4 | −1.9 | 9 (90.0) |
| Hierarchical condition categories score, mean (SD) | 3.54 (0.16) | 2.79 (0.09) | 46.7 | 3.43 | 3.36 | 4.5 | 4.67 (1.44) |
| Diagnoses, mean (SD), No. | 12.76 (0.45) | 10.79 (0.23) | 45.1 | 12.29 | 12.67 | −8.9 | 17.40 (4.27) |
Abbreviations: HIH, Hospital in Home; NA, not applicable; T-HIH, Transfer Hospital in Home; VHA, Veterans Healthcare Administration.
a
For the matched sample, SDs are not available for the means and counts are not available for the proportions because we used inverse probability weighting, rather than individual matching.