Table 1. Decedent Characteristics Stratified by Association With LTC as Identified by the Natural Language Processing Algorithm (National Violent Death Reporting System, 2003 to 2015).
| Variable | All Deaths Not Associated With LTC (n = 46 722) | All Deaths Associated With LTC (n = 1037) | Living in LTC (n = 428) | Transitioning Into or Out of LTC (n = 449) | Otherwise Associated With LTC (n = 160) | χ2 Statistic | dfa | P Valuea |
|---|---|---|---|---|---|---|---|---|
| Demographic Characteristics | ||||||||
| Age, median (IQR), y | 64 (15) | 79 (17) | 79 (19) | 80 (13) | 75 (20) | 67 | 46 | .02 |
| Male, No. (%) | 36 230 (77.6) | 765 (73.8) | 277 (64.7) | 360 (80.2) | 128 (80.0) | 31 | 2 | <.001 |
| Non-Hispanic white, No. (%) | 41 983 (89.8) | 978 (94.3) | 396 (92.5) | 428 (95.3) | 154 (96.2) | 5 | 2 | .10 |
| Education, No. (%) | ||||||||
| <High school | 5199 (11.1) | 126 (12.2) | 41 (9.6) | 64 (14.3) | 21 (13.1) | 13 | 6 | .04b |
| High school or GED | 11 624 (24.9) | 255 (24.6) | 93 (21.7) | 116 (25.8) | 46 (28.7) | |||
| >High school | 12 122 (25.9) | 256 (24.7) | 114 (26.6) | 112 (24.9) | 30 (18.8) | |||
| Unknown | 16 065 (34.4) | 360 (34.7) | 165 (38.6) | 139 (31.0) | 56 (35.0) | |||
| Missing | 1712 (3.7) | 40 (3.9) | 15 (3.5) | 18 (4.0) | 7 (4.4) | |||
| Marital status, No. (%) | ||||||||
| Married/in a relationship | 20 687 (44.3) | 333 (32.1) | 100 (23.4) | 117 (26.1) | 116 (72.5) | 162 | 8 | <.001b |
| Single/never married | 5067 (10.8) | 98 (9.5) | 58 (13.6) | 28 (6.2) | 12 (7.5) | |||
| Widowed | 7117 (15.2) | 386 (37.2) | 168 (39.3) | 204 (45.4) | 14 (8.8) | |||
| Divorced/separated | 13 047 (27.9) | 214 (20.6) | 99 (23.1) | 98 (21.8) | 17 (10.6) | |||
| Unknown | 759 (1.6) | S | S | S | S | |||
| Missing | 45 (0.1) | S | S | S | S | |||
| Health and Psychosocial Circumstances, No. (%) | ||||||||
| Depressed mood | 16 789 (35.9) | 468 (45.1) | 168 (39.3) | 202 (45.0) | 98 (61.3) | 23 | 2 | <.001 |
| Physical health problem | 17 632 (37.7) | 632 (60.9) | 217 (50.7) | 351 (78.2) | 64 (40.0) | 104 | 2 | <.001 |
| History of suicide ideation | 5619 (12.0) | 130 (12.5) | 40 (9.3) | 76 (16.9) | 14 (8.8) | 14 | 2 | <.001 |
| Recent crisis | 5272 (11.3) | 205 (19.8) | 51 (11.9) | 121 (26.9) | 33 (20.6) | 31 | 2 | <.001 |
| Any crisis | 9434 (20.2) | 291 (28.1) | 84 (19.6) | 162 (36.1) | 45 (28.1) | 29 | 2 | <.001 |
| Eviction or loss of home | 1065 (2.3) | 61 (5.9) | 10 (2.3) | 41 (9.1) | 10 (6.2) | 18 | 2 | <.001 |
| Death of friend or relative | 3610 (7.7) | 110 (10.6) | 43 (10.0) | 57 (12.7) | 10 (6.2) | 5 | 2 | .07 |
| Financial problem | 4341(9.3) | 53 (5.1) | 15 (3.5) | 14 (3.1) | 24 (15.0) | 38 | 2 | <.001 |
| Family relationship | 1911 (4.1) | 85 (8.2) | 8 (1.9) | 38 (8.5) | 39 (24.4) | 78 | 2 | <.001 |
| Death Characteristics, No. (%) | ||||||||
| Cause of death | ||||||||
| Suicide | 41 596 (89.0) | 980 (94.5) | 384 (89.7) | 439 (97.8) | S | NA | NA | NA |
| Other or undetermined | 4847 (10.4) | 57 (5.5) | 44 (10.3) | S | S | |||
| Accidental firearm | 279 (0.6) | 0 | 0 | 0 | 0 | |||
| Means of injury | ||||||||
| Gun or rifle | 26 769 (57.3) | 529 (51.0) | 104 (24.3) | 319 (71.0) | 106 (66.2) | NA | NA | NA |
| Sharp or blunt instrument | 1202 (2.6) | 40 (3.9) | S | S | S | |||
| Poisoning | 9551 (20.4) | 186 (17.9) | 97 (22.7) | 62 (13.8) | 27 (16.9) | |||
| Fall | 778 (1.7) | 47 (4.5) | S | S | S | |||
| Other | 1549 (3.3) | 35 (3.4) | S | S | S | |||
| Unknown | 902 (1.9) | 11 (1.1) | S | S | S | |||
| Missing | 5971 (12.8) | 189 (18.2) | 130 (30.4) | 42 (9.4) | 17 (10.6) | |||
| SRF injury location | 149 (0.3) | 114 (11.0) | 106 (24.8) | S | S | NA | NA | NA |
| LTC/nursing home death location | 475 (1.0) | 94 (9.1) | 88 (20.6) | S | S | NA | NA | NA |
Abbreviations: GED, General Education Development; IQR, interquartile range; LTC, long-term care; NA, not applicable; S, suppressed (if <5 to protect confidentiality); SRF, supervised residential facility.
a
Test statistic and P value for Kruskal-Wallis test or Pearson χ2 test comparing across all deaths associated with LTC (living in LTC vs transitioning into or out of LTC vs otherwise associated with LTC).
b
The test of significance does not include the missing category in the calculations of the test statistic and P value.