Table 3. Multivariable Associations of Outpatient Physical and Injury-Related Health Care Visits With Suicide Attempt Among Regular Army Enlisted Soldiers With and Without a History of Mental Health Diagnosisa.
| Variableb | No Previous Mental Health Diagnosis (n = 103 525) | Any Previous Mental Health Diagnosis (n = 59 653) | Predictor by Mental Health Historyc | ||
|---|---|---|---|---|---|
| OR (95% CI) | SREd | OR (95% CI) | SREd | ||
| I. Days with an outpatient physical health care visit in the past 2 mo | χ24 = 21.1; P < .001 | ||||
| 0 | 1 [Reference] | 155 | 1 [Reference] | 320 | |
| 1 | 1.2 (1.0-1.3) | 177 | 1.5 (1.4-1.7) | 494 | |
| 2-4 | 1.5 (1.4-1.6) | 231 | 2.1 (1.9-2.3) | 673 | |
| 5-7 | 2.4 (2.1-2.7) | 369 | 3.6 (3.3-4.0) | 1156 | |
| ≥8 | 3.3 (2.9-3.8) | 515 | 5.1 (4.6-5.6) | 1602 | |
| Statistics | χ24 = 427.9; P < .001 | χ24 = 1555.5; P < .001 | |||
| II. Previous injury-related health care visite | χ22 = 38.9; P < .001 | ||||
| Any inpatient visit for previous injury | 2.7 (2.1-3.4) | 409 | 3.1 (2.8-3.4) | 1562 | |
| Only outpatient visit for previous injury | 1.7 (1.6-1.8) | 269 | 1.4 (1.3-1.5) | 692 | |
| No visit for previous injury | 1 [Reference] | 153 | 1 [Reference] | 466 | |
| Statistics | χ22 = 232.3; P < .001 | χ22 = 527.8; P < .001 | |||
| III. Recency of last injury-related outpatient health care visit, moe | χ24 = 2.1; P = .71 | ||||
| 1 | 3.0 (2.8-3.3) | 426 | 3.0 (2.8-3.3) | 1182 | |
| 2 | 2.4 (2.1-2.7) | 337 | 2.6 (2.4-2.9) | 966 | |
| 3 | 2.2 (1.9-2.6) | 313 | 2.3 (2.0-2.6) | 864 | |
| ≥4 | 1.6 (1.5-1.8) | 233 | 1.7 (1.5-1.8) | 619 | |
| No outpatient visit for injury | 1 [Reference] | 139 | 1 [Reference] | 337 | |
| Statistics | χ24 = 658.1; P < .001 | χ24 = 836.8; P < .001 | |||
| IV. Recency of last injury-related inpatient health care visit, moe | χ24 = 11.3; P = .02 | ||||
| 1 | 3.8 (2.3-6.3) | 669 | 6.7 (5.7-7.8) | 3992 | |
| 2 | 2.4 (1.2-5.1) | 485 | 4.1 (3.2-5.1) | 2574 | |
| 3 | 1.6 (0.7-3.5) | 333 | 4.0 (3.1-5.0) | 2559 | |
| ≥4 | 1.6 (1.2-2.1) | 340 | 1.7 (1.5-1.8) | 1102 | |
| No inpatient visit for injury | 1 [Reference] | 208 | 1 [Reference] | 640 | |
| Statistics | χ24 = 41.5; P < .001 | χ24 = 899.0; P < .001 | |||
| V. Any previous combat injury | χ21 = 3.5; P = .06 | ||||
| Yes | 1.6 (1.0-2.4) | 337 | 1.0 (0.8-1.2) | 647 | |
| No | 1 [Reference] | 210 | 1 [Reference] | 689 | |
| Statistics | χ21 = 4.0; P = .047 | χ21 = 0.1; P = .72 | |||
Abbreviations: OR, odds ratio; SRE, standardized risk estimate.
The sample of Regular Army enlisted soldiers (9650 cases [153 528 control person-months]) is a subset of the total Regular Army sample (193 617 person-months) from the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) Historical Administrative Data Study. Control person-months were assigned a weight of 200 to adjust for undersampling.
Each predictor (I-V) was examined in a separate logistic regression model that adjusted for sociodemographics (sex, current age, race/ethnicity, education, and marital status) and service-related variables, including age at Army entry, time in service (1, 2, 3-4, 5-10, or >10 years), deployment status (never, currently, or previously deployed), delayed promotion, demotion, and military occupation (combat arms, special forces, combat medic, or other). Models also included a dummy predictor variable for calendar month and year to control for secular trends.
Each 2-way interaction was examined separately in a model that included all of the sociodemographic and service-related variables but not the other 2-way interactions.
The SREs (number of soldiers with a suicide attempt per 100 000 person-years) were calculated assuming that other predictors were at their samplewide means.
These previous injuries were not suicide related according to information available in the administrative records.