Discharge from military service because of behavior (for example, problems with motivation, misconduct, discreditable occurrences, or a series of minor discipline problems), criminal conviction, alcohol or drug abuse, or medical disability were examined by TBI severity as classified with the Abbreviated Injury Scale (AIS) head score. Compared with the total discharge population, discharge due to alcohol or drug abuse was more frequent in those with moderate TBI (odds ratio [OR], 5.4; 95% confidence interval [CI], 1.7–16.9) and those with mild TBI (OR, 2.6; 95% CI, 1.6–4.3) but not in those with severe TBI. Discharge due to behavior was no different in those with moderate or severe TBI and 1.8 times greater in those with mild TBI (95% CI, 1.4–2.2). Discharge due to criminal conviction was 2.7 times higher in those with mild TBI (95% CI, 1.9–3.9) and no different in those with moderate or severe TBI. Discharge due to medical disability was 7.5 times higher in those with mild TBI (95% CI, 6.0–9.3), 25.2 times higher in those with moderate TBI (95% CI, 16.2–39.2), and 40.4 times higher in those with severe TBI (95% CI, 30.0–54.4). The authors note, however, that because the risk of medical discharge is directly related to the severity of the injury, these individuals may be receiving medical discharges rather than other types of discharges (such as behavioral). A limitation of the study is that it did not take into account pre-existing factors, such as aggressive tendencies or preinjury alcohol abuse, which may have played a role in discharge outcome.
In a related study, Ommaya (1996) examined 1,617 patients admitted to hospitals for TBI in 1992 and 1993, 4,626 patients admitted for orthopedic or internal injuries, and a random sample of 9,997 active-duty Army subjects to compare rates of discharge from military service based on behavioral criteria. After adjustment for confounders (age, sex, marital status, educational level, pay grade, months in current grade, years of active-duty service, injury severity, and preinjury “adverse action,” disciplinary action recorded in a soldier’s personnel file), head injury was related to an increased risk of behavioral separation (relative rate [RR], 4.01; 95% CI, 3.54–4.94) and criminal conviction (RR, 4.99; 95% CI, 3.62–6.87) compared with the random sample of active-duty Army personnel. Head injury also was related to an increased risk of postinjury adverse action (RR, 1.31; 95% CI, 1.14–1.51). In addition, the risk of medical discharge was lower in the head-injured group than in the orthopedic- or internal-injury group (RR, 0.64; 95% CI, 0.51–0.80).
McLeod et al. (2004) examined employment retention in the British Army in a group of 564 British Army personnel who had sustained a TBI in 1994, a group of 368 British Army personnel who had a lower-limb fracture in 1994 (and did not sustain any other injuries), and a group of 25,575 healthy army personnel. All those with TBI were admitted to the hospital or medical center and were selected if they had International Classification of Diseases (ICD) codes indicating TBI and did not have other ICD injury codes. Employment retention in the Army was examined with Kaplan-Meier survival analysis, stratifying for age (16–24, 25–28, 29–33, and >34 years), which roughly paralleled career steps in the Army. The results indicated that in the youngest group (16–24 years old), healthy subjects left the Army earlier than subjects in either injury group: in a median of 1.74 years, compared with 3.91 years for those with TBI and 4.39 years for those with lower-limb fractures. An opposite pattern was observed in the oldest group (34 years old and older): healthy subjects served the longest: a median of 5.55 years, compared with 3.33 years for those with TBI and 3.75 years for those with lower-limb fractures. Subjects 34 years old or older had the lowest employment retention: 69% of them in the TBI group continued in the Army beyond year 1, and 19% were still employed at year 6, compared with 85% of the fracture group employed at year 1 and 26% at year 6 and 80% of the healthy group at year 1 and 48% at year 6 (p < 0.001). The authors discuss the possibility that the greater drop in