Prevalence of Traumatic Brain Injury Among Prisoners in South Carolina

Abstract

Objective: 

To determine lifetime prevalence of traumatic brain injury (TBI) in a statewide sample of prisoners.

Design: 

Retrospective and cross-sectional cohort study.

Participants: 

Stratified random sample of prisoners scheduled to be released by release type for each gender (275 men and 267 women with completed sentences and 19 men and 15 women granted parole) and a random sample of prisoners by gender with lifetime or death sentences (26 men and 34 women).

Primary Measure: 

Weighted prevalence of TBI. In-person interviews ascertained history of TBI using a customized version of the Ohio State University TBI Identification Method.

Results: 

Sixty-five percent of male releases and nonreleases, and 72% and 73% of female releases and nonreleases, reported at least 1 TBI with an alteration of consciousness. Forty-two percent of male releases and 50% of nonreleases, and 50% of female releases and 33% of nonreleases, reported at least 1 TBI with loss of consciousness. Thirty-five percent of male releases and 42% of nonreleases, and 55% of female releases and 58% of nonreleases, reported ongoing symptoms from TBI.

Conclusions: 

A substantial proportion of prisoners reported having experienced a TBI during their lifetime. The ongoing TBI symptoms reported by prisoners may impact their success on release.

TRAUMATIC BRAIN INJURY (TBI) continues to be a major public health concern in the United States.¹ Although some survivors of TBI fully recover, others experience a multitude of cognitive, emotional, physical, and behavioral disabilities. Personality changes are common, and after experiencing TBI, a person may become irritable or aggressive, exhibited by explosive outbursts that can occur without warning or be initiated by minimal provocation.² These events may occur more often as the person ages.^3,4 Without successful early intervention, these persons may pose a threat to others and themselves that results in their becoming involved with the criminal justice system,^5–11 even at an early age.^12–17 Once incarcerated, the behavioral and cognitive sequelae associated with TBI can make adjustment to prison life more difficult, as demonstrated by increased disciplinary incidents in some limited studies.^18–20 Furthermore, TBI-associated co-occurring adverse health conditions such as chemical dependency^21,22 and psychiatric problems can greatly affect an offender's ability to function both while incarcerated and on return to the community.²³

The reported TBI prevalence among convenience samples of mainly male offender populations range from 10% to 100%,^19,24–45 with 2 recent meta-analyses suggesting an overall TBI prevalence rate of 51% to 60%.^46,47 Yet, the majority of these study findings are not generalizable due to differing definitions and determination of head injury and TBI, varied source populations, small proportions of women in the samples, and exclusions. Few studies of offenders with TBI have been population based,^19,35 and none have included a randomized statewide prison population of both men and women. A clearer understanding of the burden of TBI in offender populations could inform public efforts and lead to TBI-specific prison rehabilitation programs.

Although a high prevalence of TBI is hypothesized among offender populations, not all TBIs result in persistent effects or long-term disability. The prevalence of ongoing sequelae resulting from these injuries needs to be established. Accurate estimates of the prevalence of offenders with ongoing consequences from TBI may encourage screening and effective treatment to become standard of care in criminal justice systems nationwide. Implementation of behavioral and cognitive strategies to inhibit aggressive and other behavioral disturbances⁴⁸ could help prevent injuries to the offender, other inmates, and corrections management. An additional potential benefit is that better prison identification may allow better planning for community reentry so that offenders with TBI would be less likely to become homeless⁴⁹ or to recidivate.²³ To our knowledge, this manuscript is the first to report TBI prevalence rates and ongoing TBI symptoms among a stratified random sample of male and female offenders in a statewide prison population.

METHODS

Participants and sampling

The South Carolina Department of Corrections (SCDC) and the institutional review boards of the Medical University of South Carolina, RTI International (the Research Triangle Institute, an independent, nonprofit research and development organization), and the Centers for Disease Control and Prevention (CDC) approved the study. The Office of Health Research Protection verified that it complied with federal regulations (CFR §45 46.306(a)(i)) regarding prisoner research. The Office of Scientific Regulatory Services of the CDC issued the study a Confidentiality Certificate. The SCDC has 27 male facilities and 3 female facilities, and it is the sole provider of state-secured adult correctional confinement for offenders who have sentences of 91 days or more. The SCDC administrative records provided initial sampling data, including demographics, intake assessment, and administrative information. The completion rate of most variables was 97% or greater unless a variable was recently introduced by the SCDC. The primary aim of the larger study from which the data are used was to assess the effect of TBI on recidivism, with a secondary goal of determining TBI prevalence. Although the analyses presented here address the secondary goal, the sampling strategy reflects the primary aim and thus the majority of the sampling focused on inmates just prior to release.

The SCDC provided an initial list of inmates with life or death sentences and provided lists of inmates scheduled for release or a parole hearing every 2 months until the goals of interview were met. Three strata were sampled separately by men and women: (1) max outs, who were inmates within 3 to 4 months of completing their prison time; (2) inmates just granted parole release; and (3) long-term inmates who would not soon be released (sentenced to death or life) (see Method Note 1, Supplemental Digital Content, available at https://links.lww.com/JHTR/A61 for details regarding sampling strategy). The goal was to interview 293 inmates with imminent release and 25 nonrelease inmates of each gender, for a total of 636 interviews over the course of approximately 1 year. Because release type and gender may affect recidivism, the release sample was a proportionately stratified sample by type of release (ie, max out and parolee) for each gender. We attempted to obtain 6% of male and 9% of female imminent releases as parolees, representing the most recent SCDC release statistics.

Presampling study exclusions included arrests outside of South Carolina, younger than 18 years at time of interview, difficulty comprehending informed consent (eg, due to intellectual disability or mental illness requiring hospitalization), housing in special management units, detainers (inmate to be released to law enforcement officials), release to a mental health facility, non–English speaking, presence in a previously sampled list from the SCDC, or previously released offenders who participated in the pretesting of the questionnaire. Postsampling, inmates were excluded if a corrections officer thought an inmate posed a danger to an interviewer. Because of the small size of the female inmate and parolee source population, all female inmates and all male and female parolees who met the selection criteria during the study period were approached. Male inmates who either completed their sentences or were not going to be released were randomly sampled. Additional description of the sample selection methodology can be found in Method Note 1, Supplemental Digital Content, available at https://links.lww.com/JHTR/A61.

Interviews were conducted from April 2009 through April 2010. The total numbers of male inmates eligible for the study during the 10 months of male recruitment were 3177 max outs, 53 parolees, and 775 nonreleases. The total numbers of female inmates eligible for the study during the 12 months of female recruitment were 461 max outs, 32 parolees, and 25 life without parole and 20 life with parole with 20 or more years (not counted for these analyses were 18 with non–life sentences of more than 30 years). Included in the max out population were 381 men and 77 women who were effectively excluded because they were initially in one of the SCDC lists as due for a parole hearing, did not receive parole, and then appeared in a later SCDC list as max outs (and thus ineligible due to being in a previously sampled list). Forty-one percent of inmates were not eligible for interview, mainly due to detainers or presence in a previous list from the SCDC. Six percent of male and 5% of female releases interviewed were due to be released as parolees. Inmates interviewed were 275 male max outs, 19 male parolees, 26 male nonreleases, 267 female max outs, 15 female parolees, and 33 female long term (plus the 1 women with more than 30-year sentence), for a total of 636 interviews. Strata were assigned on the basis of initial assignment of release by the SCDC. Ultimately, 7 inmates were released in a manner different from the initial assignment and 1 inmate had not been released at time of analysis. The numbers of inmates from each sampling strata sent to RTI for possible interview and their outcome are shown in Figure 1. The majority in each stratum was interviewed, yielding an overall 72% response rate.

Figure 1: Sample composition. Admin indicates inmates unable to be interviewed because of administrative reasons (the majority due to already having been released). Two additional inmates were excluded postsampling: one did not speak English and the other had been transferred out of the South Carolina Department of Corrections facility. Included among refusals are 3 inmates who began the interview but stopped shortly after beginning.

Comparison of inmates interviewed with those eligible but not interviewed was conducted within each of the 3 sampled strata by gender (male max out, parole, and nonrelease, and female max out, parole, and long term). Characteristics compared were age, race, sentence length (releases), violent conviction, chemical dependency status at intake, mental health status, level of security of facility, sexual predator status (nonreleases), and reading level. Only 2 significant differences were found. Female max outs interviewed were significantly younger (35.3 years) than those not interviewed (37.6 years), and female parolees interviewed were significantly less likely to have been chemically dependent on intake (33%) compared with those not interviewed (76%). Female max outs were not assigned a poststratification weight on the basis of age because the age difference was small, and there was no association between age and TBI presence. Female parolees were assigned a poststratification weight on the basis of chemical dependency. The characteristics of those interviewed are shown in Table 1 by strata and gender. Chemical dependence was high in all groups. Mental illness was far more prevalent among women than men. Reading levels were also higher on average in women than in men.

TABLE 1: Demographics of interviewed sample (N = 636)^a

Interviewing

RTI International conducted structured face-to-face interviews with the inmates in private areas within the prisons. Nine interviewers, the majority with experience interviewing in SCDC for another project, were trained for this study. Time to explain the study, obtain informed consent, administer a consent screener, and conduct an interview ranged from 60 to 120 minutes. The questionnaire included a comprehensive array of standardized and nonstandardized measures and a tool to identify TBI. The questionnaire was pretested with 6 male and 4 female parolees.

TBI identification

The Ohio State University TBI Identification Method (OSU-TBI-ID),²⁹ customized for this study, was used to obtain a history of TBI from each participant during the structured in-person interview. The OSU-TBI-ID has been shown to have good interrater and test-reretest reliability and predictive validity for use with prisoners,²⁹ and is explained in more detail elsewhere.⁵⁰ Briefly, the tool is designed to obtain self-report of TBI occurring over the individual's lifetime by eliciting recall of all injuries requiring medical attention, whether medical attention was received or not. Progressively, more distant time frames cue recall of injuries. The elicitation method subsequently concentrates on injuries involving the head or the neck with potentially high-velocity forces capable of causing shear injury to the brain. The term “injury” was chosen to avoid ambiguity and dependence on individual definitions of “head injury,” “traumatic brain injury,” or “concussion.” Etiology of injury and age at occurrence were also recorded, as well as episodes of multiple mild injuries to the head such as those that might occur in contact sports or domestic violence. Those episodes of multiple mild TBI were recorded as one TBI.

The interviewer determined whether loss of consciousness (LOC) occurred (“Were you knocked out or did you lose consciousness from this injury?”), and its nature, to distinguish LOC secondary to TBI from that arising from causes such as substance misuse. Each incident was queried for length of LOC. If no LOC was reported, other alterations of consciousness were explored (“Did the injury cause you to become dazed or confused?” “Did you forget what happened before or after?”). In addition, interviewers asked what level of medical attention was received for each injury and clarified that they were interested in both formal and informal medical attention and in injuries that “should have” received medical attention but did not. Up to 12 possible TBI-related injuries were recorded with the last 2 being restricted to what the respondent felt were the worst ones (ie, with longest LOC) of any remaining. In a final step, each injury with any type of altered consciousness was probed for the onset and persistence of 19 frequently cited sequelae of TBI, an open-ended “other” symptoms question, and whether they were told by others that they acted differently or had more trouble in school after the injury. See Method Note 2, Supplemental Digital Content, available at https://links.lww.com/JHTR/A62 for additional details regarding coding of TBI.

Statistical analysis

The SAS version 9.2 was used for analyses (SAS Institute, Inc, Cary, North Carolina). Demographic frequencies were calculated for the sample. Comparisons between those interviewed and the eligible populations by strata and gender were analyzed by the χ² test, the Fisher exact test, the Wilcoxon signed rank test, or the Kruskal-Wallis test, as appropriate, for selected variables available from the SCDC's administrative data. These comparisons included race, chemical dependency, violent crime status, mental health status, sexual predator status (nonrelease/long term only), security level, age, sentence length (release only), and reading level. Sampling weights were calculated for male and female releases (max outs and paroles) and nonreleases (ie, long term for females), and poststratification weights were calculated as needed on the basis of the SCDC variables listed previously.

TBI prevalence was determined for release and nonrelease groups by gender. In addition, prevalence was determined for a number of TBI subgroups, on the basis of the extent of alteration of consciousness, age at TBI, and presence of current, ongoing symptoms resulting from a TBI. Symptoms reported when asked about “other” sequelae were categorized. Total number of ongoing symptoms reported was summed, counting each type of symptom once. Responses to whether the inmate was told they acted differently or had more trouble in school after a TBI were analyzed separately from the rest of the more subjective, “first-hand” sequelae. All prevalence calculations were weighted to each sampling strata and included poststratification weights where necessary. In addition, percentages endorsing various ongoing symptoms by gender were calculated.

RESULTS

Four hundred thirty-one of the inmates reported experiencing a total of 950 TBIs. Information on TBI was, for the most part, quite complete. One inmate reported a TBI for which no age at occurrence was given, and 18 did not respond to any questions concerning ongoing symptoms for at least 1 TBI. Only 5 of the latter did not answer questions on ongoing symptoms for any TBI. There were 450 incidents of TBI with LOC reported, 81 (18%) of which had an unknown length of LOC. TBI prevalence by gender and release status is shown in Table 2. Approximately 65% of men, whether to be released or not, reported having experienced a TBI, whereas approximately 72% of women reported a TBI. Of male releases, 42% reported a TBI with LOC, compared with 50% of males with life or death sentences. Although 50% of female releases reported TBI with LOC, only 33% of women with life sentences reported similarly. From one-third to one-half of inmates reported experiencing more than 1 TBI. Of men, to be released and not, 15% and 23%, respectively, and 20% and 12% of women reported more than 1 TBI with LOC. Almost one-third of men reported having had a TBI prior to age 15 years, compared with approximately 25% of to-be-released women and 42% of women with life sentences. Almost twice as many to-be-released men reported a TBI with LOC prior to age 15 than to-be-released women. However, more women reported TBI with ongoing symptoms than men, with a significant difference between to-be-released men and women. Average age at first TBI, with or without LOC, was younger for to-be-released men than women, but this was opposite for those not being released.

TABLE 2: Prevalence of TBI and mean summary indices

Inmates reported a lifetime history of 0 to 11 TBIs. Table 2 shows the average number of TBIs reported by gender and release status. Inmates with TBI reported 2.1 to 2.7 TBIs, on average, and number of TBIs with LOC averaged from 1.5 to 2.2. Inmates who reported early TBI had experienced an average of 1.2 to 1.6 TBIs before the age of 15 years. Of the 421 inmates reporting TBI and answering questions regarding ongoing symptoms, 281 reported 1 to 19 symptoms. Table 2 also shows that to-be-released women reported a higher mean number of ongoing symptoms than to-be-released men, but the opposite was true for those not being released, although there were no statistically significant differences. Figure 2 shows average number of symptoms reported by inmates' highest TBI severity reported. An increasing average number of persistent symptoms were reported as TBI severity increased.

Figure 2: Average number of ongoing symptoms of TBI reported by severity of TBI. Conc indicates consciousness; LOC, loss of consciousness.

The most common persistent symptom reported by men and women with TBI was headache (see Fig. 3). Many of the symptoms showed similar percentages of endorsement by men and women. Eight of the top 10 symptoms endorsed by each gender were the same. The only differences were women endorsing trouble sleeping and dizziness or balance problems and men endorsing vision problems and reacting slowly or feeling foggy. In response to the question “After the injury, did your friends or family members say you acted differently or did you have more trouble in school?” among inmates with TBI, 29% of male releases, 41% of male nonreleases, 47% of female releases, and 25% of female nonreleases responded ‘yes'. Of these, 153 inmates provided details about how they were told they acted differently. Responses were sorted into 26 categories. The most common responses were being told that their behavior was different (35%), usually in a negative way. Twenty-six percent reported being told they became withdrawn, 16% showed mood changes, 11% became fearful, 10% were irritable, 9% developed mental health problems, 6.5% became aggressive or violent, and 6.5% developed problems with their temper.

Figure 3: Most common symptoms among prisoners reporting TBI with ongoing symptoms.

DISCUSSION

This study estimated prevalence of TBI in a statewide representative sample of inmates about to be released and inmates without forthcoming releases, by gender. The prevalence of TBI among male and female inmates was 65% and 72%, respectively. Other studies have reported from 10% to 100% prevalence in offenders.^19,24–45 It is difficult to compare prevalence rates between studies of offenders due to varying definitions of TBI, ways of collecting information, and differing offender populations studied. A random effects meta-analyses of adult offenders by various subgroups reported a strikingly similar 64% prevalence of TBI in all male offender populations, 70% in all female offenders.⁴⁶ When examining the TBI subgroup who had LOC, the meta-analysis prevalence results were slightly higher than in this study, reporting prevalence of TBI with LOC in all male offenders of 52%; TBI with LOC in only nonspecific male offenders of 59%, and TBI in female nonspecific offenders of 55%.⁴⁶ A subsequent meta-analysis calculated prevalence among adults and youth in the criminal system (convicted and not) in developed countries.⁴⁷ They reported a random-effects overall TBI prevalence of 51%. This lower rate is likely due to their inclusion of youth and nonconvict studies with low TBI prevalence. A recent study of lifetime TBI in a prison population not included in the meta-analyses discussed earlier examined male prisoners, aged 18 to 54 years, in the United Kingdom.⁴⁵ The results of that study revealed that 61% responded that they had “... had a head injury or been concussed (knocked out) for a period of time,” quite similar to this study's male prevalence, despite the different methodology. The study closest to our methodology was conducted by Bogner and Corrigan²⁹ with equal numbers of male and female prisoners in Ohio (105 each). Using the OSU-TBI-ID, they found that 85% of men and 71% of women had experienced a TBI in their lifetime and 64% of men and 51% of women had had a LOC with a TBI. The prevalence for women is very close to our study, especially for the released women (the more stable number since the sample was larger than for women not being released). However, the male prevalence is higher than our results for reasons that are not apparent, perhaps due to different demographics.

Because this was a study of lifetime history of TBI, the total number of TBIs, and when they occurred, was also determined. Among the 4 inmate groups, proportions reporting more than 1 TBI ranged from 35% to 48%. Other studies have reported prevalence of multiple TBI ranging from 3% in offenders (no gender specified) in a Midwestern US prison system¹⁹ to 61% among male prisoners in Australia.²⁴ Other study estimates include 10% among male forensic psychiatric patients,³⁷ 17% in a small sample of mostly male nonincarcerated offenders,³⁹ 23% among mostly male forensic patients in Canada,⁵¹ 31% among incarcerated male substance abusers,⁴⁴ 44% among male death row inmates,³² and 57% among male New Zealand prisoners,²⁵ male prisoners in Spain,¹² and male and female prisoners in Ohio using the OSU-TBI-ID.²⁹ Except for the latter study in Ohio, virtually all reported samples are men, so there is little research against which to compare information on women, although they are at a higher risk of repeat TBI from domestic violence.^29,52–55 Number of TBIs experienced is less often recorded. In this study, inmates with TBI averaged between 2 and 3 TBI. In the Ohio study the average among those reporting TBI was 3.2.²⁹ Evidence of the danger of repeat TBI is growing. A study of male and female inmates at federal prisons found that those with multiple mild TBIs had similar cognitive and behavioral profiles (depression, aggressiveness, impulsivity, violence, and substance abuse) as those with more severe TBI.³¹

Age at TBI, especially during childhood, may also play an important role in whether there are ongoing symptoms from TBI.⁵⁶ In addition, Bogner and Corrigan found that increased severity of TBIs prior to age 15 years was associated with worse memory and increased risk taking.²⁹ One prospective cohort has shown hospitalization for mild TBI prior to age of 5 years is associated with symptoms of attention deficit/hyperactivity disorder, conduct disorder, substance abuse, and mood disorder.⁵⁷ Another cohort that followed children with primarily mild head injury for 23 years found that long-term subjective reports of intellectual or emotional effects from the head injury were associated with a lower IQ measured after the post–acute phase of the childhood TBI.⁵⁸ In addition, reports of any symptoms from head injury was associated with a more severe head injury, grade failure/retention, unemployment, current psychological problems, and strained family relationships. Our study also showed an increase in the average number of ongoing symptoms reported as TBI severity, based on LOC, increased.

Information on ongoing symptoms is often not included in studies of TBI prevalence, but 20% of offenders with TBI (unspecified gender) in the Midwestern prison system,¹⁹ 47% of male inmates with TBI (with LOC) in a California prison,⁴¹ and 52% of male prisoners with TBI in Australia²⁴ had ongoing sequelae, which covers the range of prevalence in our study. In the Australian study of male prisoners, the most common ongoing symptoms among those with TBI were headache (33%), personality change (22%), anxiety/depression (22%), memory loss (21%), and uncontrollable anger (20%).²⁴ The overall percentage of men with TBI affected by symptoms is quite similar for the 2 studies, and 21% of men with TBI in this study also reported ongoing problems with controlling their temper. In addition, headaches and memory problems were among the top 3 complaints among men with TBI in both studies.

The impact of TBI on prison inmates has not received much research. Investigators have, however, examined brain injury as a potential cause for criminal behavior, especially violent crime, as well as a consequence from engaging in criminal behavior.^8,42 Among male offenders, a history of TBI has been reported to be strongly associated with perpetration of domestic and other types of violence.⁴⁸ Likewise, female violent offenders have often experienced a precrime TBI.²⁸ A limited number of studies reporting the relationship of in-prison behavior and brain injury suggests that offenders with TBI have a higher rate of disciplinary incidents and are slower to adapt to prison life and comply with prison rules.^18–20 It could be assumed that these offenders have persisting consequences of TBI that make their compliance with prison rules difficult. Without a clear understanding of how TBI may impact a person, prison officials may be misled to believe that the offender is deliberately defiant.^{18,24,59–61} When this happens, disciplinary action may be taken toward an offender with an “invisible disability” who has limited awareness of and/or control over his actions.¹⁸ Prison staff training concerning the consequences of TBI,^60,61 similar to that being done in some prisons regarding mental illness and epilepsy,^62–64 could be conducted. The education of prison staff in management strategies of inmates with TBI could benefit both the inmates, by potentially decreasing infraction penalties, and staff, by potentially decreasing negative or even dangerous encounters with inmates.

This study shows a high prevalence of TBI, repetitive TBI, early TBI, and TBI with ongoing symptoms, all of which can interfere with how an individual functions from day to day. Ideally, individuals receive rehabilitation services after experiencing a TBI that results in symptoms. However, we postulate that services would be rare among individuals with TBI who later enter the criminal justice system. Therefore, screening and subsequent treatment after initial incarceration could provide an avenue to assist offenders in not returning to prison after release.¹² Traumatic brain injury screening within the prison setting has been tested^29,32 and recommended to help inform improvement of offender safety and corrections facility management,²⁴ as well as a means of informing more effective intervention for conditions such as substance abuse.⁶⁵ To date, however, the relationship between TBI and co-occurring conditions among offenders has not been well-established. In addition, information on how TBI effects treatment for substance abuse, posttraumatic stress disorder, and other mental health conditions (eg, learning disabilities, pain, depression) is needed.^33,66–69 Without effective interventions, these conditions are likely to pose public health and social challenges that continue when the offenders return to the community⁴⁹ and hence predispose them to return to prison.²³

This study had a number of limitations. The prevalence of TBI relied on retrospective self-report and covered a lifetime of memory. The inmates may have faulty memory, especially concerning early TBIs, and varying definitions of “traumatic brain injury” or “head injury,” most likely leading to underestimation of prevalence. Despite this, research on the validity of self-report of TBI in both the general population and in prisoners has shown good reliability in both groups.^{29,50,70–72} Information on the 19 specific symptoms listed in our version of the OSU-TBI-ID was not collected from inmates without TBI, so we are unable to determine whether those with TBI report a significantly higher prevalence of the symptoms in comparison with the rest of the participants. Our cohort was limited to prisoners within South Carolina, and the generalizability of the findings even within South Carolina is limited because of the inaccessibility of Youthful Offender Act prisoners (special program for young offenders younger than 26 years, with indeterminate release dates), the exclusion of prisoners with detainers, and the 28% of the sample overall not interviewed because of refusal or administrative issues. Strengths of this study include using a standardized, in-depth questionnaire concerning TBI rather than 1 or 2 questions, utilizing the standard definition of TBI,^73,74 including an equivalent proportion of female inmates, and using a statewide stratified random sample of inmates.

This study has shown a high prevalence of TBI in a stratified random sample of state inmates, both those being released and those remaining in prison. These results are generally in concordance with those of other studies, where comparable, but this study also provides prevalence for subgroups of TBI and a lifetime history of TBI. Not only is TBI common among inmates, but early TBI, repeat TBI, and ongoing TBI symptoms are also quite prevalent. In addition, although there are fewer women in prison, they report more TBI in general and more TBI with ongoing symptoms, yet far less research attention has been directed toward this group. It has been shown that female prisoners with acquired brain injury present with different cognitive and behavioral impairments than male prisoners so inflicted and thus may require different treatment and management.³⁵ Treatment of TBI in the inmate population, as well as prevention in the next generation, would most likely not only be cost-effective but also beneficial to both individuals with TBI and society.