Lit Review

Zach Stednick -1-

With increasing age of Vietnam era veterans and the current conflicts in Afghanistan and Iraq, the amount of veterans depending on medical care from the Department of Veterans Affairs shows little sign of diminishing. With improving research, studies are beginning to show that psychiatric injuries are just as common as physical injuries. Many veterans of the US Armed Forces are returning home with little to no Post Traumatic Stress Disorder (PTSD) counseling and this is causing many of them to have difficulty adjusting to life in the United States. Much of this PTSD can be traced to mild to severe head injuries that occurred during service and may or may not have been examined within a reasonable amount of time. The difficulty in adjusting can set them up for further social problems, which may include homelessness or a brief transient lifestyle. Rosenheck and Frisman (1994) state that 38% of homeless men are veterans compared to only 30% in the general population. Because of the numerous government benefits that veterans receive when they return home, including housing and medical care, it would seem likely that the rate of homeless veterans would be much lower than the baseline average in the United States. The 1994 Rosenheck paper reviewed the 1987 Current Population Survey that is conducted by the Bureau of the Census and used to estimate the proportion of male veterans in the general population by age-race cohorts. The survey reported that of white males in the 20-34 age group, veterans were 4.8 times more likely to be homeless than their non-veteran counterparts in the general population (Rosenheck 1994). The study also showed that both Caucasian and African Americans had a significantly higher propensity towards substance abuse and that Caucasians in particular had a 2-3 times higher chance of developing psychiatric disorders (Rosenheck 1994).

Zach Stednick -2The combined effect of both substance abuse and psychiatric disorders are likely the driving forces behind veteran homelessness. Research has indicated that veterans with psychiatric disorders, specifically PTSD, are directly related to brain injuries. Doll and Bowley (2008) report that until recently there has been little research done into blunt force injuries that commonly lead to brain injury and eventual discharge. Recent improvements in care has made it more likely that the subject will be less likely to die from an attack and more likely to survive an attack by a blunt force object or an improvised explosive device. Doll and Bowley state that up to two thirds of veterans returning from Afghanistan and Iraq and treated at Walter Reed Army Medical Center were diagnosed with traumatic brain injury (Doll 2008). From this same veteran population, up to a third of all Iraq and Afghanistan refugees report stress or mental disorder and seek some sort of treatment from the Veteran's Administration (Doll 2008). While there are many veterans who do not need care, the Veteran's Administration (VA) has repeatedly stated that it is often difficult to reach all those on the periphery. Doll and Bowley conclude that the contribution of intense traumatic events and exposure to blast injury and traumatic brain injury is difficult to fully understand. It does seem clear that veterans with brain injuries have an increased likelihood to be vulnerable to social problems such as homelessness. Vasterling et al performed a prospective analysis to evaluate PTSD on health-related functioning of 800 Army soldiers before and after a one-year military deployment to Iraq. Both surveys were done with an interviewer and the same interviewer was used to help control for introduced bias. The results showed that post deployment PTSD was only marginally related to day-to-day health related functioning controlling for other health symptoms (Vasterling 2008). Instead, PTSD symptoms seem to adversely impact physical health functioning via their negative

Zach Stednick -3effect on health symptoms, which in turn negatively influence day-to-day functioning (Vasterling 2008). Finally, this study was uniquely able to predict future outcomes because of the nature of the study. Based on the evidence, the authors recommended attempting to deal with PTSD earlier in military service to increase productivity and to lower the overall costs which may be associated with future treatment for PTSD (Vasterling 2008). The results found by Doll and Bowley were further supported by Hoge et al (2008) who conducted a large cross-sectional study on 2714 soldiers from two different military units who had similar times of service and saw similar levels of activity. Soldiers with mild traumatic brain injury, primarily those who had loss of consciousness, were significantly more likely to report poor general health, missed workdays, medical visits, and a high number of somatic and post concussive symptoms than were soldiers with other injuries (Hoge 2008). These soldiers were interviewed 3-4 months after returning from a yearlong service duty in Iraq (Hoge 2008). The 3-4 month period was chosen to minimize recall bias and to allow symptoms of post-discharge conditions to develop further. The data was collected via voluntary survey and the survey was generally administered at a meeting or other military function where many members were present. Of the 2525 soldiers whose data was analyzed, 124 (4.9%) reported injuries with loss of consciousness, 260 (10.3%) reported injuries with altered mental status, and 435 (17.2%) reported other injuries during deployment (Hoge 20008). Of those reporting loss of consciousness, 43.9% met criteria for PTSD, as compared with 27.3% of those reporting altered mental status, 16.2% with other injuries, and 9.1% with no injury. This sample chose a non-random study group consisting of two military brigades and approximately 95% of the sample was male. Together both of these make the study not very generalizable to the general population but will certainly make the study more generalizable to

Zach Stednick -4the veteran population. In this study as well as in others, the results are only as valid as the administration of the test and the wording of the questions on the questionnaire that was given to the military personnel. Because this study was reported based on self-reported data that was collected 3-4 months after discharge and not the event itself, there is a significant possibility that recall bias was introduced. However, the results still stand and make a strong case that head injuries and more specifically PTSD need more attention at all levels of medical care. Resnick et al took this issue further by using multivariate modeling to examine the relationship between PTSD, other aspects of military service, and employment among 5,862 veterans in a national VA vocational rehabilitation program. A Veteran’s Health Association (VHA) program called Compensated Work Therapy (CWT) aims to provide diverse vocational approaches and strong collaborative methods with the ultimate goal being vocational rehabilitation for those enrolled in the program (Resnick 2008). In one component of this program, employers work with the CWT program to find veterans to perform the various tasks and jobs. Those enrolled in one of the 122 CWT programs nationwide were eligible to be subjects in the research study. The study was able to analyze a unique subset of all discharged veterans because it focused solely on those who had been discharged from the service and were now trying to re-enter society with some assistance from the Department of Veteran’s Affairs. The analysis observed that those who were competitively employed were more likely to be younger, earned less public support income, and had fewer medical conditions (Resnick 2008). The rate of employment at discharge from the program was 30% for veterans with PTSD and 36% for those without PTSD (Resnick 2008). It was observed that veterans with PTSD were 19% less likely to be employed at discharge (odds ratio = 0.81, p = 0.02) after controlling for potentially confounding variables (Resnick 2008). While this may be significant, it can still be

Zach Stednick -5observed that both groups have very low rates of employment in general. Even with government help, PTSD remains a potentially significant obstacle to employment. This issue was further examined by Rosenheck who compared the effects of race on patient outcome for the CWT program. Rosenheck performed a similar study to the one done by Resnick but instead made sure that out of 972 veterans surveyed, 470 of them were African American (Rosenheck 1998). For both Caucasians and African Americans, both had very similar rates for almost all categories including percent that successfully completed the program (49 vs. 52.1%), average hours worked per week (33.5 vs. 31.8 hours), and average length of stay in the program (188 vs. 199.7 days) (Rosenheck 1998). Although no statistically significant trends were observed, it was noted that there were clear preferences for both to remain with people of their own race and especially important to be seen by a doctor of their own race. These are interesting observations, but they likely do not translate to any observation of interest in the general population. Because the VA system is a government-operated system that is often the only choice for many of its patients, it seems to be somewhat immune to the issue of bias towards one particular race. Because of the nature of the relationship between the military and the US Government, there are few programs that can effectively bridge the gap outside of the military. One governmental program that has experienced success is the CWT program, which was analyzed by Resnick. The CWT program has many sub-programs such as Transitional Work Experience which aims to use research and previous evidence-based services to make it easier for veterans to return to employment in the civilian sector. The current programs are being expanded both to make room for more veterans and because of past success with evidence-based Supported

Zach Stednick -6Employment (SE) programs (Resnick 2008). It is hoped that these programs will become increasingly more integral in a “recovery-oriented” service system. Even after finding a job it can still be difficult for veterans as well as the general population to keep stable housing and keep that job. O’Connell et al studied this and determined that an individual will be much more successful if they are actively followed and mentored. O’Connell performed a secondary analysis of data from the Housing and Urban Development – Veterans Affairs Supported Housing (HUD-VASH) and attempted to observe what predictors were present to determine if a veteran would return to being homeless (O’Connell 2008). Of 392 formerly homeless veterans enrolled at the program at baseline, approximately 44% lived homeless for at least one day after successfully being placed in housing (O’Connell 2008). There were many reasons for this, but it was observed that PTSD resulted in an 85% higher risk of reduced housing tenure (O’Connell 2008). Although the study was well designed to be randomized and cover a wide geographic area, there are still some issues with generalizability. Most the most significant issue is that the study only loosely defines homelessness as one day of being homeless after being placed into housing. Also, the study falls into the area of bias that seems to affect many other studies in this area that is the significantly disproportionate population of males in the sample study. A similar study of client services measurement was conducted by Mares to determine client-level measures of service integration. Mares used data collected from different surveys of service delivery and interagency trust and respect by 734 chronically homeless adults in 11 cities across the country (Mares 2007). Analysis of the data showed that veterans were more likely to visit outpatient clinics, have serious drug problems, have diagnoses of PTSD, but to have less serious alcohol problems (Mares 2007). Although this agrees with previous studies, Mares

Zach Stednick -7argues that although there are positive associations between client-level measures of integration and health status, the negative relationships that exist may represent greater frustrations among homeless clients (Mares 2007). The study by Blue-Howells attempted to study the issue of program fragmentation by analyzing recently implemented programs. The study followed clients from the greater Los Angeles area and attempted to gauge their response to a new program which integrated patient care by co-locating programs and services in one building (Blue-Howells 2007). With the integrated services, veterans were easier to care for because they could have more medical needs met at one place. It is hoped that the VA system integrates the model developed in LA into their overall plan for other VA facilities across the country. Although PTSD as a condition in veterans has been well studied, few have studied the types of treatment used such as Mohamed and Rosenheck. Mohamed used data from 274,297 veterans whom were all diagnosed with PTSD in the year 2004. After multivariate analysis, it was observed that 80% of veterans were prescribed psychotropic medication that likely seems to be targeted at specific symptoms such as flashbacks and insomnia (Mohamed 2008). Mohamed also suggests that further research may be needed to better diagnose and treat PTSD instead of the current method, which aims to treat the peripheral conditions in the hope that it will treat the main problem. All studies were subject to some sort of bias, especially recall bias occurring because the subjects may have had difficulty recalling all the events required by the survey. Many of these surveys were administered a certain amount of time after their discharge date. Many injuries were likely sustained during the time of service and using the discharge date is often the only possible date that researchers can use. Another type of bias present in many of these studies is

Zach Stednick -8information bias, which is likely common in a diagnosis as wide-ranging as PTSD. The other major deficiency with many of these studies is the lack of generalizability to the non-military world. Because most of the study populations are more than 90% male, it makes the results of these studies slightly more difficult to apply to the entire population. However, because this is typical of the military population, the results only need to be specifically applied to the military. All of the studies were fairly similar in design with the exception of a few such as the Rosenheck study. The 1994 Rosenheck paper is more of an incidence survey than a prevalence survey and no military records were examined to confirm the individual was present in the military. This could potentially be biasing the actual results in terms of length of that individual being homeless, however it is effective in generating a snapshot of trends at that particular time. All of these papers demonstrate that vocational deficits exist for those veterans who return home with PTSD. Although almost all of the veterans seek some sort of care or treatment from the Veterans Affairs, there is much room for improvement within these programs. Hopefully the Department of Veterans Affairs will take steps to improve their re-establishment programs to make them more reflective of current research. Because homelessness is a multifaceted problem related to PTSD and lack of vocational training, both problems must be addressed to begin to observe progress on the issue. With earlier intervention and better diagnoses of PTSD, it is hoped that re-establishment programs will be better able to deal with these issues and the programs will be better positioned to improve the lives of veterans.

Zach Stednick -9References Blue-Howells J, McGuire J, Nakashima J. Co-location of health care services for homeless veterans: a case study of innovation in program implementation. Soc Work Health Care. 2008;47(3):219-31. Doll, D., Bowley, D. Veterans' health--surviving acute injuries is not enough, The Lancet. 2008. Volume 371, Issue 9618. Page: 1053-1055. Hoge, Charles W., McGurk, Dennis, Thomas, Jeffrey L., Cox, Anthony L., Engel, Charles C., Castro, Carl A. Mild Traumatic Brain Injury in U.S. Soldiers Returning from Iraq. N Engl J Med 2008 358: 453-463 Mares AS, Greenberg GA, Rosenheck RA. Client-level measures of services integration among chronically homeless adults. Community Ment Health J. 2008 Oct;44(5):367-76. Epub 2008 May 1. Mohamed, S; Rosenheck, R. Pharmacotherapy of PTSD in the US Department of Veterans Affairs: Diagnostic and Symptom-Guided Drug Selection. J Clinical Psychiatry June 2008 69:6. O'Connell MJ, Kasprow W, Rosenheck RA. Rates and risk factors for homelessness after successful housing in a sample of formerly homeless veterans. Psychiatr Serv. 2008 Mar;59(3):268-75. Resnick, Sandra G., Rosenheck, Robert A. Posttraumatic Stress Disorder and Employment in Veterans participating in Veterans Health Administration Compensated Work Therapy. Journal of Rehabilitation Research and Development. 2008. Volume: 45. Number: 3. Page: 427-436 Rosenheck, R, Frisman, L, Chung, A M. The proportion of veterans among homeless men. Am J Public Health 1994 84: 466-469 Rosenheck, Robert, Seibyl, Catherine Leda Participation and Outcome in a Residential Treatment and Work Therapy Program for Addictive Disorders: The Effects of Race. Am J Psychiatry 1998 155: 1029-1034 Vasterling JJ, Schumm J, Proctor SP, Gentry E, King DW, King LA. Posttraumatic stress disorder and health functioning in a non-treatment-seeking sample of Iraq war veterans: a prospective analysis. J Rehabil Res Dev. 2008;45(3):347-58.