This systematic review of research studies examining the link between TBI and homelessness identified very few articles on the topic. In comparison, a search of the terms “homeless” AND “substance abuse” conducted at the same time yielded almost 2,000 articles in MEDLINE/Pubmed. We are confident that our review has captured all of the information available to date, as we used comprehensive search strategies and contacted authors of published articles to solicit any additional studies. Care was taken to thoroughly review each identified source and extract pre-determined data elements.
The findings of the few studies identified pointed to high rates of TBI among people who are homeless. Moreover, the two studies that explored the temporal relationship between homelessness and TBI [
7,
25] revealed that for the majority of participants, the first incidence of TBI occurred before the onset of homelessness, suggesting that TBI may be a risk factor for homelessness. Although not widely explored among the research reviewed, it is also possible that another factor, such as impulse control disorders, may predispose individuals to both TBI and homelessness. For example, Hwang and colleagues (2008) found that alcohol and drug problems were significantly more common among those with a TBI in their sample. Similarly, and in line with findings from the general public, the current review also suggests that homeless individuals with a history of TBI have an increased susceptibility to subsequent injuries [
7,
25] as is illustrated by Oddy et al. and Hwang et al.’s findings that among those with a history of TBI, over 60% had experienced more than one. This is noteworthy as previous literature points to the potential for multiple TBIs to have cumulative adverse impacts on cognitive functioning [
28]. Previous literature also demonstrates that cognitive impairment may contribute to the chronicity of homelessness [
12]. When reported, injury severity tended to be mild in the homeless samples studied in the shelter and meal program settings [
7,
25] and moderate to severe in the neurosurgical department setting [
27]. However, this finding was only examined in three studies, and in one of the studies severity of injury was only determined for first injury, not for the most severe injury. It is possible that a participant from this study may have had a mild TBI and subsequently incurred more severe injury. Furthermore, it is possible that those with less severe TBIs do not access emergency rooms as frequently as those with more severe injury, explaining why those examined in the neurosurgical department setting may have been found to have more severe TBIs.
The findings of this review point to sex-related differences in the prevalence of TBI among individuals who are homeless as the majority of studies that included mixed sex samples illustrated a greater likelihood of TBI in males who are homeless [
7,
25]. Although Hux and colleagues (
2009) did not explore sex differences among their homeless sample, [
23], their sample was predominantly female and yielded one of the lowest prevalence rates of TBI as compared with samples of homeless people from the other studies. It is also important to note that Oddy and colleagues (2012) found no differences between males and females in terms of the number of TBIs acquired. However, they reported a greater prevalence of TBI among males (83%) as compared to females (17%). These findings align with the increased rates of TBI amongst males in the general population [
29].
The methodological quality of the studies included in this review, as assessed by the Downs and Black (
1998) checklist, was generally moderate. The most commonly identified limitation of the studies was limited external validity and reporting of power calculations. Specifically, information was frequently missing to enable the reader to determine whether the study sample was representative of the entire source population and whether or not the study was sufficiently powered to detect differences in homeless samples with a history of TBI and without a history of TBI. These findings suggest that there is a need for more representative sampling in future studies. This can be achieved through random stratified sampling and demonstrating statistically that included participants are similar in demographics to the target population. Power calculations should also be included in future studies.
Limitations of the evidence base
The majority of the studies did not primarily aim to address TBI among people who are homeless. As a result, history of TBI was assessed through very basic questions (see Table
1) and the questions were only asked using a validated measurement tool in one of the eight identified studies [
23]. Validated screening tools for TBI such as the Ohio State University Traumatic Brain Injury Identification Method
(OSU TBI-ID) and the Brain Injury Screening Questionnaire (BISQ) should be used in future studies to provide a more reliable estimate of rate of TBI. Ideally, studies could have objective measures of TBI such as advanced neuro-imaging. TBI was not consistently defined across the studies. For example, in some studies [
21,
24] the determination of injury was dependent on whether an individual had lost consciousness. However, it is generally accepted that loss of consciousness is not necessary for a diagnosis of TBI [
9]. Consequently, estimates of the rates of TBI among certain study samples may have been conservative. In future studies TBI should be conceptualized in keeping with standard diagnostic criteria. Studies may also not have been adequately powered to examine questions related to history of TBI. As well, a majority of the studies did not use randomized approaches to recruitment, leaving the possibility of selection biases.
Homeless persons who were not sheltered were only included in four studies [
7,
25‐
27]. Hwang and colleagues (
2008) published one of the largest studies, which recruited homeless persons from both shelters and meal programs. Oddy and colleagues (
2012) recruited from both day centres and hostels, Kim et al. recruited for a neurosurgery department, and Gonzalez and colleagues (
2001) recruited from a health care clinic for homeless people. This variation in sampling is noteworthy, as there may be significant differences in terms of physical and mental health between those who access shelters and those who do not [
30,
31]. Although it may be harder to access homeless persons living on the streets, future studies are necessary in order to examine the occurrence of TBI among unsheltered homeless persons. It may be useful to examine the difference in reported TBI among those who access shelters as compared with homeless individuals who do not in order to determine the risk factors for TBI among people who are homeless. This information may also be valuable in guiding prevention and interventions strategies to the groups of homeless individuals who are most at risk of suffering a TBI.
Few of the reviewed studies provided injury related details such as severity or symptoms related to TBI. None of the studies reported on mechanism of injury. Only three studies provided details on the severity of injury, and age at injury [
7,
25,
27], and only two of these studies reported on the temporal relationship between TBI and onset of homelessness [
7,
25]. As previously mentioned, gender differences were not widely explored in these studies. There were also no studies that empirically assessed those who provide services to the homeless (e.g. shelter staff, clinicians, nurse practitioners) on their current practices or knowledge related to TBI.
The existing body of literature also lacked qualitative explorations of homelessness and TBI. Providing evidence from qualitative research has been identified as an important means of tailoring intervention programs in order to increase the likelihood for successful uptake [
32]. Such research provides personal insight into TBI, emphasizes the importance of persons with TBI receiving empathy and respect, and increases professionals’ understanding of the challenges that survivors face [
33]. TBI research should include qualitative methods in order to gain deeper insight into the experience of being homeless with a brain injury. Qualitative data could be used to evaluate and develop sustainable and appropriate interventions.
Outside of peer-reviewed publications, there is also a lack of information on the prevalence of TBI among people who are homeless. A scoping review of grey literature found few sources available online or in print on this topic. For example, the few resources found included an article featured in the Wall Street Journal [
34], a four part series in a newspaper sold by individuals who were homeless in Portland, Oregon [
35‐
38], a series of articles available through the National Health Care for the Homeless Network (e.g.[
6]) and previous conference presentations that discuss the issue of TBI among homeless persons (e.g. [
39]). These limited sources pointed to a clear demand for more information on TBI and homelessness and highlight the lack of empirical data to support educational and awareness initiatives. The few identified sources that provide information on TBI and homelessness were mostly geared towards care providers, clinicians, and academics. Very little information was available for homeless persons themselves or for policy makers who influence funding decisions. In the future, networks such as the Health Care for the Homeless Clinics’ Network that provides many resources on this topic, could be leveraged for knowledge translation. As well, newspaper articles geared towards the general public and to homeless populations are a valuable way to disseminate knowledge.