Introduction
Marginally housed individuals experience high lifetime prevalence rates of traumatic brain injury (TBI). Although newer research has highlighted the vulnerability of this population to acquire incident TBI (incidence proportion of 17,100 per 100,000 population; Nikoo et al., 2017), reported rates are likely underestimations of the true rate due to limited and distal screening methodologies. The aim of this study was to improve the methodology of TBI screening through repeated and proximal assessments, in order to acquire a more accurate estimate of the rate of TBI in marginally housed persons. 

Methods
Participants recruited from single-room occupancy hotels (N = 326) met monthly for TBI education and screening over a one participant-year period using the Ohio State University TBI Identification Method and an additional questionnaire. TBI was defined as a head injury from a known cause with associated loss of consciousness, post-traumatic amnesia, and/or daze and confusion. Interviews were performed by research assistants trained in the sequelae of traumatic brain injury and common issues in screening for TBI in this population.

Findings
Across 326 participants, 2520 screening assessments were completed (mean = 7.7 screens, median = 8, mode = 10, SD = 3.6). Across the possible 326 person-years, 185.6 person-years of data were present, with data found to be missing completely at random. Across all 326 people, 100 people acquired 175 TBI events. The incidence proportion (100 of 326 people) was 30.7 percent, or 30674.9 TBI per 100,000 population. The incidence attack rate (100 people over 185.6 person-years) was 0.538842 persons per year. The incidence event proportion (175 events out of 326 people) was 53.7 percent, or 53680.98 TBI per 100,000 population. The incidence event rate (175 events over 185.6 person-years) was 0.94292 injuries per year. Of the 100 people who acquired at least one TBI, 61 reported only one TBI, while 39 reported two or more TBI (range 0 to 6). Of 175 TBI events, severity classifications were 142 mild (81.1%), 32 moderate to severe (18.3%), and 1 unknown (0.6%). The most common mechanisms of injury were falls (45.1 %), assaults (25.1 %), and blunt force trauma on/by an object (18.9 %). Of these, 17 (9.7 %) occurred in the context of a drug overdose, and at least 48 (27.4%) were acquired when the participant was acutely intoxicated by drugs or alcohol.

Conclusions
Among marginally housed persons, rates of incident TBI are remarkably higher than previously reported. The incidence proportion and rate observed in this vulnerable sample were almost double those previously reported in marginally housed persons, and more than 50 times the pooled North American estimates (Nguyen et al., 2016). Prevention strategies targeted towards these individuals may decrease the burden of traumatic brain injury in this vulnerable population.

Rates of Incident Traumatic Brain Injury in Marginally Housed Persons