Trauma in adults experiencing homelessness

Unexpected Admissions and Patient-Directed Discharges - Understanding Trauma Patients: Scoping Review

INTRODUCTION

Few studies have explored patient-directed discharges (PDDs), previously referred to as leaving against medical advice, in trauma patients, a unique patient population given the unanticipated nature of their injury and suddenness of hospital admission. This scoping review focuses on PDDs in trauma patients and aims to explore who may be at risk and what factors influence this decision.

METHODS

A literature search was performed using five databases for publications between 1956 and 2023. Articles were included if they discussed admitted trauma patients who underwent a PDD and/or if they proposed why trauma patients choose to undergo a PDD. Articles that solely discussed PDDs from the emergency department were excluded as were PDDs for patients presenting with traumatic brain injuries. Studies were screened by two blinded independent reviewers with a third, tiebreaker reviewer if needed.

RESULTS

7931 articles were screened. Eleven (0.1%) articles were included. Demographic information associated with increased odds of PDD were as follows: younger age, lower socioeconomic status, and undomiciled. Those with substance use disorders and pre-existing psychiatric history had a higher rate of PDD. Stab and gunshot wounds, injuries to the upper and lower extremities and face were associated with increased odds of PDD. Hypotheses for PDDs were distrust in the health-care system, injury characteristics and workup (e.g., diagnostic testing versus direct therapeutic care), and impulsive decision-making secondary to substance intoxication or withdrawal.

CONCLUSIONS

Further research is required to better understand what interventions are successful for patients at risk of discharging themselves and implementation of these interventions.

Patient-Directed Discharge in Hand Infection Cases: Interactions Between Intravenous Drug Use, Socioeconomic Factors, and Subsequent Readmissions

BACKGROUND

Patient-directed discharge (PDD) poses a significant challenge in healthcare. Prior studies have shown associations of PDD with factors like race, housing, psychiatric illness, socioeconomic status, and intravenous drug use (IVDU). This study aims to identify factors contributing to PDD in hand infection patients at a public safety-net hospital and to investigate the long-term consequences through readmissions or returns to the emergency department (ED).

METHODS

A retrospective analysis was conducted on adult patients presenting with hand infections at San Francisco's main public hospital over 1 year. Data collected included demographics, housing status, social support, psychiatric diagnoses, and IVDU. Statistical analysis involved Mann-Whitney U tests, Fisher's exact tests, and logistic regression with odds ratios (ORs).

RESULTS

A total of 131 patients were included, comprising 95 (73%) conventionally discharged and 36 (27%) PDD patients. Positive correlations were found between PDD and several factors, including unemployment, unstable housing, living alone, lack of a phone number on file, alcohol use, and IVDU. However, in the multivariate analysis, IVDU emerged as the sole statistically significant predictor (OR, 4.22; CI, 1.18-15.05; P = 0.026). Further regression analysis identified unstable housing (OR, 4.39; CI, 1.17-16.44; P = 0.028) and living alone (OR, 4.45; CI, 1.25-15.89; P = 0.021) to be positively correlated with IVDU. PDD had higher ED revisits ( P = 0.025) and readmission rates ( P = 0.014).

CONCLUSIONS

This study underscores the critical role of socioeconomic factors, particularly IVDU, in influencing PDD among hand infection patients. The findings highlight the need for integrated healthcare strategies addressing medical and social determinants to reduce PDD and improve patient outcomes.

Drop-In Wound Care: Calgary's Wound Care Model Centred Around People Experiencing Homelessness

People experiencing housing insecurities or homelessness face significant barriers to equitable healthcare. A drop-in wound care service was established to mitigate social barriers and improve accessibility. This model facilitates direct access to a multidisciplinary team of trauma-informed medical staff on a walk-in basis. A retrospective chart review was performed on patients seen at the drop-in clinic from January 2021 to December 2021. A total of 119 patients were serviced over 798 visits, with 254 unique wounds managed. 82.8% of patients were living unsheltered, in emergency shelters or in provisional accommodation at the time of assessment. Trauma wounds, lower leg ulcers and frostbites represented the top three complaints. 69.7% of all patients returned to service for at least a second visit, with a median of 4 visits per patient over 42.5 days. Unsheltered patients were most likely to return to service (87.5%) but were most likely to be lost prior to wound closure (68.8%). Timely access to care with consistent follow-up is essential for quality wound care. Our drop-in service presents a working model for providing equitable wound care to socially disadvantaged patient populations. The effectiveness of this model is highlighted by the continual expansion serving 909 and 1029 visits in subsequent years.

Association of houselessness and outcomes after traumatic injury: A retrospective, matched cohort study at an urban, academic level-one trauma center

BACKGROUND

Houselessness is associated with increased mortality and unmet health needs. Current understanding of traumatic injury in houseless patients is limited.

METHODS

This is a retrospective matched cohort study among houseless and housed adults, admitted to an urban, safety net, level I trauma center from 1/1/2018-12/31/2021. Houseless patients were matched with their housed counterparts 1:2 based on age, sex, injury severity score (ISS) and nature of injury. The primary outcome was in-hospital adverse events. Secondary outcomes included hospital length of stay (LOS), outpatient follow-up, emergency department (ED) utilization post-injury, and readmission. Conditional multivariable regression was used to determine associations between the exposure and outcomes.

RESULTS

1413 patients were included; 471 houseless patients and 942 matched controls. Median [IQR] age was 42 years [31-58] and median [IQR] ISS was 9 [5-13] for all patients. About 30 % of traumatic injuries were violent in nature. Median [IQR] total LOS was longer for houseless patients (4.4 days [2.0-8.3] vs. 3.1 days [1.4-6.5], p < 0.001). Houseless patients were more frequently admitted to the ICU (5 % versus 3 %, p = 0.045). The rate of any in-hospital adverse event was similar (houseless 17 % vs. housed 16 %, p = 0.537). Adjusting for age, sex, language, insurance, ISS, nature of injury, injury mechanism, ICU admission, and operative intervention, houselessness was inversely associated with outpatient follow-up (OR 0.60, 95 % CI 0.46-0.79) and positively associated with ED representation (OR 2.49, 95 % CI 1.64-3.78) and hospital readmission (OR 4.35, 95 % CI 3.19-5.92).

CONCLUSIONS

Housing status was not associated with increased in-hospital morbidity or mortality in trauma patients in a single institution cohort of trauma patients. Unhoused patients had lower odds of completing outpatient injury-specific follow-up and higher odds of utilizing the ED within 30 days of discharge. These findings highlight gaps in post-discharge care coordination and underscore opportunities to improve discharge services for this population.

Outcomes and characteristics differ between homeless and housed trauma patients following the COVID-19 pandemic

BACKGROUND

Americans experiencing homelessness are uniquely vulnerable to traumatic injuries and morbidity. Despite a high and increasing number of persons experiencing homelessness (PEH), American researchers have not comprehensively described the impact of this social problem on trauma patients in recent years.

STUDY DESIGN

Retrospective cohort study using the American College of Surgeons TQIP 2021-2022 data. We performed descriptive statistics and multivariable modeling to test for differences in clinical characteristics and discharge dispositions between adult trauma patients experiencing homelessness (n = 20,692) and housed trauma patients (n = 1,927,159).

RESULTS

Trauma patients experiencing homelessness are more likely to be male, younger, and belong to different racial / ethnic groups. Homeless trauma patients have 37 % longer hospital stays than housed trauma patients and are more likely to experience an assault (adjusted OR: 2.92) or self-inflicted injury (adjusted OR: 1.50) and less likely to experience an unintentional injury (adjusted OR: 0.33). Homeless trauma patients' mechanisms of injury differ from those of housed trauma patients. They are similarly likely to have an ISS score of 12 or higher (adjusted OR: 1.01). They are slightly less likely to experience in-hospital mortality than housed trauma patients (adjusted OR: 0.85). They are more likely to be discharged to court or law enforcement (adjusted OR: 1.89); to a psychiatric facility or unit (adjusted OR: 2.99); leave against medical advice (adjusted OR: 3.89); or to a skilled nursing facility (adjusted OR: 2.01) than housed trauma patients. They are less likely to be discharged to inpatient rehab or home health than housed trauma patients.

CONCLUSIONS

This study describes differences in injury outcomes and clinical characteristics affecting homeless trauma patients compared to housed trauma patients since the COVID-19 pandemic, such as longer lengths of hospital stay, greater propensity to have violent injuries, and different discharge dispositions.

Oral Health Epidemiological Investigation in an Urban Homeless Population

The purpose of this clinical epidemiological investigation was to examine the oral health conditions of homeless people in the city of Rome, Italy. A total of 157 homeless subjects were subjected to a first dental visit, during which anamnestic information was recorded in a digital medical record. A diagnosis of dental disorders was performed by assessing oral hygiene conditions, periodontal health, the presence of caries and/or root residues, and the presence of partial and/or total edentulousness. Caries and missing teeth were evaluated by the DMFT index. The first major criticality was represented by poor or absent oral hygiene. The examined sample showed a major percentage of high DMFT (63.0%); the most common clinical condition was the presence of numerous root residues. Regarding periodontal health, 73.2% of patients had gingivitis, 21.6% periodontitis, while 11 patients had periodontal pathologies with tooth mobility (7%). Finally, 8.9% of patients had one or two missing elements, 22.9% had partial edentulousness, and 8.9% of the sample had total edentulism. This analysis provides an important basis for strengthening health promotion and the importance of accessible and effective care for this population. It will therefore be necessary to continue to adopt a patient-centered approach geared towards addressing the demands that this population faces in maintaining their oral health.

Unhoused and Injured: Injury Characteristics and Outcomes in Unhoused Trauma Patients

INTRODUCTION

The unhoused population is known to be at high risk for traumatic injury. However, there are scarce data regarding injury patterns and outcomes for this patient group. This study aims to investigate any differences in injury characteristics and hospital outcomes between unhoused and housed patients presenting with traumatic injuries.

METHODS

We conducted a 3-y retrospective cohort study at a level 1 trauma center in a metropolitan area with a large unhoused population. All adult trauma patients who were identified as unhoused or housed underinsured (HUI) were included in the study. Injury characteristics, comorbidities, and hospital outcomes were compared between the two groups.

RESULTS

A total of 8450 patients were identified, of which 7.5% were unhoused. Compared to HUI patients, unhoused patients were more likely to sustain minor injuries (65.2% versus 59.1%, P = 0.003) and more likely to be injured by assault (17.9% versus 12.4%, P < 0.001), stab wound (17.7% versus 10.8%, P < 0.001), and automobile versus pedestrian or bike (21.0% versus 15.8% P < 0.001). We found that unhoused patients had higher odds of mortality (adjusted odds ratio [AOR]: 1.93, 95% confidence interval [CI]: 1.10-3.36, P = 0.021), brain death (AOR: 5.40, 95% CI: 2.11-13.83, P < 0.001), bacteremia/sepsis (AOR: 4.36, 95% CI: 1.20-15.81, P = 0.025), and increased hospital length of stay (regression coefficient: 0.08, 95% CI: 0.03-0.12, P = 0.003).

CONCLUSIONS

This study observed significant disparities in injury characteristics and hospital outcomes between the unhoused and HUI groups. Our results suggest that these disparities are impacted by social determinants of health unique to the unhoused population.

Association of Homelessness with Before Medically Advised Discharge After Surgery

BACKGROUND

Before medically advised (BMA) discharge, which refers to patients leaving the hospital at their own discretion, is associated with higher rates of readmission and death in other settings. It is not known if housing status is associated with this phenomenon after surgery.

METHODS

We identified all admitted adults who underwent an operation by one of 11 different surgical services at a single tertiary care hospital between January 2013 and June 2022. Chi-square tests and t-tests were used to compare demographic and clinical features between BMA discharges and standard discharges. Multivariable logistic regression was used to evaluate the association between housing status and BMA discharge, adjusting for demographic and admission characteristics. Documented reasons for BMA discharge were also abstracted from the medical record.

RESULTS

Of 111,036 patient admissions, 242 resulted in BMA discharge (0.2%). After adjusting for observable confounders, patients experiencing homelessness had substantially higher odds of BMA discharge after surgery (adjusted odds ratio 4.4, 95% confidence interval 3.0-6.4; p < 0.001) when compared to housed. Patients who underwent emergency surgery, patients with a documented substance use disorder, and those insured by Medicaid also had significantly higher odds of BMA discharge. System- or provider-related reasons (including patient frustration with the hospital environment, challenges in managing substance dependence, and perceived inadequacy of paint control) were documented in 96% of BMA discharges for patients experiencing homelessness (vs. 66% in housed patients).

CONCLUSION

BMA discharge is more common in patients experiencing homelessness after surgery even after adjusting for observable confounding characteristics. Deeper understanding of the drivers of BMA discharge in patients experiencing homelessness through qualitative methods are critical to promote more equitable and effective care.

A national analysis of burn injuries among homeless persons presenting to emergency departments

INTRODUCTION

Burn injuries among the homeless are increasing as record numbers of people are unsheltered and resort to unsafe heating practices. This study characterizes burns in homeless encounters presenting to US emergency departments (EDs).

METHODS

Burn encounters in the 2019 Nationwide Emergency Department Sample (NEDS) were queried. ICD-10 and CPT codes identified homelessness, injury regions, depths, total body surface area (TBSA %), and treatment plans. Demographics, comorbidities, and charges were analyzed. Discharge weights generated national estimates. Statistical analysis included univariate testing and multivariate modeling.

RESULTS

Of 316,344 weighted ED visits meeting criteria, 1919 (0.6%) were homeless. Homeless encounters were older (mean age 44.83 vs. 32.39 years), male-predominant (71% vs. 52%), and had more comorbidities, and were more often White or Black race (p < 0.001). They more commonly presented to EDs in the West and were covered by Medicaid (51% vs. 33%) (p < 0.001). 12% and 5% of homeless burn injuries were related to self-harm and assault, respectively (p < 0.001). Homeless encounters experienced more third-degree burns (13% vs. 4%; p < 0.001), though TBSA % deciles were not significantly different (34% vs. 33% had TBSA % of ten or lower; p = 0.516). Homeless encounters were more often admitted (49% vs. 7%; p < 0.001), and homelessness increased odds of admission (OR 4.779; p < 0.001). Odds of transfer were significantly lower (OR 0.405; p = 0.021).

CONCLUSION

Homeless burn ED encounters were more likely due to assault and self-inflicted injuries, and more severe. ED practitioners should be aware of these patients' unique presentation and triage to burn centers accordingly.

Approaches and reporting of alcohol and other drug testing for injured patients in hospital-based studies: A systematic review

ISSUE

Hospital alcohol and/or other drug (AOD) testing is important for identifying AOD-related injuries; however, testing methods vary. This systematic review aimed to examine biological AOD testing methods from hospital-based studies of injured patients and quantify what proportion reported key information on those testing methods.

APPROACH

Observational studies published in English from 2010 onwards involving biological AOD testing for injured patients presenting to hospital were included. Studies examining single injury causes were excluded. Extracted data included concentration thresholds for AOD detection (e.g., lower limits of detection, author-defined cut-offs), test type (e.g., immunoassay, breathalyser) and approach (e.g., routine, clinical discretion), timing of testing, sample type and the proportion of injured cases tested for AODs.

KEY FINDINGS

Of 83 included studies, 76 measured alcohol and 37 other drugs. Forty-nine studies defined blood alcohol concentration thresholds (ranging from 0 to 0.1 g/100 mL). Seven studies defined concentration thresholds for other drugs. Testing approach was reported in 39/76 alcohol and 18/37 other drug studies. Sample type was commonly reported (alcohol: n = 69/76; other drugs: n = 28/37); alcohol was typically measured using blood (n = 60) and other drugs using urine (n = 20). Studies that reported the proportion of cases tested (alcohol: n = 53/76; other drugs: n = 28/37), reported that between 0% and 89% of cases were not tested for alcohol and 0% and 91% for other drugs. Timing of testing was often unreported (alcohol: n = 61; other drugs: n = 30).

IMPLICATIONS AND CONCLUSION

Variation in AOD testing methods alongside incomplete reporting of those methods limits data comparability and interpretation. Standardised reporting of testing methods will assist AOD-related injury surveillance and prevention.

Examining the Role of Self-Harm in the Relationship between Emergency Department Service Utilization and Trauma-Induced Homelessness among Homeless Individuals in Texas

Homelessness is a complex public health problem in the United States. Current or ongoing history of trauma among individuals adds to the complexity and challenges of homelessness. Our study assessed the moderating role of self-harm in the association between emergency department (ED) service utilization and trauma-induced homelessness (TIH) among adults in Texas. Homeless adults (N = 282) who completed their baseline Vulnerability Index Service Prioritization Decision Assistance Prescreen Tool survey between February 2021 and February 2022 at a Local Mental Health Authority in Texas were selected. The outcome variable, TIH, was assessed by current period of homelessness due to experiencing trauma or abuse. The main independent variable was ED utilization, while self-harm in the past year was assessed as the moderating variable. A multivariate logistic regression with a moderation analysis was conducted while adjusting for the covariates. Individuals who utilized ED services and engaged in self-harm and risky behaviors had greater odds of experiencing current period of TIH. Male respondents were less likely to experience TIH. Finally, engaging in self-harm significantly moderated the association between ED service use and TIH. This study may help inform efforts to develop tailored interventions and promote resilience-based approaches to improve health outcomes among individuals experiencing homelessness due to TIH.

Morbidity and Length of Stay After Injury Among People Experiencing Homelessness in North America

Importance

Traumatic injury is a leading cause of hospitalization among people experiencing homelessness. However, hospital course among this population is unknown.

Objective

To evaluate whether homelessness was associated with increased morbidity and length of stay (LOS) after hospitalization for traumatic injury and whether associations between homelessness and LOS were moderated by age and/or Injury Severity Score (ISS).

Design, Setting, and Participants

This retrospective cohort study of the American College of Surgeons Trauma Quality Programs (TQP) included patients 18 years or older who were hospitalized after an injury and discharged alive from 787 hospitals in North America from January 1, 2017, to December 31, 2018. People experiencing homelessness were propensity matched to housed patients for hospital, sex, insurance type, comorbidity, injury mechanism type, injury body region, and Glasgow Coma Scale score. Data were analyzed from February 1, 2022, to May 31, 2023.

Exposures

People experiencing homelessness were identified using the TQP's alternate home residence variable.

Main Outcomes and Measures

Morbidity, hemorrhage control surgery, and intensive care unit (ICU) admission were assessed. Associations between homelessness and LOS (in days) were tested with hierarchical multivariable negative bionomial regression. Moderation effects of age and ISS on the association between homelessness and LOS were evaluated with interaction terms.

Results

Of 1 441 982 patients (mean [SD] age, 55.1 [21.1] years; (822 491 [57.0%] men, 619 337 [43.0%] women, and 154 [0.01%] missing), 9065 (0.6%) were people experiencing homelessness. Unmatched people experiencing homelessness demonstrated higher rates of morbidity (221 [2.4%] vs 25 134 [1.8%]; P < .001), hemorrhage control surgery (289 [3.2%] vs 20 331 [1.4%]; P < .001), and ICU admission (2353 [26.0%] vs 307 714 [21.5%]; P < .001) compared with housed patients. The matched cohort comprised 8665 pairs at 378 hospitals. Differences in rates of morbidity, hemorrhage control surgery, and ICU admission between people experiencing homelessness and matched housed patients were not statistically significant. The median unadjusted LOS was 5 (IQR, 3-10) days among people experiencing homelessness and 4 (IQR, 2-8) days among matched housed patients (P < .001). People experiencing homelessness experienced a 22.1% longer adjusted LOS (incident rate ratio [IRR], 1.22 [95% CI, 1.19-1.25]). The greatest increase in adjusted LOS was observed among people experiencing homelessness who were 65 years or older (IRR, 1.42 [95% CI, 1.32-1.54]). People experiencing homelessness with minor injury (ISS, 1-8) had the greatest relative increase in adjusted LOS (IRR, 1.30 [95% CI, 1.25-1.35]) compared with people experiencing homelessness with severe injury (ISS ≥16; IRR, 1.14 [95% CI, 1.09-1.20]).

Conclusions and Relevance

The findings of this cohort study suggest that challenges in providing safe discharge to people experiencing homelessness after injury may lead to prolonged LOS. These findings underscore the need to reduce disparities in trauma outcomes and improve hospital resource use among people experiencing homelessness.

What medical-legal partnerships can do for trauma patients and trauma care

ABSTRACT

Trauma patients are particularly vulnerable to the impact of preexisting social and legal determinants of health postinjury. Trauma patients have a wide range of legal needs, including housing, employment, debt, insurance coverage, and access to federal and state benefits. Legal support could provide vital assistance to address the social determinants of health for injured patients. Medical legal partnerships (MLPs) embed legal professionals within health care teams to improve health by addressing legal needs that affect health. Medical legal partnerships have a successful track record in oncology, human immunodeficiency virus/acquired immune deficiency syndrome, and pediatrics, but have been little used in trauma. We conducted a scoping review to describe the role of MLPs and their potential to improve health outcomes for patients with traumatic injuries. We found that MLPs use legal remedies to address a variety of social and structural conditions that could affect patient health across several patient populations, such as children with asthma and patients with cancer. Legal intervention can assist patients in obtaining stable and healthy housing, employment opportunities, debt relief, access to public benefits, and immigration assistance. Medical legal partnership structure varies across institutions. In some, MLP lawyers are employed directly by a health care institution. In others, MLPs function as partnerships between a health system and an external legal organization. Medical legal partnerships have been found to reduce hospital readmissions, increase treatment utilization by patients, decrease patient stress levels, and benefit health systems financially. This scoping review outlines the potential of MLPs to improve outcomes for injured patients. Establishing trauma-focused MLPs could be a feasible intervention for trauma centers around the country seeking to improve health outcomes and reduce disparities for injured patients.

Injury Patterns and Hospital Admission After Trauma Among People Experiencing Homelessness

Importance

Traumatic injury is a major cause of morbidity for people experiencing homelessness (PEH). However, injury patterns and subsequent hospitalization among PEH have not been studied on a national scale.

Objective

To evaluate whether differences in mechanisms of injury exist between PEH and housed trauma patients in North America and whether the lack of housing is associated with increased adjusted odds of hospital admission.

Design, Setting, and Participants

This was a retrospective observational cohort study of participants in the 2017 to 2018 American College of Surgeons' Trauma Quality Improvement Program. Hospitals across the US and Canada were queried. Participants were patients aged 18 years or older presenting to an emergency department after injury. Data were analyzed from December 2021 to November 2022.

Exposures

PEH were identified using the Trauma Quality Improvement Program's alternate home residence variable.

Main Outcomes and Measures

The primary outcome was hospital admission. Subgroup analysis was used to compared PEH with low-income housed patients (defined by Medicaid enrollment).

Results

A total of 1 738 992 patients (mean [SD] age, 53.6 [21.2] years; 712 120 [41.0%] female; 97 910 [5.9%] Hispanic, 227 638 [13.7%] non-Hispanic Black, and 1 157 950 [69.6%] non-Hispanic White) presented to 790 hospitals with trauma, including 12 266 PEH (0.7%) and 1 726 726 housed patients (99.3%). Compared with housed patients, PEH were younger (mean [SD] age, 45.2 [13.6] years vs 53.7 [21.3] years), more often male (10 343 patients [84.3%] vs 1 016 310 patients [58.9%]), and had higher rates of behavioral comorbidity (2884 patients [23.5%] vs 191 425 patients [11.1%]). PEH sustained different injury patterns, including higher proportions of injuries due to assault (4417 patients [36.0%] vs 165 666 patients [9.6%]), pedestrian-strike (1891 patients [15.4%] vs 55 533 patients [3.2%]), and head injury (8041 patients [65.6%] vs 851 823 patients [49.3%]), compared with housed patients. On multivariable analysis, PEH experienced increased adjusted odds of hospitalization (adjusted odds ratio [aOR], 1.33; 95% CI, 1.24-1.43) compared with housed patients. The association of lacking housing with hospital admission persisted on subgroup comparison of PEH with low-income housed patients (aOR, 1.10; 95% CI, 1.03-1.19).

Conclusions and Relevance

Injured PEH had significantly greater adjusted odds of hospital admission. These findings suggest that tailored programs for PEH are needed to prevent their injury patterns and facilitate safe discharge after injury.

Total Joint Arthroplasty in Homeless Patients at an Urban Safety Net Hospital

INTRODUCTION

Homelessness is a key social determinant of health, and the patient population has grown to over 580,000. Total joint arthroplasty (TJA) is an effective treatment of symptomatic end-stage osteoarthritis of the hip and knee and has been shown to improve health-related quality of life in the general population. However, the literature on the outcomes of TJA among homeless patients is limited.

METHODS

We retrospectively reviewed 442 patients who underwent primary, unilateral TJA between June 1, 2016, and August 31, 2017, at an urban, tertiary, academic safety net hospital. Based on self-reported living status, we classified 28 homeless patients and 414 control nonhomeless patients. Fisher exact tests, Student t-tests, and multivariate logistic regression were used to compare the demographics, preoperative conditions, and surgical outcomes between the two groups.

RESULTS

The homeless group were younger, more often male, and smokers; had alcohol use disorder; and used illicit drugs. After controlling for age, sex, and preoperative medical and social conditions, homeless patients were 15.83 times more likely to have an emergency department visit (adjusted odds ratio, 15.83; 95% confidence interval, 5.05 to 49.59; P < 0.0001) within 90 days but had similar rates of readmission (P = 0.25), revision surgery (P = 0.38), and prosthetic joint infection (P = 0.25) when compared with nonhomeless patients.

DISCUSSION

Although homeless patients did not have higher rates of readmission or revision surgery, homelessness still presents unique challenges for the TJA patients and providers. With careful preoperative optimization and collaborative support, however, the benefits of TJA may outweigh the risk of poor outcomes for these patients.

Prevalence of Housing Insecurity in Survivors of Traumatic Injury

BACKGROUND

Almost 20% of the US population are housing insecure (HI) due to inadequate living conditions, imbalanced costs to income, nonsustainable shelter, or overcrowding. Housing insecure and physical trauma share similar risk factors, but their direct association is not well elucidated. The objective of our study was to determine the prevalence of HI in survivors of traumatic injury.

METHODS

We conducted a prospective cohort study at an urban, level 1 trauma center. A survey on social determinants of health was administered to adult patients, and demographic, injury specifics, and clinical outcomes data were collected. HI was defined by affirmative answers to questions related to history of homelessness or concern about sustainable shelter. The cohort was stratified by HI; groups were compared using the Mann-Whitney U and Fisher exact tests, as appropriate.

RESULTS

Of 116 study participants, four were excluded due to missing data. Of the 112, 42 (37.8%) reported HI and most were black (69%) males (73.8%). There were no demographic differences between groups. Conversely, HI patients had a higher rate of penetrating traumatic injury (34.1% vs 14.5%, P-value = .03) and were significantly less educated compared to secure participants (P-value = .03) [no degree (26.2% vs 10.3%), high school degrees (21.4% vs 41.2%)] with concomitant illicit drug use (63.4% vs 27.9%, P < .001), and history of addiction (52.4% vs 7.2%, P < .001).

CONCLUSION

HI far exceeds national averages in our cohort. Although difficult to ascertain a cause-effect relationship, HI may be a modifiable risk factor for trauma that negatively influences outcomes.

The Association Between Interpersonal Violence and Unstable Housing Among Veterans

BACKGROUND

Despite programs to address housing for Veterans, they continue to be at high risk of unstable housing. Interpersonal violence is also highly prevalent among Veterans and may contribute to unstable housing. Our study aimed to determine whether interpersonal violence was associated with unstable housing among Veterans, and how this association was influenced by common co-occurring conditions such as substance use and mental illness.

METHODS

Veterans in the Mind Your Heart Study (N = 741) completed survey data on history of interpersonal violence and access to housing in the prior year. Interpersonal violence was defined as experiencing sexual violence, physical violence, or mugging/physical attack using the Brief Trauma Questionnaire. Multivariable models examined associations between interpersonal violence and unstable housing. Primary models were adjusted for age and sex. Potential explanatory factors were added in subsequent models, including marital status, education, income, substance use disorder, PTSD, and other mental illness.

RESULTS

Veterans who had experienced interpersonal violence had almost twice the odds of unstable housing after adjustment for age and sex (AOR 1.9, 95% CI 1.2-3.0). This association was attenuated in the fully adjusted model including substance use, PTSD, and other mental illness, illustrating the interdependence of these factors (AOR 1.5, 95% CI 0.91-2.5). Subtypes of interpersonal violence were individually associated with increased odds of unstable housing after adjustment for age and sex (physical abuse AOR 1.7, 95% CI 1.2-2.5; mugging/physical attack AOR 1.8, 95% CI 1.2-2.7; sexual violence AOR 1.4, 95% CI 0.89-2.2), but were no longer significant in the fully adjusted model.

CONCLUSIONS

Previous experiences of interpersonal violence were associated with unstable housing among Veterans. Substance use, PTSD, and other mental illness played an important role in this relationship-highlighting the potential to improve health outcomes through trauma informed approaches that address mental health, substance use, and housing concurrently.

Association of Psychosocial Factors and Hospital Complications with Risk for Readmission After Trauma

BACKGROUND

Unplanned hospital readmissions are associated with morbidity and high cost. Existing literature on readmission after trauma has focused on how injury characteristics are associated with readmission. We aimed to evaluate how psychosocial determinants of health and complications of hospitalization combined with injury characteristics affect risk of readmission after trauma.

MATERIALS AND METHODS

We conducted a retrospective cohort study of adult trauma admissions from July 2015 to September 2017 to Harborview Medical Center in Seattle, Washington. We assessed patient, injury, and hospitalization characteristics and estimated associations between risk factors and unplanned 30-d readmission using multivariable generalized linear Poisson regression models.

RESULTS

Of 8916 discharged trauma patients, 330 (3.7%) had an unplanned 30-d readmission. Patients were most commonly readmitted with infection (41.5%). Independent risk factors for readmission among postoperative patients included public insurance (adjusted Relative Risk (aRR) 1.34, 95% CI 1.02-1.76), mental illness (aRR 1.39, 1.04-1.85), and chronic renal failure (aRR 2.17, 1.39-3.39); undergoing abdominal, thoracic, or neurosurgical procedures; experiencing an index hospitalization surgical site infection (aRR 4.74, 3.00-7.50), pulmonary embolism (aRR 3.38, 2.04-5.60), or unplanned ICU readmission (aRR 1.74, 1.16-2.62); shorter hospital stay (aRR 0.98/d, 0.97-0.99), and discharge to jail (aRR 4.68, 2.63-8.35) or a shelter (aRR 4.32, 2.58-7.21). Risk factors varied by reason for readmission. Injury severity, trauma mechanism, and body region were not independently associated with readmission risk.

CONCLUSIONS

Psychosocial factors and hospital complications were more strongly associated with readmission after trauma than injury characteristics. Improved social support and follow-up after discharge for high-risk patients may facilitate earlier identification of postdischarge complications.