Learning from suicide deaths in Harris County, Texas.

This study examined individual and community demographic characteristics surrounding suicides in one of the most populous counties in the United States. We paired medical examiner records with U.S. Census data and analyzed them using geospatial software. The majority of decedents were non-Hispanic, white males who died primarily of gunshot wounds. Salient age characteristics included interpersonal violence and depression among ages younger than 40. Despite lower incomes and education levels, areas with higher population density and racial/ethnic minorities had fewer suicides. Additional research should address depression among males and the elderly, interpersonal violence, firearm access, and culture.

Efficacy of a Just-in-Time Adaptive Intervention to Promote HIV Risk Reduction Behaviors Among Young Adults Experiencing Homelessness: Pilot Randomized Controlled Trial.

BACKGROUND: People experiencing homelessness have higher rates of HIV than those who are stably housed. Mental health needs, substance use problems, and issues unique to homelessness such as lack of shelter and transiency need to be considered with regard to HIV prevention. To date, HIV prevention interventions for young adults experiencing homelessness have not specifically addressed modifiable real-time factors such as stress, sexual or drug use urge, or substance use, or been delivered at the time of heightened risk. Real-time, personalized HIV prevention messages may reduce HIV risk behaviors. OBJECTIVE: This pilot study tested the initial efficacy of an innovative, smartphone-based, just-in-time adaptive intervention that assessed predictors of HIV risk behaviors in real time and automatically provided behavioral feedback and goal attainment information. METHODS: A randomized attention control design was used among young adults experiencing homelessness, aged 18-25 years, recruited from shelters and drop-in centers in May 2019. Participants were randomized to either a control or an intervention group. The intervention (called MY-RID [Motivating Youth to Reduce Infection and Disconnection]) consisted of brief messages delivered via smartphone over 6 weeks in response to preidentified predictors that were assessed using ecological momentary assessments. Bayesian hierarchical regression models were used to assess intervention effects on sexual activity, drug use, alcohol use, and their corresponding urges. RESULTS: Participants (N=97) were predominantly youth (mean age 21.2, SD 2.1 years) who identified as heterosexual (n=51, 52%), male (n=56, 57%), and African American (n=56, 57%). Reports of sexual activity, drug use, alcohol use, stress, and all urges (ie, sexual, drug, alcohol) reduced over time in both groups. Daily drug use reduced by a factor of 13.8 times over 6 weeks in the intervention group relative to the control group (Multimedia Appendix 4). Lower urges for sex were found in the intervention group relative to the control group over the duration of the study. Finally, there was a statistically significant reduction in reports of feeling stressed the day before between the intervention and control conditions (P=.03). CONCLUSIONS: Findings indicate promising intervention effects on drug use, stress, and urges for sex in a hard-to-reach, high-risk population. The MY-RID intervention should be further tested in a larger randomized controlled trial to further investigate its efficacy and impact on sexual risk behaviors. TRIAL REGISTRATION: ClinicalTrials.gov NCT03911024; https://clinicaltrials.gov/ct2/show/NCT03911024.

Establishing a Regional Trauma Preventable/Potentially Preventable Death Rate.

OBJECTIVE: To establish a trauma preventable/potentially preventable death rate (PPPDR) within a heavily populated county in Texas. SUMMARY: The National Academies of Sciences estimated the trauma preventable death rate in the United States to be 20%, issued a call for zero preventable deaths, while acknowledging that an accurate preventable death rate was lacking. In this absence, effective strategies to improve quality of care across trauma systems will remain difficult. METHODS: A retrospective review of death-related records that occurred during 2014 in Harris County, TX, a diverse population of 4.4 million. Patient demographics, mechanism of injury, cause, timing, and location of deaths were assessed. Deaths were categorized using uniform criteria and recorded as preventable, potentially preventable or nonpreventable. RESULTS: Of 1848 deaths, 85% had an autopsy and 99.7% were assigned a level of preventability, resulting in a trauma PPPDR of 36.2%. Sex, age, and race/ethnicity varied across preventability categories (P < 0.01). Of 847 prehospital deaths, 758 (89.5%) were nonpreventable. Among 89 prehospital preventable/potentially preventable (P/PP) deaths, hemorrhage accounted for 55.1%. Of the 657 initial acute care setting deaths, 292 (44.4%) were P/PP; of these, hemorrhage, sepsis, and traumatic brain injury accounted for 73.3%. Of 339 deaths occurring after initial hospitalization, 287 (84.7%) were P/PP, of these 117 resulted from sepsis and 31 from pulmonary thromboembolism, accounted for 51.6%. CONCLUSIONS: The trauma PPPDR was almost double that estimated by the National Academies of Sciences. Data regarding P/PP deaths offers opportunity to target research, prevention, intervention, and treatment corresponding to all phases of the trauma system.

Establishing a regional pediatric trauma preventable/potentially preventable death rate.

PURPOSE: Although trauma is the leading cause of death for the pediatric population, few studies have addressed the preventable/potentially preventable death rate (PPPDR) attributable to trauma. METHODS: This is a retrospective study of trauma-related death records occurring in Harris County, Texas in 2014. Descriptive and Chi-squared tests were conducted for two groups, pediatric and adult trauma deaths in relation to demographic characteristics, mechanism of injury, death location and survival time. RESULTS: There were 105 pediatric (age < 18 years) and 1738 adult patients. The PPPDR for the pediatric group was 21.0%, whereas the PPPDR for the adult group was 37.2% (p = 0.001). Analysis showed fewer preventable/potentially preventable (P/PP) deaths resulting from any blunt trauma mechanism in the pediatric population than in the adult population (19.6% vs. 48.4%, p < 0.001). Amongst the pediatric population, P/PP traumatic brain injury (TBI) were more common in the youngest age range (age 0-5) vs. the older (6-12 years) pediatric and adolescent (13-17 years) patients. CONCLUSION: Our results identify areas of opportunities for improving pediatric trauma care. Although the overall P/PP death rate is lower in the pediatric population than the adult, opportunities for improving initial acute care, particularly TBI, exist.

A Descriptive and Geospatial Analysis of Environmental Factors Attributing to Sudden Unexpected Infant Death.

This study examined medicolegal death investigation records and autopsy reports of a medical examiner's office to identify the circumstances surrounding sudden unexpected infant deaths (SUID) and geospatial analyses to pinpoint areas of infant death concentration. Analysis of 732 records of SUID deaths occurring in a 10-year span resulted in the conclusion that environmental factors associated with the sudden death were to some extent modifiable. Co-sleeping (sharing a sleeping surface, or bed-sharing) on various surfaces (mattress, pallet, couch) occurred in 53.4% of the infant deaths. Geographic areas where the largest number of deaths occurred were characterized as areas of high poverty level. The inclusion of additional information at the time of investigation (eg, alcohol and tobacco use of co-sleepers, illness of others in household, exceptions to normal sleep routine of infant) may aid in identifying modifiable circumstances to reduce infant mortality attributable to sudden infant death.

Drug use patterns and predictors among homeless youth: Results of an ecological momentary assessment.

BACKGROUND: Alcohol and drug use is associated with significant morbidity and mortality and is highly prevalent among homeless youth. Ecological Momentary Assessments (EMA) have been used to examine the effect of urges on drug use, though not among homeless youth. OBJECTIVES: We assessed the patterns of drug use and the correlation between real-time contextual factors and drug use using EMA collected daily. We identified predictors of drug use among a sample of homeless youth 18-25 years old in Houston, Texas. METHODS: Homeless youth (n = 66, 62% male) were recruited from a drop-in center between September 2015 and May 2016. We used generalized linear mixed models and cross-validation methods to determine the best predictive model of drug use. RESULTS: The overall drug use was high: 61% and 32% of participants reported using drugs or alcohol at least one day, respectively. Marijuana and synthetic marijuana use (i.e., Kush, K2, incense packs) were reported most frequently; 86% and 13% of the total drug use EMAs, respectfully. Drug use urge was reported on 26% of days and was the highest on drug use days. Drug use was predicted by discrimination, pornography use, alcohol use, and urges for drugs, alcohol, and to steal. CONCLUSIONS: EMA can be used to predict drug use among homeless youth. Drug use treatment among homeless youth should address the role of experiencing discrimination, pornography and alcohol use, and urge management strategies on drug use. Research is needed to determine if EMA informed just-in-time interventions targeting these predictors can reduce use.

Comparing COVID-19 vaccination coverage, adverse reactions and impact of social determinants of health on vaccine hesitancy in ADRD/MCI and non-ADRD/MCI population: protocol for a retrospective cross-sectional study.

INTRODUCTION: COVID-19 vaccination is crucial for vulnerable people with underlying chronic conditions such as Alzheimer's disease and related dementias (ADRD) and mild cognitive impairment (MCI). These individuals face unique challenges, including higher risk of COVID-19, difficulties in adopting preventive behaviours and vaccine hesitancy due to concerns about adverse reactions. Therefore, efforts to promote vaccination, including boosters tailored to the currently circulating virus, are essential for people with ADRD/MCI. OBJECTIVE: The primary purpose of this study protocol is to conduct a comprehensive analysis of COVID-19 vaccination coverage and adverse reactions among individuals with ADRD/MCI in comparison to those without ADRD/MCI. Additionally, the proposed study aims to investigate the impact of social determinants of health on COVID-19 vaccination and vaccine hesitancy in individuals with ADRD/MCI. METHODS AND ANALYSIS: A retrospective cross-sectional study will be conducted utilising data from the All of Us (AoU) Researcher Workbench. Relevant data fields are extracted from sources including demographic information, COVID-19 Vaccine Survey, Basic Survey, Health Access & Utilization, Social Determinants of Health, and Electronic Health Record (EHR) data. Data on vaccination, adverse reactions and vaccine hesitancy will be collected through COVID-19 vaccine survey questionnaires. Propensity score matching and binary logistic regression will be applied to assess the vaccination rates and vaccine hesitancy, while controlling for demographic characteristics and social determinants of health factors. ETHICS AND DISSEMINATION: This study protocol received approval from the Institutional Review Board at Florida State University (STUDY00004571). Results will be disseminated through publication in peer-reviewed journals and presented at scientific conferences.

Comparing survival rates and mortality in operative versus nonoperative treatment for femoral neck fractures among Alzheimer's disease patients: A retrospective cohort study.

Introduction: Addressing femoral neck fractures resulting from ground-level falls in older adults with Alzheimer's disease (AD) involves a personalized treatment plan. There is considerable ongoing debate concerning the relative advantages and disadvantages of surgical treatment (internal fixation or arthroplasty) vs nonoperative treatment for femoral neck fractures in older persons with AD. Methods: This retrospective cohort study compared the mortality, hazard ratio, and survival rate between operative and nonoperative treatments, controlling for patients' demographic information and baseline health status. The study population consisted of Optum beneficiaries diagnosed with AD who experienced an initial femoral neck fracture claim between January 1, 2012, and December 31, 2017. Kaplan-Meier survival curves were applied to compare the treatment groups' post-fracture survival rates and mortality. Cox regression was used to examine the survival period by controlling the covariates. Results: Out of the 4157 patients with AD with femoral neck fractures, 59.8% were women (n = 2487). The median age was 81 years. The 1-year survival rate for nonoperative treatment (70.19%) was lower than that for internal fixation (75.27%) and arthroplasty treatment (82.32%). Compared with the nonoperative group, arthroplasty surgical treatment had significant lower hazard risk of death (arthroplasty hazard ratio: 0.850, 95% CI: 0.728-0.991, P < 0.05). Discussion: The findings suggest that the operative treatment group experiences higher survival rates and lower mortality rates than the nonoperative group. This paper provides insights into treatment outcomes of older adults with AD receiving medical care for femoral neck fractures.

The silent killer: Previously undetected pulmonary emboli that result in death after discharge.

INTRODUCTION: Pulmonary embolism (PE) is a recognized cause of death in hospitalized trauma patients, yet less is known about PE after discharge. PATIENTS & METHODS: All post-discharge, autopsy-demonstrated, fatal PE resulting from trauma within a large US county over six years were analyzed. Counts, percentages, mean values, SD, and IQR were calculated for all variables. RESULTS: 1848 trauma deaths were reviewed, of which 85% had an autopsy. Eighty-five patients died from PE after discharge from their initial injury. 53% were initially treated at non-trauma centers, and 9% did not seek medical assistance. 75% were injured by falling, and most injuries occurred in the lower extremities. 86% had an ISS <16, but 87% needed assistance or were bed-bound after injury, despite 75% having no mobility limitations before the injury. 53% died within one month of injury, and 91% within the first year. Before death, only 11% were prescribed chemical thromboprophylaxis or an antiplatelet agent, and only 8% were diagnosed with venous thromboembolism before death. CONCLUSIONS: Fatal PE after discharge typically occurred following activity-limiting lower extremity injuries with an ISS<16.

Geospatial Analysis of Prehospital Triage and Early Potential Preventable Traumatic Deaths.

Severely injured patients often depend on prompt prehospital triage for survival. This study aimed to examine the under-triage of preventable or potentially preventable traumatic deaths. A retrospective review of Harris County, TX, revealed 1848 deaths within 24 hours of injury, with 186 being preventable or potentially preventable (P/PP). The analysis evaluated the geospatial relationship between each death and the receiving hospital. Out of the 186 P/PP deaths, these were more commonly male, minority, and penetrating mechanisms when compared with NP deaths. Of the 186 PP/P, 97 patients were transported to hospital care, 35 (36%) were transported to Level III, IV, or non-designated hospitals. Geospatial analysis revealed an association between the location of initial injury and proximity to receiving Level III, IV, and non-designated centers. Geospatial analysis supports proximity to the nearest hospital as one of the primary reasons for under-triage.

Exploring the complexity of firearm homicides in Harris County, Texas, from 2009 to 2021: Implications for theory and prevention.

Firearm violence is a major health problem in the United States that clusters asymmetrically across geographic and demographic lines, and the persistence and unequal distribution of firearm violence suggests that novel causal explanations and theoretical frameworks may be warranted to guide preventive strategies. Thus, this study explores the following three hypotheses that are grounded in complex systems theory: 1) trends in firearm homicides risks have shifted heterogeneously in Harris County across endemic degree of risk; 2) firearm homicides clusters have remained resilient in Harris County across the study time period; and 3), the associations between known contextual correlates of firearm homicides and the distribution of firearm homicides risks in Harris County have manifested as nonlinear. Using a retrospective study design (n = 4,397) from January 1, 2009-June 31, 2021, medicolegal death investigation data from the Harris County Institute of Forensic Sciences and estimates of community characteristics from the American Community Survey were analyzed using Joinpoint trend analysis, kernel density geospatial analysis, and proportion tests. Trend analyses revealed that firearm homicides risks shifted heterogeneously across endemic degree of risk, with geographical areas with lower initial firearm homicides risks experiencing more profound upward shifts across the time period of the study. Geospatial analyses identified the resiliency of firearm homicides clusters across the study period, particularly in central, southern, and south-western districts of the city. Finally, the relationships between known contextual correlates and the distribution of firearm homicides risks in Harris County appeared to be nonlinear, particularly regarding ethnicity. This study provides data-driven results that suggest the plausibility of complex systems theory in advancing the understanding of causality in firearm homicides. Further, these findings support the urgent need for complex systems-informed preventive efforts that account for spatiotemporal heterogeneity, key interactions that generate nonlinearity, and latent feedback loops that underlie resiliency in firearm homicides.

Rural-urban, age, and gender disparities and trends in suicide and homicide: Multistate evidence across 12 years.

PURPOSE: Few studies have simultaneously assessed age and gender trends in homicide and suicide across the rural-urban continuum. Herein, we examine geographic and demographic trends in suicide and homicide death rates by: (1) determining overall macro and disaggregated trends; (2) examining differences in trends based on rural-urban county classification; and (3) identifying differences in stratified trends among age and gender classifications. METHODS: A retrospective study design used suicide and homicide data (n = 199,456) from years 2005to 2017 across 16 US states. Suicide and homicide deaths were grouped by age, gender, and rural-urban classification for descriptive analyses, and trends were analyzed using Joinpoint trend analysis software. FINDINGS: Violence resulted in 142,470 suicide and 56,986 homicide deaths between 2005 and 2017. Among both males and females, overall macro trends of suicide and homicide rates generally increased with greater rurality, and trends in rural rates differed from those in nonrural areas. Joinpoint trend analysis revealed significant increases in male suicide rates in large metropolitan (1.66%), micropolitan (1.78%), and rural areas (1.77%); female suicide rates in large metropolitan (2.17%), small metropolitan (3.25%), and micropolitan areas (3.26%); male homicide rates in large metropolitan areas (10.19%); and female homicide rates in rural areas (8.29%). Finally, when stratified by age, several significant trends were found, including increases in suicide rates among females aged 64 and older in rural areas (11.71%). CONCLUSIONS: Heterogeneous trends were found in suicide and homicide rates within specific rural-urban, age, and gender subgroups. Prevention efforts should proactively target those subgroups identified herein as most at-risk of violence.

Contextual information in medicolegal death investigation decision-making: Manner of death determination for cases of a single gunshot wound.

To explore the role of contextual information in determining manner of death, four cases involving single gunshot wounds were presented to participants (n = 252) involved in medicolegal death investigation. The participants received identical autopsy information but different contextual information. The data demonstrated that participants tended to rely on contextual information more than autopsy information: In the suicide context, participants across the four cases reached 153 final decisions of suicide (and 25 of homicide), whereas in the homicide context, participants reached only 10 final decisions of suicide (and 181 of homicide) --all while examining identical autopsy information. The impact of the contextual information was so powerful that many participants changed initial autopsy-based conclusions to align with the contextual information. Given the significant role and impact that contextual information has on expert decision making, one must consider what, how, and when contextual information should be used.

An Analysis of Traumatic and Non-Traumatic Hemopericardium Using Autopsy Data From a County Medical Examiner.

The volume of hemopericardium requiring hemodynamic changes in the trauma patient is not well understood. We performed a study using autopsy data from trauma patients who died with hemopericardium (>20 mL). Of 1848 traumatic deaths, 54 had hemopericardium at autopsy. The median pericardial blood in this group was 150 mL, which is more than the previously assumed volume to be lethal in trauma patients. Therefore, it may be appropriate to redefine the estimated volume required to cause lethal hemopericardium in trauma patients.

Evaluating the robustness of the CMS Hospital Value-Based Purchasing measurement system.

BACKGROUND: The Hospital Value-Based Purchasing Program (HVBP) is a pay for performance system that impacts traditional Medicare fee-for-service payments to hospitals through rewards and penalties. OBJECTIVES: To explore variation in overall and individual-hospital total performance score (TPS) and embedded domains for hospitals during 2014-2018. DATA SOURCE: Hospital data were retrieved from the publicly available HOSArchive dataset. STUDY DESIGN: Distribution of annual TPS and HVBP domain scores for 2014-2018 was evaluated using descriptive statistics. Transitional probabilities were analyzed to evaluate annual movement in the TPS ranking for outlier hospitals in the Top and Bottom 5%. PRINCIPAL FINDINGS: TPS scores are positively skewed while the distribution of domain scores vary with patient experience, (clinical) outcome, and efficiency domains having a large number of (positive) outliers. Mean TPS score decreased from 40.54 in 2014 to 38.04 by 2018. Improvement was shown in mean domain scores for clinical process of care and clinical outcome using 95% confidence intervals, with hospitals gaining 10 points over the study period in clinical outcome. Changes in the mean scores for other domains did not show consistent increases or decreases. Chi-square analyses of hospital ranking categories showed some evidence that, as a group, hospitals initially ranked in the Bottom 5% are making consistent annual movements to higher categories. In contrast, over half of the hospitals ranking in the initial Top 5% remained in the top category across all study years. CONCLUSIONS: It may be time for CMS to redesign the HVBP incentive program to assure the measures accurately demonstrate sustained improvement, the domain weights appropriately reflect the level of importance, and the TPS comparative ranking methodology does not discourage lower-performing hospitals from actively improving the care they deliver and achieving top ranks.

Challenges in the implementation of electronic systems for patient report of symptoms in oncology: a scoping review.

Background: Under-recognition and under-treatment of symptoms are prevalent throughout the health care system in the United States. While the reasons for this are complex, it is widely recognized that electronic symptom reports can improve clinicians' ability to manage symptoms. However, electronic symptom reporting has yet to be widely implemented. Electronic systems are most effective when tailored to the specific patient population or clinical setting. For example, numerous oncology-focused electronic symptom reporting systems have been developed for patients with cancer undergoing treatment in the United States. The objective of this scoping review was to identify challenges that arose in the implementation of electronic systems for patient-reported symptoms in oncology clinical practice, and approaches that were taken or recommended to overcome those challenges. Methods: This scoping review involved comprehensive searches of Medline, CINAHL, and the Cochrane Central Register of Controlled Trials, which yielded 3,133 articles. Following screening, 20 research studies met the inclusion criteria and were included in this review. Data were systematically extracted from the articles using a qualitative content analysis. Results: Challenges identified were thematically categorized as technical issues, system usability issues, patient lack of comfort/knowledge of technology, incomplete/missing data, lack of patient use of the system, other patient issues, difficulties timing completion with clinical processes, lack of clinic staff involvement/engagement, and lack of clinician comfort/knowledge regarding the use of patient-reported outcome data. Discussion: The findings of this review highlight challenges that need to be addressed when implementing an electronic symptom reporting system for patients with cancer, and potential strategies for overcoming these challenges. This review may help hospital administrators and clinicians prepare for and improve the implementation of electronic symptom reporting systems into clinical practice, thereby providing evidence to enable their broader use.

Identifying gaps and improving investigation of fatal elder abuse and/or neglect.

Death investigator and autopsy reports for decedents 65 years and older within a major metropolitan area over a five-year period were assessed for the possibility of elder abuse and/or neglect. The study consisted of two stages. A simple two-question screening criteria was used to determine whether the decedent was (1) dependent on another for at least one activity of daily living and (2) had a presence of at least one indicator of abuse and/or neglect. Second, only cases with affirmative criteria responses were reviewed to identify inconsistent or deficient variables that precluded (or if present, allowed) determination of abuse and/or neglect. A multidisciplinary panel of local and national experts, including forensic pathologists, law enforcement, and geriatricians assessed these indicators as indicative of presence of abuse/neglect, and these indicators were subsequently developed as a supplemental data collection tool. Of a possible 2798 cases, 2324 (83%) were excluded using the screening criteria. This reduced the number of cases that warranted further investigation to 474 (17% of elderly deaths in this timeframe). All 474 decedents were dependent on another for at least one ADL and 322 (68%) had unexplained injuries. In 180 (38%) cases had recorded notation of a suspicion of abuse and/or neglect at the time of death. The results support the premise that a simple, two-criterion screening can effectively identify cases of potential abuse and/or neglect and, when followed by a supplemental data collection tool, cases can be efficiently evaluated.

Fatal and Nonfatal Injury Patterns of Stranded Motorists.

ABSTRACT: Because nonfatal and fatal incidents for stranded motorists are not separated from vehicular accidents, little data are available on incident characteristics. To close this gap, data fields were inserted into databases at a medical examiner's office and two trauma centers to collect injury-related information. Forensic nurses and pathologists aided in forming a collaboration among the agencies involved and supported data collection efforts. Data collected over a 5-year period were examined for injury patterns to determine risk factors affecting these patterns. Of the total sample (N = 219), 24.7% had spinal injuries resulting in fatal injuries for 46 of 54 individuals. The odds were stranded motorists with spine-related injuries (C1-C7) had 9.13 times higher risk for a fatal outcome compared with those without spine-related injuries. Severe injuries (Abbreviated Injury Scale scores ≥ 4) noted for head/neck (29.7%) and chest (24.2%) were significantly associated with fatality. Of the 219 cases, 22.8% were inside of a stopped vehicle, and 77.2% were outside a vehicle at the time of injury. Outcomes illustrated the success of the interprofessional collaboration between trauma centers and a medicolegal death investigation agency that resulted in data useful for forensic nurses and pathologists documenting evidence, emergency and trauma responding personnel in patient priority stabilization, and injury prevention specialists for highway safety programs.

When falls become fatal-Clinical care sequence.

OBJECTIVES: This study encompassed fall-related deaths, including those who died prior to medical care, that were admitted to multiple healthcare institutions, regardless of whether they died at home, in long-term care, or in hospice. The common element was that all deaths resulted directly or indirectly from injuries sustained during a fall, regardless of the temporal relationship. This comprehensive approach provides an unusual illustration of the clinical sequence of fall-related deaths. Understanding this pathway lays the groundwork for identification of gaps in healthcare needs. DESIGN: This is a retrospective study of 2014 fall-related deaths recorded by one medical examiner's office (n = 511) within a larger dataset of all trauma related deaths (n = 1848). Decedent demographic characteristics and fall-related variables associated with the deaths were coded and described. RESULTS: Of those falling, 483 (94.5%) were from heights less than 10 feet and 394 (77.1%) were aged 65+. The largest proportion of deaths (n = 267, 52.3%) occurred post-discharge from an acute care setting. Of those who had a documented prior fall, 216 (42.3%) had a history of one fall while 31 (6.1%) had ≥2 falls prior to their fatal incident. For the 267 post-acute care deaths, 440 healthcare admissions were involved in their care. Of 267 deaths occurring post-acute care, 129 (48.3%) were readmitted within 30 days. Preventability, defined as opportunities for improvement in care that may have influenced the outcome, was assessed. Of the 1848 trauma deaths, 511 (27.7%) were due to falls of which 361 (70.6%) were determined to be preventable or potentially preventable. CONCLUSION: Our data show that readmissions and repeated falls are frequent events in the clinical sequence of fall fatalities. Efforts to prevent fall-related readmissions should be a top priority for improving fall outcomes and increasing the quality of life among those at risk of falling.

Predicting Daily Sheltering Arrangements among Youth Experiencing Homelessness Using Diary Measurements Collected by Ecological Momentary Assessment.

Youths experiencing homelessness (YEH) often cycle between various sheltering locations including spending nights on the streets, in shelters and with others. Few studies have explored the patterns of daily sheltering over time. A total of 66 participants completed 724 ecological momentary assessments that assessed daily sleeping arrangements. Analyses applied a hypothesis-generating machine learning algorithm (component-wise gradient boosting) to build interpretable models that would select only the best predictors of daily sheltering from a large set of 92 variables while accounting for the correlated nature of the data. Sheltering was examined as a three-category outcome comparing nights spent literally homeless, unstably housed or at a shelter. The final model retained 15 predictors. These predictors included (among others) specific stressors (e.g., not having a place to stay, parenting and hunger), discrimination (by a friend or nonspecified other; due to race or homelessness), being arrested and synthetic cannabinoids use (a.k.a., "kush"). The final model demonstrated success in classifying the categorical outcome. These results have implications for developing just-in-time adaptive interventions for improving the lives of YEH.