Background: Traumatic brain injury (TBI) patients exhibit variable post-injury recovery trajectories. Days at Home (DAH) is a patient-centered measure that captures healthcare transitions and offers a more nuanced understanding of recovery. Here, we use DAH to characterize longterm recovery trajectories for moderate to severe TBI (msTBI) survivors. Methods: This multicenter retrospective cohort study utilized population health data from Ontario to identify adults sustaining isolated msTBI hospitalized between 2009-2021. DAH were calculated in distinct 30-day intervals from index admission to 3 years post-injury; latent class mixed modeling identified unique recovery trajectories and trajectory attributes were quantified. Results: There were 2,510 patients eligible for latent class analysis. Four DAH trajectories were identified: early recovery (69.9%), intermediate recovery (11.4%), late recovery (2.9%), and poor recovery (15.8%). Patients in the poor recovery group were older, more frail, and had lower admission GCS scores, while those in early recovery exhibited lower acute care needs. Intermediate and late recovery groups exhibited protracted transitions home, with near-complete reintegration by 24 months. A prediction model distinguished unfavorable trajectories with good accuracy (C-index=0.824). Conclusions: Despite high initial institutional care requirements, 85% of patients reintegrated into the community within three years of msTBI. These findings shed light on post-injury care requirements for brain-injured patients.

Background: Artificial intelligence (AI) holds promise to predict outcomes for patients sustaining moderate to severe traumatic brain injury (msTBI). This systematic review sought to identify studies utilizing AI-based methods to predict mortality and functional outcomes after msTBI, where prognostic uncertainty is highest. Methods: The APPRAISE-AI quantitative evidence appraisal tool was used to evaluate methodological quality of included studies by determining overall scores and domain-specific scores. We constructed a multivariable linear regression model using study sample size, country of data collection, publication year and journal impact factor to quantify associations with overall APPRAISE-AI scores. Results: We identified 38 studies comprising 591,234 patients with msTBI. Median APPRAISE-AI score was 45.5 (/100 points), corresponding to moderate study quality. There were 13 low-quality studies (34%) and only 5 high-quality studies (13%). Weakest domains were methodological conduct, robustness of results and reproducibility. Multivariable linear regression highlighted that higher journal impact factor, larger sample size, more recent publication year and use of data that were collected in a high-income country were associated with higher APPRAISE-AI overall scores. Conclusions: We identified several study weaknesses of existing AI-based prediction models for msTBI; this work highlights methodological domains that require quality improvement to ultimately ensure safety and effiicacy of clinical AI models.

The proliferation of Artificial Intelligence (AI) is significantly transforming conventional legal practice. The integration of AI into legal services is still in its infancy and faces challenges such as privacy concerns, bias, and the risk of fabricated responses. This research evaluates the performance of the following AI tools: (1) ChatGPT-4, (2) Copilot, (3) DeepSeek, (4) Lexis+ AI, and (5) Llama 3. Based on their comparison, the research demonstrates that Lexis+ AI outperforms the other AI solutions. All these tools still encounter hallucinations, despite claims that utilizing the Retrieval-Augmented Generation (RAG) model has resolved this issue. The RAG system is not the driving force behind the results; it is one component of the AI architecture that influences but does not solely account for the problems associated with the AI tools. This research explores RAG architecture and its inherent complexities, offering viable solutions for improving the performance of AI-powered solutions.

Background: Despite the utility of administrative health data, there remains a lack of patient-centered outcome measures to meaningfully capture morbidity after traumatic brain injury (TBI). We sought to characterize and validate days at home (DAH) as a feasible measure to assess population-level moderate to severe TBI (msTBI) outcomes and health resource utilization. Methods: We utilized linked health administrative data sources to identify adults with msTBI patients presenting to trauma centers in Ontario injured between 2009-2021. DAH at 180 days reflects the total number of days spent alive and at home excluding the days spent institutionalized in acute care, rehabilitation, inpatient mental health settings or post-acute readmissions. Construct and predictive validity were determined; we additionally estimated minimally important difference (MID) in DAH180days. Results: There were 6340 patients that met inclusion criteria. Median DAH180days were 70 days (interquartile range 0-144). Increased health resource utilization at baseline, older age, increasing cranial injury severity and major extracranial injuries were significantly associated with fewer DAH180days. DAH180days was correlated to DAH counts at 1-3 years. The average MID estimate from anchor-based and distribution-based methods was 18 days. Conclusions: We introduce DAH180days as a feasible and sufficiently responsive patient-centered outcome measure with construct, predictive and face validity in an msTBI population.

Background: Employment and personal income loss after traumatic brain injury (TBI) is a major source of post-injury stress and barrier to societal reintegration for affected patients. We sought to quantify the labor market implications for tax-filing adult TBI survivors. Methods: We performed a matched difference-in-difference analysis using a national retrospective cohort of working adult TBI survivors injured between 2007-2017. Linear and logistic mixed effects regressions were used to estimate the magnitude of personal income loss and proportion of patients displaced from the workforce in the three post-injury years (Y+1 to Y+3). Results: Among 18,050 patients identified with TBI, the adjusted average loss of personal annual income was $-7,635 dollars in Y+1 and $-5,000 in Y+3. An additional -7.8% individuals were newly unemployed compared to the pre-injury baseline. For mild, moderate, and severe TBI subgroups, income loss was $-3354, $-6750, and $-17375 respectively in Y+3; the proportion of newly unemployed individuals in Y+3 was 5.8%, 9.2%, and 20% lower than baseline. We estimated 500 million dollars of incurred labor markets losses related to TBI in Canada. Conclusions: This work represents the first national cohort data quantifying the labor market implications of TBI. These results may be used to inform post-injury care pathways and vocational rehabilitation.

In 2016, the National Center for Advancing Translational Science launched the Trial Innovation Network (TIN) to address barriers to efficient and informative multicenter trials. The TIN provides a national platform, working in partnership with 60+ Clinical and Translational Science Award (CTSA) hubs across the country to support the design and conduct of successful multicenter trials. A dedicated Hub Liaison Team (HLT) was established within each CTSA to facilitate connection between the hubs and the newly launched Trial and Recruitment Innovation Centers. Each HLT serves as an expert intermediary, connecting CTSA Hub investigators with TIN support, and connecting TIN research teams with potential multicenter trial site investigators. The cross-consortium Liaison Team network was developed during the first TIN funding cycle, and it is now a mature national network at the cutting edge of team science in clinical and translational research. The CTSA-based HLT structures and the external network structure have been developed in collaborative and iterative ways, with methods for shared learning and continuous process improvement. In this paper, we review the structure, function, and development of the Liaison Team network, discuss lessons learned during the first TIN funding cycle, and outline a path toward further network maturity.

An understanding of child psychopathology and resilience requires attention to the nested and interconnected systems and contexts that shape children’s experiences and health outcomes. In this study, we draw on data from the National Survey of Children’s Health, 2016 to 2021 (n = 182,375 children, ages 3– to 17 years) to examine associations between community social capital and neighborhood resources and children’s internalizing and externalizing problems, and whether these associations were moderated by experiences of racial discrimination. Study outcomes were caregiver-report of current internalizing and externalizing problems. Using logistic regression models adjusted for sociodemographic characteristics of the child and household, higher levels of community social capital were associated with a lower risk of children’s depression, anxiety, and behaviors. Notably, we observed similar associations between neighborhood resources and child mental health for depression only. In models stratified by the child’s experience of racial/ethnic discrimination, the protective benefits of community social capital were specific to those children who did not experience racial discrimination. Our results illustrate heterogeneous associations between community social capital and children’s mental health that differ based on interpersonal experiences of racial/ethnic discrimination, illustrating the importance of a multilevel framework to promote child wellbeing.

Background: We aimed to develop an efficient and reliable artificial intelligence solution to automate prediction of neurosurgical intervention using acute traumatic brain injury computed tomography (CT) scans. Methods: TBI patients were identified from 2005 - 2022 at a Level 1 Canadian trauma center. Model training, validation, and testing was performed using head CT scans with patient-level labels corresponding to whether the patient received neurosurgical intervention. The finalized model was then deployed in a simulated prospective fashion on all TBI patients presenting to our center over an 18-month epoch. Results: 2,806 TBI scans were utilized for development of the Automated Surgical Intervention Support Tool (ASIST-TBI). 612 additional consecutive scans were used for simulated prospective model deployment. Prediction of neurosurgical intervention exhibited an area under receiver operating curve (AUC) of 0.92, accuracy of 0.87, sensitivity of 0.87, and specificity of 0.88 on the test dataset. On simulated prospective data, the results were: AUC 0.89, sensitivity 0.85, specificity 0.84 and accuracy of 0.84. Conclusions: We demonstrate the development and validation of ASIST-TBI, a machine learning model that accurately predicts whether TBI patients will need neurosurgical intervention. This model has potential application to optimize decision support and province-wide efficiency of inter-facility TBI triage to tertiary care centers.

Background: Despite growing evidence for early surgical decompression for traumatic cervical spinal cord injury(tCSCI) patients, controversy surrounds the efficacy of early surgical decompression on patients with a complete (ASIA A) cervical injury. Methods: Patients with ASIA A cervical tCSCI were isolated from 4 prospective, multi-center datasets. Patients who had a Glasgow coma scale of less than 13, were over the age of 70 or under 16 were excluded. Significant gain was defined to include those that recovered more than two muscle groups (greater than 3/5 power) below their level of injury. Analysis of variance (ANOVA) was then done to compare significant gain over the 1 year follow-up period for patients with and without early decompressive surgery (<24hrs). Results: We identified 420 cervical ASIA A tCSCI patients. The mean number of muscle groups gained was 2.69 (SD 2.3.12) for those who had early surgery compared to 2.37 (SD 3.38) for those with late surgery. Of those patients who had early surgery 39.67% had a significant improvement vs. 28.76% of those who did not have early surgery (P = 0.030). Conclusions: For the first time, we have shown a clear therapeutic benefit of early surgical decompression within 24 hrs in ASIA A tCSCI patients.

Background: The semantic variant of primary progressive aphasia (svPPA) is a form of dementia, mainly featuring language impairment, for which the extent of white matter (WM) damage is less described than its associated grey matter (GM) atrophy. Our study aimed to characterise the extent of this damage using a sensitive and unbiased approach. Methods: We conducted a between-group study comparing 10 patients with a clinical diagnosis of svPPA, recruited between 2011 and 2014 at a tertiary reference centre, with 9 cognitively healthy, age-matched controls. From diffusion tensor imaging (DTI) data, we extracted fractional anisotropy (FA) values using a tract-based spatial statistics approach. We further obtained GM volumetric data using the Freesurfer automated segmentation tool. We compared both groups using non-parametric Wilcoxon rank-sum tests, correcting for multiple comparisons. Results: Demographic data showed that patients and controls were comparable. As expected, clinical data showed lower results in svPPA than controls on cognitive screening tests. Tractography showed impaired diffusion in svPPA patients, with FA mostly decreased in the longitudinal, uncinate, cingulum and external capsule fasciculi. Volumetric data show significant atrophy in svPPA patients, mostly in the left entorhinal, amygdala, inferior temporal, middle temporal, superior temporal and temporal pole cortices, and bilateral fusiform gyri. Conclusions: This syndrome appears to be associated not only with GM but also significant WM degeneration. Thus, DTI could play a role in the differential diagnosis of atypical dementia by specifying WM damage specific to svPPA.