Stakeholder Engagement to Identify Implementation Strategies to Overcome Barriers to Delivering Chronic Pain Treatments: A NIDILRR and VA TBI Model Systems Collaborative Project.

OBJECTIVE: The purpose of this article is to illustrate the process of stakeholder-engaged intervention mapping approach to identify implementation strategies to overcome data-driven prioritized barriers to receiving chronic pain services for persons with traumatic brain injury (TBI). SETTING: Community. PARTICIPANTS: Healthcare providers (n = 63) with 2 or more years' experience treating persons with TBI, interviewed between October 2020 and November 2021 provided data for identification of barriers. TBI, chronic pain, and qualitative research subject matter experts (SMEs) participated in the mapping approach. DESIGN: Participatory-based research design, using descriptive and intervention mapping approaches. RESULTS: Four barriers to accessing chronic pain treatment by persons with TBI which emerged from provider interviews were prioritized for intervention mapping: cognitive deficits of patients (67%); patient comorbidities (63%); mental health and/or substance abuse issues (59%); and patient participation (62%). SMEs used prioritized barriers to develop 4 primary objectives and implementation strategies designed to: (1) engage consumers to validate and identify strategies; (2) tailor pain treatment and delivery to overcome barriers; (3) develop and disseminate guidelines and best practices when delivering care to persons with TBI to support spread; and (4) increase awareness, skills, and readiness of workforce to deliver pain treatment to persons with TBI. SMEs used an evidence-based approach to develop a mapping matrix of the prioritized barriers, implementation objectives, and aligned implementation strategies to impact change. CONCLUSION: Implementation science is needed to facilitate knowledge translation into practice for this complex population to overcome barriers to care. Implementation strategies to address barriers to accessing chronic pain care for individuals with TBI were chosen through a participatory approach to engaging SMEs to support these rehabilitation implementation efforts. Future work includes gathering input from individuals with TBI and chronic pain and to move the intervention (implementation) mapping matrix forward to inform future implementation research, policy, and practice.

Provider Perspectives of Facilitators and Barriers to Reaching and Utilizing Chronic Pain Healthcare for Persons With Traumatic Brain Injury: A Qualitative NIDILRR and VA TBI Model Systems Collaborative Project.

OBJECTIVE: To identify facilitators and barriers to reaching and utilizing chronic pain treatments for persons with traumatic brain injury (TBI) organized around an Access to Care framework, which includes dimensions of access to healthcare as a function of supply (ie, provider/system) and demand (ie, patient) factors for a specified patient population. SETTING: Community. PARTICIPANTS: Clinicians (n = 63) with experience treating persons with TBI were interviewed between October 2020 and November 2021. DESIGN: Descriptive, qualitative study. MAIN MEASURES: Semistructured open-ended interview of chronic pain management for persons with TBI. Informed by the Access to Care framework, responses were coded by and categorized within the core domains (reaching care, utilizing care) and relevant subdimensions from the supply (affordability of providing care, quality, coordination/continuity, adequacy) and demand (ability to pay, adherence, empowerment, caregiver support) perspective. RESULTS: Themes from provider interviews focused on healthcare reaching and healthcare utilization resulted in 19 facilitators and 9 barriers reaching saturation. The most themes fell under the utilization core domain, with themes identified that impact the technical and interpersonal quality of care and care coordination/continuity. Accessibility and availability of specialty care and use of interdisciplinary team that permitted matching patients to treatments were leading thematic facilitators. The leading thematic barrier identified primarily by medical providers was cognitive disability, which is likely directly linked with other leading barriers including high rates of noncompliance and poor follow-up in health care. Medical and behavioral health complexity was also a leading barrier to care and potentially interrelated to other themes identified. CONCLUSION: This is the first evidence-based study to inform policy and planning for this complex population to improve access to high-quality chronic pain treatment. Further research is needed to gain a better understanding of the perspectives of individuals with TBI/caregivers to inform interventions to improve access to chronic pain treatment for persons with TBI.

Provider Perceived Facilitators and Barriers to Identifying, Perceiving, and Seeking Healthcare for Chronic Pain After TBI: A Qualitative NIDILRR and VA TBI Model Systems Collaborative Project.

OBJECTIVE: Identify determinants to chronic pain healthcare for persons with traumatic brain injury (TBI) informed by an Access to Care Framework. Findings related to the Access Framework's core domains of identifying a need, perceptions of the need, and seeking healthcare are reported. SETTING: Community. PARTICIPANTS: Healthcare providers (n = 63) with 2 or more years of experience treating persons with TBI interviewed between October 2020 and November 2021. DESIGN: Descriptive, qualitative study. MAIN MEASURES: Semi-structured interviews with open-ended questions of chronic pain management for persons with TBI. Informed by the Access Framework, responses were coded by and categorized within the domains of identifying healthcare needs, perceptions of needs, and factors related to healthcare seeking from the supply and demand perspective. RESULTS: For the overall sample, 14 facilitators and 6 barriers were endorsed by more than 20% of the provider cohort. Top facilitators included on-site availability of needed resources and treatments (94%), adequate time and provider capability to ensure patient comprehension of diagnosis and treatment plans (83%), and establishing patient motivation and buy-in with the treatment plan (75%). Barriers most endorsed included policies impacting access (46%), wait times for services (41%), and patient uncertainty regarding telehealth commonly due to cognitive and physical challenges (37%). Unique determinants are reported across civilian versus Department of Veterans Affairs (VA) healthcare systems and different provider types. CONCLUSION: This is the first evidence-based study to inform policy and planning to improve access to high-quality chronic pain treatments for persons with TBI. Results will inform future interventions at the systems, patient, and policy levels of healthcare that can be tailored to healthcare settings (VA, Civilian) and types of providers (rehabilitation therapists, psychologists, and medical). Evidence-informed interventions may help minimize healthcare disparities experienced by persons with TBI and facilitate access to high-quality, evidence-informed chronic pain care.

Rehabilitation of Neuropsychiatric Symptoms in Patients With Long COVID: Position Statement.

Long COVID, a term used to describe ongoing symptoms after COVID-19 infection, parallels the course of other postviral syndromes. Neuropsychiatric symptoms of long COVID can be persistent and interfere with quality of life and functioning. Within the biopsychosocial framework of chronic illness, rehabilitation professionals can address the neuropsychiatric sequelae of long COVID. However, current practice models are not designed to address concurrent psychiatric and cognitive symptoms in adults living with long COVID. Thus, we present a biopsychosocial framework for long COVID and provide treatment strategies based on evidence from current literature of postviral chronic illness. These recommendations will guide rehabilitation professionals in identifying common neuropsychiatric symptoms in long COVID that can be targeted for intervention and addressing these symptoms via integrative interventions taking into account the biopsychosocial presentation of long COVID symptoms.

Relations Among Suicidal Ideation, Depressive Symptoms, and Functional Independence During the 10 Years After Traumatic Brain Injury: A Model Systems Study.

OBJECTIVE: To investigate relative causality in relations among suicidal ideation (SI), depressive symptoms, and functional independence over the first 10 years after traumatic brain injury (TBI). DESIGN: Prospective longitudinal design with data collected through the Traumatic Brain Injury Model Systems (TBIMS) network at acute rehabilitation hospitalization as well as 1, 2, 5, and 10 years after injury. SETTING: United States Level I/II trauma centers and inpatient rehabilitation centers with telephone follow-up. PARTICIPANTS: Individuals enrolled into the TBIMS National Database (N=9539) with at least 1 SI score at any follow-up data collection (72.1% male; mean age, 39.39y). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Patient Health Questionnaire-9 and FIM at years 1, 2, 5, and 10 post injury. RESULTS: A cross-lagged panel structural equation model, which is meant to indirectly infer causality through longitudinal correlational data, suggested that SI, depressive symptoms, and functional independence each significantly predicted themselves over time. Within the model, bivariate correlations among variables were all significant within each time point. Between years 1 and 2 and between years 2 and 5, depressive symptoms had a larger effect on SI than SI had on depressive symptoms. Between years 5 and 10, there was reciprocal causality between the 2 variables. Functional independence more strongly predicted depressive symptoms than the reverse between years 1 and 2 as well as years 2 and 5, but its unique effects on SI over time were extremely marginal or absent after controlling for depressive symptoms. CONCLUSIONS: A primary goal for rehabilitation and mental health providers should be to monitor and address elevated symptoms of depression as quickly as possible before they translate into SI, particularly for individuals with TBI who have reduced functional independence. Doing so may be a key to breaking the connection between low functional independence and SI.

CONNECT: A pragmatic clinical trial testing a remotely provided linkage to service coordination after hospitalization for TBI.

OBJECTIVE: To test whether a complex behavioral intervention delivered remotely to connect individuals to clinical resources after hospitalization for TBI improved their quality of life. DESIGN/METHODS: Community-based randomized pragmatic clinical trial. Main measures TBI-QOL, Activity Measure for Post-Acute Care (AM-PAC), Clinical Satisfaction and Competency Rating Scale. RESULTS: 332 individuals ≥18 years-old hospitalized for TBI in four upper Midwest states were randomized to Remote (n = 166) and Usual Care (n = 166) groups. The groups were equivalent and representative of their state population's racial and ethnic composition, age, and proportion living in rural communities. There were no significant differences within or between experimental groups over the study period in TBI-QOL t-scores. There was a significant improvement in AM-PAC Daily Activities within the Remote group and a significant between-group improvement in clinical satisfaction for the Remote group. CONCLUSION: Enrolling a representative, regional community-based sample of individuals with TBI can be successful, and delivering a customized complex behavioral intervention remotely is feasible. The overall lack of intervention effectiveness was likely due to enrolling individuals without pre-identified clinical needs, initiating intervention after the immediate post-acute phase when needs are often highest, inability to provide direct clinical care remotely, and potential lack of outcome measure responsiveness in our sample.

Identifying group-based patterns of suicidal ideation over the first 10 years after moderate-to-severe TBI.

OBJECTIVE: To identify group-based patterns in suicidal ideation (SI) over the first 10 years after traumatic brain injury (TBI). METHODS: Participants included 9539 individuals in the TBI Model Systems National Database who responded to Patient Health Questionnaire-9 Item 9 assessing SI at 1, 2, 5, and/or 10 years post-injury. A k-means cluster analysis was conducted to determine group-based patterns of SI, and pre-injury variables were compared with ANOVAs and chi-square tests. RESULTS: SI and attempts decreased over time. Four group-based patterns emerged: Low, increasing, moderate, and decreasing SI. The low SI group comprised 89% of the sample, had the highest pre-injury employment, fewer mental health vulnerabilities, least severe injuries, and were oldest. The increasing SI group had the most severe TBIs, were youngest, and disproportionately Black or Asian/Pacific Islander. CONCLUSION: These findings reinforce the importance of mental health and suicide risk assessment during chronic recovery from TBI.

The Association Between Community Participation and Social Internet Use Among Adults With Traumatic Brain Injury.

OBJECTIVE: To examine the association between social Internet use and real-world societal participation in survivors of moderate-severe traumatic brain injury. DESIGN: Prospective cross-sectional observational study. SETTING: Ten Traumatic Brain Injury Model Systems Centers. PARTICIPANTS: A total of 331 participants in the Traumatic Brain Injury Model Systems, interviewed at any follow-up year between April 2014 and March 2015. MAIN MEASURES: Survey on Internet use, including social media and other online socialization; Participation Assessment with Recombined Tools-Objective with separate analyses of Productivity, Social Relations, Out and About subscales; covariates included demographics, injury variables, and functional and emotional status at follow-up. RESULTS: Participants were classified as social Internet users (N = 232) or nonusers (N = 99). Users had significantly higher Participation Assessment with Recombined Tools-Objective Social Relations scores than nonusers. A similar finding pertained to Out and About scores, with the between-group difference significantly greater for those with greater depressive symptoms severity. Users and nonusers did not differ significantly on Productivity subscale. CONCLUSIONS: The positive association between social Internet use and real-world social participation suggests that people with traumatic brain injury do not use social media as an alternative to real-world socialization. Rather, it is likely that similar barriers and facilitators affect both online and real-world social participation following traumatic brain injury. Emotional function should be considered as a moderating factor in further studies.

Evidence-Based Cognitive Rehabilitation: Systematic Review of the Literature From 2009 Through 2014.

OBJECTIVES: To conduct an updated, systematic review of the clinical literature, classify studies based on the strength of research design, and derive consensual, evidence-based clinical recommendations for cognitive rehabilitation of people with traumatic brain injury (TBI) or stroke. DATA SOURCES: Online PubMed and print journal searches identified citations for 250 articles published from 2009 through 2014. STUDY SELECTION: Selected for inclusion were 186 articles after initial screening. Fifty articles were initially excluded (24 focusing on patients without neurologic diagnoses, pediatric patients, or other patients with neurologic diagnoses, 10 noncognitive interventions, 13 descriptive protocols or studies, 3 nontreatment studies). Fifteen articles were excluded after complete review (1 other neurologic diagnosis, 2 nontreatment studies, 1 qualitative study, 4 descriptive articles, 7 secondary analyses). 121 studies were fully reviewed. DATA EXTRACTION: Articles were reviewed by the Cognitive Rehabilitation Task Force (CRTF) members according to specific criteria for study design and quality, and classified as providing class I, class II, or class III evidence. Articles were assigned to 1 of 6 possible categories (based on interventions for attention, vision and neglect, language and communication skills, memory, executive function, or comprehensive-integrated interventions). DATA SYNTHESIS: Of 121 studies, 41 were rated as class I, 3 as class Ia, 14 as class II, and 63 as class III. Recommendations were derived by CRTF consensus from the relative strengths of the evidence, based on the decision rules applied in prior reviews. CONCLUSIONS: CRTF has now evaluated 491 articles (109 class I or Ia, 68 class II, and 314 class III) and makes 29 recommendations for evidence-based practice of cognitive rehabilitation (9 Practice Standards, 9 Practice Guidelines, 11 Practice Options). Evidence supports Practice Standards for (1) attention deficits after TBI or stroke; (2) visual scanning for neglect after right-hemisphere stroke; (3) compensatory strategies for mild memory deficits; (4) language deficits after left-hemisphere stroke; (5) social-communication deficits after TBI; (6) metacognitive strategy training for deficits in executive functioning; and (7) comprehensive-holistic neuropsychological rehabilitation to reduce cognitive and functional disability after TBI or stroke.

Injury Severity and Depressive Symptoms in a Post-acute Brain Injury Rehabilitation Sample.

This study explored the relationship between injury severity and depressive symptoms for treatment-seeking individuals with traumatic brain injury (TBI). The Mayo Classification System was used to classify TBI severity in 72 participants who completed the Patient Health Questionnaire at admission and at dismissal from rehabilitation. Patients with mild TBI reported more depressive symptoms than those with moderate or severe TBI at admission and at dismissal. Although injury severity groups differed by gender composition, gender had no effect on severity of depressive symptoms. All participants reported fewer depressive symptoms at dismissal from rehabilitation, including lower endorsement of dysphoria by discharge. Participants with mild TBI, however, continued to report depressive symptoms of a mild severity at dismissal, with residual problems with anhedonia. These findings underscore the benefit of interdisciplinary post-acute rehabilitation services for persons with TBI of any severity, including those with mild injury.

Longitudinal Examination of Resilience After Traumatic Brain Injury: A Traumatic Brain Injury Model Systems Study.

OBJECTIVES: To evaluate (1) the trajectory of resilience during the first year after a moderate-severe traumatic brain injury (TBI); (2) factors associated with resilience at 3, 6, and 12 months postinjury; and (3) changing relationships over time between resilience and other factors. DESIGN: Longitudinal analysis of an observational cohort. SETTING: Five inpatient rehabilitation centers. PARTICIPANTS: Patients with TBI (N=195) enrolled in the resilience module of the TBI Model Systems study with data collected at 3-, 6-, and 12-month follow-up. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Connor-Davidson Resilience Scale. RESULTS: Initially, resilience levels appeared to be stable during the first year postinjury. Individual growth curve models were used to examine resilience over time in relation to demographic, psychosocial, and injury characteristics. After adjusting for these characteristics, resilience actually declined over time. Higher levels of resilience were related to nonminority status, absence of preinjury substance abuse, lower anxiety and disability level, and greater life satisfaction. CONCLUSIONS: Resilience is a construct that is relevant to understanding brain injury outcomes and has potential value in planning clinical interventions.

Characteristics of Firearm Brain Injury Survivors in the Traumatic Brain Injury Model Systems (TBIMS) National Database: A Comparison of Assault and Self-Inflicted Injury Survivors.

OBJECTIVE: To characterize and compare subgroups of survivors with assault-related versus self-inflicted traumatic brain injuries (TBIs) via firearms at the time of inpatient rehabilitation and at 1-, 2-, and 5-year follow-up. DESIGN: Secondary analysis of data from the Traumatic Brain Injury Model Systems National Database (TBIMS NDB), a multicenter, longitudinal cohort study. SETTING: Retrospective analyses of a subset of individuals enrolled in the TBIMS NDB. PARTICIPANTS: Individuals 16 years and older (N=399; 310 via assault, 89 via self-inflicted injury) with a primary diagnosis of TBI caused by firearm injury enrolled in the TBIMS NDB. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Disability Rating Scale, Glasgow Outcome Scale-Extended, sociodemographic variables (sex, age, race, marital status), injury-related/acute care information (posttraumatic amnesia, loss of consciousness, time from injury to acute hospital discharge), and mental health variables (substance use history, psychiatric hospitalizations, suicide history, incarcerations). RESULTS: Individuals who survived TBI secondary to a firearm injury differed by injury mechanism (assault vs self-inflicted) on critical demographic, injury-related/acute care, and mental health variables at inpatient rehabilitation and across long-term recovery. Groups differed in terms of geographic area, age, ethnicity, education, marital status, admission Glasgow Coma Scale score, and alcohol abuse, suicide attempts, and psychiatric hospitalizations at various time points. CONCLUSIONS: These findings have implications for prevention (eg, mental health programming and access to firearms in targeted areas) and for rehabilitation planning (eg, by incorporating training with coping strategies and implementation of addictions-related services) for firearm-related TBI, based on subtype of injury.

Resilience Following Traumatic Brain Injury: A Traumatic Brain Injury Model Systems Study.

OBJECTIVE: To examine resilience at 3 months after traumatic brain injury (TBI). DESIGN: Cross-sectional analysis of an ongoing observational cohort. SETTING: Five inpatient rehabilitation centers, with 3-month follow-up conducted primarily by telephone. PARTICIPANTS: Persons with TBI (N=160) enrolled in the resilience module of the TBI Model System study with 3-month follow-up completed. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Connor-Davidson Resilience Scale. RESULTS: Resilience scores were lower than those of the general population. A multivariable regression model, adjusting for other predictors, showed that higher education, absence of preinjury substance abuse, and less anxiety at follow-up were significantly related to greater resilience. CONCLUSIONS: Analysis suggests that lack of resilience may be an issue for some individuals after moderate to severe TBI. Identifying persons most likely at risk for low resilience may be useful in planning clinical interventions.

Prevalence of suicidal behaviour following traumatic brain injury: Longitudinal follow-up data from the NIDRR Traumatic Brain Injury Model Systems.

OBJECTIVE: This study utilized the Traumatic Brain Injury Model Systems (TBIMS) National Database to examine the prevalence of depression and suicidal behaviour in a large cohort of patients who sustained moderate-to-severe TBI. METHOD: Participants presented to a TBIMS acute care hospital within 72 hours of injury and received acute care and comprehensive rehabilitation in a TBIMS designated brain injury inpatient rehabilitation programme. Depression and suicidal ideation were measured with the Patient Health Questionnaire (PHQ-9). Self-reported suicide attempts during the past year were recorded at each follow-up examination, at 1, 2, 3, 10, 15 and 20 years post-injury. RESULTS: Throughout the 20 years of follow-up, rates of depression ranged from 24.8-28.1%, suicidal ideation ranged from 7.0-10.1% and suicide attempts (past year) ranged from 0.8-1.7%. Participants who endorsed depression and/or suicidal behaviour at year 1 demonstrated consistently elevated rates of depression and suicidal behaviour 5 years after TBI. CONCLUSION: Compared to the general population, individuals with TBI are at greater risk for depression and suicidal behaviour many years after TBI. The significant psychiatric symptoms evidenced by individuals with TBI highlight the need for routine screening and mental health treatment in this population.

A randomized practical behavioural trial of curriculum-based advocacy training for individuals with traumatic brain injury and their families.

PRIMARY OBJECTIVE: To test whether a curriculum-based advocacy training programme improves advocacy behaviour when compared to a matched group engaged in self-directed advocacy activities. RESEARCH DESIGN: Community-based randomized practical behavioural trial. METHODS AND PROCEDURES: Adults with moderate-severe TBI 1 or more years post-injury and their family members were recruited in Minnesota (4 years), Iowa and Wisconsin (each 3 years) and randomized into a curriculum-based or self-directed advocacy training group. Both groups met on the same day, at separate locations in the same city, once per month for 4 consecutive months. MAIN OUTCOMES AND RESULTS: Pre-post written and video testimony were rated using the Advocacy Behaviour Rating Scale (ABRS). Mean ABRS scores increased after intervention in both groups (curriculum n = 129, self-directed n = 128), but there was no significant difference in this increase between groups. When groups were combined, a significant pre-post improvement in mean ABRS scores was observed. CONCLUSIONS: Curriculum-based advocacy training was not superior to a self-directed approach in improving ABRS scores. A significant improvement in expression of an advocacy message was observed when intervention groups were combined. These findings suggest that bringing together like-minded motivated individuals is more important than programme structure or content in changing advocacy behaviour.

Multidisciplinary Management Strategies for Long COVID: A Narrative Review.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused millions of infections to date and has led to a worldwide pandemic. Most patients had a complete recovery from the acute infection, however, a large number of the affected individuals experienced symptoms that persisted more than 3 months after diagnosis. These symptoms most commonly include fatigue, memory difficulties, brain fog, dyspnea, cough, and other less common ones such as headache, chest pain, paresthesias, mood changes, muscle pain, and weakness, skin rashes, and cardiac, endocrine, renal and hepatic manifestations. The treatment of this syndrome remains challenging. A multidisciplinary approach to address combinations of symptoms affecting multiple organ systems has been widely adopted. This narrative review aims to bridge the gap surrounding the broad treatment approaches by providing an overview of multidisciplinary management strategies for the most common long COVID conditions.

Caregiver resilience following traumatic brain injury: Findings at six months postinjury.

PURPOSE/OBJECTIVE: Examine contributors to resilience among caregivers of individuals who have sustained a moderate-to-severe traumatic brain injury (TBI), with the goal of identifying important targets for an intervention to improve caregiver resilience as well as outcomes for people with TBI. RESEARCH METHOD/DESIGN: Participants were adult caregivers (n = 176) and individuals with TBI who required inpatient rehabilitation at six TBI Model System sites. Measures included the Connor-Davidson Resilience Scale-10, Family Needs Questionnaire, Zarit Burden Interview, Patient Health Questionnaire-9, and Generalized Anxiety Disorder-7. Data were collected between September 2018 and June 2021. RESULTS: Caregivers endorsed levels of personal resilience that were comparable to norms for community samples and slightly higher than groups under stress or with medical illness. Reports of the burden associated with the caregiving role were relatively low, as was reported psychological distress. In a multivariable model, higher proportions of met emotional support needs were associated with increased resilience. CONCLUSIONS/IMPLICATIONS: Resilience may be strengthened by emotional support networks, including friends or family who may not already be directly involved in the provision of care. Supporting engagement with community agencies, peer mentors, or other informal resources within the family system that provide emotional support may bolster resilience outcomes for caregivers. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Evolution of rehabilitation services in response to a global pandemic: reflection on opportunities and challenges ahead.

The rapidly evolving COVID-19 public health emergency has disrupted and challenged traditional healthcare, rehabilitation services, and treatment delivery worldwide. This perspective paper aimed to unite experiences and perspectives from an international group of rehabilitation providers while reflecting on the lessons learned from the challenges and opportunities raised during the COVID-19 pandemic. We discuss the global appreciation for rehabilitation services and changes in access to healthcare, including virtual, home-based rehabilitation, and long-term care rehabilitation. We illustrate lessons learned by highlighting successful rehabilitation approaches from the US, Belgium, and Japan.

Using Machine Learning to Examine Suicidal Ideation After Traumatic Brain Injury: A Traumatic Brain Injury Model Systems National Database Study.

OBJECTIVE: The aim of the study was to predict suicidal ideation 1 yr after moderate to severe traumatic brain injury. DESIGN: This study used a cross-sectional design with data collected through the prospective, longitudinal Traumatic Brain Injury Model Systems network at hospitalization and 1 yr after injury. Participants who completed the Patient Health Questionnaire-9 suicide item at year 1 follow-up ( N = 4328) were included. RESULTS: A gradient boosting machine algorithm demonstrated the best performance in predicting suicidal ideation 1 yr after traumatic brain injury. Predictors were Patient Health Questionnaire-9 items (except suicidality), Generalized Anxiety Disorder-7 items, and a measure of heavy drinking. Results of the 10-fold cross-validation gradient boosting machine analysis indicated excellent classification performance with an area under the curve of 0.882. Sensitivity was 0.85 and specificity was 0.77. Accuracy was 0.78 (95% confidence interval, 0.77-0.79). Feature importance analyses revealed that depressed mood and guilt were the most important predictors of suicidal ideation, followed by anhedonia, concentration difficulties, and psychomotor disturbance. CONCLUSIONS: Overall, depression symptoms were most predictive of suicidal ideation. Despite the limited clinical impact of the present findings, machine learning has potential to improve prediction of suicidal behavior, leveraging electronic health record data, to identify individuals at greatest risk, thereby facilitating intervention and optimization of long-term outcomes after traumatic brain injury.