Abstract
Background: Traumatic brain injury (TBI) in older adults is a growing public health concern, yet outcome research has primarily focused on moderate-to-severe injuries or aggregated severity groups. Small-volume intracranial lesions, subdural hematomas (SDH ≤ 8mm), epidural hematomas (EDH ≤ 8mm), and cerebral contusions (≤ 2cm), are often classified as mild, but this may underestimate the functional burden in aging populations. Evidence is lacking on discharge outcomes in patients with isolated small-volume TBIs, particularly across age strata.
Objective: To evaluate age-related differences in hospital discharge disposition, specifically to rehabilitation or hospice care, among adults aged ≥40 years with isolated, small-volume blunt TBIs, stratified by TBI type and age group.
Methods: A retrospective cohort study was conducted using American College of Surgeons-Trauma Quality Programs-Participant Use Files (ACS-TQIP-PUF) data (2017–2022). Adults ≥40 years with isolated blunt TBIs, without skull fractures or polytrauma, were included. Lesion size was defined radiographically via abbreviated injury scale codes. Logistic regression models assessed associations between age group (middle-aged [40–64] vs. elderly [≥65]) and discharge disposition, adjusting for TBI subtype, comorbidities, injury severity, and facility factors.
Results: Among 135,343 patients with isolated small-volume TBIs, elderly individuals were more likely to be discharged to rehabilitation (EDH: OR 1.601, 95% CI 1.234–2.076, P <0.001) and hospice care (SDH: OR 2.184, 95% CI 1.630–2.927, P <0.001) compared to middle-aged patients. These age-based differences persisted across TBI subtypes.
Conclusion: Despite being classified as “mild” by volume, small TBIs are associated with significant discharge care needs among older adults. Elderly patients face markedly higher odds of requiring rehabilitation or end-of-life care. These findings challenge traditional severity frameworks and highlight the need for age-adapted discharge planning in the context of small-volume intracranial injuries.
- Epidural hematoma
- Hospice care
- Hospital discharge disposition
- Retrospective cohort study
- Subdural hematoma
- Traumatic brain injury
Traumatic brain injury (TBI) is a major contributor to hospitalization, disability, and death in older adults.1 While much of the existing literature has focused on moderate-to-severe injuries, an important but underrecognized population includes patients with small-volume intracranial hemorrhages, such as subdural hematomas (SDH ≤ 8 mm), epidural hematomas (EDH ≤ 8 mm), and contusions (≤ 2 cm).2 These injuries are often categorized as “mild” based on radiographic size alone, yet they may carry significant clinical consequences, particularly among elderly patients with reduced physiological reserve, brain atrophy, and multimorbidity.3,4 Current prognostic frameworks typically aggregate TBIs by severity using the Glasgow Coma Scale (GCS) or abbreviated injury score (AIS), without accounting for lesion-specific size or age-related vulnerability. As a result, patients with small-volume lesions may be misclassified as low-risk, leading to under-preparedness in discharge planning. Moreover, there is little literature stratifying outcomes in this population by age group and TBI subtype. Understanding whether small-volume TBIs in elderly patients result in disproportionate functional decline or palliative care needs is essential to guide post-acute care decisions.
Hospital discharge disposition is a critical marker of recovery trajectory.5 Discharge to inpatient rehabilitation suggests ongoing functional impairment, while hospice referral reflects poor prognosis and limited recovery potential. Prior studies have shown elderly patients with TBI are more likely to experience poor outcomes; however, few studies have examined whether isolated small-volume TBIs in this population confer higher odds of discharge to rehabilitation or hospice.6,7 This study addresses this gap by using a large, multi-center trauma dataset to assess whether adults aged ≥65 years with small-volume isolated blunt TBIs are more likely to require rehabilitation or hospice care than middle-aged adults (40–64 years). Through stratifying by both TBI subtype (SDH, EDH, contusion) and age group, we provide a more nuanced analysis of how seemingly minor intracranial injuries may disproportionately affect elderly patients. These findings aim to inform age-specific discharge strategies and challenge prevailing assumptions about the benign nature of small-volume TBIs.
Objective
The primary objective of this study was to evaluate whether elderly patients (≥65 years) with isolated small-volume blunt TBIs, defined as SDH or EDH ≤8 mm or contusions ≤2 cm, are more likely to be discharged to rehabilitation or hospice care compared to middle-aged adults (40–64 years). By stratifying outcomes by both age and TBI subtype, this study aimed to challenge conventional assumptions about the benign nature of small-volume lesions and identify age-related disparities in post-acute care needs.
Methods
Study design and data source
This retrospective cohort study utilized the American College of Surgeons-Trauma Quality Programs-Participant Use Files (ACS-TQIP-PUF) dataset from 2017 to 2022. The study was deemed exempt from oversight by the Institutional Review Board [RHO10224] in compliance with current regulations with waived consent. Results were reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (see Supplementary Table 1; available online). The ACS-TQIP-PUF dataset includes anonymized data from over 796 trauma centers across the United States, encompassing Level I–V and undesignated centers, with all records submitted to the National Trauma Data Bank (NTDB).8 Data in the ACS-TQIP PUF are abstracted by trained trauma registrars using the American College of Surgeons National Trauma Data Standard (NTDS) and coded according to the Abbreviated Injury Scale (AIS 2015 revision). Intracranial hemorrhage type and size were derived directly from AIS codes linked to computed tomography (CT) imaging obtained within the first 24 hours. The specific AIS codes used to classify SDH ≤8 mm, EDH ≤8 mm, and contusions ≤2 cm are provided in Supplementary Table 2 (available online), which serves as the code-mapping tool for TBI size and type classifications used in this study. The study retrospectively analyzed various factors, including TBI type and size, demographics (age, sex, race, ethnicity, insurance status, center specifics, and transportation mode), injury characteristics (midline shift, injury type, Injury Severity Score [ISS], AIS body regions, Glasgow Coma Scale [GCS], and injury mechanism), neurosurgical interventions (intracranial pressure [ICP] monitoring and craniotomy), venous thrombosis prophylaxis low molecular weight heparin (VTEp LMWH), comorbidities, and hospital disposition (rehabilitation and hospice care).
Study population and eligibility criteria
The study population consisted of adult patients (≥40 years) with isolated blunt TBI, as identified in the ACS-TQIP dataset (see Figure 1). Patients with skull fractures were excluded to reduce heterogeneity and isolate the effect of intracranial hemorrhage volume on discharge outcomes. Skull fractures may independently alter prognosis and clinical management, particularly regarding neurosurgical intervention and bleeding progression, thereby confounding associations between TBI subtype and disposition. Patients with AIS injury scores >2 in other body regions (face, chest, abdomen/pelvis, extremities, or external) were also excluded to ensure a cohort without moderate-to-severe polytrauma.
Participant Flowchart Among Middle Aged to Elderly Adults (≥ 40 Years) Who Had Smaller Isolated Blunt Traumatic Brain Injury (SDH, SAH, EDH) Without a Skull Fracture.
Note: ACS-TQIP=American College of Surgeons Trauma Quality Improvement; TBI=traumatic brain injury; ISS=injury severity score; AIS=abbreviated injury severity score; SDH=subdural hematoma; EDH=epidural hematoma; SAH=subarachnoid hemorrhage
TBI type and lesion size were identified using AIS codes, which reflect radiographic findings from CT scans obtained within the first 24 hours (see Supplementary Table 2, available online).9 The study focused on small-volume injuries, defined as SDH or EDH ≤8 mm in thickness and contusions ≤2 cm in diameter. These thresholds align with the Modified Berne Norwood Criteria, which were originally developed to stratify patients by risk of hemorrhage progression and safety of early anticoagulation.10 The “low rebleeding risk” group, characterized by small, isolated hemorrhages without midline shift or neurological deterioration, has been shown to have a low likelihood of clinical decompensation and is frequently used as a surrogate for mild TBI in both observational studies and clinical protocols. By limiting inclusion to this low-risk subset, our study isolates a population traditionally considered low severity but potentially underrecognized in terms of post-discharge care needs. Patients were then categorized into six exposure groups based on TBI type (SDH, EDH, contusion) and age group (middle-aged [40–64 years] and elderly [≥65 years]).
Outcomes of interest
The primary outcomes were hospital discharge to rehabilitation and hospice care among patients aged ≥40 years with isolated small-volume blunt TBIs. Rehabilitation discharge indicates ongoing functional impairment requiring post-acute therapy, while hospice discharge reflects limited recovery potential. Secondary analyses examined whether TBI subtype, comorbidities, injury severity, and clinical management factors (e.g., anticoagulation use, trauma center level) influenced these outcomes.
Statistical analysis
Descriptive statistics analyzing characteristics of the study sample were reported as means, medians, and frequencies. Continuous variables were compared using the Kruskal-Wallis test, and categorical variables using Pearson’s chi-square test, as appropriate, in R-4.4.1 (R Core Team, 2023). Predictor variables for the regression models were selected based on multicollinearity tolerance (>0.5) to ensure model stability and interpretability. In accordance with contemporary guidance distinguishing descriptive, predictive, and causal modeling approaches, the logistic regression models in this study were constructed as predictive models, aimed at estimating the probability of discharge to rehabilitation or hospice based on patient demographic, clinical, and facility characteristics, rather than inferring causal effects of these variables. Each disposition outcome was modeled as a binary variable. For the rehabilitation model, the dependent variable was coded as rehabilitation = 1 versus all other discharge locations = 0 (including home, skilled nursing facility, long-term acute care, hospice, or death). For the hospice model, the dependent variable was coded as hospice = 1 versus all other discharge locations = 0. Rehabilitation and hospice outcomes were not treated as mutually exclusive categories; rather, each outcome was evaluated in a separate logistic regression model. The primary independent variable was age group (middle-aged [40–64 years] vs. elderly [≥65 years]), stratified by TBI subtype. In each model, the middle-aged group served as the reference category, allowing for estimation of the relative odds of discharge to rehabilitation or hospice for elderly patients. Secondary covariates included sex, race, ethnicity, mode of transportation, GCS, trauma center verification level, anticoagulation use (LMWH), neurosurgical interventions, and comorbidities. Covariates were selected based on clinical relevance and model stability rather than significance testing. Multicollinearity was assessed using tolerance (>0.5), and injury type (SDH, EDH, contusion) and age group (40–64 vs ≥65) were modeled as categorical factor variables consistent with Figure 1. Reference categories are identified in Tables 1 and 2. Model adequacy was evaluated using multicollinearity diagnostics, McFadden’s pseudo-R2 (reported by SAS as the max-rescaled R2), and discrimination metrics (ROC/AUC), with results reported in Supplementary Table 3 (available online) and Tables 1 and 2. All regression analyses were conducted using SPSS version 28 (Armonk, NY). Statistical significance was defined as P < .05.
Multivariable logistic regression analysis of factors associated with discharge to inpatient rehabilitation among adults aged ≥40 years with isolated small-volume blunt traumatic brain injury (subdural hematoma ≤8 mm, epidural hematoma ≤8 mm, or contusion ≤2 cm) and no skull fracture.
Multivariable logistic regression analysis of factors associated with discharge to hospice among adults aged ≥40 years with isolated small-volume blunt traumatic brain injury (subdural hematoma ≤8 mm, epidural hematoma ≤8 mm, or contusion ≤2 cm) and no skull fracture.
Results
Section I: Descriptive analysis of the study cohort
This retrospective cohort study utilized data from the ACS-TQIP spanning 2017–2022. The initial dataset included 6,714,002 patient records. After excluding cases with missing AIS, ISS, and ISS Region values, the dataset was reduced to 3,013,607 adult patients. To isolate TBI cases, 2,393,462 patients with moderate-to-severe polytrauma were excluded, resulting in 620,145 eligible cases. Further exclusions were applied to remove 31,694 penetrating injuries and 136,697 cases with skull fractures, refining the study population to 451,754 patients. To focus on middle-aged and elderly populations, 55,154 young adults (18–39 years) were excluded, leaving a cohort of 396,600 patients aged 40 years and older. To ensure analysis of smaller TBIs, 261,257 patients with larger injuries were excluded, resulting in a final study population of 135,343 middle-aged and elderly adults with moderate-to-critical TBI. Patients were then categorized into six exposure groups based on TBI type and age: SDH ≤8mm (24% in 40–64 years, 68% in ≥65 years), EDH ≤8mm (0.6% in 40–64 years, 1.3% in ≥65 years), and isolated contusion ≤2cm (1.9% in 40–64 years, 2.8% in ≥65 years). This structured selection process ensured a homogenous study population, allowing for a robust analysis of the impact of TBI size and type on midline shift and neurosurgical interventions across different age groups.
Descriptive statistics of the total sample
Descriptive statistics were performed for the entire study sample to characterize the demographic, clinical, and injury-related features of patients with isolated small-volume blunt TBIs. The complete results of these descriptive analyses are provided in Supplementary Tables 4–6 (available online), including breakdowns by age group, TBI subtype, comorbidities, and injury severity indicators. The median age of the sample was 74 years (IQR = 19). Most patients were Caucasian (81%) males (52%) with Medicare (63%) as the primary insurer. The majority arrived via ground ambulance (79%), with 41% transported to Level I trauma centers. Pre-existing anticoagulation therapy was present in 24% of patients. Common comorbidities included hypertension (64%), diabetes (30%), congestive heart failure (8%), renal failure (3%), and cerebrovascular disease (6%). The median Injury Severity Score (ISS) was 14 (moderate injury), and the maximum head AIS score was 3 (serious TBI). The mean GCS score was 14, consistent with mild TBI classification. Falls were the predominant mechanism of injury (82%). Overall, 9.66% of patients were discharged to rehabilitation and 2% to hospice care.
Non-parametric and chi-square tests for sample
Statistical analyses were conducted using non-parametric tests and chi-square tests to evaluate associations between patient characteristics and mortality outcomes. Due to the non-normal distribution of continuous variables, the Kruskal-Wallis test was applied, while Pearson’s chi-square test was used to assess categorical variable distributions. These tests helped identify differences across TBI subtypes before conducting logistic regression analysis. Elderly patients (≥ 65 years) with isolated small SDH (≤ 8 mm) were (48%) male Caucasian (84%) with a Medicare (80%) insurance payor (all values P < .01). Elderly TBI patients with smaller SDH were most likely to have pre-existing advanced directive (7%), chronic obstructive pulmonary disease (10%), diabetes (30%), hypertension (71%), and renal failure (3%; all values P < .01). Isolated smaller SDH who were elderly typically had a mean ISS of 14 (moderate; SD = 4, P < .01) and AIS max head mean of 3.4 (serious; SD =.5, P < .01). This SDH subgroup typically hospital discharged to hospice at a rate of 2% and rehabilitation at 10% (all values P < .01). Among elderly patients with smaller EDH, the median age was most likely to be older at 78 years (IQR = 12, P < .01). This group was more likely to have pre-existing anticoagulation therapy (29%, P < .01) and renal failure (3%, P < .01). Elderly (≥ 65 years) smaller EDH had a mean ISS of 13 (moderate; SD = 6, P < .01) and AIS max head mean of 3.1 (serious; SD =10, P < .01). This SDH subgroup typically hospital discharged to hospice at a rate of 2% and rehabilitation at 11% (all values P < .01).
Section II: Predictors of discharge status
All elderly (≥65 years) patients were most likely to hospital discharge to rehabilitation with smaller EDH ≤8mm (OR 1.601, 95% CI 1.234–2.076, P <0.001; Table 1) most associated with hospital discharge disposition to rehabilitation. All elderly groups ≥65 years were more likely to hospital discharge disposition to hospice care with elderly SDH (≤8 mm) most likely compared to the adult cohort (40-64 years; OR 2.184, 95% CI 1.630–2.927, P <0.001; Table 2).
Discussion
This study highlights an underrecognized dimension of post-acute care needs in patients with small-volume TBIs, particularly among the elderly. Despite the radiographic classification of SDH, EDH, and contusions as “minor” injuries based on volume, our findings reveal elderly patients (≥65 years) are more likely to require hospital discharge to rehabilitation or hospice care than their middle-aged counterparts. These results challenge the assumption that small-volume TBIs invariably carry a benign clinical trajectory and underscore the need to reconsider how “mild” injuries are operationalized in trauma care, particularly in the context of aging physiology and complex comorbidity profiles.
One critical contribution of this study is the demonstration that discharge disposition, often used as a surrogate for functional status and recovery potential, is not solely driven by lesion size or injury classification, but it is profoundly shaped by age-related vulnerability. While intracranial lesions ≤8 mm or contusions ≤2 cm may meet criteria for low rebleeding risk under the Modified Berne Norwood framework, their impact on older adults can be disproportionately severe due to age-associated factors such as brain atrophy, reduced neuroplasticity, impaired cerebral autoregulation, and higher baseline functional dependence.11 These physiological and psychosocial contributors are not reflected in conventional severity metrics, which may lead to systematic under recognition of discharge planning needs in this population.
Importantly, this analysis avoids a homogenized view of older adults. The elderly are a diverse group, with wide variability in baseline function, cognition, resilience, and social support. While age ≥65 remains a standard threshold in research and policy, clinicians and researchers must be cautious not to overgeneralize or assume uniform decline. Functional outcomes in elderly patients following TBI are not determined by age alone but by the interplay of pre-injury status, comorbidities, injury mechanism, and access to timely rehabilitative care. A more individualized approach, potentially incorporating frailty and cognitive screening, may improve prognostication and resource allocation.
Another key consideration is the global variability in post-TBI rehabilitation systems. In the United States, discharge to inpatient rehabilitation typically reflects a transition to a distinct facility with defined functional thresholds. In contrast, many other countries deliver rehabilitation services within the hospital or community-based settings, potentially blurring distinctions between discharge destinations. These structural differences must be considered when interpreting and applying our findings internationally. Furthermore, access to rehabilitation resources may vary by region, insurance coverage, or hospital type, raising additional questions of equity in TBI care that merit further investigation.
The increased likelihood of hospice discharge among elderly patients with isolated small-volume SDH also warrants attention. While hospice utilization in TBI care remains low overall, our findings suggest even seemingly minor injuries can portend poor outcomes in older patients, particularly those with advanced comorbidity or pre-existing directives. Early identification of patients at risk for unfavorable recovery trajectories can facilitate timely discussions around goals of care, palliative interventions, and family support, thereby improving quality of care at the end of life.
In summary, this study contributes novel insights into the often-overlooked impact of small-volume TBIs in elderly patients. It calls attention to the mismatch between radiographic injury classification and real-world discharge outcomes and advocates for an age-informed, individualized, and globally contextualized approach to post-TBI care. Future research should expand on these findings by incorporating long-term functional outcomes, frailty indices, and international comparisons to develop evidence-based, equitable discharge planning models for older adults with TBI.
Limitations
This study has several limitations. First, while we examined discharge disposition as a proxy for functional recovery, we lacked post-discharge follow-up data on long-term outcomes such as return to independence, quality of life, or survival. Second, we did not account for variation in rehabilitation access, caregiver support, or facility-level practices, which may influence disposition decisions and introduce regional bias. Third, although the use of AIS codes allowed for objective lesion size classification, potential inter-observer variability in imaging interpretation and coding across centers cannot be ruled out. Fourth, we excluded patients with skull fractures and polytrauma to isolate the effect of small-volume intracranial hemorrhage, but this may limit generalizability to broader TBI populations. Lastly, while we stratified by age, we did not directly assess frailty, pre-injury functional status, or cognitive impairment, factors likely to influence both prognosis and discharge planning. Despite these limitations, the large national dataset and strict inclusion criteria strengthen the internal validity of our findings and offer important insights into discharge outcomes for an often-overlooked TBI subgroup.
Conclusion
This study is among the first to systematically evaluate hospital discharge outcomes in older adults with isolated small-volume traumatic brain injuries, a subgroup often labeled as “mild” yet shown here to have substantial post-acute care needs. By leveraging a large national trauma registry and stratifying outcomes by both age and specific TBI subtype, we demonstrate elderly patients face higher odds of discharge to rehabilitation or hospice care compared to middle-aged adults, even when injuries are radiographically small. These findings challenge conventional assumptions about the benign nature of small TBIs and reveal a critical gap in recognition and planning for post-hospital care. Our results support the need for age-informed discharge strategies and early intervention for elderly patients, even in cases traditionally viewed as low-risk. Future work should integrate frailty, functional status, and long-term outcomes to refine prognostication and improve care pathways for this vulnerable but underappreciated patient population.
Acknowledgments
TQIP-PUF Admission 2017 to 2022 Version (e.g., 2020.1.0), Chicago, IL, 20XX. The content reproduced from the TQIP PUF remains the full and exclusive copyrighted property of the American College of Surgeons. The American College of Surgeons is not responsible for any claims arising from works based on the original data, text, tables, or figures.
Footnotes
Funding: This project was funded in part through philanthropic support of Marshfield Clinic Foundation (255800-00-RES SUPPT TRAUMA).
Disclosures: There are no conflicts of interest to share. This research was submitted as an Abstract (#2046214) to the 2025 American College of Surgeons Trauma Quality and Safety Conference in San Diego, CA from July 17-20, 2025.
Author Contributions: This research is in compliance with ICMJE authorship guidelines (1) substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; (2) drafting the work or revisit it critically for important intellectual content; (3) final approval of the version to be published; (4) agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The following types of contributions were:
Conception and Study Design: HR, AE, DM, AP, LM
Literature Review: HR, AE, DM, AP, LM
Data Acquisition: HR, DM
Data Analysis and Interpretation: HR, DM, LM
Drafting of the Manuscript: HR, AE, DM, AP, LM
Critical Revision: HR, AE, DM, AP, LM
- Received April 15, 2025.
- Revision received November 18, 2025.
- Revision received January 7, 2026.
- Accepted January 19, 2026.
References
- 1.↵Centers for Disease Control and Prevention (CDC). (2024). Traumatic brain injury & concussion: Data & statistics. Available at: https://www.cdc.gov/traumatic-brain-injury/data-research/index.html#:~:-text=Based%20on%20the%20most%20recent%20data:%20There,hospitalizations%20and%20190%20TBI%2Drelated%20deaths%20per%20day. Accessed February 20, 2025
- 2.↵Thompson HJ, McCormick WC, Kagan SH. Traumatic brain injury in older adults: epidemiology, outcomes, and future implications. J Am Geriatr Soc. 2006;54(10):1590-1595. doi:10.1111/j.1532-5415.2006.00894.x.
- 3.↵Yee G, Jain A. Geriatric Head Injury(Archived). In: StatPearls. Treasure Island (FL): StatPearls Publishing; August 8, 2023.
- 4.↵Tarumi T, Zhang R. Cerebral blood flow in normal aging adults: cardiovascular determinants, clinical implications, and aerobic fitness. J Neurochem. 2018;144(5):595-608. doi:10.1111/jnc.14234.
- 5.↵National Academies of Sciences, Engineering and Medicine, Health and Medicine Division, . Traumatic brain injury: A roadmap for accelerating progress. Washington (DC): National Academies Press (US); 2022 Feb 1.
- 6.↵Maas AIR, Menon DK, Manley GT, Traumatic brain injury: progress and challenges in prevention, clinical care, and research. Lancet Neurol. 2022;21(11):1004-1060. doi:10.1016/S1474-4422(22)00309-X
- 7.↵Waltzman D, Haarbauer-Krupa J, Womack LS. Traumatic Brain Injury in Older Adults-A Public Health Perspective. JAMA Neurol. 2022;79(5):437-438. doi:10.1001/jamaneurol.2022.0114
- 8.↵American College of Surgeons. Trauma quality programs participant use file. Trauma Quality Programs Participant Use File. Available at: https://www.facs.org/quality-programs/trauma/quality/national-trauma-da-ta-bank/datasets/ Accessed February 20, 2025
- 9.↵Abbreviated Injury Scale: 2015 Revision. 6th ed. Chicago, IL: Association for the Advancement of Automotive Medicine.
- 10.↵Pastorek RA, Cripps MW, Bernstein IH, The Parkland Protocol’s modified Berne-Norwood criteria predict two tiers of risk for traumatic brain injury progression. J Neurotrauma. 2014;31(20):1737-1743. doi:10.1089/neu.2014.3366.
- 11.↵De Guio F, Duering M, Fazekas F, . Brain atrophy in cerebral small vessel diseases: Extent, consequences, technical limitations and perspectives: The HARNESS initiative. J Cereb Blood Flow Metab. 2020;40(2):231-245. doi:10.1177/0271678X19888967.
