Predictive models for the occurrence of expansive intracranial hematomas and outcomes after surgical evacuation in patients with traumatic brain injury in Uganda: a prospective cohort study - Report - MDSpire
Coming Soon: Introducing MDSpire News. Learn more
Conexiant’s news site is now MDSpire News. Learn more

Exploratory Predictive Models for Expansive Intracranial Hematoma Development and Surgical Outcomes in Traumatic Brain Injury Patients in Uganda: A Prospective Cohort Analysis

  • By

  • Larrey Kasereka Kamabu

  • Godfrey S. Bbosa

  • Ronald Oboth

  • Ssenyondwa John Baptist

  • Martin N. Kaddumukasa

  • Daniel Deng

  • Hervé Monka Lekuya

  • Louange Maha Kataka

  • Joel Kiryabwire

  • Moses Galukande

  • Martha Sajatovic

  • Mark Kaddumukasa

  • Anthony T. Fuller

  • Michael M. Haglund

  • September 15, 2026

Share

Clinical Report: Predictive Models for Expansive Intracranial Hematoma in TBI

Overview

This study assessed the prevalence of expansive intracranial hematomas (EIH) in Ugandan TBI patients and developed predictive models for EIH occurrence and surgical outcomes.

Background

Traumatic brain injury (TBI) is a significant global health issue, particularly in low-resource settings like Uganda, where it leads to high mortality and disability rates. Expansive intracranial hematomas (EIH) are a critical complication of TBI, associated with increased intracranial pressure and worse functional outcomes. Early identification of patients at risk for EIH is essential for improving management in these settings.

Data Highlights

MetricValue
Patients with EIH192 (59.3%)
AUC for EIH prediction (Firth model)0.854
Bootstrap optimism corrected AUC0.845
Pooled out-of-fold AUC (LASSO)0.835
Pooled out-of-fold AUC (Elastic Net)0.837
Pooled out-of-fold AUC (Random Forest)0.804

Key Findings

  • 59.3% of TBI patients experienced expansive intracranial hematomas (EIH).
  • EIH was associated with significantly worse quality-of-life measures at 3 and 6 months postoperatively (p < 0.010).
  • The Firth bias-reduced model showed strong discrimination for EIH prediction with an AUC of 0.854.
  • Penalized logistic regression models (LASSO and elastic net) performed comparably to random forest in predicting EIH.
  • Predictive variables included subdural hematoma, diffuse axonal injury, and blood pressure measurements.

Clinical Implications

The prevalence of EIH in TBI patients in Uganda highlights the need for effective risk stratification tools in low-resource settings.

Conclusion

EIH is common among TBI patients in Uganda and correlates with poorer postoperative outcomes.

Related Resources & Content

  1. npj Digital Medicine, 2025 -- Interpretable Multiomics Models for Predicting Surgical Interventions and Blood Transfusion Requirements in Traumatic Brain Injury
  2. Frontiers in Neurology, 2026 -- A predictive decision tree model for hematoma expansion in females following spontaneous intracerebral hemorrhage
  3. npj Digital Medicine, 2025 -- Prognostic Assessment and U-Shaped Relationship Between SBP and Risk in Patients with Unstable Pelvic Fractures and Traumatic Brain Injury
  4. Frontiers in Neurology, 2026 -- Development and internal validation of a machine learning model for predicting intracranial infection after spontaneous intracerebral hemorrhage: a two-center retrospective study
  5. BTF Management of Penetrating Traumatic Brain Injury Guideline Summary - Guideline Central, 2026
  6. Imaging predictors of hemorrhagic progression of a contusion after traumatic brain injury: a systematic review and meta-analysis - PMC, 2024
  7. BTF Management of Penetrating Traumatic Brain Injury Guideline Summary - Guideline Central
  8. Imaging predictors of hemorrhagic progression of a contusion after traumatic brain injury: a systematic review and meta-analysis - PMC
  9. Trial of Decompressive Craniectomy for Traumatic Intracranial Hypertension | New England Journal of Medicine

Original Source(s)

Related Content