Predicting in-hospital mortality after traumatic brain injury: External validation of CRASH-basic and IMPACT-core in the national trauma data bank

Ratliff, Hunter T.; Korst, Genevieve; Hrushka, Jaron M; Jupiter, Daniel C.

doi:10.1016/j.injury.2020.10.051

Cited by 14 publications

(5 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate possible rehabilitation referral bias, we performed a systematic comparison of the TBIMS database sample to patients with moderate or severe TBI included in published studies that used 3 other large databases . This comparison included patient and clinical characteristics and outcome measures.…”

Section: Methodssupporting

confidence: 92%

Recovery of Consciousness and Functional Outcome in Moderate and Severe Traumatic Brain Injury

et al. 2021

View full text Add to dashboard Cite

IMPORTANCE Traumatic brain injury (TBI) leads to 2.9 million visits to US emergency departments annually and frequently involves a disorder of consciousness (DOC). Early treatment, including withdrawal of life-sustaining therapies and rehabilitation, is often predicated on the assumed worse outcome of disrupted consciousness.OBJECTIVE To quantify the loss of consciousness, factors associated with recovery, and return to functional independence in a 31-year sample of patients with moderate or severe brain trauma. DESIGN, SETTING, AND PARTICIPANTSThis cohort study analyzed patients with TBI who were enrolled in the Traumatic Brain Injury Model Systems National Database, a prospective, multiyear, longitudinal database. Patients were survivors of moderate or severe TBI who were discharged from acute hospitalization and admitted to inpatient rehabilitation from January 4, 1989, to June 19, 2019, at 1 of 23 inpatient rehabilitation centers that participated in the Traumatic Brain Injury Model Systems program. Follow-up for the study was through completion of inpatient rehabilitation. EXPOSURES Traumatic brain injury.MAIN OUTCOMES AND MEASURES Outcome measures were Glasgow Coma Scale in the emergency department, Disability Rating Scale, posttraumatic amnesia, and Functional Independence Measure. Patient-related data included demographic characteristics, injury cause, and brain computed tomography findings. RESULTSThe 17 470 patients with TBI analyzed in this study had a median (interquartile range [IQR]) age at injury of 39 (25-56) years and included 12 854 male individuals (74%). Of these patients, 7547 (57%) experienced initial loss of consciousness, which persisted to rehabilitation in 2058 patients (12%). Those with persisting DOC were younger; had more high-velocity injuries; had intracranial mass effect, intraventricular hemorrhage, and subcortical contusion; and had longer acute care than patients without DOC. Eighty-two percent (n = 1674) of comatose patients recovered consciousness during inpatient rehabilitation. In a multivariable analysis, the factors associated with consciousness recovery were absence of intraventricular hemorrhage (adjusted odds ratio [OR], 0.678; 95% CI, 0.532-0.863; P = .002) and intracranial mass effect (adjusted OR, 0.759; 95% CI, 0.595-0.968; P = .03). Functional improvement (change in total functional independence score from admission to discharge) was +43 for patients with DOC and +37 for those without DOC (P = .002), and 803 of 2013 patients with DOC (40%) became partially or fully independent. Younger age, male sex, and absence of intraventricular hemorrhage, intracranial mass effect, and subcortical contusion were associated with better functional outcome. Findings were consistent across the 3 decades of the database.CONCLUSIONS AND RELEVANCE This study found that DOC occurred initially in most patients with TBI and persisted in some patients after rehabilitation, but most patients with persisting DOC recovered consciousness during rehabilitation. This recovery trajectory may inf...

show abstract

Section: Methodssupporting

confidence: 92%

Recovery of Consciousness and Functional Outcome in Moderate and Severe Traumatic Brain Injury

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The new combined prediction models showed excellent discrimination for the 6-month unfavorable outcome and the 6-month mortality when they were internally revalidated for such TBI patients, which was higher than the previous report and showed good goodness-of-fit. The calibration chart in this paper shows that the IMPACT Core+FIB model is overestimated at low scores and underestimated at higher scores when predicting 6-month unfavorable outcome but generally has a good calibration, which is consistent with the Camarano et al's [22] calibration results of the IMPACT Core model recently. The slope of the IMPACT Core+APTT model is close to 1 when predicting 6-month mortality, indicating that there is a strong consistency between the observed results and the predicted results.…”

Section: Discussionsupporting

confidence: 89%

Development and Verification of Prognostic Prediction Models for Patients with Brain Trauma Based on Coagulation Function Indexes

et al. 2022

Journal of Immunology Research

View full text Add to dashboard Cite

Objective. To assess the effect of adding coagulation indices to the currently existing prognostic prediction models of traumatic brain injury (TBI) in the prediction of outcome. Methods. A total of 210 TBI patients from 2017 to 2019 and 131 TBI patients in 2020 were selected for development and internal verification of the new model. The primary outcomes include death at 14 days and Glasgow Outcome Score (GOS) at 6 months. The performance of each model is evaluated by means of discrimination (area under the curve (AUC)), calibration (Hosmer-Lemeshow (H-L) goodness-of-fit test), and precision (Brier score). Results. The IMPACT Core model showed better prediction ability than the CRASH Basic model. Adding one coagulation index at a time to the IMPACT Core model, the new combined models IMPACT Core+FIB and IMPACT Core+APTT are optimal for the 6-month unfavorable outcome and 6-month mortality, respectively (AUC, 0.830 and 0.878). The new models were built based on the regression coefficients of the models. Internal verification indicated that for the prediction of 6-month unfavorable outcome and 6-month mortality, both the IMPACT Core+FIB model and the IMPACT Core+APTT model show better discrimination (AUC, 0.823 vs. 0.818 and 0.853 vs. 0.837), better calibration (HL, p = 0.114 and p = 0.317 ) and higher precision (Brier score, 0.148 vs. 0.141 and 0.147 vs. 0.164), respectively, than the original models. Conclusion. Our research shows that the combination of the traumatic brain injury prognostic models and coagulation indices can improve the 6-month outcome prediction of patients with TBI.

show abstract

“…Furthermore, the previous validation of IMPACT and CRASH on the nationwide NTDB determined using calibration curves showed that IMPACT had a mix of overpredictions and underpredictions while CRASH underpredicted mortality. 15,18,29 In comparison, our HIC models were shown to be well-calibrated and neither overpredict nor underpredict mortality. Although our HIC models seem to be superior to the previously reported validation of CRASH and IMPACT on NTDB, the lack of pupil data in our NTDB cohort prevents us from performing hypothesis testing to assess for statistical significance.…”

Section: Discussionmentioning

confidence: 88%

Machine Learning for Predicting In-Hospital Mortality After Traumatic Brain Injury in Both High-Income and Low- and Middle-Income Countries

et al. 2022

View full text Add to dashboard Cite

BACKGROUND:Machine learning (ML) holds promise as a tool to guide clinical decision making by predicting in-hospital mortality for patients with traumatic brain injury (TBI). Previous models such as the international mission for prognosis and clinical trials in TBI (IMPACT) and the corticosteroid randomization after significant head injury (CRASH) prognosis calculators can potentially be improved with expanded clinical features and newer ML approaches.OBJECTIVE:To develop ML models to predict in-hospital mortality for both the high-income country (HIC) and the low- and middle-income country (LMIC) settings.METHODS:We used the Duke University Medical Center National Trauma Data Bank and Mulago National Referral Hospital (MNRH) registry to predict in-hospital mortality for the HIC and LMIC settings, respectively. Six ML models were built on each data set, and the best model was chosen through nested cross-validation. The CRASH and IMPACT models were externally validated on the MNRH database.RESULTS:ML models built on National Trauma Data Bank (n = 5393, 84 predictors) demonstrated an area under the receiver operating curve (AUROC) of 0.91 (95% CI: 0.85-0.97) while models constructed on MNRH (n = 877, 31 predictors) demonstrated an AUROC of 0.89 (95% CI: 0.81-0.97). Direct comparison with CRASH and IMPACT models showed significant improvement of the proposed LMIC models regarding AUROC (P = .038).CONCLUSION:We developed high-performing well-calibrated ML models for predicting in-hospital mortality for both the HIC and LMIC settings that have the potential to influence clinical management and traumatic brain injury patient trajectories.

show abstract

Predicting in-hospital mortality after traumatic brain injury: External validation of CRASH-basic and IMPACT-core in the national trauma data bank

Cited by 14 publications

References 40 publications

Recovery of Consciousness and Functional Outcome in Moderate and Severe Traumatic Brain Injury

Recovery of Consciousness and Functional Outcome in Moderate and Severe Traumatic Brain Injury

Development and Verification of Prognostic Prediction Models for Patients with Brain Trauma Based on Coagulation Function Indexes

Machine Learning for Predicting In-Hospital Mortality After Traumatic Brain Injury in Both High-Income and Low- and Middle-Income Countries

Contact Info

Product

Resources

About