Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients

• Published in: BMJ Vol. 336; no. 7641; pp. 425 - 429
• Main Authors: Perel, Pablo, Arango, Miguel, Clayton, Tim, Edwards, Phil, Komolafe, Edward, Poccock, Stuart, Roberts, Ian, Shakur, Haleema, Steyerberg, Ewout, Yutthakasemsunt, Surakrant
• Format: Journal Article
• Language: English
• Published: London British Medical Journal Publishing Group 23.02.2008; British Medical Association; BMJ Publishing Group LTD; BMJ Publishing Group; BMJ Publishing Group Ltd
• Edition: International edition
• Subjects: Accidents, Traffic - statistics & numerical data; Adult and adolescent clinical studies; Age; Biological and medical sciences; Brain damage; Brain Injuries - diagnostic imaging; Brain Injuries - mortality; Calibration; Clinical outcomes; Clinical trials; Cohort Studies; Computed tomography; Computerized axial tomography; Database models; Death & dying; Disabilities; General aspects; Glasgow Coma Scale; Head injuries; Health outcomes; Hematoma; Hemorrhage; High income; Humans; Income; Injuries of the nervous system and the skull. Diseases due to physical agents; Low income; Low income groups; Medical prognosis; Medical research; Medical sciences; Modeling; Mortality; Multivariate Analysis; Organic mental disorders. Neuropsychology; Outcome Assessment (Health Care) - standards; Patients; Physical trauma; Prognosis; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Regression analysis; Tomography; Tomography, X-Ray Computed; Traffic; Traumas. Diseases due to physical agents; Traumatic brain injury; Variables; Ventricles (cerebral); World; Human; Medicine; Nervous system diseases; Prognosis; Cohort study; Evolution; Patient; Models; Head trauma; Public health; International
• ISSN: 0959-8138; 0959-8146; 1756-1833; 1756-1833; 1468-5833
• DOI: 10.1136/bmj.39461.643438.25

Objective To develop and validate practical prognostic models for death at 14 days and for death or severe disability six months after traumatic brain injury.Design Multivariable logistic regression to select variables that were independently associated with two patient outcomes. Two models designed: “basic” model (demographic and clinical variables only) and “CT” model (basic model plus results of computed tomography). The models were subsequently developed for high and low-middle income countries separately. Setting Medical Research Council (MRC) CRASH Trial.Subjects 10 008 patients with traumatic brain injury. Models externally validated in a cohort of 8509.Results The basic model included four predictors: age, Glasgow coma scale, pupil reactivity, and the presence of major extracranial injury. The CT model also included the presence of petechial haemorrhages, obliteration of the third ventricle or basal cisterns, subarachnoid bleeding, midline shift, and non-evacuated haematoma. In the derivation sample the models showed excellent discrimination (C statistic above 0.80). The models showed good calibration graphically. The Hosmer-Lemeshow test also indicated good calibration, except for the CT model in low-middle income countries. External validation for unfavourable outcome at six months in high income countries showed that basic and CT models had good discrimination (C statistic 0.77 for both models) but poorer calibration.Conclusion Simple prognostic models can be used to obtain valid predictions of relevant outcomes in patients with traumatic brain injury.