Prediction of outcome after moderate and severe traumatic brain injury

Abstract

Objective The International Mission on Prognosis and Analysis of Clinical Trials (IMPACT) and Corticoid Randomisation After Significant Head injury (CRASH) prognostic models predict outcome after traumatic brain injury (TBI) but have not been compared in large datasets. The objective of this is study is to validate externally and compare the IMPACT and CRASH prognostic models for prediction of outcome after moderate or severe TBI. Design External validation study. Patients We considered 5 new datasets with a total of 9036 patients, comprising three randomized trials and two observational series, containing prospectively collected individual TBI patient data. Measurements Outcomes were mortality and unfavourable outcome, based on the Glasgow Outcome Score (GOS) at six months after injury. To assess performance, we studied the discrimination of the models (by AUCs), and calibration (by comparison of the mean observed to predicted outcomes and calibration slopes). Main Results The highest discrimination was found in the TARN trauma registry (AUCs between 0.83 and 0.87), and the lowest discrimination in the Pharmos trial (AUCs between 0.65 and 0.71). Although differences in predictor effects between development and validation populations were found (calibration slopes varying between 0.58 and 1.53), the differences in discrimination were largely explained by differences in case-mix in the validation studies. Calibration was good, the fraction of observed outcomes generally agreed well with the mean predicted outcome. No meaningful differences were noted in performance between the IMPACT and CRASH models. More complex models discriminated slightly better than simpler variants. Conclusions Since both the IMPACT and the CRASH prognostic models show good generalizability to more recent data, they are valid instruments to quantify prognosis in TBI.

Predicting Long-Term Outcome After Traumatic Brain Injury Using Repeated Measurements of Glasgow Coma Scale and Data Mining Methods

“…We propose that the ANN model with 96.13 % AUC is optimal for predicting 6-month functional outcome and the NB with 90.14 % AUC is the best algorithm for predicting 6-month mortality. Several studies have evaluated the ability of prediction models to correctly predict outcomes for patients with TBI and showed a wide range of AUCs (65 to 85 %) [17,18,22]. We found that our data mining models (ANN, NB, DT) accurately discriminated functional outcome with an AUC of 92-96 % and mortality with an AUC of 78-91 %, which compared favorably with AUCs reported in previous studies.…”

“…The results from the analysis of important attributes (Table 5) indicate that the GCS scores at different time points (ER, 7th, and 14th day), GCS changes between time points (from ER to 7th day and from ER to 14th day), and age were the most influential variables in the performance of our prediction models. As the data were collected in a tertiary trauma center in an agricultural county where persons 65 years or older represent more than 14 % of the population and motorcycle is the dominant method of transportation, the mean age of these patients with head injury, 55.5 years, is older than in most previous studies of adults with TBI [9,18,[61][62][63]. Pupillary size was shown to be a non-significant predictor in our study.…”

“…GCS is the most widely implemented clinical scoring system to assess level of consciousness after TBI, which classifies the severity as mild (GCS [13][14][15], moderate (GCS 9-12), or severe (GCS < 9) [14][15][16]. Studies have demonstrated that GCS, which is easy to assess and reproducible, is an important prognostic factor in outcome prediction models [16][17][18][19] and may influence treatment strategies [20,21]. However, how changes in GCS scores at various time points after hospital admission relate to outcome is unknown.…”

The Glasgow Outcome Scale — 40 years of application and refinement

“…This involves dividing the study population into subgroups according to the early severity of their injury. This is done by taking into account factors such as clinical and imaging findings and patient age 57 . The point of dichotomization is adjusted according to the subgroup.…”

The first hours of traumatic brain injury. Evaluation and management in emergency room

“…La TAC rappresenta uno strumento di grande utilità per stabilire l' evoluzione e la prognosi del paziente con TBI, anche se non è stata stabilita una relazione statisticamente significativa [45]. Dai dati ottenuti dai referti dei pazienti dello studio CRASH è stato evidenziato che la compressione delle cisterne perimesencefaliche e la presenza di emorragia subaracnoidea traumatica si associa a prognosi negativa [50]. La proteina S100β presente nelle cellule gliali e e di Schwann è elevato nel sangue e nel liquor in risposta al danno.…”