practical prognostic models based on large cohort of international patients Medical Research Council (MRC) CRASH Trial Collabolators Editorial by Menon and Harrison BMJ/ February 23 rd 2008/ Vol. 336/ page: 425-429
Presented by: Zulhijrian Noor Irana Priska
Advisor: Ahmad Zuhro Maruf, dr., SpBS JOURNAL READING Jurnal INTRODUCTION Traumatic brain injury is a leading cause of death and disability worldwide 1.5 million / year Most of the burden (90%) low and middle income countries Clinicians treating by assesment of prognosis, 80% believed the accurate assesment of prognosis was important decisions to specific treatment such as hyperventilation,barbiturates and mannitol ( by survey, 2005 ) Jurnal Prognostic models statistic models that combine data from patients to predict outcame more accurate then simple clinical predictions Computers based prediction of outcome Increase certain therapeutic interventions in predicted good outcome, reduces it in poor outcome Jurnal Many prognostic model have been reported but non are widely used No were developed in populations from low and middle income countries MRC CRASH trial, the cohort study prospectively included patient within 8 hours of the injury and achieved almost complete follow up at 6 month Jurnal Develoved and validated prognostic models for death at 14 dayy and death and disabilityat 6 month in patient with traumatic brain injury Jurnal METHODS Patients 10.008 adult patients with traumatic brain injury ( GCS 14 ), within 8 hours of injury Outcomes death of a patient was recorded on a early outcome form that was completed at hospital discharge,death, or 14 days after randomisation. Unfavourable outcome ( death or severe disability ) at 6 months was defined with Glasgow outcome scale Jurnal Glasgow outcome scale 5 categories 1. Good recovery: able to return to work or school 2. Moderate disability: able to live indipendently; unable to return to work or school 3. Severe disability : able to follow commands/unable to live independently 4. Persistent vegetative state : unable to interact with environment; unresponsive 5. dead Jurnal prognostic variables age,sex,cause of injury,time from injury to randomisation,Glasgow coma score at randomisation,pupil reactivity,result of CT,level income in country. analysis included all of variables in a first multivariable logistic regression analysis. Explored liniearity between age and mortality at 14 days prognostic models developed different models for each of the two outcomes: a basic models (only clinical and demographic variables ),CT model ( result CT ). Jurnal Performance of models assessed performance of the models in term of calibration with the Hosmer-Lemeshow and discrimination was assessed with the C statistic. Internal validation the internal validity of the final models was assessed by the bootstrap re-sampling technique Jurnal External validation externally validated the model in an external cohort of 8509 patients with moderate and severe traumatic brain injury from 11 studies conduted in high income countries. Score development a clinical score base on regression coefficient Jurnal
Jurnal General characteristics more patients were men (81%), more come from low-middle income countries (75%) 58% of participants were included within three hours of injury. Road traffic crashes were the most common cause of injury (65%) 79% underwent computed tomography 1948 patient (19%) died in 2 weeks,2323 (24%) dead at 6 month,3556 (37%) were dead or severly dependent at 6 month Jurnal
Jurnal Low middle vs high income countries in comparison from low-middle income countries were younger,more men,were recruited later, had less severeTBI ( as defined by GCS and pupil reactivity ), abnormal result on CT. Older age was a stronger predictor of 14 day mortality in high income countries, also obliteration 3th ventricle and non-evacuated haematoma. Lower GCS was a stronger predictor in low- middle countries
Jurnal Multivariable predictive models o Basic models 4 predictors : age,GCS, pupil reactivity and the presence of major extracranial injury. o CT models characteristics on CT were strongly assosiated with the outcomes. Petechial haemorrhages,obliteration of the third ventricle or basal cystern,SAH,Midline shift,and non-evacuated haematoma.
Jurnal o obliteration of the third ventricle and midline shift strongest predictor of mortality at 14 days o non-evacuated haematoma strongest predictor of mortality at 14 unfavourable outcome at 6 months Jurnal o Performance of models good calibration when evaluated with the Hosmer-Lemesshow test. o Clinical score for example : a 26 year old patient froom low-middle income countries with GCS 11,one pupil reactive,and absen of a major extracranial injury, according to basic models : probably dead at 14 days of 10% and 23.9% risk of death or severe disability at 6 month. Jurnal
Jurnal
Jurnal DISCUSSION There are differences in outcomes and on the strenght of predictors of outcomes on patient from high and middle-low income countries. Older age, Low GCS, absent pupil reactivity, absent major extracranial injury poor prognosis
Jurnal GCS showed a clear linier relation with mortality GCS 3 was lower than in patient with a score of 4 may be because scores of sedated patients are reported as 3 Jurnal Increasing age was associated with worse outcome but this association was apparent only after age 40 Plausible explanation extracranial comorbidities, changes in brain plastisity, differences in clinical management associated with increasing age Jurnal Present of obliteration of 3 rd ventricle or basal cistern as on Ct Scan associated with the worse prognosis at 14 th days Recent findings absnce of basal cistern is a strongets predictors of sixth month mortality Jurnal Patient from low-middle income countries had worse early prognosis than those from high income countries The strength of association between some predictors and outcomes differed by region: Low GCS (poorer in low-middle income countries) quality of care and greater use of sedation Incresing age (poorer in high income countries) CT-Scan technology and accurate diagnosis
Jurnal STRENGTHS AND WEAKNESS Strength The use of a well described cohort of patients Prospective and standadised collection of data on prognostic factor Low loss to follow up The use of a validated outcome measure at a fixed time after the injury The large sample size
Jurnal STRENGTHS AND WEAKNESS Weakness Data from wich models were developed come from a clinicl trial and this could therefore limit external validity For the validation they were forced to exclude the variabels major extracranial injury and petechial haemorrages because they were not available in the IMPACT sample
Jurnal IMPLICATIONS They have developed a methodology valid, simple, accurate model that may help decisions about health care for individual patients Help in the design and analysis of clinical trials, through prognostic stratification. Can be used in clinical audit by allowing adjustment for case mix Jurnal FUTURE RESEARCHE Future research could also evaluate different ways, or formats, for presenting the models to physicians; their use in clinical practice; and whether ultimately they have any impact on the management and outcomes of patients with traumatic brain injury. Jurnal SUMMARY Traumatic brain injury is a leading cause of death and disability worldwide with most cases occurring in lowmiddle income countries Prognostic models may improve predictions of outcome and help in clinical research Many prognostic models have been published but methodological quality is generally poor, sample sizes small, and only a few models have included patients from low-middle income countries Jurnal