Groot YJ, Lingsma HF, Bakker J, et al. External validation of a prognostic model predicting time of death after withdrawal of life support in neurocritical patients

Department of Intensive Care, Center for Medical Decision Making, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands.
Critical care medicine (Impact Factor: 6.31). 09/2011; 40(1):233-8. DOI: 10.1097/CCM.0b013e31822f0633
Source: PubMed


The ability to predict the time of death after withdrawal of life support is of specific interest for organ donation after cardiac death. We aimed to externally validate a previously developed model to predict the probability of death within the time constraint of 60 mins after withdrawal of life-sustaining measures.
The probability to die within 60 mins for each patient in this validation sample was calculated based on the model developed by Yee et al, which includes four variables (absent corneal reflex, absent cough reflex, extensor or absent motor response, and an oxygenation index >4.2). Analyses included logistic regression modeling with bootstrapping to adjust for overoptimism. Performance was assessed by calibration (agreement between observed and predicted outcomes) and discrimination (distinction of those patients who die within 60 mins from those who do not, expressed by the area under the receiver operating characteristic curve).
Mixed intensive care unit in The Netherlands.
We analyzed data from 152 patients who died as a result of a neurologic condition between 2007 and 2009.
A total of 82 patients had sufficient data. Fifty (61%) died within 60 mins. Univariable and multivariable odds ratios of the predictors were very similar between the development and validation sample. The prediction model showed good discrimination with an area under the receiver operating characteristic curve of 0.75 (95% confidence interval [CI] 0.63-0.87) but calibration was modest. The mean predicted probability was 80%, overestimating the 61% overall observed risk of death within 60 mins. Modeling oxygenation index as a linear term led to an improved version of the Mayo NICU model. (area under the receiver operating characteristic curve [95% CI] = 0.774 [0.69-0.90], bootstrap-validated area under the receiver operating characteristic curve [95% CI] = 0.74 [0.66-0.87]).
The model discriminated well between patients who died within 60 mins after withdrawal of life support and those who did not. Further prospective validation is needed.

Download full-text


Available from: Erwin J O Kompanje

  • No preview · Article · Jan 2012 · Critical care medicine
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Successful donation of organs after cardiac death (DCD) requires identification of patients who will die within 60 min of withdrawal of life-sustaining treatment (WLST). We aimed to validate a straightforward model to predict the likelihood of death within 60 min of WLST in patients with irreversible brain injury. In this multicentre, observational study, we prospectively enrolled consecutive comatose patients with irreversible brain injury undergoing WLST at six medical centres in the USA and the Netherlands. We assessed four clinical characteristics (corneal reflex, cough reflex, best motor response, and oxygenation index) as predictor variables, which were selected on the basis of previous findings. We excluded patients who had brain death or were not intubated. The primary endpoint was death within 60 min of WLST. We used univariate and multivariable logistic regression analyses to assess associations with predictor variables. Points attributed to each variable were summed to create a predictive score for cardiac death in patients in neurocritical state (the DCD-N score). We assessed performance of the score using area under the curve analysis. We included 178 patients, 82 (46%) of whom died within 60 min of WLST. Absent corneal reflexes (odds ratio [OR] 2·67, 95% CI 1·19-6·01; p=0·0173; 1 point), absent cough reflex (4·16, 1·79-9·70; p=0·0009; 2 points), extensor or absent motor responses (2·99, 1·22-7·34; p=0·0168; 1 point), and an oxygenation index score of more than 3·0 (2·31, 1·10-4·88; p=0·0276; 1 point) were predictive of death within 60 min of WLST. 59 of 82 patients who died within 60 min of WLST had DCD-N scores of 3 or more (72% sensitivity), and 75 of 96 of those who did not die within this interval had scores of 0-2 (78% specificity); taking into account the prevalence of death within 60 min in this population, a score of 3 or more was translated into a 74% chance of death within 60 min (positive predictive value) and a score of 0-2 translated into a 77% chance of survival beyond 60 min (negative predictive value). The DCD-N score can be used to predict potential candidates for DCD in patients with non-survivable brain injury. However, this score needs to be tested specifically in a cohort of potential donors participating in DCD protocols. None.
    Full-text · Article · Apr 2012 · The Lancet Neurology
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Summary - The aim of this study was to develop and validate a clinical tool to calculate the probabilities of asystole following withdrawal of life-sustaining treatment (WLST) in potential donation after circulatory death (DCD) donors in the United Kingdom, which is imperative to shared decision-making. A two-stage prospective observational cohort study was undertaken in multi-centre mixed and neurological adult intensive care units in the United Kingdom. One hundred and sixty three potential DCD donors who underwent WLST were included in this study between 2010-2011. An asystole prediction-scoring (APS) tool incorporating clinical variables, assimilated to a score on the basis of derived severity was validated. Data were collected at two time points: initial referral and 60 minutes prior to WLST. Cox regression analysis determined overall probabilities of asystole following WLST. Cox regression demonstrated statistically significant (p<0.05) probabilities of asystole using the APS tool. Probabilities of asystole were produced for time points between 0 and 240 minutes. Lower scores have a low probability of asystole occurring within 180 minutes while higher scores have a high probability of asystole occurring at all time points. Potential donors with APS total scores greater than 30 all died within 180 minutes of WLST. The APS tool provides important information on the likelihood of asystole within defined time lines. This ability to predict time lines could be used by clinicians in decision-making in referring potential DCD donors, and through sharing probabilities of donation occurring with family members thereby facilitating shared decision-making to underpin informed consent.
    Full-text · Article · Jan 2013 · Cells Tissues Organs
Show more