London, United Kingdom 2013 Cerebrovasc Dis 2013; 35 (suppl 3)1-854 15 8:30-10:00 Oral Session Room 2,3,4 Stroke prognosis Chairs: D. McCabe, Ireland and A. Czlonkowska, Poland 1 Stroke prognosis 8:30 - 8:40 Prediction of disability and functional outcome in patients with ischemic stroke for effi-cient discharge planning I.R. de Ridder1, H.F. Lingsma2, M. Scheele3, H.M. den Hertog4, M. Dirks5, E.W. Steyerberg6, D.W.J. Dippel7 Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHER-LANDS1, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETH-ERLANDS2, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHERLANDS3, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHERLANDS4, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHERLANDS5, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHERLANDS6, Erasmus MC University Medical Center Rotterdam, Rotterdam, THE NETHERLANDS7 Background A timely start of rehabilitation in acute ischemic stroke depends on early planning of hospital discharge. Efficient discharge planning by nurses requires early and accurate prediction of dis-ability and functional recovery. We developed and validated a prognostic chart for this purpose based on clinical information available on admission. Methods We used data from the Paracetamol (Acetaminophen) in Stroke (PAIS) study, a placebo-con-trolled multicenter randomized clinical trial, for model development. With multivariable ordi-nal logistic regression analysis with backward selection with p<0.2 for inclusion, we identified predictors of Barthel Index (BI) at 1 week and modified Rankin Scale (mRS) at 3 months. The model was internally validated with bootstrap resampling. External validation was performed on the data from the PRACTISE study, a multicenter cluster-randomized trial designed to eval-uate an implementation strategy to increase the proportion of patients treated with intravenous alteplase. Results The development data included 1227 patients with ischemic stroke. The multivariable model for prediction of a higher BI at 1 week after stroke included age, score on NIHSS and diabetes mellitus. The model for mRS at 3 months consisted of the same variables (age, NIHSS score and diabetes mellitus) and additionally included previous stroke and atrial fibrillation . The in-ternally validated AUC was 0.764 for prediction of BI and 0.748 for mRS. External validation of the model for mRS on a set of 1581 patients yielded similar performance (AUC=0.744). Conclusion We developed a prediction model that accurately predicts BI at 1 week and mRS at 3 months in patients with acute ischemic stroke. Since the model can be used on admission, it may be a valuable tool for early prognostication and discharge planning. Table: multivariable ordered logistic regression model for prediction of higher Barthel Index (BI) at 1 week and lower modified Rankin Score (mRS) at 3 months in patients with acute isch-emic stroke. BI at 1 wk mRS at 3 months OR p OR p Age per year >60 0.96 <0.01 0.94 <0.01 NIHSS per point 0.81 <0.01 0.84 <0.01 Diabetes mellitus 0.61 <0.01 0.54 <0.01 Previous stroke - - 0.61 <0.01 Atrial fibrillati-on - - 0.70 <0.02
Karger_ESC London_2013
To see the actual publication please follow the link above