Validation of Stroke Prognostic Scores: What Do Clinicians Need to Know?

Saposnik, Gustavo

doi:10.1159/000354780

Cited by 1 publication

(2 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While discrimination and calibration are essential properties of any prognostic model, they are uninformative as to clinical value. What a clinician needs to know is the proportion of the patients who will die or survive correctly identified 30 . According to Pfeiffer and Gall 31 , the concept of “concentration of risk” (i.e., the proportion of individuals who will develop the event of interest and who are included in the proportion of individuals with a risk exceeding a certain threshold) is more directly relevant to decision making.…”

Section: Discussionmentioning

confidence: 99%

“…The covariates examined included age (per 5-year increase), marital status (married/not married), hypertension, diabetes, COPD, history of CAD, AF, anemia, renal dysfunction, time from stroke onset to rehabilitation admission, ischemic stroke, dysphagia, neglect, and motor and cognitive FIM scores at admission. These variables were selected based on prior studies showing an association with the outcomes of interest 6,30,[42][43][44][45][46][47][48][49][50][51][52][53][54][55][56] . A multivariate logistic regression analysis with backward stepwise selection (p > 0.20 for exclusion) was performed to assess the association of covariates with 3-year mortality.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Machine learning to predict mortality after rehabilitation among patients with severe stroke

Scrutinio

Ricciardi

Donisi

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Stroke is among the leading causes of death and disability worldwide. Approximately 20–25% of stroke survivors present severe disability, which is associated with increased mortality risk. Prognostication is inherent in the process of clinical decision-making. Machine learning (ML) methods have gained increasing popularity in the setting of biomedical research. The aim of this study was twofold: assessing the performance of ML tree-based algorithms for predicting three-year mortality model in 1207 stroke patients with severe disability who completed rehabilitation and comparing the performance of ML algorithms to that of a standard logistic regression. The logistic regression model achieved an area under the Receiver Operating Characteristics curve (AUC) of 0.745 and was well calibrated. At the optimal risk threshold, the model had an accuracy of 75.7%, a positive predictive value (PPV) of 33.9%, and a negative predictive value (NPV) of 91.0%. The ML algorithm outperformed the logistic regression model through the implementation of synthetic minority oversampling technique and the Random Forests, achieving an AUC of 0.928 and an accuracy of 86.3%. The PPV was 84.6% and the NPV 87.5%. This study introduced a step forward in the creation of standardisable tools for predicting health outcomes in individuals affected by stroke.

show abstract

Section: Discussionmentioning

confidence: 99%