Machine learning-based scoring model for predicting mortality in ICU-admitted ischemic stroke patients with moderate to severe consciousness disorders

 

What possible use of predicting failure to recover helps survivors? WHY THE FUCK AREN'T YOU DOING RESEARCH THAT DELIVERS RECOVERY?

Machine learning-based scoring model for predicting mortality in ICU-admitted ischemic stroke patients with moderate to severe consciousness disorders

Zhou Zhou^1,2Bo Chen²Zhao-Jun Mei^1,2Wei Chen^1,2Wei Cao³En-Xi Xu²Jun Wang²Lei Ye^1*Hong-Wei Cheng^1*

• ¹Department of Neurosurgery, The First Affiliated Hospital of Anhui Medical University, Hefei, China
• ²Department of Neurosurgery, Affiliated People’s Hospital of Jiangsu University, Jiangsu, China
• ³Department of Neurology, Affiliated People’s Hospital of Jiangsu University, Jiangsu, China

Background: Stroke is a leading cause of mortality and disability globally. Among ischemic stroke patients, those with moderate to severe consciousness disorders constitute a particularly high-risk subgroup. Accurate predictive models are essential for guiding clinical decisions in this population. This study aimed to develop and validate an automated scoring system using machine learning algorithms for predicting short-term (3- and 7-day) and relatively long-term (30- and 90-day) mortality in this population.

Methods: This retrospective observational study utilized data from the MIMIC-IV database, including 648 ischemic stroke patients with Glasgow Coma Scale (GCS) scores ≤12, admitted to the ICU between 2008 and 2019. Patients with GCS scores indicating speech dysfunction but clear consciousness were excluded. A total of 47 candidate variables were evaluated, and the top six predictors for each mortality model were identified using the AutoScore framework. Model performance was assessed using the area under the curve (AUC) from receiver operating characteristic (ROC) analyses.

Results: The median age of the cohort was 76.8 years (IQR, 64.97–86.34), with mortality rates of 8.02% at 3 days, 18.67% at 7 days, 33.49% at 30 days, and 38.89% at 90 days. The AUCs for the test cohort’s 3-, 7-, 30-, and 90-day mortality prediction models were 0.698, 0.678, 0.724, and 0.730, respectively.

Conclusion: We developed and validated a novel machine learning-based scoring tool that effectively predicts both short-term and relatively long-term mortality in ischemic stroke patients with moderate to severe consciousness disorders. This tool has the potential to enhance clinical decision-making and resource allocation for these patients in the ICU.(So, you can quickly triage them to the death intervention?)

Introduction

Stroke, including both ischemic and hemorrhagic types, remains one of the leading causes of mortality and long-term disability worldwide (1). Stroke mortality is projected to increase by 50% from 2020 to 2050 (2), significantly adding to the disease burden. The burden is particularly severe among patients who experience both severe ischemic stroke and consciousness disorders (3, 4), involving prolonged hospital stays, intensive rehabilitation efforts, and significant caregiver support (5). Consciousness disorders encompass a range of conditions, including coma, vegetative state, and minimally conscious state (6, 7), and are associated with significantly worse prognoses compared to ischemic stroke patients without consciousness disorders (8).

In this study, we focus on a distinct and challenging subgroup: ischemic stroke patients with moderate to severe consciousness disorders (GCS ≤ 12) at admission, excluding those with a GCS score of 4-1-6 or 4-2-6, as they are classified as having speech dysfunction with clear consciousness (9, 10). All these severe ischemic stroke patients were admitted to the ICU (11).

Patients in this category are typically incapable of independently deciding on interventions such as mechanical ventilation, artificial nutrition, surgical decompression, or even the withdrawal of life-sustaining treatment. In many severe stroke cases, however, physicians and patient surrogates must make decisions under conditions of prognostic uncertainty and ambiguous definitions of acceptable outcomes (12). Accurate prediction of outcomes in these patients is essential for guiding clinical decisions, managing resources, and providing appropriate counseling for patients’ families. Prognostic models that accurately predict outcomes for patients with severe stroke are currently insufficient. Traditional assessment tools, such as the GCS and the Modified Rankin Scale (mRS), often overlook the complexities inherent in these patients’ conditions. Moreover, these models tend to rely on static clinical evaluations and do not take advantage of the massive data available from modern healthcare databases. Recent advancements in machine learning (ML) have shown potential in developing more precise and individualized prognostic models (13, 14). ML techniques can analyze large datasets to identify patterns often missed by traditional methods, enhancing prognostic accuracy for patients (15, 16). Despite its potential, research applying machine learning to predict outcomes in severe ischemic stroke patients remains limited. This gap underscores the need for innovative approaches to improve prognostic accuracy in this high-risk population.

Therefore, the primary objective of this study is to develop an automated scoring model using machine learning techniques to estimate mortality for severe ischemic stroke patients with moderate to severe consciousness disorders. By enhancing the interpretability and accuracy of the predictive model, we aim to facilitate its integration into clinical workflows and decision-making processes.

More at link.