Donna Armaignac, CCNS, CCRN-K, PhD
Director Center Advanced Analytics; Director Best Practice Tele Critical Care
Baptist Health South Florida
Pompano Beach, Florida, United States
Disclosure information not submitted.
Venkataraghavan Ramamoorthy
Baptist Health South Florida
Coral Gables, Florida
Disclosure information not submitted.
Munir B. Rubens, PhD
Biostatistician
Baptist Health South Florida, United States
Disclosure information not submitted.
Title: Association between ICU Risk Prediction Scores and ICU Needs among Acute Ischemic Stroke Patients
Introduction:
Current best practice standards for patients with acute ischemic stroke (AIS) receiving intravenous thrombolysis (IVT) and endovascular mechanical reperfusion therapy (EVT) treatments entail ICU level care to monitor and prevent complications of treatments. Risk prediction scores are used to predict potential severity of illness. This study sought to identify the relationship between risk prediction scores and ICU needs of AIS patients.
Methods:
The current study was retrospective analysis of electronic healthcare data from a Regional Comprehensive Stroke Center of Excellence and included AIS patients ≥18 years of age. The data was collected from patients admitted to ICU between October 2016 and February 2021. Descriptive statistics were used to describe the demographics and quantify treatments. ICU needs were measured in terms of actual ICU procedures and treatments received by the patients. Risk predictor variables collected in the study included Acute Physiology Score (APS), National Institute of Health Stroke Score (NIHSS), and Intensive Care after Thrombolytic (ICAT) scores. Correlation and regression analysis were performed in order to understand the relation between risk predictors and ICU needs.
Results:
There were 4129 AIS patients and 51.6% were females. The mean (SD) age of the patient was 71.2 (14.9) years. More than half of the patients were white Hispanic (60.0%), followed by whites (22.2%), and African American (12.3%). Nearly 10.3% patients received IVT, 6.0% EVT, and 5.3% both treatments. Correlation analysis showed that there was significant association between ICU needs APS (r2=0.336, P< 0.001), NIHSS (r2=0.326, P< 0.001), and ICAT (r2=0.199, P< 0.001) scores. Regression analysis showed that ICAT (β=0.336, P=0.029), NIHSS (β=0.195, P< 0.001), and APS (β=0.248, P< 0.001) scores were significantly associated with ICU needs.
Conclusion:
Risk predictors such as APS, NIHSS, and ICAT scores significantly predicted ICU needs in AIS patients. ICU needs should be validated in studies with larger sample size and heterogonous group of AIS patients. Additional risk predictors should be researched and incorporated for risk modeling and automation of ICU needs assessments, for improved prediction of ICU needs among AIS patients.