February 23, 2018Open Access

A dual boundary classifier for predicting acute hypotensive episodes in critical care

Q: What is the clinical evidence from this study?

Study design: Other. Population: Acute Hypotensive Episodes in critical care (n=4593). Intervention: Dual Boundary Classifier (DBC) vs. Standard machine learning classifiers (e.g., SVM, KNN, Random Forest, Mean-based method). Primary outcome: Prediction of Acute Hypotensive Episodes (p=<0.05).

SBSakyajit BhattacharyaTata Consultancy Services (India)VHVijay HuddarAmazon (United States)VRVaibhav RajanGoogle (United States)

Key Result

The Dual Boundary Classifier algorithm predicted acute hypotensive episodes up to two hours in advance with nearly 80% sensitivity and 95% specificity, significantly outperforming previous approaches.

Structured PICO

Does the Dual Boundary Classifier (DBC) algorithm improve the early prediction of acute hypotensive episodes in ICU patients compared to standard classifiers?

Population

4,593 critical care patient records from the MIMIC II database with at least 6 hours of mean arterial pressure measurements, evaluated for the prediction of acute hypotensive episodes.

Intervention

Dual Boundary Classifier (DBC) algorithm using mean and standard deviation of past Mean Arterial Pressure (MAP) measurements to predict Acute Hypotensive Episodes.

Comparator

Standard classifiers including mean-based prediction, Support Vector Machines (SVM), Random Forest (RF), K-Nearest Neighbors (KNN), and Adaboost (ADA).

Outcome

Prediction of Acute Hypotensive Episodes (AHE) measured by sensitivity, specificity, and classification accuracy.

A novel dual-boundary classification algorithm using simple blood pressure summary statistics can accurately predict acute hypotensive episodes in ICU patients up to two hours in advance.

Main Result

p-value: p=<0.05

Limitations

The model remains oblivious to the underlying causes of acute hypotensive episodes and patient diversity.
Data is limited to a single tertiary hospital (MIMIC II database), requiring multi-center studies for further validation.
Performance of the classifier on features that combine measurements from different vitals was not tested.

Abstract

An Acute Hypotensive Episode (AHE) is the sudden onset of a sustained period of low blood pressure and is one among the most critical conditions in Intensive Care Units (ICU). Without timely medical care, it can lead to an irreversible organ damage and death. By identifying patients at risk for AHE early, adequate medical intervention can save lives and improve patient outcomes. In this paper, we design a novel dual-boundary classification based approach for identifying patients at risk for AHE. Our algorithm uses only simple summary statistics of past Blood Pressure measurements and can be used in an online environment facilitating real-time updates and prediction. We perform extensive experiments with more than 4,500 patient records and demonstrate that our method outperforms the previous best approaches of AHE prediction. Our method can identify AHE patients two hours in advance of the onset, giving sufficient time for appropriate clinical intervention with nearly 80% sensitivity and at 95% specificity, thus having very few false positives.

Ask AI

Helpful

Bookmark

View Full Paper

Ask AI

Helpful

Bookmark

View Full Paper

Cite This Study

Bhattacharya et al. (Fri,) conducted a other in Acute Hypotensive Episodes in critical care (n=4,593). Dual Boundary Classifier (DBC) vs. Standard machine learning classifiers (e.g., SVM, KNN, Random Forest, Mean-based method) was evaluated on Prediction of Acute Hypotensive Episodes (p=<0.05). The Dual Boundary Classifier algorithm predicted acute hypotensive episodes up to two hours in advance with nearly 80% sensitivity and 95% specificity, significantly outperforming previous approaches.

synapsesocial.com/papers/6a21420fa16f1d2b6a5acb73 https://doi.org/https://doi.org/10.1371/journal.pone.0193259