Predicting hospital mortality for ICU patients: time series analysis

Aya Awad, Mohamed Bader-El-Den, James McNicholas, Jim Briggs, Yasser El-Sonbaty

Research output: Contribution to journalArticlepeer-review

447 Downloads (Pure)

Abstract

Current mortality prediction models and scoring systems for Intensive Care Unit (ICU) patients are generally usable only after at least 24 or 48 hours of admission as some parameters are unclear at admission. However, some of the most relevant measurements are available shortly following admission. It is hypothesized that outcome prediction may be made using information available in the earliest phase of ICU admission. This study aims to investigate how early hospital mortality can be predicted for ICU patients. We conducted a thorough time-series analysis on the performance of different data mining methods during the first 48 hours of ICU admission. The results showed that the discrimination power of the machine learning classification methods after 6 hours of admission outperformed the main scoring systems used in intensive care medicine (Acute Physiology and Chronic Health Evaluation, Simplified Acute Physiology Score and Sequential Organ Failure Assessment) after 48 hours of admission.
Original languageEnglish
Pages (from-to)1043-1059
JournalHealth Informatics Journal
Volume26
Issue number2
Early online date26 Jul 2019
DOIs
Publication statusEarly online - 26 Jul 2019

Keywords

  • time-series analysis
  • mortality prediction
  • missing values
  • patient mortality
  • classification
  • machine learning
  • critically ill

Fingerprint

Dive into the research topics of 'Predicting hospital mortality for ICU patients: time series analysis'. Together they form a unique fingerprint.

Cite this