Photo Credit: praetorianphoto
The following is a summary of “Roadmap for the evolution of monitoring: developing and evaluating waveform-based variability-derived artificial intelligence-powered predictive clinical decision support software tools,” published in the December 2024 issue of Critical Care by Seely et al.
Continuous waveform monitoring is standard-of-care for patients with critically illness, with derived metrics such as heart rate, respiratory rate, and blood pressure variability offering diagnostic and prognostic value, and when combined with machine learning, could enhance predictive models for illness severity and physiological reserve, potentially advancing clinical decision support systems (CDSS).
Researchers conducted a retrospective study to review the use of continuous waveform monitoring combined with machine learning for predicting illness severity and reduced physiologic reserve.
They reviewed and analyzed the multidisciplinary steps involved in developing and rigorously evaluating predictive CDSS tool derived from monitoring data.
The results showed that the development and evaluation of waveform-based variability-derived predictive models required a multi-phase, multidisciplinary approach. The process encompassed data science (data collection, waveform processing, variability analysis, statistical analysis, machine learning, predictive modeling), CDSS development (iterative prototype research leading to a commercial tool), and clinical research (observational and interventional studies, followed by feasibility and definitive randomized controlled trials). This approach presented unique challenges, including technical, analytical, psychological, regulatory, and commercial hurdles.
Investigators concluded the proposed roadmap could guide the development and evaluation of predictive CDSS tools to improve monitoring and care.
Source: cforum.biomedcentral.com/articles/10.1186/s13054-024-05140-6
