ICUs have been severely strained by the pandemic, with shortages of staff, beds, personal protective equipment, and ventilators. Traditional predictive algorithms used to predict patient outcomes, manage capacity, and drive triage decisions have also been exposed as having shortcomings. Artificial intelligence can assist refine raw data and deliver more usable knowledge, particularly in the ICUs, by extracting clinically significant information from noise in a data-rich environment.
Given the massive amount of data available per patient, Frownfelter believes AI has the potential to organise that data and provide meaningful knowledge that allows clinicians to act, rather than reams of “raw data” that can often distract from what’s important. He will speak about the impact of AI on predicting ventilator utilisation at HIMSS21 next week. You may improve outcomes and minimise problems while allowing staff to be more efficient and effective in their care by focusing the necessary resources on the patients who are most at risk of deterioration.
This intelligence allows care teams to better allocate resources and guarantee that ventilators and ICUs beds are available for the patients whose lives are on the line. AI has the potential to be the mechanism by which we learn about new problems and challenges in the therapeutic setting quickly and accurately. Unfortunately, we’ve also found that AI may be done properly or badly, resulting in omission errors, bias magnifying, and a loss of faith in the medical community.