AI Tool Help Predict COVID-related Death Risk

Artificial intelligence could provide healthcare professionals with additional information about COVID-19 patient outcomes. According to a research article published in BMJ Healthcare & Informatics , researchers from the Veterans Affairs Medical center at Washington DC developed an artificial intelligence tool to calculate the risk of death among COVID-19 patients. Tim Strebel, a computer programmer at the DC VA, has designed this tool in collaboration with VA's National Artificial Intelligence Institute (NAII).

While Artificial intelligence is increasingly being used in health care, this new tool is designed in such a way that the risk of death due to COVID-19 can be predicted earlier by means of using patient details admitted in the healthcare settings.

Pilot study with two models of AI tool

Researchers conducted a pilot study using this new tool at 13 VA medical centers. The tool prepares a report with 120-day mortality risk scores for patients in both inpatient and outpatient settings. Two models of the tool were tested of which the first one uses factors like age, body mass index (BMI), and comorbidities from the Veteran's electronic health record, and the second one uses the same along with Veterans' lab findings and vital signs evaluated at the time of admission.

The second model showed a significant accuracy in predicting COVID-related mortality. "While a few comorbidities on their own are predictive of mortality, such as diabetes and dementia, we've found that the amount of and severity of comorbidities in a patient is the best way to use them to predict mortality," said Strebel.

Limitations of the AI tool

Advertisement

Although the tool helps predict COVID-related mortality, the study team highlighted the potential risk of bias that can occur while using AI. As an example, Strebel said, "While age is no doubt one of the leading predictors of death, there are always exceptions. Some patients in their 90s survive COVID-19. Conversely, some really young patients die from COVID-19. In those rare cases, our models may show that older Veterans are at higher risk of death than they may actually be simply because of their age." As the tool shows significant accuracy, these limitations can be overcome by taking other factors and laboratory investigations into account.

Apart from this, the team opines that this AI tool could help doctors identify Veterans at high risk of suicide, interpret laboratory findings of cancer patients, and choose the best therapy as per the patient's health condition.

Advertisement

Source-Medindia