How AI could exacerbate existing health disparities

AI
Algorithms are inherently objective, but not if the data is skewed.

Although machines lack the biases humans possess, artificial intelligence could inadvertently aggravate existing health disparities without data that accounts for underrepresented populations. 

Research shows the healthcare industry is already overrun with healthcare inequality. Studies show black Medicare patients are 33% more likely than whites to be readmitted to the hospital after surgery, minorities frequently receive more low-value services than their white counterparts and the perception of the healthcare system varies significantly between high- and low-income patients.

Meanwhile, payers and providers have yet to identify payment models that can close some of those well-documented gaps in care.  

RELATED: High-end, high-cost concierge medicine for America’s wealthiest widens healthcare divide

Those disparities that have taken root in the healthcare industry could carry over as artificial intelligence and machine learning take on a bigger role, according to Quartz. Data that disproportionately favors white males could perpetuate the same problems that healthcare is already battling.

Plus, as the publication points out, low-income areas of the country with less access to digital devices may be gradually left behind. This is particularly concerning given that 60% of rural counties experience high rates of chronic illness coupled with low rates of broadband connectivity.

The American Medical Informatics Association (AMIA) has circled lack of internet access as growing concern. The organization told the Federal Communications Commission that access to broadband “is, or soon will become, a social determinant of health.” Restricted access to digital tools means fewer options for patients in areas where broadband is scarce, which translates to fewer data points to feed into AI algorithms.