AI in Healthcare 2020 Leadership Survey Report: Health systems are both buying and developing AI apps

Standard practice seems to be a blend of commercial AI solutions and teams at healthcare facilities and systems developing their own, with half adopting this strategy. Exclusively purchasing commercial apps is the plan of 38% of facilities, while 13% will take a develop in-house-only strategy. When commercial apps are the plan, three quarters are validating their accuracy with local patient data.

Some 60% of respondents say they’ll add 1-10 AI apps over the next 18 months, with 6 percent pushing that number to 11 to 50. 1% say they’ll add more than 50, while 33 percent say none. 37% of respondents say they are developing AI apps.

With the C-suite at the AI helm, it’s not surprising that almost 60 percent of healthcare organizations report having a data governance policy providing guidance for the utilization of PHI (EHR data, imaging and reports) in AI research and application development. 

Data scientists are increasingly part of the team too, so say 60% of facilities who employ them. Of those, about 38% have fewer than 10, while 17% employ 11 to 50. About 4% report having more than 50. Physician data-scientists also are more likely to be pitching in on clinical AI projects like automated brain bleed and AFib detection.

Beyond planning strategy, 50% of facilities plan to use AI as a competitive advantage, marketing its use and emphasizing innovation.