Artificial Intelligence in Healthcare

The US healthcare system was pushed to its limits in 2020. Covid-19 filled-up hospital beds. Elective procedures and in-person visits were put on hold…”and more than 14.6 million Americans lost their employer-sponsored health insurance.”(1) Providers had to adapt on the fly, pivot while caring for patients, and find new ways to keep their doors open. Telehealth emerged as a major tool, and clinical artificial intelligence (AI) earned a great deal of attention from hospital administrations. “A recent survey found 56 percent of healthcare executives accelerated their AI deployment plans in response to the pandemic.”(1) 

AI has been bubbling-up within the healthcare community since the 1980’s. Beginning with software that relied on “if-then” rules and progressing to more sophisticated algorithms. “Early rule-based systems had the potential to accurately diagnose and treat disease but were not totally accepted for clinical practice.”(2) A major stumbling block has been the failure to integrate AI systems with clinical workflows and Electronic Healthcare Records (EHR). “Integration issues have been a greater barrier to widespread adoption of AI in healthcare when compared to the accuracy of suggestions.”(2)

Today, EHR systems are beginning to offer some AI capabilities. And AI systems are digesting huge amounts of information, identifying trends, and connecting the dots between apparently disparate data sets. Machine Learning enables the software to improve efficiency without any human interaction. And AI predictive analytics arrives at conclusions that would elude human discernment.

There are dozens of current and potential uses for AI in healthcare,but a recent Optum survey found "...the top three applications health care execs plan to tap AI for Include:"(3) 

 


Wearable technologyMore and more consumers have access to devices with sensors that can collect valuable data about the individual’s health. We have smartphones with step trackers, wearables that can track a heartbeat around the clock. “Collecting and analyzing this data – and supplementing it with patient-provided information through apps and other home monitoring devices – can offer a unique perspective into individual and population health.”(4) The technology exists that can gather and analyze this data, but the biggest hurdle to implementation is consumer acceptance of this intimate continual monitoring. The hope is, that patients tend to trust their physicians more than they might trust Facebook or Google. “There’s a good chance [wearable data will have a major impact] because our care is very episodic and the data in a continuous fashion, there’s a greater likelihood that the data will help us take better care of patients.”(4)

 
Therapeutic and clinical technology AI has already helped hospitals manage capacity, triage Covid-19 patients and target outreach to vulnerable patients and populations to prevent avoidable hospitalizations. Today, “Under the direction of Joe Biden’s pick for Secretary of HHS, Xavier Becerra, we will almost certainly see the shift of value-based models of care accelerate. Expect a push for greater participation and more mandatory value-based payment programs.”(1). Prescriptive clinical AI will help hospitals leverage data to guide clinical decisions for better patient outcomes and pave the way to value-based pricing. This will be particularly important when elective procedures are again scheduled following their delay during the pandemic.  

Coding and reimbursement 
If we learned anything from the pandemic, it was how social and economic factors impact public health. Low-income essential workers had the most exposure to coronavirus, increasing the risk for severe infection, and they had the least ability to cover the costs of care. This dynamic had a twofold impact on hospitals: (1) infection spikes among this population flooded hospitals, and (2) they had limited means to pay for care – increasing the hospital’s already burdened Charity Care budget. “AI’s ability to make connections in data can help health systems better account for these social determinants of health (SDOH) in their population health investments and individual care plans.”(1)

Ontash, Inc. operates in this healthcare space with its Disability Workbench AI tool. The tool is used by Quality Reimbursement Services, to identify candidates for Disability Benefits (SSI/SSDI), enabling the community’s most vulnerable to qualify for Medicaid and/or Medicare health insurance. But perhaps the greatest impact AI will have on healthcare, has yet to be fully realized, and that is in the area of Population Health. “This approach utilizes non-traditional partnerships among different sectors of the community – public health, industry, academia, health care, local government entities, etc. – to achieve positive health outcomes.”(5)  The technology exists to draw data from these sectors - identify individuals at risk, enable pro-active measures to stem outbreaks and insure more positive health outcomes for the entire community. The problem isn’t technology or software, or even privacy issues (see HIPAA compliance). The problem lies in bringing community sectors together to share data and take appropriate actions.  Unfortunately, getting humans to change their behavior remains a far greater challenge, than engineering life-saving software systems. 

(1)  4 Ways Clinical AI will Transform Healthcare in 2021, healthcare innovation, Dec. 29, 2021, Zenobia Brown and John Frownfelter, M.D.

(2)   AI in healthcare, foresee medical(3)   THIRD OPTUM SURVEY ON AI IN HEALTH CARE

(4)   Top 12 Ways Artificial Intelligence Will Impact Healthcare, Health IT Analytics, Jennifer Bresnick

(5)   What is Population Health? Centers for Disease Control and Prevention, Oct. 6, 2020