Well that's a bad look as the Senators contemplate filling in the House gaps in the VA Bill
Readers Write: AI is Essential to Stopping Further COVID-19 Spread and Limiting Future Pandemics — Here’s Why
AI is Essential to Stopping Further COVID-19 Spread and Limiting Future Pandemics — Here’s Why
By Sally Embrey
Sally Embrey, MSPH, MS is VP of public health and health technologies of DataRobot of Boston, MA.
It is safe to say that 2020 showed us the limitations of the US healthcare structure and a long-antiquated approach to public health and emergency preparedness. Under lockdown, it became clear how unequipped the world was to address the scope of the COVID-19 pandemic. Since the onset of the greatest healthcare crisis in our lifetimes, the strengths and weaknesses of our healthcare system came into clear focus, while the consequences of our failures will be felt for years to come.
But how do we move forward? In the United States, COVID-19 spread has varied widely, from states to cities and counties. This spring, as COVID-19 cases in Michigan, New York, and New Jersey were slowly declining, cases in Oregon were increasing at higher rates than anywhere else in the United States, according to The New York Times.
One thing that has been emphasized time and time again is that the absence of more complete, accurate, and representative data was a key and often missing factor in our ability to effectively respond to the COVID-19 pandemic. We also learned that the systems required to process that data were as equally important to delivering the insights needed. The pandemic has made the role of AI in the healthcare field essential to preventing and mitigating future pandemics.
Research groups worldwide built and deployed various AI-driven systems that sought to fight the pandemic. For example, researchers developed systems that automatically analyzed CT images to provide the probability of COVID-19 infection to rapidly detect COVID-19-related pneumonia. Since AI can locate lesions in seconds instead of hours, it can significantly reduce the workload for already overburdened physicians. Other models were developed and deployed throughout COVID-19 to help understand clinical severity and identify the patients most at risk of serious illness and even death. By deploying AI, healthcare systems could prioritize which patients needed to be hospitalized and provided immediate care, and early care was shown numerous times to help with health outcomes.
At the same time, AI systems gave the federal government and state governments insight into where resources were needed most critically. They utilized AI-driven long-term forecasting models to understand the scope and spread of COVID-19, as well as drive site selection during the vaccine trials by predicting where outbreaks were most likely to occur up to eight weeks before cases increased. This could forever change how we enroll individuals into clinical trials, which are typically constrained to research hospitals or highly manual processes. Improving and streamlining the approval of vaccinations is the golden ticket to infectious disease prevention.
Organizations across the healthcare and technology industries also stepped outside of the box to create at-home COVID-19 antigen tests, many of which have an accompanying gamified platform. By combining physical antigen tests with AI and an accessible digital platform, patients are better able to understand their risk of being contagious with COVID-19. Arming people with information about their COVID-19 risk through innovative solutions powered by AI is the solution for slowing and preventing future pandemics.
As a leading nation in health research and technology, we have a responsibility to do better, and we must ensure we can more quickly contain this type of outbreak in the future. By leveraging the importance of complete, accurate, and representative data and combining it with the power of AI and public-private coordination, we can and will be ready to stop future pandemics.
You forgot that politics will have to be programmed in your AI. On both sides of the aisle, misinformation, doing things for show and not effectiveness, was rampant. Good luck trying to extricate politics from public health now.
I dunno. Seeking an answer in AI for America’s healthcare woes seems a little desperate. LIke, adding one magic new ingredient will suddenly “transform public health”.
Has that ever worked?
It seems to me that politicians who are trying to make impossible choices between the economy and public health demands, won’t be much impressed with AI recommendations. Unless the AI recommends Unobtanium!
Will AI (I prefer the term Machine Learning – ML) magically fix all the incentives that have been created in the US health system due to FFS? No. But that’s not what Sally is implying.
She gives some specific examples of how AI/ML helped drive efficiencies and made care better during this pandemic (within the constraints of current system).
That’s been our experience as well at Parkland and broadly with Dallas County, TX public health department. Specifically related to COVID-19, we were strategically able to utilize ML derived insights to improve many aspects of COVID-19 related workflows including patient triage, testing site optimization, identification of COVID-19 hot spots, and vaccination prioritization and understanding of herd immunity to just name a few. And we continue to build on it to help with other public health issues such as access & equity, other communicable diseases etc. AI/ML is a useful tool and is constantly getting better and when aligned with strategic goals and with the right moral compass can be really useful.