Artificial intelligence has it all to greatly give advantage to the healthcare sector by way of streamlining clinicians’ administrative tasks, decreasing the costs involved, as well as speeding up the process of drug discovery.
As part of a Senate Health Education, Labor, and Pensions subcommittee hearing held on November 8, lawmakers as well as witnesses emphasized the need of implementing adequate precautions for technology to make sure that its effective and safe use. They went on to express concerns that, without these guardrails, AI could go on to harm patients or even worsen existing inequities.
The only scenario in which the needs of the little guy are actually taken care of is when the big guys have got their interests at stake, stated Senator Ed Markey. As on date, there happens to be a strong demand among major players in AI, and it is very important that one give priority to implementing all necessary protections for those who may be at a disadvantage. Besides this, one must ensure that these measures go on to be executed simultaneously in place of one after another.
Throughout the hearing, the senators, many of whom happened to be from the Democratic Party, did show their interest when it came to the potential of AI in healthcare, but at the same time also raised their concerns about the risks that get associated with this technology.
Senator Roger Marshall pushed the need for caution in case of formulating regulations for AI and also emphasized the significance when it came to avoiding any measures that could hinder innovation or even enable individuals with malicious intent to gain an edge. All across the last five decades, AI & ML have gone on to make unmatched discoveries and, at the same time, significantly enhance healthcare, that too without significant government intervention.
One emerging as well as exciting application of tech happens to be generative AI, which is a cutting-edge tool that generates fresh content such as text or images. Microsoft, Oracle, and Amazon have gone on to create notetaking products with the focus on minimizing the time healthcare professionals spend when it comes to documenting patient visits.
Lawmakers as well as those who have witnessed the tool have applauded the potential of similar products, stating that they have the ability to decrease the heavy administrative burdens that the providers face.
Keith Sale, the University of Kansas Health System’s VP and Chief Physician Executive of Ambulatory Services, went to inform that the application of AI documentation has gone on to significantly reduce the time he spends in the clinic. It is well to be noted that in the past, he would have to spend hours reviewing notes and manually input them into the electronic health record.
He also says that it is a tool and should not replace his practical judgment or the way he makes decisions that happen to impact his patients. So, its purpose is to enhance his practice rather than replace him in his profession.
Professor of Harvard and director of the computational health informatics program at Boston Children’s Hospital, Kenneth Mandl, says that AI has in it to enable clinicians to access and evaluate data that would otherwise be beyond their reach, and it can also serve as an avenue for sharing information and also analyzing the performance of the healthcare system.
This concept happens to be truly remarkable. Transitioning from allotting 80% of the GDP to potentially only 8% or 10%, thereby aligning with the international average.
According to the supervising attorney at the Greater Wisconsin Agency on Aging Resources, Christine Huberty, AI’s certain applications, like the utilization of predictive algorithms by insurers to gauge coverage for care, at present operate with minimal human oversight, unless decisions are contested.
Her agency, which is into offering legal assistance to seniors dealing with coverage denials, used to handle just one to two cases of this nature in a year, but now manages that number in just one week.
One of the recent investigations published by Stat revealed that Medicare Advantage insurers utilized algorithms so as to forecast the amount of care a patient would require, and this resulted in coverage denials and at the same time compelled patients to either pay for their own care or else go through lengthy appeals processes.
Apart from this, Senator Ben Ray Lujan, who happens to be a representative from New Mexico, pointed out that the effectiveness in case of AI is heavily dependent on the quality of the data that is inputted. If in case, algorithms are primarily trained on data from white as well as male patients, there is a risk of biases taking place that could negatively impact other groups.
He put forth a study from 2020, published in JAMA, which went on to explore that deep learning algorithms that use patient data from the US happened to be primarily trained on groups from California, Massachusetts, as well as New York and lacked representation from other parts of the country.
He added that, from his perspective, it is crucial to make use of technology so as to enhance health outcomes and, at the same time, minimize health disparities rather than worsening them.
According to Harvard’s Mandl, designing specific regulations for the enormously changing field of technology is quite challenging, and the government will need to come up with long-lasting approaches in order to effectively monitor the tech.