Modeling the pandemic risk
How to model the COVID-19 pandemic and forge a path to better risk insurance.
Artificial intelligence (AI) has slowly yet surely pervaded our daily lives; we delegate mundane, thoughtless tasks to automated systems, attempting to purge some of the busy work and free up physical time and mental space to either be more productive at work or indulge in a little more play. We use Alexa, Siri, or other smart assistants on our personal devices to help check off our shopping lists, update reminders, and make home security safer. AI algorithms are becoming increasingly sophisticated and more ubiquitous all to do what human beings can do today – except faster, cheaper, and better. From programming to cleaning our floors with smart vacuums to a not-so-distant future of self-driving cars chauffeuring us to personalized co-working spaces, the progress of AI is everywhere and likely here to stay.
Not so surprisingly, AI has found its way into healthcare and is making noticeable inroads in several areas. When it comes to general health and wellness, AI in combination with the Internet of Things (IoT) empowers us, the consumers, to take charge of our health and pay attention to our vitals, fitness, and diet. Smart medical devices and/or wearables have the ability to continuously track health conditions. There are many examples of AI-based healthcare solutions. Google has DeepMind Health, which combines machine learning systems and neuroscience to build algorithms that imitate the human brain. IBM’s Watson for Health, a form of AI, can store far more health and treatment information and review it exponentially faster than any human being. In 2016, a partnership between Barrow Neurological Institute and IBM Watson made a groundbreaking discovery that was possible by Watson’s review of thousands of pieces of research to identify new genes linked to ALS (amyotrophic lateral sclerosis (ALS).1 As an example of a more tactical application of machine learning, Milliman developed a model2 to predict an individual’s likelihood of being diagnosed with opioid use disorder, thus enabling healthcare organizations to engage in a more timely manner with at-risk individuals. In 2018, researchers from the U.S., Germany, and France trained a deep learning AI platform to identify skin cancer with greater accuracy than human beings. The result was that the AI platform found 95% of the melanomas, outperforming most dermatologists at 86.6%.3
The potential benefits of integrating AI into our healthcare ecosystem and devices are numerous, including automation of tasks, analyses of patient data, and delivery of better healthcare. Analyses of massive amounts of data can lead to patient insights and predictive analytics, facilitating healthcare interventions to occur at more appropriate times and potentially avoiding expensive medical situations for patients, payers, and providers. Simply put, AI and predictive analytics have the potential to create better outcomes for the healthcare community as a whole.
However, there are some challenges that AI faces in terms of adoption and widespread use in healthcare.
The top concern raised by far is the handling and privacy of patient data. Even though technology companies, device manufacturers, and other stakeholders are increasingly investing in state-of-the-art technology to address this concern, protecting patient data remains a high priority.
Another challenge is to form meaningful and marketable insights. That is, while the technology exists to capture and analyze large amounts of data, the inferences themselves may not make sense without the right format, context, and key metrics for the right audience to take actionable steps.
Finally, there is a notion that AI may eventually replace care providers, when in reality, AI technology and health professionals will work hand-in-hand to create the best possible patient outcomes. This in large part is due to the fact that predictive analytics and the use of AI in healthcare are relatively new and yet to gain the trust of professionals in the medical community and consumers.
Despite these potential pitfalls, public and private sector investment in healthcare AI has been remarkable. According to a February 2019 Forbes article, PricewaterhouseCoopers says more than a third of provider executives were investing in AI, machine learning, and predictive analytics in 2018. And some estimates predict total healthcare AI investment will reach $6.6 billion by 2021. Accenture predicts AI can help address 20% of unmet clinical need. Another analysis predicts that top AI applications could result in annual savings of $150 billion across the healthcare industry by 2026.4
With the potential to pare back healthcare expenditures and create better outcomes for participants in the healthcare community, the future of AI in healthcare is looking bright.
1IBM Watson Health (December 14, 2016). Barrow Identifies New Genes Responsible for ALS using IBM Watson Health. Press release. Retrieved on November 13, 2020, from https://www.prnewswire.com/news-releases/barrow-identifies-new-genes-responsible-for-als-using-ibm-watson-health-300378211.html.
2Boschert, J. et al. (January 2020). More accurately assessing opioid risk in Medicare members. milliman.com. Retrieved on November 13, 2020, from https://us.milliman.com/-/media/milliman/pdfs/articles/more_accurately_assessing_opioid_risk_in_medicare_members.ashx
3Computer learns to detect skin cancer more accurately than doctors. (May 28, 2018). The Guardian. Retrieved on November 13, 2020, from https://www.theguardian.com/society/2018/may/29/skin-cancer-computer-learns-to-detect-skin-cancer-more-accurately-than-a-doctor.
4AI And Healthcare. (February 11, 2019). Forbes. Retrieved on November 13, 2020, from https://www.forbes.com/sites/insights-intelai/2019/02/11/ai-and-healthcare-a-giant-opportunity/#156b16f64c68.
AI and predictive analytics are transforming healthcare. Here’s how.
The potential benefits of integrating artificial intelligence into our healthcare ecosystem and devices are numerous, including automation of tasks, analyses of patient data, and delivery of better healthcare.