In this year’s AIMed Europe took place in central London between 17 and 19 September, representatives from two large insurance firms – Reza Khorshidi, Chief Scientist, Global, AIG and Dr. Sneh Khemka, President of International Population Health solutions, Aetna International, detailed how artificial intelligence (AI) is not only changing healthcare but also health insurance, even though meaningful adoption is still yet to happen. 

A prediction problem 

Khorshidi said photography turned from a chemistry problem into a computing problem because people no longer rely on films but digital cameras or their smartphones. Likewise for AI, it is creating excitement because it has the potential to turn many current problems into a prediction problem. For example, in radiology, it’s no longer about an expert looking at so many images over time, machine learning algorithms are going to study these images and make the predictions. 

In the case of insurance, Khorshidi explained, the idea of life insurance originated from the Churches of Scotland, whereby priests contribute the same amount of money to a fund, so that their families have some sort of financial support when they passed away. Its foundation lies in the use of Bernoulli’s Law of Odd Number, whereby one can predict how many tosses are going to be head or tail, if they throw a coin a large amount of time. When this idea is being applied to healthcare, we are able to estimate roughly 20-30% of the people have or are going to have diabetes based on the population of the country. 

Reza Khorshidi presenting on stage.

However, the challenge is, given at any instance, it is hard to predict whether a 20-year-old healthy individual is likely to develop diabetes or across the UK, how many grownups are going to have diabetes in the coming year. This is where precision medicine will be handy and when insurance turns its business model from preparatory to anticipatory, as a result of predictions made by AI. 

Khorshidi cited Amazon’s “anticipatory shipping” idea, in which consumers not only receive purchase recommendations online, but as long as an Amazon delivery agent is nearby, they could go up to the consumers right away and suggest if they are interested to buy something. All based on the algorithmic prediction that made up a knowledge of you and what you are going to buy next. 

A new primary care provider 

Dr. Khemka believes the health insurance industry is entering the door of augmented intelligence but in general, it still fails to adopt true AI. Nevertheless, health insurance no longer hide behind screen to reject people’s medical claims. They are more likely to partner with healthcare companies and venture into wellness or even primary care.

For example, when consumers walk into one of these so-called health hubs, they are likely to be assessed by a set of sensors, coupled together with health information obtained from smartphones or wearables, predictive algorithms are going to forewarn and help insurance to get in from of what is right for the consumers, to avoid any unnecessary claims. In fact, as Dr. Khemka pointed out, about 50% of the customers undergone Aetna’s digital primary care do not need to go on further for physical primary care visits. Over the two years of Aetna’s merger with CVS, there was a 23% reduction of unnecessary medical claims. 

Dr. Sneh Khemka presenting on stage.

Most importantly, AI is more efficient in sifting out real medical claims from the deceiving ones. Presently, fraud made up of 7-8% of the overall medical claims. Some of these deceits are done systematically by overseas hospitals with formal websites but non-genuine doctors and patient accounts. As such, it is challenging for a human to verify each and every of these claims, but with pattern recognition, details that previously went undetected by a human are now clearly presented. 

In short, both presenters thought as insurers, digital software is capable of providing care in a much more meaningful way outside the traditional hospital environment by cutting down on the cost and going extra miles to cater for better services and experiences. 

Author Bio
synthetic gene empathy chinese artificial intelligence data medicine healthcare ai

Hazel Tang

A science writer with data background and an interest in the current affair, culture, and arts; a no-med from an (almost) all-med family. Follow on Twitter.