Finance and insurance companies have been leveraging synthetic data for many years to improve their workflows while ensuring information confidentiality. With the COVID-19 pandemic, scientists who are striving to find ways to combat the virus have considered synthetic data. How can this technology be of use in healthcare, and how does it help to cope with the pandemic?
What Synthetic Data Is
Without going into convoluted definitions, synthetic data is artificially generated data. It is similar to real data but doesn’t copy it. Synthetic data is generated automatically with the help of dedicated algorithms. It can be in the form of text, video, image, audio or information from tables.
Synthetic data can be applied in various areas. Waymo uses it to train its driverless cars. American Express uses artificially generated financial information to improve its fraud detection system. Synthetic data helps companies calculate risk accurately while protecting real customers’ data. The OpenAI team has taught the language model GPT-3 to compose texts similar to those that a human would write. A program belonging to Nvidia creates photos of people based on images of real individuals.
In healthcare, using synthetic data means that important analysis can be done without associating particular people with their medical records. After the outbreak of the coronavirus pandemic, the need for applying such data in healthcare has increased.
Synthetic Data in Healthcare
Secure data exchange is one of the major concerns in healthcare. According to the General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA), any confidential information can’t be disclosed without the consent of the person it belongs to.
Information from patient records must be stored and transferred securely Using the data without specifying the name of the patient is prohibited, too, as it is possible to identify an individual based on the data set.
That’s why it is more lawful and secure for researchers to create synthetic data as they conduct studies crucial for humanity. Prototypes of training software for machine learning models are trained with synthetic data so they can work with real patient data later. Developers don’t have access to the real information — they can’t read it, extract it from software or use it in any other way.
If the true patient is not at risk of being identified, information from real medical records can serve as the basis of synthetic data, though joint case records are much more commonly used. There is also the sort of approach that Mitre offers via Synthea, an open-source tool that allows for creating fictional patients based on publicly available information: scientific research data, disease statistics, demographics and so on. Although the generated dataset is not as reliable as “fakes” of the real medical records, the platform continues to be improved under the auspices of the U.S. government.
Although synthetic data is not suitable for studying real diseases and treatment methods, it can be the basis for the development of applications that allow for using real data without breaking the law.
Thus, synthetic data opens access to research and development of new technologies in healthcare.
Practical Applications of Synthetic Data
Soon after the pandemic outbreak, Israeli scientists began testing synthetic data technology based on EMRs from the last 20 years. Sheba Medical Center — the country’s largest hospital — used the MDClone platform to synthesize the data of its coronavirus patients.
The healthcare facility invited analysts who collected all the information about the virus from the data set. The result of the cooperation of medical researchers and software developers was an algorithm that helps the hospital staff decide when to prescribe medications or when inpatient treatment is needed.
The software allowed Sheba Medical Center to combine the data from its EMRs with the data belonging to another Israeli healthcare facility — Maccabi HealthCare Services. This provided scientists with a broad view of the course of the individual disease, helping estimate coronavirus outcomes for each person. Without synthetic data technology, the project would have taken much longer as permission to use confidential information would have been required.
Of course, medical scientists can’t rely solely on synthetic data in their research, but the data lets them easily analyze an unlimited number of hypotheses that can lead to significant time savings during the approval of new drugs for real patients.
Although some data security experts doubt that synthetic data in healthcare can ensure patients’ anonymity, this data is extremely useful in prognostications, survival analysis, clinical trials, decision-making and more. Such technologies will accelerate innovation in healthcare while helping scientists comply with legislation.
The FDA approved the first new drug to treat Alzheimer’s in nearly 20 years in recent weeks, and it is a prime example of why our spending on healthcare is so unnecessarily high and not slowing down anytime soon. The new drug is Aduhelm, an infusion therapy developed by Biogen.
A panel of 11 scientists reviewed the research and science for the FDA, and 10 voted against approval of the new treatment, while one was undecided. In fact, three of the scientists have resigned over the approval. The FDA moved to approve Aduhelm despite the lack of evidence that it either cures or slows the progression of the Alzheimer’s and has given the company nine years to conduct a confirmatory trial.
What is the actual cost of this treatment? The drug price is set at a whopping $56,000/year, and the overall costs will be much higher. According to the Kaiser Family Foundation, when related costs are included (the testing to monitor for brain bleeds and other possible side effects, outpatient facilities and staff, etc.) the real price tag will be more like $100,000/year.
What is the cost to the individual? Copays will be as much as $11,500 – nearly 40% of the average income for a Medicare, enrollee according to the Kaiser Family Foundation.
How will this affect Medicare Part B premiums? As an infusion given in a provider’s office, administering Aduhelm will be covered by Medicare Part B. The current average premium for this coverage is just under $150/month. This will almost certainly have to be increased, so the impact will be widespread across Medicare enrollees.
What are the potential impacts to our overall healthcare spending? Biogen estimates that 1 million to 2 million Alzheimer’s patients match the patients studied in the clinical trials. Overall, we have approximately 6 million people who have Alzheimer’s in the U.S., and most are enrolled in Medicare. Interestingly, the FDA has approved the medication widely not just for those who are in the early stages of the disease with mild symptoms like those in the trial. If 1 million people are given this treatment, it could cost Medicare $56 billion annually. Medicare Part B spent $37 billion in total on drugs in 2019. Total outpatient Medicare drug spending with pharmacy prescriptions was $136 billion for 2019.
Biogen’s estimates of future sales are seemingly conservative. The company and other analysts are expecting $103 million in sales this year, about $1 billion in 2022 and $5 billion-plus in 2023.
When you factor in the incentives paid to prescribing physicians by Medicare ($3,360 for each prescription in this case), it seems we have a real problem on our hands.
The Centers for Medicare and Medicaid Services could decide not to cover Aduhelm, but if the past is any indicator this is not likely. Private insurers that provide Part B benefits could also place some limitations on the drug’s use. But, all things considered, It is clearly time for us to take a serious look at how we have allowed a fifth of our economy to get to this point.
The U.S. has grown to become the second-largest Spanish-speaking country in the world, only behind Mexico. The impact the Hispanic population has had — and will continue to have — on this country is quite vast, especially with purchasing power nearing $2 trillion. This serves as an extraordinary opportunity across a variety of markets, insurance being one of them.
While this demographic shift has created a once-in-a-generation opportunity to serve a historically underserved market, many in the Hispanic community continue to struggle when it comes to getting insurance, which goes for both personal and commercial insurance. Not only is it challenging for this customer to find and evaluate options in Spanish, but the customer is also often forced to brick-and-mortar brokerages, two inconveniences that could easily be solved.
Conning Research predicts that there will be more 30 million new Hispanic drivers searching for insurance throughout the next 30 years. Why not meet the need?
To do so, it’s important to understand the distinctions in purchasing behavior and preference when comparing the Hispanic market with the broader insurance market. These differences are critical in effectively servicing this market.
Younger Customer = Opportunity for Longer-Term Value
Because the Latino consumer is younger than the general population, with a median age of 28, this younger customer base translates into a longer average lifetime as they enter their prime earning years. The trend in increasing purchasing power also reflects the increasing needs for personal insurance, from car insurance today to homeowner’s insurance in the future.
This is also an opportunity to build a customer base that over-indexes on brand loyalty, as more than 50% of Latinos are “likely to find a good source and stick with it,” compared with just over a third of the broader market. A relationship can unlock long-term value and pay dividends over time.
Being Available in Spanish Is Non-Negotiable
Most of the Latino market prefers to speak Spanish, so, if you’re interacting with prospective Latino customers, you should consider providing Spanish language materials or support. You can hire employees who speak Spanish or contractors who can translate your marketing materials. This will open up access to a market that cannot be uniformly served in English.
Understand What’s Needed and Prioritize Relationship-Building
Most U.S. Hispanic consumers today primarily purchase auto and health insurance. Given the socioeconomic and cultural characteristics of our community, many Latinos focus exclusively on buying the insurance they need.
While many insurance companies want to serve the Latino market, they often ignore the fact that customers are primarily purchasing basic auto insurance and basic health insurance. While maybe not the ideal purchase for insurers right off the bat, these products can serve as an entryway to serving the future insurance needs of the customer, or even providing other valuable and complex products. In many working-class neighborhoods, for example, auto insurance is purchased at the same place where you do your taxes or get loans, so by building a relationship with the Latino consumer you will have the opportunity to earn more of their business in the future.
Building a relationship with your Latino customer first requires meeting them where they are. Latinos are extremely tech-savvy and spend more time on their mobile phones than any other demographic. Latinos do research, watch television and purchase all over their phone. Using mobile channels, like social media and text messaging, to reach and service our community will help lower your costs and increase chances of success.
Putting the proper measures in place to find and attract Latino customers is a no-brainer. It’s not only the right thing to do to provide this underserved demographic with easy-to-access and fair-priced insurance, but it also means increased customer acquisition and revenue numbers for you. To move beyond intention and into implementation, work with your Latino employees to better understand the community, hire employees who know the market and explore partnering with other trusted service providers in the community.
By taking advantage of this opportunity, insurers will reap the rewards throughout the next decade. And if you don’t meet the market need, someone else certainly will.
Warning: While this is not quite as disturbing as the poor woman with a spider in her ear on Twitter, I’m still sleeping with one eye open.
Also, TLDR: Prices play a key role in free markets, right? They help set supply and demand and indicate value. My guess is that the vast majority of healthcare and by extension health insurance in the U.S. is a badly functioning market in part because prices are nearly irrelevant.
The story, in which I bravely warn my family of danger.
Not long ago, I went to bed early and was woken up by what sounded like a very loud fly knocking against the inside of our bedroom window. I turned on the light to see a bat weaving and darting just overhead. (We have low ceilings – the bat was way closer to me than any bat should be.)
I’m happy to report that I bravely warned the rest of my family by shrieking (repetitively) at the top of my lungs; the bat got a broom-to-the-rear-end assist out the bedroom window.
Really, there was no dignity to be had for anyone that night.
But, I had a bigger problem. I had two little marks on my arm that Dr. Google suggested could be a bat bite; my primary care physician’s office told me to go to the emergency room.
Turns out, showing up as bat woman at a suburban ER late on a Monday night makes you an absolute rock star.
I ended up getting the rabies vaccine and a shot of human rabies immunoglobulin, which was absolutely the largest shot I have ever seen in my life. Follow up was three more doses of vaccine over a three-week period.
And that’s why I received an $18,289 bill: $1,120 for the ER visit, which didn’t shock me, and $17,169 for the shots, which did.
Pic, or it didn’t happen:
I looked at the bill, nearly fell over and fortunately made it as far as the “You owe” box. My share, luckily, is $300; my insurance company paid $3,986.
Then I started thinking.
Why do we pay people to make up prices for healthcare, and what does it feel like to do that job?
Is $17,169 even a real price? Does anyone actually pay that?
I’m guessing no, if my health insurance company can get the hospital to knock 74% off the tab. And, given the way health insurance works in the U.S., I’d assume that every other insurance plan has negotiated some similar discount.
So, how does it feel to sit in a room with Excel open, presumably, and figure out what the rack rate is for care? What are the numbers even based on, if anything? And how does it feel to know that the only people who may get charged that are the people who are uninsured? (More on that later.)
The Centers for Disease Control and Prevention’s page on the cost of rabies prevention (published June 2019) pegs the expense for the immunoglobulin plus the four vaccine doses at $1,200 to $6,500, for the medications alone, not the hospital charge. So, how do you get to north of $17,000? And, according to First Databank, a company that monitors drug prices, the price of the drugs I received has increased 388% in the last decade, for no obvious reason, as best I can tell. Think the CDC missed that.
What is it like to be on the opposite side of this pricing, to be negotiating on behalf of the health insurer? Does someone go back to the boss to say, “Hey, I negotiated us a 74% discount!”? Doesn’t that doesn’t just strike everyone involved as laughably bizarre?
Why do we ask people to do useless work? What does that do to them, and what are we failing to grow or build or improve by asking them to focus on this nonsense?
Was it cost-effective to treat me?
This bill got me metaphorically scribbling on the back of an envelope.
Is $17,169 a fair price? Well, maybe that’s a silly question, because that’s not what my insurer and I paid together.
Let’s just assume that the fair price is about $4,000, which is roughly what we paid for the shots.
Rabies is a nearly completely fatal disease, and the onset of symptoms to death is about seven days. Let’s say the resulting hospitalization and treatment costs $100,000 (source: wild guess). Soullessly assigning no economic value whatsoever to my life (which I, by the way, value greatly), we should be willing to pay for 25 courses of rabies shots to prevent one case, or whenever the risk of catching rabies is 4% or higher.
What was my risk of catching rabies? The estimates I saw suggested that anywhere from .1% to 10% of bats are infected with rabies, depending heavily on local conditions. It is easily transmissible, apparently, especially with a confirmed bite, so let’s assume that, if my bat friend had rabies, I would get it.
So we’re within a wide margin of error based on avoiding costs. What if we ghoulishly try to value a life? There are lots of papers on this topic. I’m no expert, but I’m going to pick $100,000 per remaining high-quality year of life left, which seems roughly reasonable based on the literature. Let’s assume I have 40 high-quality years left, or $4 million of value.
Treating a known bat bite definitely makes sense.
Maybe the cost makes sense from that perspective? But would you need to consider it as part of all of the healthcare I will receive from this point on in my lifetime? Or just at this moment, with this choice?
I don’t know, but I’m pleased that I was able to make a rational economic decision on the fly.
How did we end up in a place where life-saving, cost-effective treatment is most ruinously expensive for the people least able to afford it?
If I had been uninsured and gone to the hospital, I’m going to assume I would have been treated, at least with the first set of shots, though I don’t know for sure. If I had gotten the full series of shots, how much would I have been billed? The full rack rate? The worst negotiated rate with an insurance company? Something different?
(My $300 ER copay alone represents a significant burden for many families, including those affected by joblessness or reduced working hours because of COVID.)
I did some research – it turns out the major manufacturers of rabies treatments have programs to pick up the cost for those who can’t afford it. I’m sure this is an enormous relief for those who qualify, which I desperately hope is everyone who can’t pay the price. I also hope it’s easy to apply.
But, systemically, how does this make sense?
We set fake prices that are beyond crushing for most families. Then we don’t charge them to people who are insured. We save the worst prices for people who don’t have health insurance, who are even likely less able to pay ruinous prices than those who do have insurance. So then the manufacturers have programs (complete with separate paperwork!) to waive or minimize the cost for the uninsured, at least those who apply.
And since the manufacturers still need to make a profit, those waived costs actually get funneled back into the pricing the insurance companies negotiate with the healthcare systems.
Again, what information is in the pricing for these drugs? And in what ways are the healthcare and health insurance markets dysregulated as a result of the lack of clear pricing information?
As both the consumer of our healthcare insurance and the employer paying for it, how am I supposed to assess this situation?
My co-founder and I picked out our health insurance plan. While there seemed to be a whole lot of choice for a business of our size, the plans were really all the same… $2,000 deductible with $50 copays? $2,005 deductible with $49.75 copays? And so on.
I’m exaggerating, but not by much.
The service our healthcare plan provides is fine, and the cost seems reasonable, but how do I assess whether the plan is as efficient as it could be? The costs paid to healthcare providers get passed back to us (all of us insureds) through pricing. If the plan pays too little, it isn’t fairly compensating the medical providers, which will eventually refuse to work with our health insurance or go out of business, meaning we have an availability problem. If the plan pays too much, we pay for it.
Am I supposed to be really pleased that our health insurer negotiated a $14,000 discount? Or should I be mad it didn’t negotiate a $15,000 discount? How would I know? How could I make a more rational decision?
The answer is…I can’t. There’s so little information in the pricing, and so much opaqueness, that we’ll have to make another decision next year based on price and service when we renew our health insurance.
Why do we tolerate the healthcare and health insurance mess we have in the U.S.?
I’m willing to believe that all of these ridiculous pricing mechanics exist for a reason, but do the reasons still make sense?
There are so many distortions here…
Even when I am both the buyer and user of employer-sponsored health insurance, I don’t have the information I need to make any kind of a rational decision except what makes the most sense for us for the next year.
There’s no easy way to understand how much healthcare will cost before treatment, especially in emergency situations.
The list prices are no more than sky high caps on medical procedure prices.
I’m no expert in health insurance, but I see that same underlying issue here that I do in my own familiar property and casualty space – massive systemic complexity.
In the P&C space, I’d argue that most of the complexity derives from old court cases that created boundaries between lines of insurance. This led to technology solutions for each line and specialized staff and culture to handle input into, maintenance of and output from these ossified systems. The market need for those silos has blurred or disappeared, but thus far they’ve been indelible marks in the insurance landscape.
The same is probably true of health insurance. I’ll add the root of employer-sponsored health insurance (which usually separates the buyer from the end user), which stemmed from wage controls during World War II.
There’s an interesting historical summary in National Bureau of Economic Research Working Paper 14839. In short, fringe benefits were excluded
The issue is that systemic problems require systemic solutions. Systemic solutions are so much less attractive than quick fixes – they don’t sound catchy, they don’t boil down to one sentence, they take time to implement.
Yet, as a people, our failure to fight for systemic solutions is surely catching up with us.
from WWII wage controls, which caused employers to add more benefits to attract and retain workers. Health insurance existed in a fragmented way before this, but really came into its own and was firmly established as an employer benefit in the period. And now, 55% of Americans get their health insurance through their jobs, according to the Census Bureau. This has to be a piece of the complexity.
Public Service Announcement:Rabies is really nasty. Nearly always fatal, and really, really nasty – it kills tens of thousands of people worldwide each year, mostly in places where rabid dogs are common. It’s also not specifically spread by animal bites; it’s spread in the saliva of ill animals, so a scratch or a lick to a mucus membrane can also spread it. If you may have been exposed (including if you wake up in a room with a bat — bat bites don’t necessarily hurt and can be nearly invisible), you really need to talk to a medical professional.
Postscript: I had a long conversation about CEO pay at pharmaceutical companies with someone regarding this whole situation. CEO pay can be grotesque, especially at companies that do not pay their workers living wages and don’t provide decent benefits. However, cutting CEO pay is largely an issue of equity, not much of a solution to this cost issue. I looked up the 2019 salary of the immediate past CEO of Sanofi Pasteur, which manufactured the shots I was given. If he had worked totally for free, applying the savings evenly over their revenue, my bill would have been $4 lower.
COVID-19 is, in many ways, still a disease of uncertainty, both in the severity of its symptoms and the scale of financial hardship it could ultimately cause. And no business is better equipped to confront uncertainty than insurers.
So it is perhaps unsurprising that life and health carriers are responding to the pandemic with a variety of consumer offerings, from complimentary, compassionate benefits to new protection products and services.
After performing a global review of responses to COVID-19, RGA identified six primary trends:
Compassionate and Complimentary Benefits
Insurers have sought to build trust and goodwill by offering complimentary COVID-19 coverage as a compassionate benefit or marketing expense, with an emphasis on policies with lower face value to manage overall exposure. A multinational insurer in Hong Kong, for example, has begun offering an additional hospital cash benefit of HK$600 (equivalent to US$77) per day for covered clients who may be required to undergo a mandatory quarantine in a hospital or isolation center. Similarly, the local branch of another major insurer in Thailand partnered with a leading telecom operator to offer a market-first, free-of-charge COVID-19 coverage benefit to customers. If a customer requires treatment in a hospital, he or she will receive a hospital indemnity benefit of up to THB1,000 (equivalent to US$31) per day. Still another multinational partnered with a Singapore-based insurer serving private hire drivers to offer a complimentary benefit for all of these essential workers as part of the company’s Group Prolonged Medical Leave insurance policy.
Other forms of consumer relief have proved popular, including premium holidays, grace periods and reductions to policy/premium amounts. Many insurers globally have offered grace periods for premium payments either voluntarily or at the request of local governments and regulators. On the health insurance side, some insurers have waived cost-sharing, co-pay and other deductibles for inpatient hospital admissions due to COVID-19. Some have also sought to support healthcare first responders and other frontline workers through donations of personal protective equipment (PPE) and other charitable efforts.
COVID-19 emerged in Asia, so it stands to reason that regional insurers have been first to develop hospital cash and comprehensive care products offering standalone protection in case of diagnosis. For example, one Bangkok-based broker and insurer teamed up to offer Thailand’s first policy that provides cash upon diagnosis with the coronavirus. Similarly, a major Indian insurer offers a COVID-19 support plan, providing end-to-end treatment services, including consultations with qualified doctors, to policyholders who become infected. In Malaysia, a major multinational has introduced a COVID-19 hospital assistance program with an upfront one-time cash payment upon hospitalization with the disease. The program includes a one-time cash payout for family assistance should dependents also be diagnosed. It also includes other assistance for consultation and treatment costs while in isolation or intensive care. Similar products are now emerging across Europe and North America.
The word pandemic derives from the combination of two Greek words: pan (“all”) and dēmos (“people”). But, while all are at risk of contracting the coronavirus, a few face far greater danger due to the essential public services they must perform. Insurers are customizing certain offerings to serve these frontline workers. In China, a first-in-market COVID-19 medical worker insurance program pays cash compensation upon diagnosis. Similarly, healthcare personnel at specified primary and secondary public hospitals, treatment centers, and pharmacies are eligible to sign up for another Chinese insurer’s COVID-19 coverage for free. Another insurer launched COVID-19 coverage targeting shopkeepers in India, with the product paying 100% of the sum insured, irrespective of hospitalization expense, upon diagnosis.
Health and Wellbeing
The coronavirus not only co-opts our cells, it exploits our fears. A lack of clear information and shortages of available testing have compounded the problem in some locations. U.S. insurers responded with new consumer plans that seek to bundle mental wellness services with physician care to address public anxiety with clear and actionable medical guidance. One U.S.-based healthcare carrier repurposed its existing telehealth application for mobile devices. The app now provides a coronavirus assessment based on guidelines from the U.S. Centers for Disease Control and Prevention and the U.S. National Institutes of Health. Customers can connect directly to a board-certified doctor via text or secure two-way video call and use the app to discuss the assessment results. Another U.S-based provider of healthcare IT solutions and services launched a new telehealth product to help physicians and patients stay connected during COVID-19 through real-time video technology. A number of mental health schemes have also emerged around the world with an emphasis on technology to address social isolation.
Artificial Intelligence (AI) has been coming to medicine, and insurance, for some time. Now the spread of COVID-19 may present a new opportunity to increase use of smart apps and chatbots. A number of insurers are relaunching and rebranding existing AI applications to meet surging diagnosis and informational needs in an era of social distancing and staffing shortages. In China, one major insurer launched a smart, AI-based audio screening system for COVID-19 to strengthen epidemic control and prevention through automated interviewing and risk assessment. Another U.S.-based case manager launched a coronavirus chatbot to answer questions related to COVID-19 and assist in diagnosis, and a multinational in Hong Kong retooled its mobile application to assist in coronavirus contact tracing. Much remains unknown about the overall effectiveness of these emerging technologies, but the increasing use of AI is a trend that merits monitoring.
New Approaches to Sales Operations
COVID-19 has been dubbed an “invisible enemy,” but its effect on the insurance industry has been very apparent. As traditional evidence and sales channels have been disrupted by lockdowns, carriers have moved to accelerate a transition to alternative evidence, simplified and accelerated underwriting and digital distribution.
“Selling at a distance” is a hot industry topic, and those insurers with relatively strong digital capabilities may be best-positioned, while others are playing catch-up. A major multinational recently launched a digital enrollment system, while another unveiled “simple life insurance” to be sold online. Another insurer is now using WhatsApp to deliver policy and renewal documents. One carrier has simplified the claims process for its critical illness policyholders. Upon diagnosis of COVID-19, the policyholder needs to only submit a certificate from a government medical officer to receive a lump sum payout rather than the more copious paperwork typically required.
As the pandemic unfolds, we expect more offerings to emerge. In the medium term, it may not just be health and safety concerns that drive offering design, but the state of the overall economy. Interest rates and slowing economies are placing renewed pressure on insurers to reassess less profitable offerings, such as those with generous guarantees, and to emphasize capital efficiency in the overall product portfolio. Against this challenging backdrop, it is unclear how many product innovations of all kinds are languishing in the exploratory phase versus being introduced to consumers at this time.