Artificial Intelligence (AI) and Blockchain in Medicare: Technologies to Watch in 2019

Expect more mass production of audits.

Looking forward to the remainder of 2019, we will see a number of exciting information technologies (IT) that will reshape healthcare and Medicare. These include artificial intelligence (AI), the cloud, DNA sequencing, in silico testing of new molecular entities (NMEs), and distributed ledger systems, also known as “blockchain.”

We also will continue to see more mass production of audits. In Medicare, our government will continue to outsource its work to highly paid “for-profit” Recovery Audit Contractors (RACs). They will use big data techniques to continue to squeeze the little guy. Of course, some true fraud will be found here and there, but not enough to put even a tiny ripple in the overall amount of healthcare spending.

So we can expect the synthetic and bloated merry-go-round economy of auditors, healthcare providers, coding experts, attorneys, administrative laws, inspectors, and recoupments, while endless arguments over medical necessity and documentation will continue unabated.

The little providers will continue to get by in a type of police state for data, in which more fraud is manufactured by IT than is actually present.

Organized crime has been with us since the Phoenicians set up their first trading networks linking the Middle East to Spain. It survived Greece and Rome and Early Europe. It is with us today. Only a fool would expect this corruption to disappear in 2019. So you can expect more system hacks, more ransomware episodes, more viruses and other malware, and a continuing boom in the trafficking of illicitly obtained patient data on the dark web. Everyone has to make a living, even crooks.

Expect more medical-industrial cyber-espionage. IT will be used to steal R&D data, abscond with trade secrets, embezzle opioids, and plant counterfeited drugs into the mainstream, where no one suspects they are fake. As patients expire because the drug they were taking didn’t work, many will continue to wonder if a real drug was used or a fake.

The management information systems (MIS) in hospitals, insurance companies, and other providers will continue to be targets of information terrorism – the unauthorized change, corruption, or destruction of medical data.

Our law enforcement and government investigators will continue to be incapable of providing much help. These days, hackers have a better chance of getting a parking ticket than getting caught breaking into an information system.

Is the IT infrastructure underpinning our healthcare system considered to be an essential element of our national infrastructure? Should it receive the same level of national security protection as our financial markets, nuclear power plants, or the electrical grid? How many hospitals have studied the potential effects of electro-magnetic pulses (EMPs) on their systems? Yet in 2019, our society will continue to face Internet viruses and the Damocles Sword of widespread inter-linked IT system disasters threatening entire companies, hospital industrial networks, or complex supply chains.

Artificial Intelligence
The majority of problems in Medicare claims processing are the result of human error. Rejections often are justified, with the sub-standard and not-quite-complete documentation being entered by overworked, underpaid, and exhausted healthcare workers, for whom the idealism of helping humanity that led them into the medical field long ago has faded.

So if human error is a big part of the problem, then the best fix is to cut out the human element. Artificial intelligence promises to do just that. Compared to humans, AI systems have better memories, can think better and faster, do not get tired, and don’t have expense accounts. The need for doctors in diagnosis and specification of treatment regimens will continue to decrease. AI will allow many jobs to be “repurposed,” which is a euphemistic way of saying “humans need not apply.”

AI promises to transform claims processing by making it faster and more efficient. Results of Medicare audits should improve. AI will be used to identify errors and find the fraud for which everyone is looking. Why use humans to audit thousands of medical records when AI can audit billions? The need for sampling will disappear. AI will make it possible to audit all medical records!

Perhaps AI will reduce the 70 percent of audits overturned on appeal. After all, auditors these days have higher error rates than those they audit.

“It’s the software, stupid!” – CMS Regulation of IT
Hopefully, 2019 will see the Centers for Medicare & Medicaid Services (CMS) take more responsibility in regulating IT. It could start with creating a unified data entry logic that forces conformity with Local Coverage Determinations (LCDs) so as to eliminate provider errors. In the same way today’s IT systems will not allow entry of a letter into a numerical day-month-year field, the CMS unified system logic will not allow providers to file improper claims. Liability for errors will be transferred from the medical provider to the algorithm. In a sense, every single claim will be audited as it is entered into the information system, not years later.    

Audits will not focus on the vulnerable healthcare workers, but instead on how well the IT systems work at preventing fraud.

Distributed Ledger Systems (DLS) and “Blockchain”
Another technology to watch in 2019 is the aforementioned blockchain.

Think about the real criminals (you can see these at the OIG’s “Most Wanted Fugitives” pages online, https://oig.hhs.gov/fraud/fugitives/ and https://oig.hhs.gov/fraud/fugitives/profiles.asp#other-fugitives).

Now relaxing on a white sandy beach in sunny Cuba, the winner seems to be Angel Lagoa, who sucked out more than $30 million in fraudulent Medicare claims, then flew the coop. 

A common element of fraud often is the use of patient and doctor records purchased on the dark web. These are used to file claims to Medicare. Doctors find out years later that they have been “prescribing” durable medical equipment (DME), drugs, and treatments to patients they never have seen or heard of. After getting caught, somehow the perps manage to run off to Pakistan, or Nigeria, or Cuba, or any number of other places, after ferreting away the cash. The princely sum of $30 million goes a long way in Cuba, where the cost of living is 42 percent less than in the United States, in Nigeria, where it is 55 percent lower, or in Pakistan, where it is 71 percent lower.

This is where blockchain comes in.

Blockchain is being applied to medical records. It will integrate patient data. It also will record information from any medical device or testing equipment that touches the patient in any way. Testing of tissue, fluids, X-Rays, or other imaging will be recorded. The use of blockchain is not unlike a patient record carrying around the entire complex medical history of the patient. And it can’t be changed or accessed improperly, because it is heavily encrypted. 

In blockchain, a private key is used for encryption of transactions and a public key is used for decryption. Keys involve very large prime numbers that are multiplied together. The “secret” is that given a very large number, it is computationally challenging to find (“factor”) the two primes that were multiplied to get it. Example: 19 x 31 = 589. Perhaps you can find those two prime factors of 589. But what about finding that 999,962,000,357 is the prime multiple of 999,983 x 999,979 – and those are small numbers in the encryption world!

Putting all of the mathematics aside, the use of blockchain will make much of the criminal fraud we see today impossible. Watch for the first pilots.

Other Technologies to Watch in 2019
There is so much more to look for in 2019 – for example, quantum computing will make simulation and imaging possible at the molecular level, leading to better immunotherapy against cancer, and other develpments, including the following:

• The cloud will improve the security, reliability, and standardization of data.
• DNA sequencing will become even faster, perhaps almost in real time, and certainly more routine.
• In silico field trials will accelerate screening of new molecular entities.
• Testing of all types will speed up and drop in cost.
• Robots will improve care for the elderly – and continue to revolutionize non-invasive surgery. 

Every lonely patient will love talking with their virtual pet. What will be the coding for that?

Program Note:

Listen to Edward Roche on Monitor Monday, Jan. 28, 10-10:30 a.m. EST.

Comment on this article