Compliance Concerns for Artificial Intelligence and Telemedicine

Both AI and telemedicine have some common goals. They both aim to provide medical care without a face to face consultation. They both aim to cut down on costs. They also both require the advice of experienced healthcare regulation lawyers who understand the legal concerns. The main concern is that they may cross the line into giving unlicensed medical advice. There are privacy issues. They both may require informed consent from the patient.

At the core, telemedicine is the ability to offer medical advice through technology. The technology allows a patient to speak with doctors and health providers through a mobile device or a desktop instead of through a face-to-face visit. Telemedicine is used, where authorized, between patient and doctor. It can also be used between healthcare providers such as when a doctor sends out an imaging test such as an MRI for review.

AI can make telemedicine work better by:

• AI is being used in combination with telemedicine to make a better diagnosis. Clinicians are already using telemedicine to diagnose, monitor, and treat diabetic retinopathy. AI algorithms are being used to improve the diagnosis.

According to Fortune Magazine, FDNA is a startup which is aiming to help doctors figure out what kind of disease someone has by just analyzing their face.

“Doctors can snap pictures of their patients and upload them to FDNA’s mobile app, which then spits out a list of disorders they might have by analyzing telltale facial features associated with those conditions (the tech is not a diagnostic tool, but rather a way to narrow down the list of possible genetic suspects). And the company hopes the system can drastically improve the “diagnostic journeys” that those with rare diseases typically face: Such patients, on average, are seen by seven doctors before the correct diagnosis is made.”

With the new AI software, the goal is that no doctor consultations will be required – the software will do the job.

• Better hospital administration and logistics. Fortune also reported that, “Last October, GE Healthcare and the Johns Hopkins Hospital launched a fully digital hub to better manage everyday operations. The Judy Reitz Capacity Command Center gets a constant influx of data about important events at the hospital; it receives about 500 messages every minute from more than a dozen different Hopkins IT systems and with the help of predictive analytics turns this swamp of data into suggestions for action that prevent bottlenecks and get patients both into and out of the hospital faster.

And, according to Johns Hopkins at least, it’s showing impressive early results. The hospital says the command center has shaved more than an hour off the time it takes to dispatch an ambulance to another facility and that emergency room patients are assigned a bed 30% faster than before.”

• Recommending treatments. Memorial Sloan Kettering Cancer Center and a hospital in Florida have teamed up with IBM Watson Health to recommend treatment plans for cancer patients. According to Fortune,

“an astounding 90% [of companies] said they expect to make at least some telemedicine services available to their workers this year. By 2019, nearly all of them will.” Telemedicine allows the worker to be examined by a physician online instead of in the office.

Gartner believes that robotics and other forms of smart machines will be able to provide elder care and other types of home healthcare.

AI may even be used to help doctors input patient data quickly and more efficiently so the doctors can spend more time with patients.

Additional AI uses for telemedicine

Emerj reports that there additional AI uses for telemedicine:

• Virtual consultations. The technology does more than just help a patient talk with a physician. The technology uses machine learning to help analyze clinical data in a patient’s electronic health or medical record (EHR/EMR) to provide patient care recommendations.
• InfiniteMD. This company, based in Boston, recently began a trial phase to add AI features to its previous use of matching cancer patients with health providers and clinical trials.

“When using InfiniteMD and its new AI features, a patient begins by submitting a video consultation request on the InfiniteMD website along with their medical records. Internally, the goal is for the AI platform to recommend treatment options specific to the user’s needs and provide information on treatment availability (i.e. the patient’s country or the United States). The AI platform can also determine if the user meets the patient profile criteria for any clinical trials, according to the company.

According to the company, data used to generate these recommendations and train algorithms includes previous recommendations from medical specialists who’ve partnered with InfiniteMD, other data from medical cases managed by the startup, the latest information on clinical trials and current professional guidelines for oncologists.

Before a patient gets their recommendations, the AI-driven suggestions are sent to a human InfiniteMD admin who reviews this data as well as submitted documents, like medical records. Once the application’s suggestions are verified, the human admin will use the overall data to match the patient with a specialist who most closely meets their needs.

The video consultation is scheduled and takes place using a desktop or mobile device with a camera. The health provider then sends a summary of the consultation to the patient following the session.”

• Diagnostic Support. An example of diagnostic support is the following:

HealthTap, a company based in Palo Alto, California in 2010, claims it uses AI and machine learning to help users identify their potential ailment and its probable causes.

The patient first submits a patient health record which includes such information as the patient’s prior medical history, current medications, age, and gender. A chatbot (software that uses text or audio to conduct a conversation) is used to access the initial patient information.

“User responses are compared with doctor responses to similar cases in the database to generate recommendations for care”

Who is using telemedicine?

Telemedicine is being used for many purposes including radiology and digital imaging. Patients can go to a local facility. The test results can be then be uploaded and sent to a radiology specialist who can read the images and inform a physician what the results are. Psychiatrists can treat speak and talk to patients while avoiding having the patient go their office. Telemedicine also includes prenatal counseling, nutrition therapy, alcohol abuse assessment, and other health matters.

Issues in telemedicine

The ability of AI software to help upload and review images and speak with patients through videoconferencing offers many advantages. Just as with telemedicine though, the ability of the AI to adapt poses legal and ethical issues. An experienced healthcare lawyer can explain what federal and state laws apply. Some of the laws and questions that apply that relate to AI and telemedicine include:

• The unauthorized practice of medicine. These are generally governed by local state law and by the local, state, and federal medical societies.
• The need to obtain informed consent.
• HIPAA (Health Insurance Portability Accountability Act) which governs the security and privacy of electronic patient health information. There are other state, federal, and international privacy laws that need to be reviewed regarding AI devices/software depending on where the data is collected and analyzed, where the patient is located, and other factors.
• FDA rules and regulations for the legal manufacture of medical devices.
• OSHA (Occupational and Safety Health Administration) laws for workplace safety.
• Possible fee-splitting issues if there is a relationship between the hospital or medical practice and the developer of the AI being used for telemedicine? Fee-splitting may run afoul of both Stark Law and the Anti-Kickback Statute
• What issues are involved if the AI is developed or used in another state?
• Will new laws be required to address the aggregation of genomic and clinical data?
• Will there need to be any laws that govern the review of the AI algorithms to ensure they are working? Most software is tested before it is released but as AI learns, there usually aren’t new tests done before the AI is used.
• How does California’s new Bot Disclosure Law (BDL) apply. Much telemedicine uses bots to gather initial patient information. The BDL “With certain exceptions, makes it unlawful for any person to use a bot to communicate or interact with another person in California online with the intent to mislead the other person about its artificial identity for the purpose of knowingly deceiving the person about the content of the communication in order to incentivize a purchase or sale of goods or services in a commercial transaction…” Generally, the bot can take patient information, if it makes clear it is indeed a bot and not a real person.

In addition, ownership issues may arise around who owns the original version of the software and the adapted versions.

Additional complications when AI is used for telemedicine

As AI advances, the lines between what nurses who rely on AI for information can tell patients and what physicians can tell patients – may blur.

One futuristic use that will create all sorts of legal issues is the use of robotics to do surgeries through telemedicine. Already, robotics is used in some surgeries to guide and help surgeons to do very minute delicate tasks. If the software that guides the surgery is at a remote site, then many of the legal issues discussed – such as the unauthorized practice of medicine will come into focus.

What complicates the legal issues of any artificial software, including AI software used in conjunction or as a part of telemedicine, is that AI constantly relearns based on new patient information. So, the question begs, do all the relevant legal and ethical issues need to be reevaluated each time the AI adapts to new data. In general, when drugs and medical devices are remodeled, they need to be approved by the FDA (Federal Drug Administration) again. If the same principles apply to AI, the FDA will be incredibly busy because AI constantly charges the original model, the original software.

Many of the legal and regulatory issues involving artificial intelligence and telemedicine are evolving. What’s certain is that there are many new applications that are and will be developed because AI offers the unique advantage of learning and relearning the more it reviews patient data and data from other resources.  A core concern with AI is whether the regulation issues need to continually be monitored as the AI software accesses new information

Contact Cohen Healthcare Law Group, PC, our healthcare compliance lawyers understand the core legal issues involved with telemedicine. We are working to keep current with the ever-changing artificial intelligence landscape so we can advise developers, medical practices, and others about the relevant legal compliance issues.