How AI in Healthcare is Revolutionizing (and improving) Patient Engagement and Adherence.

Big data, artificial intelligence (AI) and machine learning (ML) are already impacting healthcare in numerous ways – from diagnostics to treatment, to everyday administrative processes, including scheduling or tracking regulatory compliance. But perhaps the most noticeable, at least for patients, is how AI is helping to revolutionize patient engagement and adherence.

Patient engagement has been around since the days of Hippocrates. Everyone who has undergone a medical treatment is familiar with the paper handouts and pamphlets so common in traditional patient engagement – experts say roughly eight of 10 hospitals still use paper handouts for patient education upon discharge. But there are, of course, other types. These include other low-tech methods such as face-to-face conversations, letters, and phone calls, to more tech-savvy approaches, including emails, SMS messages, online patient portals, patient mobile apps, and patient television systems.

But in recent years, conventional, one-way patient engagement has been found somewhat wanting. Research has shown that patients now expect more – and with more and more industry competition, including from e-commerce and franchised multinationals, healthcare providers need to offer better patient experiences to stay competitive.

It’s all why many experts say AI/ML models fed by big data (generated by electronic health records (EHRs), mobile devices, wearable technology, and other next-generation data sources) is the next frontier in patient engagement.

Most healthcare providers need to improve patient engagement

According to a recent survey of 2,000 healthcare consumers and 200 business decision-makers (BDMs), health providers’ engagement efforts fall short of patient expectations. The survey shows that even though 100 percent of BDMs say they’re doing a good job engaging patients, only 35 percent say they feel valued by their doctor’s office.

A further 25 percent said doctors and insurers don’t do a good job with care coordination. Interestingly, nearly 50 percent of patient respondents in the same survey said they’re comfortable with doctors using AI to aid their healthcare decisions.

One way to improve flagging patient satisfaction and engagement numbers is through the use of data and AI, says Joe Greskoviak, president and chief operating officer at Indiana-based health services provider Press Ganey Associates. “I think healthcare consumers have become much more demanding relative to the type of information that they require to better understand some of their healthcare choices,” he says, adding that patient loyalty primarily boils down to three things: Communication, provider empathy, and care coordination, with an emphasis on the former.

“What we find is that loyalty is primarily being driven by number one, communication,” says Greskoviak. “There’s a difference between waiting and not knowing why you’re waiting.”

How AI can help with patient engagement

AI and ML can help in potentially profound ways with patient engagement, often seen as the “last mile” of healthcare delivery – a crucial component that can make all the difference between good and adverse health outcomes and client satisfaction. AI allows providers to more deeply engage patients in the right ways, while filling care gaps and nudging patients to improve their behavior – all without adding to the workloads of healthcare providers (and, in some cases, lessening those workloads).

AI is helping to improve patient engagement in several ways, including:

  1. Taking a page from digital marketing best practices by using ML to determine the best ways to reach patients, at the right time, and with the right message. Perhaps some patients have a history of responding only to text messages at a certain time of day – in those cases, ML can learn from patient behavior to engage them more effectively. This also saves providers money and lowers engagement costs since it streamlines communications.
  2. Natural language processing (NLP)-powered chatbots that respond to patient queries quickly and accurately, and guide patients through everyday processes previously performed by staff with a level of empathy not possible even just a few years ago. Because patients will quickly turn off a provider with little to no digital bedside manner, emotional appropriateness is key when communicating with patients. Indeed, providers have discovered that building algorithmic empathy into chatbot ontologies helps improve engagement and, ultimately, health outcomes. And using an AI-powered chatbot instead of a human can save overhead costs and free up resources for other tasks.
  3. Tailoring recommended treatment plans and other follow-up actions by mining large amounts of past treatment and patient data, such as EHRs, for similar patient cohorts. Patient noncompliance is an ongoing problem that negatively contributes to health outcomes, but providers can use ML to create treatment plans with the best chance of success for particular individuals. Providers can also subtly influence patient behavior through appropriate messaging and content, as mentioned above, at the right times and through the right delivery vehicles.

AI and patient engagement: Ethics and implementation

None of this is to say that AI, while tremendously promising, is a patient engagement panacea. Ethical implications can come into play when technology makes decisions or has conversations traditionally handled by humans, especially when people’s health (and lives) are at stake. 

Transparency can sometimes be an issue for AI-driven diagnostics – how does a doctor explain to a patient how an algorithm diagnosed them? Indeed, issues of privacy, accountability, patient autonomy, and informed consent can crop up when AI drives patient engagement. Some patients aren’t comfortable giving personal health information to an algorithm, one which may or may not have similar sensitivities to privacy and confidentiality as a physician. And then, of course, there’s the issue of having a computer communicate bad or unpleasant news to a patient – not an ideal scenario, at least for most patients.

For these and other reasons, the American Medical Association (AMA) Journal of Ethics advises practitioners to exercise caution when implementing AI into clinical practice. It recommends using AI as a complementary tool – not as a replacement for a physician – and highlights the importance of skilled human supervision “to identify possible ethical dilemmas.”

Implementation of AI on a large scale remains elusive, however, at least for now. “AI systems must be approved by regulators, integrated with EHR systems, standardized to a sufficient degree that similar products work in a similar fashion, taught to clinicians, paid for by public or private payer organizations and updated over time in the field,” according to a study by the Royal College of Physicians. The study notes these challenges will be overcome, but that it will take time.

Despite these issues, AI remains a tool that will continue to play an increasingly important role in patient engagement and healthcare in general. While AI and ML models likely won’t end up replacing physicians and their administrators anytime soon, they have already proven themselves a valuable tool for patient engagement – a trend that will no doubt continue.

Contact Us.