Chatbots in the exam room: Are doctors ready?
Key Takeaways
Chatbots and digital voice assistants are a ubiquitous presence in our personal lives—raise your hand if you don't have Siri or Alexa in your home, or on your smartphone. Not too many hands raised, just as we suspected.
In 2011, when Apple launched Siri, its voice interface for the iPhone OS, the world began its discourse with all kinds of machines. Today, thanks to this technology, humans are using their voices to control lights and thermostats, play music, and request basic information. And, according to an article published by MachineDesign, the machines are delivering. Speech recognition platforms have become more than 95% accurate, which is on par with humans.
The use of artificial intelligence in healthcare is no longer a futuristic concept. It's already changing many aspects of medicine, from drug development to aiding radiologists and fighting opioid abuse.
Now, AI-powered chatbots are making headway in the exam room. Clinical chatbots are being tested to see if they can help doctors make more informed decisions at the point of care. They’re also being evaluated as a way to help health systems reduce the time and resources needed to support patient needs and improve their hospital stays.
But, will clinical chatbots stick, or are they just a fad? Let’s take a closer look at the inroads chatbots have made in healthcare and whether they’ll play an even bigger role in the future.
What is a clinical chatbot?
“Chatbot” is a catchy, futuristic term—but what is it exactly? Simply put, a chatbot is a broad definition for any piece of AI software that simulates human conversation with users.
It can be text, but it can also be voice. The (sometimes annoying) customer service chat box that pops up in the corner of a retail web page is a chatbot. Another example is Woebot Health, an app that you can chat to about mental health issues. A wildly popular chatbot is Florence (named after Florence Nightingale), who performs the role of a nurse in the absence of the human variety—reminding you to take pills, track body weight and activity level, and more. And Alexa is another type of chatbot.
The defining feature: A chatbot understands and uses human language, and simulates a conversation.
So how are clinical chatbots, also known as medical or healthcare chatbots, being used in medicine? A patient might encounter a clinical chatbot in an app, on a third-party chat service like Facebook Messenger, or on a medical provider’s website. Doctors, on the other hand, are beginning to find them embedded into electronic health record (EHR) systems, where, after a few verbal commands, they surface insights that would otherwise require interruptions to the clinical workflow.
An article published by Providertech in 2020 explains that chatbots use machine learning algorithms and natural language processing to comprehend and act on human commands and responses. These AI-enabled tools have progressed to the point where they can mimic conversations. Through this mechanism, chatbots can help care providers engage with more patients outside of the clinic. For example, from the clinical side, chatbots can schedule patient appointments, deliver test results, and even provide triage.
“The adoption of chatbots in healthcare is growing exponentially. Research firm Frost & Sullivan predicts that by 2025, 90 percent of hospitals in the U.S. will use artificial intelligence and chatbots to augment care delivery,” according to the article.
According to a 2019 study published by PeerJ, the healthcare market for AI has been growing by an annual rate of 40% since 2016. Authors note that machine learning algorithms can process vast amounts of data, allowing for new insights into diagnostics and treatment options. Coupled with the rise of healthcare-related big data (from sources that include EHR and wearable health tracking devices), chatbots can also analyze patient outcomes and relay key insights to medical providers. Researchers posited that AI would eventually improve patient outcomes by 30-40% and lower treatment costs by up to 50%.
And, although the terminology is often used interchangeably, clinical chatbots are not the same as virtual medical assistants (who are actually human). Read more about VMAs here on MDLinx.
The benefits of virtual healthcare
The ProviderTech article explains why investing in this type of technology for hospitals and clinics can reap great dividends for hospitals, doctors and patients. Chatbots allow patients easy access to health information, freeing up clinical staff for other tasks. They can also facilitate scheduling, billing, and registration, all of which can eliminate unnecessary office visits.
A vital part of this is the chatbot’s use of natural language processing. This piece of tech can help the bots convey compassion and adds human-like comfort to their engagement with a patient.
The benefits of chatbots are illustrated by the findings of a study published in the Journal of Medical Internet Research this month. Researchers conducted 25 patient interviews across five health centers, with a focus on assessing their experiences using a chat-based and automated medical history-taking service.
They found that patients had an overall “positive attitude toward e-visits” and were “pleased with the prospects of the digital tool tested.” Notably, the results also showed that the opportunity for patients to take their time when responding to health questions resulted in a reduction of stress and fewer errors in answers. Researchers also found that the ability to upload photographs made some visits to the clinic redundant, saving the patient and doctor time and money.
The current uses of medical chatbots
Chatbots have already made significant headway in healthcare. From an app on a patient’s phone, these bots can answer patients’ general health-related questions, provide specific information (like how to manage chronic conditions), or even offer advice on available treatments (for example, behavioral health therapies). They’re also scheduling appointments, checking symptoms, providing virtual triage, assisting with medication management, and more.
A recent Medical Futurist article provides some examples. These include:
Northwell Health’s chatbot, which was launched to address the fact that 40% of less privileged patients don’t follow through with colonoscopy appointments. According to the company, the chatbot will “encourage patients by addressing misunderstandings and concerns about the exam, delivering information in a responsive, conversational way over email or text.”
OneRemission’s chatbot, which provides cancer patients with lists of useful diets, exercises, and post-cancer practices. While the purpose of the app is to provide information to patients in order to free up physicians’ time, the bot also has a feature that allows patients to consult with an online oncologist on demand.
Babylon Health is a medical consultation bot that allows users to report the symptoms of an illness and then advises them on next steps. The bot uses a patient’s medical history and checks symptoms against a medical database. Launched in 2013 and now valued at more than $2 billion, Babylon Health has already partnered with the UK’s National Health Service and is now extending into the United States.
GYANT, is a chatbot designed to help patients understand their symptoms. The bot sends health data to doctors, who can then provide diagnoses and prescribe medicine in real-time. According to the Medical Futurist article, as of 2019, the bot had helped more than 785,000 people in Latin America to undergo a pre-diabetes screening, leading roughly 74,000 at-risk people from low-income populations to seek preventative treatment.
What does the future hold?
The examples mentioned above are just the beginning. As the technology progresses, chatbots will appear in places where they stand to make a big difference. With their touchless interface, there is potential for use in operating rooms and examination rooms. For those who are recovering and may have lost mobility, they can be used in long-term care or assisted living facilities, as well as hospital wards, to provide more autonomy for patients.
During the pandemic, healthcare chatbots demonstrated their value as a way to provide care remotely. As a result, researchers are anticipating an adoption boom.
A recent analysis found that venture capitalists have invested more than $800 million in at least 14 health chatbot start-ups. This is hardly surprising given how much money bots are predicted to save care providers: Consulting firm Accenture posited that AI may save $150 billion annually across US healthcare by the year 2026.
The bottom line? This technology is here to stay—and evolve. Don’t panic. Chatbots aren’t here to replace you, they’re here to help you become a better doctor.
Read more about healthcare technology on MDLinx: