Loading Results
We have updated our Online Services Terms of Use and Privacy Policy. See our Cookies Notice for information concerning our use of cookies and similar technologies. By using this website or clicking “I ACCEPT”, you consent to our Online Services Terms of Use.

Texting for Better Health

Share
Facebook
X
Pinterest
LinkedIn
Email
Print

UH a leader in using chatbots to help patients manage chronic conditions and avoid hospital readmission

UH Clinical Update | August 2022

Can texting with an artificial intelligence “bot” lead to better health outcomes for patients? The Population Health team at University Hospitals is working to find out. Since late 2019, its care managers have enrolled patients recently discharged from the hospital for congestive heart failure, heart attack, stroke, pneumonia and chronic obstructive pulmonary disease (COPD) in post-acute automated chats, done via email or text. The project also recently expanded to helping patients manage chronic conditions, such as high blood pressure, asthma, congestive heart failure and type 2 diabetes.

How does it work? All the chatbot requires is a cellphone and the ability to email or text. It does not require a user ID, password or downloading of an app, making the process accessible even for patients with limited technology experience. Patients with congestive heart failure, for example, receive an automated daily text asking them to text back their weight and blood pressure. Patients with COPD receive a text asking how their breathing seems that day and to text back a response. The texts and responses from the bot are in simple, conversational language – no medical jargon allowed.

The UH perspective: The chatbot platform is not so much an elaborate remote patient monitoring solution as it is a safety net and patient education tool, says Mark Schario, Vice President of Population Health at UH. When a patient texts back a response during the chat that may indicate a health concern, the bot will note the risk, offer the patient educational resources and the care manager will follow up with a phone call to the patient.

“Chatbot technology tracks patients’ progress and offers insights to patients and clinicians to facilitate earlier interventions when problems occur,” Schario says.

Chatbot advantages: One key advantage of a chatbot over a live phone call is that patients can engage with the content on their own schedules, when it is convenient for them. It also frees up care managers from having to make dozens out outbound calls, especially in a time when many people don’t answer the phone for unknown numbers.

But perhaps the biggest advantage of chatbots, Schario says, is the beneficial effect they can have on streamlining care management workloads. This is crucial, he says, in a time of healthcare labor shortages, when Population Health programs are managing enormous numbers of increasingly complex patients.

 “You literally cannot hire enough people to manage the complex needs of patients today,” he says. “We were looking for something that would be, to use an old army term military term, a force multiplier. How do you make one person stronger? Technology is one of those things where one care manager can actually interact with many patients but they don't have to physically be doing it themselves. Chatbots make frequent patient contact to collect and provide routine information, allowing care managers to spend more time on high-value interactions that require clinical judgment.”

What about the human dimension? Doesn’t something get lost in the patient experience when the patient is texting with a bot? Not so far, according to the UH Population Health team. Callie Bahner is Manager of Care Management with UH Population Health. She says that patients seem to like chatbots and accept them on their own terms.

“The patients appreciated the chatbot’s use of conversational language, avoidance of medical jargon, and asynchronous connection, participating with the chatbot on their own time frame,” she says.

In fact, she says, chatbot technology seems to make patients more personally connected to their care management team, not less.

“Time and time again, what we hear is, ‘I knew you'd call me,’ she says. “They really feel that connection. We've had so many patients tell us that if they put this in the text, you’d call me. And we do. It really gives them a sense of comfort. The chatbot provided a safety net, a feeling of partnership, and connectedness free from judgment. The chats provided not only clinical data but also starting points for conversations. The frequency of the follow-up calls from the care managers along with the information obtained from the chats fostered conversations that brought to light information not previously shared with us.”

One other consideration: Chatbots are a good option for communicating with patients who may not respond to other forms of care manager outreach, such as outbound phone calls and home visits. Schario and Bahner tell the story of one patient with both COPD and congestive heart failure, for example, who was reluctant to have a care manager visit her in her home. Another patient was confused that outbound phone calls from the care management team were telemarketing calls “trying to sell her something” – leaving the important information of the phone call undelivered. There’s also some suggestion that patients may be more forthcoming in reporting their health data via chatbot than they might be during a live phone call.

For all these reasons, UH is all in on chatbots. In fact, UH Ventures is working with chatbot vendor Conversa Health and several UH service lines to develop more chats for UH patients, with a direct role in developing the artificial intelligence logic behind the chats.

“This type of technology is really the future,” Schario says. “Not everybody can be called every day. Chatbots help much more than you could ever imagine when you're trying to manage a large, complex panel of patients.”

For more information on the use of chatbots at UH, please email UHACO@UHhospitals.org.

Share
Facebook
X
Pinterest
LinkedIn
Email
Print