Burnout of healthcare professionals has not waned with the pandemic but assistance from artificial intelligence may offer tired workers a reprieve from tedious tasks and a chance to better connect with patients.
Half of physicians are experiencing burnout and 45% of physicians over 55 and 14% in the younger age group plan to stop seeing patients in the next three years, according to a survey from the Commonwealth Fund. Partially for this reason, primary care physicians and nurses are expected to leave healthcare in droves in the upcoming years.
A survey by Incredible Health showed that 34% of nurses planned on leaving the profession in 2022. Of those planning to leave, 44% cited high-stress conditions and burnout as the reasons for wanting to quit.
Some health systems and hospitals are turning to artificial intelligence and machine learning technologies to help tackle tedious tasks and better triage patients.
AI-enabled triage can shuttle patients to the ideal care provider
The work AI can do to reduce burnout is only just starting, according Anu Shukla, co-founder and executive chairman of health tech company Botco.ai.
Of course, chatbots are already a common tool for communicating with patients, but Botco.ai is leveraging new technology that can be implemented in the highly sensitive behavioral health specialty.
By using Botco.ai’s automated intake process, patients get to the right provider faster without wasting time at the point of care communicating things like change of address and insurance provider. The platform reports that 89% of respondents claimed that the company’s automated chat tools make it easier for patients to access mental health treatment.
“We've come across all these applications for sentiment analysis that had to judge how people are feeling and so that's important in mental health applications,” Shukla told Fierce Healthcare. “We can recognize the words that lead to you putting up a message saying, 'If you are feeling you need immediate help, please call the suicide prevention health hotline.'”
The chatbot shows its greatest advantage in performing triage informed by machine learning, Shukla said, and can direct patients toward the right provider, saving patient and doctor time and reducing frustration.
She used the example of an urgent care partner who reported 18,000 conversations moving through Botco.ai’s technology in the first three days after implementation. That provided the bot with more information to learn about the specific population accessing that provider’s website, what those patients tend to look for, what words they use and how they sound when they use them.
Botco.ai uses sentiment analysis to see what is beyond a patient’s words, which sends them to the right provider and then gives that provider insights into a patient’s mental status. Shukla says that beyond triage, the use of AI technology is limited and potentially dangerous in the sensitive behavioral health space. But having a patient shuffled through a long, analog intake process when they need immediate care is also risky, she noted.
“AI and humans working together is the right combination for many situations,” Shukla said. “We analyze the sentiment. If we get a dangerous signal, we flash the messages to call someone, or we just transfer them over with their permission and so now there’s a human being on the other end.” Depending on the time of day, that human may be in a hotline call center, or they may be a clinician.
In behavioral health, Shukla emphasizes that AI decreases shadowy barriers to care, like stigma towards therapy. Shame related to seeking behavioral health support, notably within racial minorities, can be decreased by communicating with a computer instead of another person. For people accessing therapy for the first time, who may be from a culture that deems it an unobjectionable sign of mental weakness, a bot can decrease fear.
Shukla believes that providers want to see these patients and are frustrated when they can't support the populations they personally identify with.
U.S. Surgeon General Vivek Murthy, M.D., has listed health worker burnout as one of five current priorities. His office lists possible solutions to burnout as including the reduction of administrative burdens, increased use of human-centered technology and access to mental health and substance use care.
Physician demand will grow faster than supply leading to a shortage between 54,100 and 139,000 physicians by 2033, according to the Association of American Medical Colleges (AAMC).
To avoid the upcoming drop-off, the AAMC suggests better use of technology by health organizations to improve doctor training and patient care.
Botco.ai published a survey in late August reporting that 73% of mental health facilities increased their online support tools in 2022 and 89% of mental health executives say AI-enabled chat improved mental health care accessibility for patients.
Of the 120 executive respondents, 41% said they believe that the pandemic made mental health treatment harder to access and 55% said the pandemic further decreased access for racial minorities and LGBTQ+ populations.
Online tools like chatbots were credited with connecting patients to care 49% of the time while other forms of outreach were less common, such as word of mouth (25%), doctor recommendations (19%) or advertisements (7%).
Conversational AI tools were already in place in 61% of the facilities surveyed which included mental health facilities, behavioral health organizations, addiction treatment providers and rehabilitation centers.
The report showed that the benefits extended to health professionals who can defer tedious work to the bot and work to the top of their license, providing care.
Botco.ai found that 76% of facilities used automated chat tools to ease intake processes, with 83% using the tool to verify insurance benefits.
Decreasing and streamlining administrative burden
After the point of care, insights collected from Botco.ai are integrated into the provider’s electronic health records system to decrease administrative burden for clinicians and admin staff.
Pegasystems is an API provider that also works to decrease friction in workflow by speeding up interactions and getting patients the care they need faster through AI. Pegasystems’ vice president of healthcare and life sciences Kelli Bravo sees AI as a key tool to help healthcare employees save time on copy and pasting.
She uses the example of a client that communicates with customers on 919 different channels. Pegasystems can accommodate this high, complex volume while tailoring messages for customers.
“There are a lot of dimensions that we're seeing today,” Bravo said. “Banks and financial institutions started being smart about this; Amazon is excellent at this. What we're trying to do is adapt that capability into the healthcare realm, because it does matter. People want to be able to relate to the context, they want to know and understand who they are, and they want to have empathy, we actually can dial in empathy.”
AI can notify health system workers that a region has recently experienced a forest fire so the next time someone calls to talk about a bill, “the system will say, Look, this region needs to have a different conversation.”
Other prompts can be served up like whether a patient has searched for information related to diabetes the last five times they’ve been on the health provider’s website. That way, providers can better inform their care and answer the questions patients might not even know how to ask. Interactions are more fruitful, giving providers a sense of satisfaction, according to Bravo.
The system also analyzes workflows and prioritizes the task most valuable to the corner of the health system in question. For example, when assisting payers, Pegasystems’ AI chooses which insurance claim is the timeliest and places that at the top of the list.
“That's a huge value of AI predicting what's the value-added work and getting that work prioritized and done and then distributing it to the right people based on what they're great at,” Bravo said. “It helps with retention and also helps with training. It gets so much faster, the system can develop workflows to guide you through the process.”
AI can allow for human-to-human engagement with patients or health plan members, Bravo said.