Patient engagement is top of mind for executives and clinicians at hospitals and health systems today. While there may be a need for more constant touchpoints with patients, that would bring with it an unintended physician burden.

So, what's the solution? Some say artificial intelligence can play a role. Especially the latest and most powerful flavor – generative AI, the kind behind ChatGPT. But careful scrutiny by healthcare provider organizations of such technology would be needed.

Healthcare IT News sat down with Michael Haas to discuss patient engagement and communication and the role AI can play. Haas is a technology management consulting manager in RSM US’s healthcare industry practice. In 2022, he was selected for the firm’s Industry Eminence Program as a senior analyst covering the healthcare industry.

Q. Many experts say hospitals and health systems need more touchpoints with patients – which would leave physicians with the burden. But this is a key aspect of patient engagement. How could ChatGPT help out?

A. In the healthcare industry, ChatGPT has been used to respond to patients’ questions, assist with clinical decision making and triage, extract patient data, and streamline laborious administrative processes such as clinician scheduling, patient referrals and credentialing.

These all are monotonous and repetitive tasks that can be automated and streamlined on a day-to-day basis. According to Medical Economics, physicians spend more than three hours a day on documentation and inbox management. Three hours a day and fifteen hours a week spent on administrative tasks is leading to large burnout with physicians and medical staff.

With this extra time, they can focus more on patient care and perhaps even just take a mental break to recharge. Currently, most healthcare practices have a fee-for-service business model that requires providers to see multiple patients to generate revenue.

When providers only have so much time in a day to see patients, they often are forced to complete their charting at night, and this can lead to poor documentation and medical errors. Adopting ChatGPT can help streamline documentation and collection of data while the provider focuses on patient care.

Q. On another key patient front, how could ChatGPT help healthcare provider organizations with patient communication?

A. ChatGPT is a powerful tool for content creation, helping healthcare organizations reach the right audiences more efficiently.

ChatGPT’s biggest impact on the healthcare industry can be streamlining repetitive and time-consuming tasks, including scheduling, drafting internal communications and responding to patient questions. ChatGPT can generate human-like responses to a wide range of requests, making it an ideal tool for healthcare applications.

From personalized treatment plans to remote patient monitoring, ChatGPT is transforming the way healthcare providers deliver care to their patients. With its advanced language processing capabilities, ChatGPT can accurately and quickly translate medical and technical terms, and common expressions, allowing patients to understand their diagnosis, treatment options and medication instructions.

Q. You suggest while ChatGPT can help overburdened physicians, careful evaluation is needed. What should clinicians and health IT executives be doing before using ChatGPT?

A. ChatGPT is still very new and should be used with caution. Who is responsible for validating the efficiency and accuracy of what ChatGPT interprets or communicates? What is an organization’s risk plan to make sure a patient receives the correct care?

Since this technology is so new, no organization should just adopt it into their everyday practices. ChatGPT and other large language models give human-like responses. However, these systems do not process information like humans nor do these systems actually “know” anything.

They simply understand how language relates and can recognize patterns, which it processes as summaries of text and responses to questions. There should be a phased plan rollout over controlled groups of test patients and then real patients to validate the communication and data is accurate.

The end goal is to make a better experience for the patient and reduce burdens for the providers. By setting benchmarks and checks and balances to affirm accuracy, this will reduce the overall risk to the patient and organization, and adoption of the technology will be received more positively.

Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.