When it comes to telehealth billing, there's a conundrum.
The increasing popularity of telehealth, the use of electronic communication technologies to provide care when the doctor and patient aren't in the same place at the same time, has created the problem. When it comes to submitting a bill for the service, the current approach fails to quantify a varying level of medical expertise and experience.
That makes Ohio's telehealth billing system, like the rest of the nation's, unsustainable, said Dong-Gil Ko, PhD, an associate professor at the University of Cincinnati's Carl H. Lindner College of Business.
Ko is tackling this issue using artificial intelligence and electronic health records. His work, recently published in the Journal of the American Medical Informatics Association, aims to create a fairer, more effective billing model.
The current time-based billing model creates inequities in compensation by undervaluing experienced doctors, Ko said. Those with greater expertise, who can provide accurate answers quickly, could be compensated less than less knowledgeable doctors who take more time to respond.
This system unfairly rewards inefficiency and fails to recognize the value of cognitive judgment and expertise, Ko said, leading to skilled doctors being undercompensated despite offering higher-quality care.
To address these shortcomings, Ko is collaborating with Umberto Tachinardi, MD, UC Health's chief health digital officer, and Eric J. Warm, MD, an internal medicine physician and researcher at the UC College of Medicine. Together, they are leveraging AI and electronic health records from UC Health to develop a new billing model that incorporates doctors' clinical judgment and expertise, alongside the time spent responding to patient inquiries.
"Can we value medical doctors for their expertise and who they are?" Ko said. "Doctors undergo years of rigorous medical training to develop specialized knowledge. Let's acknowledge and recognize that and find a way to measure it. By doing so, we can create a balanced billing model that considers not just the time spent but also their medical expertise."
Ohio's current medical billing code, which went into effect in 2023, pays medical professionals based on how much time they spent answering a question via a secure messaging system. If they spend less than five minutes responding, the service is free. If they take more than five minutes to answer a question, they receive compensation, with fees increasing as their time increases.
"Experienced doctors can respond quickly due to their expertise but earn less because the current billing model prioritizes time over skill," Ko said. "In contrast, a newly minted medical resident may take much longer to answer the same question due to their limited experience and knowledge, enabling them to charge the patient."
"This creates a systemic issue. If we follow the American Medical Association's guidelines, less experienced doctors are compensated while experts may not be."
Along with potentially shortchanging experts and discounting their effort, the current model also could erode trust between doctors and their patients. It forces doctors to make billing decisions without reliable methods to measure their work.
"Doctors don't measure response time with a stopwatch, and some questions may require multiple sessions to address, making billing decisions even more complex," said Ko, who created the first research lab at the Lindner College of Business.
Additionally, the uncertainty of whether patients will get billed may discourage them from contacting a medical professional. That breaks down continuity of care, delaying treatment and potentially leading to worse health outcomes.
"We need balance," said Ko. "Both time and medical expertise must be considered in billing."
Ko anticipates telehealth billing challenges will grow as generative AI becomes more integrated into medical practice. While AI can deliver faster solutions, doctors will still need to validate its responses and invest time in maintaining and operating these systems -- efforts that must be compensated to avoid increasing burnout among medical professionals, Ko said.
"At the early stages, validating AI-assisted responses will be critical," Ko said.
Ko's AI system can use doctors' behaviors to better understand and measure their expertise, offering a framework for fairer compensation. His tests of machine learning models have delivered consistent results, demonstrating the potential to more accurately evaluate doctors' expertise and time spent on patient inquiries.
"This time-based metric is constrained, and this model is not sustainable, especially with generative AI coming into the picture," Ko said.
Looking ahead, Ko envisions broader applications for his research. He aims to create a system that predicts whether a patient will be billed before submitting a question and to uncover insights from patient data to improve care outcomes.
By combining AI and innovative research, Ko's work could transform telehealth billing, ensuring fair compensation for doctors while improving patient outcomes. He plans to pilot his program with health systems in 2025.
