For many people, a necessary and often frustrating step in accessing medical services is determining whether a healthcare provider is part of a health insurance network.
Commonwealth no surprise methodwill take effect in January 2022 and establish new rules to protect consumers from surprise medical bills and help remove consumers from payment disputes between health care providers or health care facilities and their health care plans. bottom. It also included requirements for insurance companies to maintain, verify and update their physician rosters.
However, a recent data analysis of approximately 450,000 physicians in the Medicare Provider Enrollment, Chain, and Ownership system found that only 19.4% had consistent address and specialty information across all directories listed. I understand. The survey results were published in a research paper.
“For patients, what this means is that when they go to an insurance company directory and look up a doctor, the information is likely to be incorrect,” he explains. Neil Butala, MD, MBA,Assistant professor cardiology in the University of Colorado School of Medicine Who led the data analysis? “The No Surprises Act does a lot to protect consumers, but patients can face many obstacles to getting care if the accurate information requirement is not implemented. I have.”
Big data analysis
Butala and his collaborators will search in the Medicare Provider Enrollment, Chain, and Ownership System databases in the UnitedHealth, Elevance, Cigna, Aetna, and Humana online physician directories based on physician name and zip code in September 2022. You have searched for all doctors in .
Among physicians found in multiple directories, they compared the consistency of clinic addresses across directories. They used the National Uniform Claims Committee taxonomy to assess the consistency of professional designations across directories. Physician information was considered consistent if it was the same in all directories where the physician was found.
Researchers partnered with HiLabs to aggregate and streamline this publicly available data and analyze it using the programming language Python. In doing so, they found that only 27.9% of physicians had consistent practice addresses, whereas his 67.8% of physicians had consistent specialty information.
Additionally, Butara and his collaborators found that as the number of directories in which doctors were listed increased, the information became less consistent. For a doctor with only one clinic listed, he improved his address consistency across national health registries to 84.8%.
find solutions to information problems
“Many of these errors are caused by having multiple addresses,” explains Butala. “For example, I recently moved from Boston to Denver and I’m still in directories known as ‘ghost’ entries all over Boston.
“Because providers already have a heavy administrative burden, staff may enter provider information as a batch process, especially when practicing with large groups. List all providers everywhere, regardless of whether they have
What this means for healthcare consumers is frustration. Butara said he also spends time on the internet or on the phone to find out not only if the provider is in the network, but where they actually practice. A recently published data analysis did not include phone number consistency, but could study that data across insurance directories, including information on whether doctors are accepting new patients. I’m here.
“In an ideal world, what Medicare is proposing would be for the federal government to create a national directory of providers with a uniform method for verifying information and a standardized method for transmitting it. to do,” says Butala. “I would love to see that happen, but in practice it will probably take years to implement.”
The situation is further complicated by the fact that each insurer requires different information from a provider or group of providers, he said. As such, the financial commitments for provider groups or institutions to hire and train additional administrative staff and update the directory would be substantial.
“A more short-term solution could involve using advanced analytics,” says Butala. “Machine learning techniques are great for cross-checking information from different sources and fundamentally solving the problem of information transfer across systems.
“I think this problem is not getting enough attention because it is a boring problem and a data quality problem. But it can easily exacerbate existing healthcare disparities. Those who do not have access to, or who are not familiar with, the health care system, or who are unable to take half a day off work to know what their insurance covers and make an appointment, may ultimately be unable to access the care they need. and.”
