By Janae Sharp, May 23, 2019
In 2018, the iPhone had a glitch. Many devices displayed the letter “A” with a Unicode “[?]” every time the user tried to type the letter “I.” My phone was one of those affected by the glitch, and I remember people online sharing a workaround. Users could adjust their settings and set their phones to know that whenever it saw the strange groups of symbols, it should display an “I” instead.
I decided not to use the workaround, in part because I was curious to see how long it would take Apple to fix the glitch. And it took a while.
Healthcare can be like that iPhone glitch.
Layers of Abstraction
In technology, we develop a workaround to manage a problem when we know things aren’t functioning correctly but don’t have the tools or the access to fix the problem itself. During the iPhone debacle, I didn’t think the weird “I” issue was important enough to merit changing my settings, and I didn’t especially want that manual override anyway. I mean, what if I needed to sometimes type “A[?]” and not have it mean something different?
Health information technology is facing the task of allowing patient data exchange. The Centers for Medicaid and Medicare Services (CMS) has proposed a rule that would accelerate this mission. But the challenges of data availability and workflow ignorant accounting remain as points of frustration for physicians.
The Data Behind the API
There’s a reason why so many physicians upset with the functionality of electronic health records; There are multiple bugs in the system. One is the physicians do not always have the data that they want at the point of care.
During her keynote speech at the Utah HIMSS annual conference, Barbara McAneny, M.D., president of the American Medical Association, pointed out that the point-of-care data that she wants to see is the price of a given medication. What will this drug cost a patient? She wants to see this critical information in her workflow. As an oncologist, she treats patients who have catastrophic financial losses with their illness.
I asked a few people what would need to happen to make that data available — because there are technologies that will surface insights into the workflow. My sources didn’t want to be named, and they claimed that pharmacy benefit managers are the devil and don’t want patients to see the cost of medication.
Then patients would see how much savings are passed along to the patient — hint: not much. The lack of transparency has led to accusations of price collusion and price fixing. CVS was criticized for gaps in what it charges insurance companies and what they reimburse, called spread. I imagine McAneny would want to see whether that prescription would be less expensive for a patient with cash pay.
What data structures and regulations would need to be in place to enable patient pharmacy pricing transparency at the point of care? The data access is frequently unstructured and would require significant infrastructure. For example: access to the payment agreements of every payer. CMS proposed interoperability rule holding payers accountable for creating better data access is a start. Actually getting the next step to pharmacy drug pricing transparency at the point of care? In order for this to happen, regulatory levers would need to be pulled. (This week, the HITAC task force recommended not including price transparency in the proposed ONC interoperability rule.)
The question of regulations and a free market are alive and well in the world of healthcare transparency. New interoperability measures that are criticized as layers of regulation would not create this level of transparency. So will this regulation help with the problem of drug pricing transparency? The technology exists to make it visible, but currently the data is not being shared.
Sticks and Carrots
Some criticism of interoperability efforts stems from the lack of financial incentives to share data. Having data about drug pricing is a financial advantage. Having better information about patient needs creates less competition between health systems. Data is the new currency. The healthcare data marketplace was valued at $14.25 billion in 2017 and is projected to reach $68.75 billion in 2025, according to a recent report by BIS research.
With such a strong financial disincentive to share, very few are surprised that health data does not flow at the speed of the internet. The new CMS proposed incentives address the connection points of data but lack sufficient detail about the data being passed back and forth.
An API needs to have a bit more detail. Passing information back and forth without reference to the needs of patients or physicians and what the data will be used for will mean necessary information for patient care is lost in translation. Data improvements that improve workflow are going to be a critical step toward improving interoperability.
Some of this movement of data and motivation to improve is forced, meaning if CMS didn’t make healthcare systems share their data, they would not. Electronic health record (EHR) companies would not share data, either. The business case for not data sharing is often that the people who benefit are not the ones who are paying for the data exchange. New rules alter this a bit, suggesting that insurance companies should be responsible. Proposed CMS rules might address data connection but not the larger problems of workflow and patient data matching.
“It’s unfortunate that our industry moves with either sticks or carrots. We keep hearing from different vendors in the industry that interoperability is technically there, but because of pricing strategies and shutting down market competition, those integrations are not effectively made available for mass usage. This proposed rule will eliminate that challenge,” says Saurabh Mathur, Holon’s chief technology officer.
According to Mathur, data are too disconnected from the workflow of a physician.
“The challenge with the standards is that they talk about how to deliver the information without reference to workflow,” he says. “Without specific data points, there is no way to regulate.”
In healthcare, we have the same barrier with definitions and understanding that language does, but no one wants to work on understanding. CMS proposed a rule that would require data exchange through APIs.
“There are so many unbudgeted workflows in healthcare. Doctors finish their shifts at 7 p.m., but don’t get home until midnight because they’ve been documenting patient encounters,” Dirk Stanley, M.D., MPH, chief medical information officer for UConn Health, explains. “From a mathematical standpoint, the problem is clear: When you add up the time documentation takes versus how much time doctors actually spend with patients, you can see how the current system really doesn’t work. People who work in the finance industry may find a benefit in tracking how much time documentation takes, but the insurance companies are not reimbursing doctors for the actual workflow.”
The reality of modern healthcare is that no one seems to be looking at the workflow. Operational policies should account for data moving efficiently back and forth, for physicians getting relevant data at the point of care and for comprehensibility of data between payers and providers. Graham Grieve, the founder of HL7 FHIR, gave an illustrative step-by-step of potential problems behind the FHIR interface on his blog: Not every party involved in healthcare data is looking for the same thing.
Health information exchange has demonstrated cost savings and better health outcomes. Technology has also shown that we do not have a user-friendly and workflow-informed system. It will be necessary for EHR companies and healthcare systems to have financial incentives to share data — and CMS is leading the way in determining what that data landscape will look like.
Physicians like McAneny and Stanley have valuable insights about the types of data that should be shared and the work that should be accounted for. We need to account for appropriate patient data at the point of care and workflow. McAneny has a great point: The data she wants is the pricing data for patients already drowning in the cost of cancer care, and without force, pharmacy benefit managers do not want to share that data with her at the point of care.
The comment period for proposed interoperability rules from CMS has been extended until June 3. Without additional consideration for data usability and workflow, they run the risk of being a workaround like the solution Apple gave to customers while their iPhones suffered from a bug for the letter “I.” The letter is used quite often, and so are our health data.
Disclaimer: I’ve worked with Holon Solutions for Sharp Index and received compensation for event promotion, but I’m not receiving compensation at this time. (I do really like them and their technology.)
Original Article: Inside DigitalHealth