A recent piece by two Boston Children’s Hospital informatics researchers in the New England Journal of Medicine, Kenneth Mandl, M.D. and Isaac Kohane, M.D., made the argument that EHR vendors are holding back innovation in the health IT industry. This controversial opinion comes from the duo’s belief that many vendors have failed to adopt basic Internet-era sources for their systems such as private cloud-based storage and secure communication protocols, as well as modern consumer technologies such as word processing and search engines.
In a two-part interview, HCI Associate Editor Gabriel Perna talked with Dr. Mandl in-depth about his assertion. In part one, he talked about how he and Dr. Kohane came to this conclusion, what are some of the basic innovations missing in health IT, and how a lack of integration is hurting the industry. In part two, he goes in depth about standards and interoperability, how government mandates have affected innovation, and he talks a little about the power of mobile health for EHR vendors (mHealth).
Kenneth Mandl, M.D.
How would you like to see the health IT marketplace change? What kinds of industries can you compare it to?
A common analogy has been with the banking industry, and while we know they have had their issues, they have solved some problems very elegantly. The ATM network is highly functional and effective, very usable by anybody, anywhere, and what they chose was a few data types in banking that were critical. That includes account transfers, balances, withdrawals, deposits, and they made them work. In healthcare, we have a few transactions that we could make work. The fact is we’re so bulky in our approach to these propriety HIT products, since they’re not designed to work or share information outside of their propriety databases, even for the user right there, much less for a situation where the patient shows up right there. There are successes in other industries that we should emulate.
Any industry manages to maintain secure and private information to share it across their sites of business. There are many generic solutions to do this. The way to do this is to use modern architecture, using application program interfaces.
Speaking of standards based analytic tools and interoperability in general, would you say that it’s not in EHR vendors’ best interest to adopt this kind of technology, or is it?
I think there is a fear on the part of the vendors of data leaving the system, because once it leaves, other competing systems may handle the data better. Healthcare should not be held hostage to that business model. We need the data in applications that serve our patients’ needs and the healthcare system needs. There’s no question that data analytics are sorely lacking in healthcare and the business intelligence for most healthcare systems, is not based on the data in EHRs. Only now are we beginning to see that there are approaches to exporting data from EMRs for the purposes of analytics.
For example, there’s a project that comes out of our group, called I2B2. Sixty academic medical centers nationwide, and a few globally, export data from EHRs onto a shared analytic platform. This is another example of the need to move data out. It’s very important to have a healthy ecosystem of applications. For there to be a healthy and nimble evolution and an innovative process in healthcare IT, we need to be sure we’re not stuck in these monolithic pieces of software that the EMR industry has designed.
What are the advantages for those vendors if this data were more open? The advantages to the end-user are obvious, but what about to the vendor itself?
Well, ultimately I think there’s enormous advantage for the vendors to be more open and allow third party applications to interface. I’ll give you an example based on discussions with executives at large EMR companies. The customization that’s required for individual installs is enormously burdensome, and the imagination of healthcare providers and health systems is really fueled by much greater experience in the use of consumer technologies than has ever been the case before. So, young doctors are really digital natives, and they know what they want data to do for them. If one has open-systems and enables third-party apps to run on EHR data, one has the ability to share some of that development burden with the customers. When the customers do that development, they are essentially extending the product and making it more useful for other users. You can make a strong business case for why it’s actually in the interest of the vendors to open up. I think that will take some degree of reorientation, but it’s quite necessary for a healthy system.
What, if anything, have you seen from the government, in terms of mandates, that makes you think there could be a push for innovation from the vendor side?
I believe the government is very interested in promoting innovation on the vendor side, and is looking to do that through meaningful use and through developing lightweight standards for a data exchange. An example of that is the Direct Project at direct.org, or something like Query Health. So I think they are on the money in terms of setting up a few lightweight protocols that would enable innovation, data liquidity, and the use of data across multiple applications.
We’ve seen a lot of health related apps on mobile devices – how can these EHR vendors leverage this interest to drive patient engagement, better care coordination, etc.?
Right now, most of the health-related apps are either disconnected from health system data or are very specific to a specific system. And both of those types of apps have their role, but what we really need in addition are apps that run on health system data, and that run widely, and prevent the need for reinvention for functionality at the local level.