Healthcare IT Industry Spotlight Interview, Part 3: Casey West, Managing Director at SSM Partners
You’ve made investments in companies such as Apixio and RemitData that offer solutions based on vast amounts of healthcare data. What opportunities do you see in healthcare data science and how do you evaluate risks and regulations when considering healthcare data investments?
This is one of the biggest questions in the industry. In the status quo, there is more data, both structured and unstructured, in healthcare than ever, primarily as a result of electronic health records (EHR) adoption. So, we are capturing more data than ever, yet the historical challenges around data and healthcare persist: providers are concerned about who gets to see their data, maybe rightfully so. Who knows what payors will do with provider data? Solutions providers, I won’t name names, are also not so quick to make their data available.
Frankly, the integrity of data is a big challenge. Harmonizing data between different systems is not easy. You see this in the pop health sector where the smaller companies are really challenged to standardize data and glean useful, credible insights from many different electronic medical records (EMR) systems.
This is something that I have been thinking quite a bit about lately and consider a “Black box problem.” It’s a first order data problem that we have not yet come to grips with. We are flush with data, but providers tend to struggle to determine whether it’s credible. Thinking about the application of machine learning and data analytics in non-healthcare settings, I’m reminded of when I first subscribed to Netflix. It began making recommendations: some were good, some were bad, and it improved. In a healthcare setting, when you are making treatment recommendations to a physician, you are not going to get 10 bites at the apple. The opportunity to prove yourself is limited, so if the analysis is not good, you will lose your opportunity to influence behavior. That is a big difference between healthcare and other industries. As our society brings young physicians and healthcare providers up, it does so in a pretty rigid and hierarchical fashion with oversight. In this system a lot of mistakes are made, and hopefully caught by other providers who teach their students to do better. What is a similar context for systems to do that, and how do we train systems in a safe and effective manner? I don’t think we’ve figured that out, so the promise of big data in healthcare is enormous. We are all grappling with the enormity of that opportunity, but we still have some pretty significant hills to climb.
Apixio, which is a great company, doesn’t operate in the clinical space. It focuses on coding and makes coders much more efficient and accurate, which is saving people a lot of money. This is still a setting, however, where lives are not at stake. Nobody would die as a result of what would be acceptable “learning behavior.” We like that dynamic.
Open application program interface (“open API”) has transformed technology companies in other verticals, but has yet to make any real headway in healthcare. Do you think that open API will play a role in the next wave of healthcare IT innovation and investment?
We’ve invested in at least 1 company that created APIs for its partners to pull data from within our solution with limited success. In the broader context of clinical data, the lack of EHR harmonization limits your ability to glean real insights from your data. There’s a lot of data cleaning that needs to be done before you get clinical insights.
I think that a lot of the excitement surrounding open APIs in healthcare comes down to empowering patients to access their own data. That is less encumbered by the challenges faced by larger clinical data sets. It is still in its early days, but I can certainly personally relate to the desire to have access to my information presented in a simple manner.
What advice do you have for entrepreneurs who are deciding whether to bring on a financial partner such as SSM?
Select first based on chemistry, second based on potential to add value, and third based on price (of course, those are easy words to say, if you’re in the midst of a negotiation on price). I’d maybe flip the first two, but the point is that it’s a long-term relationship that needs to be based on trust if it will be successful. Chemistry and competency are core to that trust. Taking a haircut on value could make you much better off with a trusted, value-added partner. Additionally, just like in life, when you take on a partner, things change, and you should be ready for and mindful of those changes. It is a core part of growing up as a person and as a company.
I’m in Memphis for one day. What’s not to be missed?
You have to eat barbecue. You should also go to the National Civil Rights Museum. It’s a powerful, awesome place.