Recently, I asked Northwestern University Assistant Professor and Founder of PRIMR, Dr. Nicholas Soulakis, to tell me how he plans to use data science to innovate health care. Dr. Soulakis is a public health scientist whose research focus lies at the intersection of epidemiology and informatics; he is particularly interested in understanding the expanding, data-rich environment created by health information technology, and leveraging computationally techniques to monitor and improve healthcare quality. Previously, he was a programmer, Manager at American Express, and Research Scientist and Director at The New York City Department of Health and Mental Hygiene -- these roles led him to start PRIMR.
Marquis Cabrera: Where did you get the idea for PRIMR from?
Dr. Nicholas Soulakis: A year ago, I set out to radically improve how practices use and analyze data to orchestrate care coordination, especially in community health centers, which typically have complex delivery challenges because primary care is often interlaced with social services. I spent a lot of time visiting clinics, talking to doctors, nurses, social workers and realized that very few practices have the ability to take advantage of advanced analytics. I stripped the whole thing down to the bones and realized practices needed to engage their own data, trust their numbers before they could even begin to go down that road. They basically had a lot of do-it-yourself tools, but not a lot of confidence in how to use them -- even worse, I was handing them the most complicated one. I realized what they really wanted -- simple numbers they trusted - I could probably program myself in hours with the right data sources. After that, it was just a matter of architecting a service that would scale.
Marquis Cabrera: What is PRIMR? What is the value proposition?
Dr. Nicholas Soulakis: PRIMR is a services marketplace for small ambulatory practices to connect to reliable consultants for analysis they can trust on a secure data warehouse platform. PRIMR promotes high-value, lasting relationships between small clinics and skilled analysts, and, by proxy, stronger relationships with their data.
Marquis Cabrera: How can analytics (and data science) improve primary care practices? What pain point is PRIMR solving? Why is this such a big problem?
Dr. Nicholas Soulakis: Many small practices struggle just to survive in a metrics-driven healthcare market. Our goal for PRIMR is to see those practices thrive by using better analytics to improve the quality of care for their patient populations and increase revenue by successfully pursuing quality incentives. Practices can also use analytics to build business capabilities for communicating their progress to stakeholders and the community in annual reports and presentations.
Practices of 10+ employees make up 20% of the ambulatory practices in the US, but command 80% ($700Bn) of the revenue. That leaves almost 400k small practices with considerably less revenue, yet all of the same challenges faced by large practices. Hiring an in-house analyst is expensive, consultants vary wildly in quality and credentials, and sensitive health data requires expertise for extraction and storage. EHR tools can be easy to use, but they can be expensive and also require in-house expertise. Big technology platforms can be a good fit for large enterprises, but they're often out of reach for a neighborhood practice.
Marquis Cabrera: Can you give me a use case on how your product would be used?
Dr. Nicholas Soulakis: In the pneumonia example, a practice manager would visit the PRIMR marketplace, which is personalized to their practice. If she knew that the practice priority is to improve pneumonia vaccination rates, she could access a catalog of pneumonia reports. But more likely, she would connect with her favorite PRIMR analyst who would suggest various quality improvement reports. Those reports can come from the analyst themselves, existing PRIMR templates, or from co-branded bundles from quality-improvement organizations, professional societies, or insurance companies. She chooses the report she wants, receives an estimate of how long the analysis will take to complete, and the price. She pushes the button, gets her analytics.
Marquis Cabrera: Have you tested an MVP (minimum viable product)? Have you gained traction?
Dr. Nicholas Soulakis: We are planning our proof of concept now. Over the past year, through the InciteHealth Fellowship at Harvard Medical School Center for Primary Care, we worked with some fantastic mentors and early partners. I connected with our state hospital association, which is very innovative in their use of data for quality improvement and forward thinking in how to service their members -- over 200+ hospitals, with 1000+ affiliates. The hospital association hosts both an all-payer claims database and a health information exchange, creating nearly unlimited demand for insightful analytics. PRIMR can offer a way to add capacity and sustainability to their quality programs while avoiding high personnel costs. InciteHealth seed funding provided us the opportunity to take on an ambitious first project and build out our channel partner strategy which should allow us to add about one clinic a week.
Marquis Cabrera: How would you guys make money?
Dr. Nicholas Soulakis: After the clinic is satisfied with the work, they pay the analyst on the platform. We take a percentage of the payment. Our rating system is weighted to promote repeat business and long-term relationships which also determines analyst payment rates. Analysts can expect to keep less when just getting to know a clinic, but as they establish a track record they can keep as much as 80-90% of the transaction. We provide the statistical programming environment and data storage for the analysts, making the entire service is a good deal for everyone. Also, the clinics pay a subscription fee which goes towards our operational costs.
Marquis Cabrera: What has been the feedback from clinics about your product?
Dr. Nicholas Soulakis: The product itself is still in development, but the experience of getting analytics from a credentialed analyst with a shopping cart checkout has been a runaway success. Feedback on our prototypes from testing partners has confirmed that they enjoy the simplicity of the PRIMR marketplace. They are used to 6-week implementation cycles or even worse the dreaded "Get Demo" button -- the velvet rope of expensive technology platforms. We care about population health, but we also recognize small practices are also small businesses. They have grown accustomed to having small-business options in accounting, insurance and payroll services -- all readily available through easy-to-use platforms. They are thrilled when they realize the patient population data that is so critical to the success of their practice may be just as easy-to-use through PRIMR.
Marquis Cabrera: Although you're selling B2B, how does this improve the quality of care for a patient visiting their primary care doctor?
Dr. Nicholas Soulakis: Although we promote better analytics and use of data, the goal is to promote the health of populations. Patients are more than the numbers generated by data analysis. When a provider turns on a decision support rule to improve their pneumonia vaccination rate among the seniors at their practice, better analytics allow the provider to monitor their progress and set performance goals among sub-populations. For us, success sounds like "My doctor always reminds me to get my shots" as much as it does "We went from 5% to 30% penetration among our target population in our first year and hit 55% after three years."
Marquis Cabrera: Awesome! What advice would you give to someone looking to innovate the primary (health) care field?
Dr. Nicholas Soulakis: Never lose sight of the patient perspective. Even an "enterprise solution" needs to respect the relationships of providers to their patients -- you are not going to create a new culture, but if you are lucky you will realize the sophistication, ingenuity, and creativity that happens everyday in practice settings across the country. To me, 'disruption' is overstated, 'articulation' is admitting you went out with a solution in hand and came back with a better one when you took the time to listen and observe.