There were two interesting developments in the field of social networks for healthcare practitioners last week. The first was the publication of a paper in JAMA “Variation in Patient-Sharing Networks of Physicians Across the United States”. The second was the sale of Sermo Physician Network to WorldOne for an undisclosed price. Sermo had raised $40+m in venture capital prior to sale, making a bet that social networking for physicians could drive value to pharmaceutical and financial firms based on disclosing interactions between members of the network.
If physician behavior and prescribing activity are key to your healthcare business, I think it is important to understand the relationship and differences between these two events.
Sermo bet hard on the Facebook model – physicians would interact on social networks, share knowledge and insight, and third parties could benefit from getting access to those interactions concerning their products or services. Sermo had also begun expanding its revenue model by providing paid content and sponsored education programs to network members, trying to capture “digital” dollars from life science companies. Pharma companies are desperately trying to gain advantage through digital advertising campaigns to influence physician prescribing behaviors, and multi-channel marketing efforts including the development of web sites for branded medications.
These companies also struggle to understand how useful these branded web sites are by linking on-line activity to increased sales – linking eyeballs on websites to increased prescriber activity. Return on investment is difficult to measure and pharmaceutical companies traditionally rely on labor-intensive surveys to gain insight into how these digital marketing efforts are working.
The paper in JAMA, and the related editorial, do a great job breaking down the challenges and opportunities for building businesses or marketing efforts dependent on understanding physician behavior. Here’s the gist of their findings – it’s not about SoMoLo , it’s about big data. Physicians tend to work in isolation, they influence healthcare primarily through the individual patient-physician encounter. It’s difficult to track how physicians interact with each other and the greater healthcare system because there is no true healthcare value chain in the US. Physicians are busy, they don’t tend to spend a lot of time with on-line or mobile networks. And they tend not to like surveys, even when you pay for their time and opinions.
In a recent interview with Larry Miller, CEO of Activate Networks , Larry told me that patient referrals constitute 50% of interactions between physicians, with the remaining 50% being composed of collegial interactions, friends and coworkers, and associations within practice groups or hospitals. To quantitatively measure real physician interaction and influence you need to map the flow of patient data through the system, through insurance claims data, clinical data (electronic medical records) and patient labs to create a physician-to-physician interaction network. Larry calls this an “experiential, community level network” that has much greater value than an on-line social network influence model, and Activate Networks is seeking to capitalize on this model using a big data analytics approach.
The authors of the JAMA papers recognize that the ideal way to understand true physician networks is most likely a combination of social relationships and shared patient interactions. They also correctly point out that understanding how various aspects of networks are related to better care and lower costs could be central to enabling accountable care and successful healthcare reform. To drive quality and decrease costs, we need to understand where standards of care exist and how physicians influence those standards. Physicians care, payers care, and patients care. Pharmaceutical companies desperately care. If you have ideas on how to link these kinds of quantitative and qualitative insights, I have clients that would like to talk with you.
Jim Golden is a data scientist who writes about the intersection between big data, life science and healthcare. He blogs at Hacking Health on Forbes, where this post was first published.