Darwin's Medicine blog

Professor Brian D Smith is an authority on the pharmaceutical industry and works at SDA Bocconi University and Hertfordshire Business School.

Deep Value

Value-based healthcare is difficult because it’s deep

A lot of my current work, both the academic and the advisory, centres on the creation of genuine, differentiated, customer-perceived value and how life sciences companies can evolve to create this for healthcare systems. It’s a hard, knotty question and when I’m faced with difficult issues I always like to seek inspiration from Darwinian biology.

In this particular case, my mind wandered to convergent evolution. As usual, please follow me through the scientific thinking until I arrive, before the end of this article, at the practical consequences.

Much evolution is divergent or radiative; a single species, facing differing environments, developing differentiated traits until it spawns multiple species. But we also see the opposite: very different species developing similar traits because they are trying to adapt to somewhat similar environments. The streamlined shapes of sharks and dolphins, the wings of bats and birds, and fingerprints in human beings and koala bears are my favourite examples of convergent evolution. It’s interesting because it shows that evolution is a response to environmental selection pressures, which was precisely Darwin’s point, of course.

Evolutionary processes, convergent or otherwise, are always at the front of my mind when I’m trying to understand how the life sciences industry is changing and where it might go. Darwin was in my thoughts as I struggled with something a prospective PhD student of mine was talking about. He is a market access leader and he spoke with much experience when he said: “We all know we have to deliver real and demonstrable value to healthcare providers. We just don’t know how to do it.” He wasn’t
the first person to say this to me, and to hear bright people, who work for great companies, say something like this always intrigues me. It’s like an intellectual red rag to my academic bull. I can’t resist the question. How can it be that companies that can easily apply leading-edge science to product development can’t solve this health-economic problem? I’ve researched this topic with many companies and I felt my research was whispering the answer to me: it was about not only the data but how that data was processed into the knowledge available to payers and those payers’ attitudes to value.

From simmering away at the back of my mind, the challenge of creating value jumped into the foreground of my thinking as I read a new book, Deep Medicine, by the ever-thoughtful Eric Topol. It’s a wonderful book, full of relevant and useful anecdotes, but it is centred on a simple proposition. Deep medicine, says Topol, requires deep phenotyping (ie, masses of data), deep learning (ie, processing that data into knowledge) and deep empathy (by the doctor, to the patient). In a flash, I could see that Dr Topol’s thinking was convergently evolving in the same direction as mine. Whereas he was struggling with how artificial intelligence could improve healthcare, I was struggling with how life sciences companies could improve value.

With due acknowledgement to Topol, it’s clear that creating what we might call deep value requires three ingredients that are
directly analogous to those for deep medicine. Firstly, deep value requires masses of data about outcomes and costs of all kinds, the sort that only really emerges from real-world evidence, the health-economic analogue of deep phenotyping.

Secondly, deep value requires very sophisticated learning, the sort that probably requires not only processing power but also very realistic algorithms.

Finally, deep value requires deep empathy, this time by the payer for the whole patient population, the sort of empathy that transcends mere cost-benefit comparisons of products.

This deep value perspective reveals the challenge as one of co-evolution of three separate but related systems: the system for generating data, making sense of that data, and educating and directing payers. As a co- evolutionary problem, we can’t create deep value unless these three systems evolve together. No wonder companies find it so difficult.

As I’ve written about in earlier columns, taking a co-evolutionary perspective is helpful not only to explain the problem but also to solve it. It tells us that life sciences companies must work with their customers to generate vast amounts of the right kind of data. Those two partners must also cooperate to develop the methodologies to generate valid and useful knowledge from that data. Finally, providers must help their payers to evolve from cost- focused procurers to value-based treatment- enablers. As co-evolutionary studies tell us, the speed and effectiveness with which we can create deep value depends on all three processes being synchronised by signalling between all involved.

As Randolph Nesse writes in his lovely new book, Good Reasons for Bad Feelings, doing medicine without evolution is like doing engineering without physics. I think much the same is true if one attempts to create value in healthcare.