Joe McCannon: How can we accelerate the focus on spread and scale in population health?

This content relates to the following topics:

Article information

  • Posted:Wednesday 29 November 2017

Speaking at The King's Fund Annual Conference, 29 November 2017, Joe McCannon, Co-Founder and CEO, Billions Institute, explores the practical aspects of successfully achieving scale and effectively improving population health outcomes.

Transcript

The centre for Medicare and Medicaid innovation was really kind of a lynch pin of this idea that we need to introduce new models of population health that would be more affordable and sustainable and produce high quality outcomes for the entire population, but it also said in addition to that innovation work, what we want you to do is to find existing effective models that achieve the triple aim and see if we can find new approaches to spreading those as effectively as possible and that really was my baby, my charge because what we observed across the country in the United States was variation.  The very differences that you can see in cost and quality outcomes here suggested to us that there were models of low cost, high quality care across the US.  

So to start with, what I want to point out as sort of the typical behaviour that we saw and the typical mindset that we saw was this deep belief in natural diffusion.  The idea that if we’d had a successful experiment or a successful study, just by virtue of its merit, its quality, it will spread to everyone that can benefit from it and if you remember nothing else that I say today, please remember that there’s absolutely no evidence to suggest that that’s the case.  

Instead, what we know is that in those rare cases where effective practice is taken up by broadly in the field, it’s a product of very active work and very active dissemination.  We learned that those projects are initiatives that were obsessed with fidelity to the original model, to the first prototype or the first successful experiment.  We’re much less successful in those that were interested in adaptability and actually allowing the model or the practice to be produced in different ways, in different contexts, and it’s just so critical because context is so different.  Suppose we have a breakthrough practice in the circled area here in upper Minnesota, in Duluth in Minnesota, the way that that model gets implemented there is going to look very, very different than the way it gets implement in Boise.  We need to create an environment, even when we’re trying to orchestrate this change, where we very clearly give licence and give permission to those in the field to make that kind of adaptation and the less successful initiatives in spreading models of prevention and population health are ones that have vague goals as opposed to explicit timebound and apportioned aim.  

Here’s an example; a group comes along and says to us, our aim is to spread what we know to stop the development of diabetes in Americans.  You can see how this is not galvanising and it’s quite diffuse, but what they did that I think is beautiful is they, not only did they say we want to spread our intervention to 60,000 pre-diabetics in the next three years, which really gives you sense of the big goal and the overall hill to take but then they said by helping clinics to get 170 pre-diabetics each to adopt it which worked out to be five people per month per clinic.  

The next thing that we learned is that there’s a great risk to something called theory lock.  Most of the initiatives that we either supported or observed that were trying to take practices to scale had a silver bullet solution in mind, an idea that they felt would be the absolute key, the essential, that if they could just get that right, everything else would fall into place.  In the Medicare and Medicaid where I work, the biggest insurance company in the States, you know, our view was if we could just get financial incentives right, if we could just pay doctors in exactly the right way, as if by magic behaviour would change and the system would change across the board, and again no evidence to suggest that there’s any such silver bullet and that kind of theory lock is very damaging. 

By contrast, what we learned is that it actually takes three things in combination to move from where we are to getting to that large scale aim and meeting our goal for spreading infective practice.  The first thing it requires is awareness.  People need to actually know about the thing that we’re doing, and that’s a big lift in on itself; how do we cut through the noise and get people to say, okay, I’ve heard of you and what you’re trying to accomplish, then we’ll, how do we get people to decide, yes this is the thing that I want to invest my time in, of all the many things I could invest my time in and then behaviour change, even when people are willing, how do you actually take them through the blocking and tackling of implementation and support adult learners in implementing new practice.  

The next thing we observed is that the less successful attempts to spread models of prevention and population health were oriented towards the boss.  They were oriented upstream, and the more successful models were oriented toward the field and toward those who were trying to actually implement practice in a different way.  We always say it’s the difference between asking this question which is how can I get all these people to do what I want them to do by cajoling and manipulating and tricking and embarrassing, how can I shift from that question to asking how can I help all these people to accomplish what they want to accomplish. 

The next thing we saw in the least successful attempts to scale these models of population health was that there was kind of a mindset of central office broadcast which is that from Washington DC we will have Webinars and we will develop materials and databases and we will tell the country what the best practices are.  What we realised was, that our successful initiatives were the ones where something very different was happening, where there were breakthroughs of learning happening all around the country.  There was exchange between participants in different setting and different environments and we didn’t really have very much control of it.  The best that we could do is we could observe where the breakthroughs were happening.  We could sort of harvest and distil best practices and then redistribute it throughout the network or the system.  

We saw restricted adherence to plan overall, that there was a strategy that was put in place at the start of a programme and the work that people felt they had to do was to stick to that strategy.  That was much less successful than a structure where improvisation was allowed and the specific mechanism that I want to reference here is contracting.  In our government work we let these massive contracts of millions, tens of millions, sometimes hundreds of millions of dollars to do big pieces of work we require detailed descriptions of exactly what was going to be done eighteen months from now or two years from now.  Well that’s just absurd.  It doesn’t make any sense at all so we have to have contracting structures and methods and approaches that write in the possibility of constant adaptation.   There’s a habit I think in large initiatives to wait to assess progress at spread-out intervals, you know to wait six months or a year, eighteen months for someone to report to tell us how we’re progressing and where we are in the course of the work.  

In the initiatives that succeeded for us and the times we spread models of prevention and population health successfully, we took a very different approach, almost a command centre approach.  This is the situation room at the Department of Health and Human Services.  The person in the middle of the photograph sitting down with white hair is Secretary Sebalias and you know this kind of approach to looking out at the field, looking at maps, heat maps and run charts and control charts every day and being able to study and react to what’s happening, is really the essence of management of successful large-scale change management of large scale networks. 

We didn’t just sit there analytically and kind of naval gaze about what was happening out in the field, we used those data as a springboard to get out into the field right away.  So we always said the data are only clues.  Data are just clues, the answer is out in the field and we need to do everything we can to get out in the field as rapidly as possible.

I’ll finish where I started, simply by saying you know I think the work that you’re doing is so important, it’s so critical at this moment in time and you’re really carrying forward an enterprise that I invested a big part of my life in and care about a lot, so I hope that is in someway helpful to you and I wish you the very best in what you’re setting out to accomplish.  So thank you.