Diffusion Of Innovation



good morning my name is Alan while I'm editor in chief of Health Affairs I'm really thrilled today to be the host of the event associated with the release of the February issue of the journal health affairs and it's devoted to diffusion of innovation I always say a few words at the outset of events like these just to set the stage this is a topic near and dear to my heart and it was a challenging issue to put together the but which makes me feel even more proud about how it turned out and I just want to spend a moment on that theme everyone loves innovation we're all for it I've yet to find someone who's against it diffusion of innovation however is the difference between a good idea and a good idea maybe taking root in one place and a good idea actually having a broader impact and most people who say they're in favor of innovation are in favor of it because they think good things will happen because of it and if we do not diffuse what we learn and if we don't fail to diffuse what we learn doesn't work then all of the hype and energy about innovation is for naught and so we were really happy because I would say pretty much every issue of Health Affairs has at least a few articles about innovations to pull together a theme on the on diffusion of innovation and if you're here hopefully you know a little something about the topic and you know the diffusion is hard and it's slower than we want it to be and it's less orderly than we want it to be and the effects even if positive in the first instance are not always replicated when we diffuse but we also know that there are a lot more resources put into innovation than into diffusion of innovation because the activities that are necessary to diffuse are frankly not as sexy they're not as exciting to funders to organizations to government we also know a lot about the role of culture in diffusion not just adoption of maybe a new technology like when a new phone comes out and we all want to run down and buy it but if you're talking about organizational change which is what most of the innovations people are looking at now in the delivery system organizational change has a whole cultural element that's quite different and more complex than adoption of a technology in Washington I would say that a the key lever that is used for innovation is the key lever used for most things in Washington which is money and the notion is if you either create resources for Ana for innovation or you create incentives of course for things to happen then people innovate in response to those incentives and and I think there's no doubt that that is true what it leaves behind though is all of the actual one-by-one changes that have to happen for that to occur and sometimes market forces and organizational cultures will embrace that and sometimes they will not and so we also have a whole host of intermediary organizations that have realized that they have a place in this and that's to take good ideas and try to help bring them to scale so it's it's complicated but it's really important and I that's why I'm so happy that we have the issue this month because if we continue to publish about individual innovations every month as we will do but we fail to embrace and understand the complexity of diffusion we will have suboptimal outcomes and the investments we make an innovation will not pay off so as we pulled together the issue the the general contents are the same structure we're using this morning and that's why I wanted to give you a sense of what's in the journal issue as well as what we're talking about today so there's a lot of sort of how does it happen a little bit more theoretical there is a whole theory of course around the diffuse of innovation and its application to the healthcare space is is complicated and so we need to understand the theory behind it our second panel will look at models of diffusion we particularly have specific endeavors funded in these in this case by federal agencies to take particular approaches to scaling and spreading new ideas and it's important given the investment made in those for us to understand the implications of those we have a number of what I'll loosely call case studies in the journal only one of them is what you would literally call from an academic perspective a case study but stories of innovation which again we need to read and understand because you get tremendous insights about what it takes and you get tremendous insights about what people try and feel don't work from learnt from looking at those lessons we have a number of papers then on that actually analyze specific innovations and their diffusion and look at how they affect patients providers in the health systems so this issue which as I say I think it kind of touches on a really central topic for healthcare spans all the way from theory through models through practice through impact and I'm really proud that we have such a great constellation of papers and so each of them individually can teach you lessons but if you're really trying to understand how it happens what it takes to happen the the collection of papers does that so forgive my slightly longer than usual introduction but this is a such an important topic for healthcare right now I wanted to give you a little bit more of context then we often do so as we go into our presentations today there we go we do have at the front of the issue a data graphic that pulls together a collection of the papers that in some of the highlights graphically from them I'd encourage you to look at that as well as we kick off I want to acknowledge the support of the peterson center on health the Blue Shield of California foundation the Leona and Harry Helmsley Charitable Trusts and the agency for Healthcare Research and quality for the support that it takes to put together an issue like this and an event like this and so let's move to our first panel which looks at how diffusion happens we'll be hearing from Jim Dearing professor and chair of the department of communications at Michigan State University studied and worked under Everett Rogers well known for his seminal work on the diffusion of innovation and he's extended it in some very interesting ways that you will hear about today andrew Ballas professor in the College of Allied Health Sciences at Augusta University where he recently was the Dean among his other other entries on his bio I noticed he was the wild distinguished professor of health policy at the University of Missouri that's a chair I endowed myself I'm sure you know and David door professor and vice chair of medical informatics and professor of general internal medicine and geriatrics at Oregon Health Sciences University where I was born so anyway enough about me let's turn it over to the panel and I'll give it first attempt Thank You Ellen a pleasure to be here with you today thanks also to the staff at Health Affairs for putting together this special issue it's been about a year-long journey I think a very good one and honor to be presenting here with both Andrew and David what have we collectively learned about the diffusion of innovations one of the key lessons in my opinion of this strong collection of empirical articles in the February issue of Health Affairs is the importance of partnerships and in particular the importance of partnering with intermediary organizations to extend the reach that innovations can achieve to increase the rate of adoption by individuals or organizations be they counties cities health departments healthcare systems and especially to lend credibility to the innovations themselves through association with those intermediary organizations or partnering organizations partnering helps to boost innovations up diffusion curves I think this is very well established by this current issue of Health Affairs you can see it clearly in the work of arlene berman and colleagues in their focus on learning collaboratives as a type of intermediary or organization you can see it with Rocco Perla x' work with government and CMS playing much the same role with Sarah Ono and looking at extension both Cooperative Extension and then healthcare extension as a type of intermediary organization again to broaden the reach to increase the rate and to associate high credibility intermediaries with particular innovations we've learned in the study of diffusion that innovations when you chart them either by percentage or number as this diagram suggests over time in a cumulative fashion those innovations exhibit very often a non-linear trajectory that is a curvilinear s-shaped curve typical of many diffusion studies I'll talk more about that in a moment we also know that innovations decelerate over time they're displaced displaced by newer innovations not necessarily better but certainly newer we also know that innovations when they're diffusing or not they exist in a competitive environment as this diagram also suggests that is while most studies focus on one particular innovation and chart it's spread usually that's not the perspective of the potential and actual adopters they have choices in what to adopt so it's a competitive marketplace for innovation adoption lastly most innovations don't diffuse diffusion is the rare situation usually what we see is failure to diffuse of innovations and that is not always do as you know to the quality or effectiveness of the innovations in question we've learned that there are particular attributes about innovations themselves that can sometimes explain a lot of the variants in why people do what they do and how we respond to innovations in particular cost is a key predictor as is the effectiveness perceived effectiveness of the outcomes of using and intervention as is complexity or simplicity how easy an innovation is to understand number one and number two how easy it is to implement and use on a regular basis cost-effectiveness and simplicity along with a fourth attribute or characteristic of compatibility are consistently found to explain considerable variance in adoption decisions we've also learned that the extent to which you can see the results of an innovation that we call observability and the extent to which you have to commit staff or resources to using an innovation and implementing it what we call trial ability those can be under certain circumstances very important pros and cons also explaining diffusion we've also learned that social influence matters greatly that is especially for what we call consequential innovations those that a potential adopter is is convinced is going to affect how they work or how they live or be very important for their organization or their political jurisdiction people don't typically make such a weighty decision only based on the perceived attributes of an innovation themselves they look to others they look to others as keys as a heuristic for a basis of what to do themselves in fact most of the time for most innovations we are looking to other people first rather than making careful rational so-called rational cost-benefit analyses lastly a few principles about diffusion contextual conditions political conditions real-world events can greatly affect receptivity to innovations thus it behooves innovation proponents to carefully consider the introduction and framing of innovations that is waiting can be an excellent strategy if one wants to spread in innovation secondly evidence of effectiveness can be less important to diffusion than cost complexity and compatibility this is a rather consistent finding we see plenty of ineffective innovations that have spread just like we see plenty of very good innovations that go nowhere third innovations perceived to be consequential can take years to diffuse because of careful measured pros and cons more or less rational assessments by influential members of adopting communities that is opinion leaders or influential members of a particular type let's say YMCA directors let's say City Public Health officials if they're considered to be influential by their peers those influential individuals tend to carefully assess pros and cons whereas most of us most of the time do not once informal opinion leaders adopt diffusion can be difficult to stop getting over that initial hump is key fifth policy diffusion is as Allen suggested very aided by resource allocations and in particular with policy diffusion studies we find that many innovations start locally go national and then spread back down to a state level or county level or sometimes city level for implementation sixth state adoption of innovations is often the result of bursts of federal attention perhaps no surprise but also state to state proximity states adopt sometimes based on what neighboring states do we also find that hemolysis eight states that are similar to each other in the US and political jurisdictions like counties adopt when similar counties also adopt seventh social disparities do result from the diffusion of innovations we know this well and it's a basis for then of course targeting to disadvantaged populations with evidence-based innovations eighth knowledge about diffusion processes can be used to affect the rate at which an innovation spreads and which segments benefit first from innovations whereas otherwise they would not ninth reinvention is becoming a very hot topic there is strong evidence increasingly that you can do a lot to take the externally valid innovation and tweak it change it reframe it to encourage broader adoption among a wide set of potential adopters and then lastly easily adaptable or customizable innovations are more likely to spread than those that are not thank you very much good morning Mary Alice from Augusta University thank you very much for coming and thank you Health Affairs for creating this opportunity this wonderful opportunity to talk about the dissemination of innovation and thank you for your leadership Ellen I really appreciate that and we all benefit from that I know that we just have a few minutes for this for these introductions so in the interest of time I would like to start in the 19th century then there was an interesting encounter at the Chicago conference where Robert Wood Johnson the first job went he was an exhibitor there and he went to a presentation by Lister about surgical a severe sepsis and that was the time and he kind of got the idea and started developing all the supplies that are so much essential to achieve appropriate implementation of those wonderful scientific discoveries that surgical a sepsis actually brings to all of us to our patients unfortunately this lesson but obviously the company became extremely successful and did a lot of good things but more good things for healthcare but we still did not learn as a society how much implementation is important and how new ways need to be explored continuously and there are two kinds of problems as we look at the history of dissemination efforts one is certainly when there is a delay and then we talk so much about that but there is another type of problem that we also need to recognize when we see that certain innovative ideas or discoveries actually get into practice too quickly and one example could be the intensive glucose monitoring and managing in ICU patient populations that was recommended by a very respected study in a high-impact very high-impact Journal and very very quickly spread this practice and several years later another larger clinical trial raised some concerns that maybe it is really not as helpful in preventing mortality in the ICU and at that point it was already quite widely spread and certainly this is something that we need to to be careful about that there were some studies that actually suggested that there is a possibility that because of the too fast implementation of that particular innovation the about 26,000 patients died in our ICUs so this is a very staggering number and certainly should raise concerns about the appropriate speed so it's innovation and the diffusion of innovation is the more more than more than speed we need to look at other issues as well and I can tell you that in our study we looked at a number of factors at several years ago that define how in that particular case preventive care procedures go into practice and this study they became much cited that it takes about 17 years on average to take preventive care procedures from the clinical trial so it's not even from the original basic research but from the clinical trial from the successful clinical trial into practice meaning that 50 percent of practices actually used them so I always warn people that when you go to the doctor's office don't complain about the old Better Homes and Gardens because maybe that's actually a newer thing so it is sometimes certain things got into practice very fast others very slowly and somehow the whole process needs appropriate revision and I think that the house affairs issues landmark in that regard because it really puts the spotlight on the demands of the implementation process and the complexities of that process so in this particular study we looked at various models and we looked at various malfunctions of the of this you know implementation dissemination process and you can learn from from both and obviously the Rogers already we had a major play a major role is the various ways of adoption and it really illuminated how dissemination of innovation happens and there have been a number of refinements of of the eteri over over time but I think that people realize that there are there are some waves and and certainly there is another trend that we need to recognize that the knowledge as it becomes practically valuable goes through a series of steps of transformations and I think that this table which can be found in the in the article and the journal issue gives this message that when you look at innovation dissemination you have to think about the waves of adoption but you can also and you have to think about the knowledge transformation that should parallel that because that's how it is it is becoming a harmonious and successful process obviously the first group is the the pioneers those who actually conduct the clinical trial and after that immediately after that the comes the question so how much it is sustainable what is the promise of that and how do we develop the business model how do get we get the supplies and so forth and then the next step is when it goes to their pioneering group a leading group have the larger group of institutions and then the mainstream of practice takes it over and then at that point the we need to put more emphasis on the public health impact of the innovation and also on the various ways to send the message that this is the right right practice and then they filed and if it becomes the more complete that's the point where we probably will worry about a lot about the practice variations and also the underserved communities and their access to state-of-the-art interventions so with that note I just want to mention the summary of MA our recommendations one is that the certain knowledge transformation needs much greater emphasis in everything we do and if you want to make a difference in the use of in the practical implementation of biomedical research discoveries clinical discoveries we have to invest more in knowledge transformation a large part of it will be electronics so the office of the National Coordinator already recommends a number of standards to facilitate such exchanges and we are fully supportive of that actually we would like to see that role being expanded and certainly as we move forward we need to phase these activities and help institutions to give goes to all the waves in an appropriate sequence as opposed to jumping into too quick implementation or being too long and too delayed in in in their efforts so in that process perhaps some measures of the capacity to change which is such an important characteristic of successful organizations should be developed to evaluate that as opposed to the steady measures that we use today so is that note thank you again House Affairs and I'm looking forward to the next presentation thank you so much it's a pleasure to be here and actually to follow dr. Ballas whose papers I read initially when I started out in informatics he's well known he and dr. Chapman who wrote that article so I'm David d'Or from Oregon Health and Science University and really this is a nice sequence because in our paper we looked at the degree to which data drives diffusion of innovation by focusing on three large health reform efforts mop innovative delivery models and we found really three things one that absolutely data is driving diffusion of innovation but it's also incredibly a part of the innovations themself in medicine and that creates a lot of complexity for uptake and risks and second in these delivery models that people adopted technology to try and use this information data and information more effectively they did a lot of adoption but it did not nearly meet the need that they had so there's still evolution that has to happen there and three that technical assistance was very helpful in getting these participants the people adopting these innovations unstuck from their data-driven problems which we'll see where many oops how did I do that okay we are not good at IT in informatics just FYI so I just turned off the screen so really this is a journey with two co-authors who are absolutely wonderful here dr. Julia Abner Milstein who's now at UCSF worked on a study funded by the Commonwealth Fund on accountable care organizations which are large collaborative zuv healthcare entities that take on a population the health care needs of a population in terms of both quality and cost I looked at advanced primary care models where primary care practices that want to transform their care for a population often in terms of also quality and cost of care focusing on comprehensive primary care and the the subsequent Plus model that came out of CMMI the Center for Medicare and Medicaid innovation and my work was also funded by the Commonwealth Fund and then Deb Cowan who actually you'll hear another talk about a group that she's involved with as well escalates where she wrote about the evidence now the the over the the endeavor funded by AHRQ that actually focused on improving heart health measures for the small to medium primary care practices that didn't have as much experience in quality improvement and so that we thought this was particularly good because these or these endeavors really went across this diffusion curve if you the ACO s might be thought of as early adopters or even innovators they're really trying to find new ways to do things the advanced primary care self selected so these are practices that are probably on the early phase and then the evidence now practices for a variety of reasons might be later majority in terms of adoption of Qi and as we thought about this each of the phases or issues related diffusion of innovation has data involved to some extent from knowing about the innovations to being persuaded perhaps by your own data should I do this or not and then making that decision but a lot of what we found was that in the implementation itself data drove how the innovation was taken up and how it was ultimately successful or the challenges that it brought the I just wanted to make a quick statement here that there was also a lot of you can see from this table that describes them there's a lot of technical assistance that's provided in the AC o–'s internally but for CPC and CPC plus the advanced primary care models from organizations intermediaries that actually helps to do this and they use data extensively to in evidence now was set up such that cooperative seven cooperatives actually we're helping even to extract the data in many cases to help drive the innovations forward and that's really important so what happened first we found that they use they adopted technology extensively in these models and intensified their use of it in many cases to drive the innovation both in terms of just actually doing the innovations such as adopting and intensifying or modifying EHR use to creating population-based registries that really stored all the information about particular conditions or needs for a population of patients so they could see how they're doing infill care gaps and a cos did this as well as advanced primary care that really focused on higher need populations through risk stratification sort of identifying who's at high risk and then doing complex care management for that and they to found that they needed to use a lot of health information technology to manage this data to understand it and to act on it efficiently they also reached out to patients through use of patient portals unsurprisingly evidence now expressed a number of concerns with actually adoption most of them had EHRs the vast majority most of them were participating in federal programs like Meaningful Use but they really struggled to modify their EHRs or exchange information but that wasn't unique to them even though they were later in the diffusion curve each of the initiatives reported that modifying their health information technology especially electronic health record systems the sort of point of care information systems that we use was very difficult extracting and exchanging information as dr. Ballas just referenced the work of ONC has been trying to improve this and still was a major problem and limited diffusion and then actually trying to manage those populations through the registries was difficult all of them also reported that they were absolutely exhausted from the number of changes that were expected from them and really pointed to the health information technology as a major barrier or a major source of that fatigue on the bright side the technical assistance was very helpful and we see that they used a variety of technical assistance including their peers including online portals and forums where they could go to find out how they could improve we use practice facilitation both in a general set we see this both in a general sense where they're helping the practice is to get unstuck but the practice facilitation and many of the endeavors needed more expertise from H IT experts and others to get unstuck so in all in all data-driven diffusion was and there was a lot of adoption in these but it still wasn't nearly enough and so focusing on getting this data out and making it so it can be shared or interoperable more easily is important making the programs more flexible so that you could really focus on what needs to be adopted and then these technical assistant components were often pointed to is what helped to drive the success that they saw thank you so much I'm going to begin with just a couple of questions of the panel Jim you made a quick reference to reinvention as sort of the way the early literature on diffusion was really a one-way this is now a different way of thinking about it I just wonder if you could go a little deeper in that and what we what we know about it yeah thanks Helen let's see is that on yep when I say reinvention what I mean are proposed changes made to an innovation by the developers of the innovation or other proponents like intermediaries who are trying to spread and support the innovation so there there are this is a tough one reinvention and it's and it's distinguished in part from what we also would term adaptation those changes made by users of an innovation itself that might never be understood or acknowledged by the developers or proponents things that just go on in clinical practice or maybe in state government but reinvention is done purposively for the purpose of accelerating or broadening spread and this this you know the phrase I think of most most often in this in this reference Ellen is not letting the perfect be the enemy of the good because oftentimes it's taking something from clinical trials that's been externally validated through multiple research studies and coming up reinventing that intervention such that it might achieve less positive results to a degree but it can reach and be supported and be affordable by many others thank you Andrew you you and your paper and your comments talked about two rapid adoption it's easy in retrospect to see examples of that I'm trying to figure out in real time how we can protect ourselves against that given the competitive you used a clinical example but from a delivery system example the competitive advantage of adoption if if something seems to be working well even in a clinical setting the payers are involved in diffusion so I can it as I say you can see it in the rearview mirror how can we see it better looking in forward question I believe that looking at the public health impact of innovations is a is a step that is frequently skipped and vanity skip then it really counts every everyone so this is just another example that the old concept of implementation of adaptation is that read it and do it is really flawed it never worked and it should never work it did actually the implementing and innovation is an iterative process I wholeheartedly agree with the concept of reinvention this is the scientific discoveries yeah that's that's that can be a moment then when you recognize something but you have able to develop a best practice that is a process over time and we should respect every step of the process not just a first and David you just describe the role of data and data analytics if there's any phrase hurt as often as innovation its data analytics in I I do wonder as you think particularly in some of those models where there were well you know in all senses their resource constraints but particularly when you're thinking of practice is trying to take on projects the degree to which the need or the expectation of the investment in data and data analytics crowds out some of the other things that organisations also need to do to change so I guess I'm trying to put the data in the context of the broader environment of innovation yeah so that's an incredibly important aspect especially for informatics and as we are in sort of the hype cycle of big data and analytics which is where is their real value especially given the potential cost because sometimes expertise for data analytics can be expensive especially if you're small or medium sized practice that you don't have that locally and actually acquiring it we saw in several of the endeavors was expensive and often actually the technical assistance got them unstuck by saying you don't need something really complicated to make this work your clinical and intuition can be good as a start and then you can add simpler rules or algorithm or approaches as you develop the capacity because there's also a time lag to really get those implement and working well especially when there's data extraction and getting them unstuck so they were working on it inevitably improve the quality of the of the innovation adoption overall without necessarily sacrificing that much of the quality of the insight because many of our analytics approaches don't really provide that much of a jump in terms of ability to understand and to act better for patients and these populations so you know I think their skepticism I sort of share a little bit in terms of this is expensive do I really need it and we said don't let they in it perfect be the enemy of the good and don't get stuck because you don't have access to this or you can't implement right away we have time for a couple of questions among the audience we have microphones I don't know if I'm gonna let you ask like wait wait for the microphone let everyone know no no Jeff saw Berger Peterson Center Jim I I really have been waiting to ask you this question it and you and you set me up well by showing the the social media network and I'm curious about advances in in network effects finding opinion leaders boundary spanners does that change ever Rogers theory of diffusion and the sense of find the innovators early adopters and the like how do you blend those two together or do you really good question I don't think it it changes his main theoretical components really he published he co-authored a book back in 1973 that was some people regard as the introduction of network analysis has applied to issues of diffusion for innovations into communities health innovations so f had clearly been aware and thinking and trying to figure out network analysis the tools weren't as sophisticated in the early 70s or mid 60's but you can do network analysis with paper and pencil and sometimes they did so I don't think it really changes his major components of his model but there's been so much advance in network analysis different network programs and new research findings about the role of key influentials and how to find them so I'm gonna follow up it's a little bit of a diversion but but it makes me want to ask also a question about we use this term diffusion each of you have talked some about clinical diffusion some about sort of delivery system diffusion some about policy diffusion I'd love to get each of you to give your take on commonalities differences across those so that we're not using one model of diffusion for things that are fundamentally different or maybe your sense is that there's actually mostly commonality across those about I'd be really interested to know because all of those are important types of diffusion well to me the when you talk about different diffusion challenges the the ultimate goal is to spread best practices even if they are kind of presumed best I mean not yet necessarily proven and and and spread so we have the opportunity of looking at them only at the ideal like the policy or the technology just to see how they impact the actual best practice which is the clinical practice so at the end of the day we want to see a better patient outcomes and everything should be measured in relations to that other thoughts on that yeah this is sort of interesting so I'm thinking through my experience you know one of the privileges I've had is to try and work from diffusion from policy to implementation and the technology and you know they work best when you can keep the overall goal in mind I think the improving health and and well-being in populations and especially when there's efficiency that's needed limited resources but it is true absolutely that the the way in which you shape diffusion in each of those is quite different they speak a different language the risks you know when you're trying to define a policy and then actually implement it to the point you know that people will take it different ways and may there might be gaming for technology you absolutely know that there will be unanticipated problems that will significantly sometimes stop what you're doing and so to that end I think having the expertise for people who not only know how to create but also know how to implement or adapt as you go is is really important from each of the domains and that's why sort of this multifactorial technical assistance or multi-component seemed really important to us in that you could have people who could assist at the different levels or sort of different strategies if you're doing a large-scale innovation like we saw Allen you've teed it up really nice for the next panel it's true but you get one last shot if you want it but you're right I just say that I you could read the next set of papers that speakers will introduce as having at least as many commonalities as models for diffusion as they have distinct approaches very good ok please join me in thanking our we will move directly to our next which will do just as Jim said looking at some different models come on up I'll get yes I see oh I'll give you an approach so as they're coming up this is a segment on some different examples and models of diffusion Arlene Berman is the director of the center for evidence and practice improvement of the agency for Healthcare Research and quality she's a general internist and geriatrician Rocco Perla assistant professor of health services research at the University of Massachusetts formerly president of health leads and director of the learning and diffusion group at the Innovation Center at center for medicare medicaid services Sarah oh no cultural anthropologist specialized training and feminist anthropology and assistant professor in the Department of Family Medicine at the Oregon Health and Sciences University I'll turn it over first to Arlene sorry to make you sit and then stand back up good morning it's really a pleasure to be here this morning and I really want to acknowledge my coho authors because this was a really true cross-agency effort at arc and the author's represented four different orc offices and centers so we diffused across our own centers with many decades of cumulative experience in developing implementing and evaluating learning collaborative so learning collaborative collaborative consists of multiple parties joining forces to have to accomplish your grip sorry learning collaborative multiple parties joining forces to accomplish a goal and obtain or create knowledge and they have many different purposes varying from really knowledge focus for example research collaborative stew problem focus for quality improvement collaborative and the social activity of learning as a hallmark of collaborative which leverage and support peer to peer learning expert to peer learning or both and also they they have many different names that you know knowledge networks community of practices and we used learning collaborative zzz's and umbrella terms learning collaboratives are increasingly used to hasten diffusion and implementation of innovation evidence and effective models of care and they actually the other presentations on this panel really illustrate that really well it was great to read the other papers and are contributed to the p4p the partnership for patients at CMS and for the evidence now collaborative provided technical assistance and brought together learning across the different evidence now collaborative well despite all of this use really the factors contributing to their success or failure are poorly understood we've heard some about those earlier today and the evidence on their effectiveness is really mixed so work has sponsored collaborators for nearly two decades I guess we were early adopters and for this paper we examined 15 of her collaboratives and identified the factors potentially contributing to their success or failure so we started with the practice-based research Network collaborative in the year 2000 and our analysis and more informed by literature review was used to develop a taxonomy that can be used to support the development evaluation and study of learning collaboratives and taxonomies really helped us classify a diverse knowledge and arrange in order it to really make sense out of the chaos so what I'm going to do today is provide a brief overview of our taxonomy and the next two slides show the primary and secondary elements of the arc learning collaborative taxonomy and that the taxonomy includes four primary elements innovation communication time and social systems and 19 secondary elements as well as 78 tertiary elements so we really drill down into the secondary elements in detail and that's available in an appendix online our primary elements map to Rogers diffusion of innovation framework I guess he's the Bible what we started with him and the secondary elements mapped to Wilders collaboration Factory inventory Rogers framework as well as elements we identified in our analysis so dr. Dearing already told us a lot about what is an innovation but Rogers defined innovation as an idea practice or object that is perceived as new by an individual or group and innovations may be almost entirely composed just of information so we can spread ideas so we adapted a Rogers definition by expanding it to include the non-directed organic sharing of ideas and practices that in the end might or might not be objects of diffusion we made this adaptation to allow for learning through the exchange of ideas so what are four secondary elements this slide has the first two for innovation the elements include the type of change collaborators vary by the type of change being sought to advance knowledge to improve quality or safety or to develop or sharpen skills so evidence now collaborative seek to improve cardiovascular risk management in primary care by increasing performance on a bundle of four related evidence-based practice recommendations whereas on the opposite end of the type of changes in Medicaid medical directors Learning Network served as a forum for discussing state Medicaid drug coverage policies among other things that was really directed by the members of the network so that takes us to the degree of prescription the extent to which the convener of the collaborative sets forth a predetermined agenda the convener may also predetermine the aim as well as what what is going to be diffused or it could be open so for example the community care coordination Learning Network focused on implementing the pathways models to connect vulnerable populations to primary care whereas the Medicaid medical directors coat network determined their agenda so scope refers to the breadth focus which in turn refers to collaboration aims in geographic boundaries so some have had a very narrow focus the medication therapy management learning community so it really don't prove care for patients at risk of complication from uncontrolled type 2 diabetes at federally qualified health centers in Houston Texas whereas others had very broad scope such as the chartered values exchange Learning Network which focused on improving quality in 24 regions across the u.s. representing a third of the u.s. population and again part of the innovation is the supporting tools of that available which includes resources products technology to improve the understanding of the innovation increase it in efficiency of its adoption and spread and provide other support so the second primary element is communication how when we are members of and convenors of collaborative zhalong those within invited experts or innovators share their messages knowledge resources and insight in support of ensuring influence and goal attainment so the elements the secondary elements of communication include the motor venue is it online in person the directionality is it peer to peer experts appear or both frequency how often do they meet the degree of formality that is the extent of structure of convener orchestrated diffusion or how much is you know kind of spontaneous all arc collaboratives have encouraged unstructured peer-to-peer learning which is has been both informal and organic and most our collaborators have featured both informal formal and structured components yeah yeah so I'll just be quick on the next one so the next one is the third elements are time it's important the duration how long does it last how long does it take to recruit members sometimes it's easy sometimes it's part of the process sometimes it's hard to get people on board the rate of the adoption and also sustainability is it is it sustained do people think about sustainability is there a desire to sustain the collaborative and to be able to keep that you know going explicitly or stopping it depending on your goals and some of arts collaborative stopped when funding ended and others found ways to sustain themselves and an example is we had an emergency department use learning collaborative that was sustained basically when a member of the collaborative decided to keep it going so the social we've heard a lot this morning about social systems and that's a key element of these networks and the social nature of diffusion learning and collaboration and concept of systems interrelated parts that make up a whole comprises social systems and I think these will be familiar to you the elements and for time I won't go into detail but the credibility of the convener the character is characteristics of the members are they all the same or they a diverse group governess governance purpose of a shared vision the culture of collaboration the activity level and engagement of members and also roles processes and structure so for time I can't go into all of that but I'm gonna just sum it up so calabria as really as we moved to learning health systems can play a critical role in the development and evolution of learning health systems about implementing evidence and generating evidence from practice about what works we think our taxonomy can support learning collaborative design planning budgeting and evaluation and identify characteristics of the innovation time elements for implementation communication strategies elements of the social nature and celeritas can play so the taxonomy can support learning collaborative design can identify characteristics of the can help dasari this may help decision-makers make better investment so it's expensive to do these and it costs time and resources so we think by thinking explicitly and working through these characteristics can help with design but also inform research about what what works in developing a research agenda about what we need to know about learning collaborative to make them more effective and efficient and also what's really important is you know like any other complex interventions have a standardized way of reporting on them in the literature so that we could actually start making comparisons of learning across collaborative so thanks a lot and I'm going to stop there thanks Alan and thanks to health affairs for hosting today's event when I first joined the Innovation Center back in 2011 I received one question over and over again what is the Innovation Center and my stock response was Congress established the Innovation Center for the purposes of testing new delivery and payment reform models that would reduce program expenditures without negatively impacting the quality of care that beneficiaries received and then they would say no like really what do you do and what I realized was that the Innovation Center is really a commitment that the country is making around testing any new idea insight or model before scaling it more broadly across the country now implied in that commitment is the sober realization that no model is going to get it right out of the gates until you test a model with real patients in real health systems with real payers in real communities you just have huge blind spots that you can't see and you risk scaling unanticipated consequences and so the opportunity that the Innovation Center provided to learn alongside the market allowed us to minimize those risks because let's be honest healthcare is complicated now there's been a lot written and debated about the Innovation Center and its models but the story of the early lessons learned during the startup phase has never been told and so that was a story that we wanted to tell in this issue and we convened a number of early former CMMI leaders to hold that mirror up and to be honest with ourselves about what worked what didn't and to ask the question you know was was there stuff that we learned that others might find really important especially given the future direction of the Innovation Center and we we came together and we identified four key lessons and I'm going to share those right now they're the architecture for the paper the first and I'm not sure we truly appreciated this at the time but the role that CMS carves out for itself relative to models in the market is critical it's not one size fits all and it's not black and white I think about the healthcare innovation Awards grants in that particular model CMS's role was that of an innovation catalyst we weren't prescriptive we didn't tell anyone what to do the all the ideas were market generated and to give you an idea of how receptive the market was to that model we received 7,000 letters of intent in over a thousand applications so there was this intense interest with a little bit of infusion of capital into the market now compare that to the role that CMS played on the comprehensive primary care initiative was totally different CMS in that model was a convener of multiple payers both public and private they were trying to incentivize better primary care in the country and so the that taxonomy ends up becoming a really important consideration relative to how the government interacts with the market the second learning was that CMMI is not just designed to test models it was also designed to learn and in many ways I think we were learning how to learn often in real time and some of the models did really well some did poorly others washed out in between but the one thing we never had a shortage of was direct and critical feedback and we got that feedback in a number of different ways the first was that models participants would drop out that's pretty direct and we saw that happen early on with model one of bundled payments the second way is participants would defect from one model to another and so we saw that with Pioneer ACO is transitioning to the lower risk Medicare Shared savings program and the third is that the participants would stick it out with us but all along the way continued to provide really critical feedback all that information helps shape the future generation of models and was important learning for us the role that public sector leadership was able to play to have that broader perspective and to shape some of that learning and see some of the patterns that were emerging ended up being a critical facet of the work the third lesson was around the time needed to evaluate the return on investment in the impact of models it was just too short I mean think about what these models are trying to do they're trying to transform a segment of a market there are still competing interests there are perverse incentives that are out there and one of the things that we learned is that the basic functioning operationally of the participants well we just put it this way there was a steep learning curve basic budgeting accounting many were dealing with legacy IT systems so just data procurement and reporting ended up being a real challenge so in many ways given the degree of disruption and transformation we were looking to achieve a year-to-year ROI that was going to be breakthrough just seems unrealistic even a three to four year time frame seemed unrealistic now looking back the fourth lesson to me I think was the most important and that was the idea that innovation isn't always about more models or new models sometimes it has to be about better integration and coordination of existing models now I spent four years at CMS and I did a distance commute between Boston and Baltimore I used the same Airport shuttle service every week I get to know a lot of the drivers one in particular he was a 70 year old veteran who walked with a terrible limp he'd have to stop the van halfway to CMS get out walk around the van because sitting for that long was unbearable he later told me that he needed a hip replacement but he couldn't maneuver through the VA his paperwork got lost they couldn't validate his services and he didn't have the time to deal with all of that administrative stuff if he didn't work he didn't eat today there are 48 models supported through the Innovation Center but when we started there were zero somewhere between zero and 48 models we kind of lost this guy and I think we not alone the solution to a fragmented healthcare system can't be a fragmented innovation environment and so I think there's real leadership opportunity right now to think about how we thoughtfully reduce the administrative burden in the system at the same time putting the needs of beneficiaries and patients first the federal government is always going to be a key actor in health care and I think what we realized as we went through and did some reflection on through the paper is that the question is will it and its private sector partners commit to learning their way forward for the next phase of the work ahead thank you good morning my name is Sarah Ono and I'm here as a member of the family medicine department at Oregon Health and Science University thank you to health affairs and the team for the invitation to be here today and the assistance on this article which improved with with your careful consideration and insight I am here representing an amazing team of collaborators who are led by Deborah Cohen who is the PI of Escalades which is the national evaluation for evidence now evidence now pops up a couple places in this issue and if you're not familiar with it it is an arc funded initiative and it's a big efforts evidence now is the largest study that arc has funded and it has the potential to give us information on a wide range of topics the target is cardiovascular risk reduction it focuses on aspirin blood pressure cholesterol and smoking cessation the ABC s and it is also about the challenges that are facing small rural primary care practices all over the country and I said rural but there also are a whole bunch of them in New York City so it definitely has adaptability in a number of environments these primary care practices are struggling to keep up with the changing evidence new technology payment models policy and if there are any doctors in the room you know you're still trying to have a meaningful encounter with your patients at those appointments so my training is as an anthropologist and the questions around how practices change how practice culture changes is something that is particularly interesting to me the ability to get practices to a point where they can do ongoing quality improvement is no small task oh and there's the list of the amazing authors who worked on this with me this is a map of the United States and I know it may be small for some of you who are in the but hopefully you can see that there are states that are color-blocked those are the seven regions that are participating in evidence now the dots are the practices and this is at the zip code level so if there's even one practice in a zip code there's a dot and in several of these there are many practices and I like this map I've been looking at this map for a while because it gives you a sense of what kind of diffusion we're talking about and this is in the single initiative of evidence now this is an opportunity to in some cases cover an entire state if you look at at North Carolina at Virginia at Oklahoma they really have dots just about everywhere so what this is it's seven cooperatives the eighth grantee is the evaluation that I'm here representing seven cooperatives touching twelve states they've recruited and are working with over 1500 practices which translates to about 5000 clinicians and eight million patients so we are working on this diffusion and it is it is showing itself to be possible in each of these X in each of these areas external support needed to be provided to practices in order to support the technology that David Orr has talked about the ability to access extract data practice facilitators to help with quality improvement how to take data understand it get new evidence and figure out what to do with it to build its building skills and another piece of this that has become very interesting to me personally is this idea of the model of healthcare extension it's not an entirely new idea it's something that we have taken from agriculture and the Cooperative Extension service model that exists in every state has sustained for a hundred years and has had a number of impacts more than than I can fully address today or in this paper it is building on the work of big thinkers who will hopefully continue to write about this and pull out some of the historical pieces of how we got to this point because it's it's incredibly interesting and I was told my slides might be slow well if it comes up one of the things we did was we looked at the empirical data there it is the empirical data that was generated through these seven examples these seven cooperatives because in order for them to reach over 200 practices in their state or region they had to put in place something that was a healthcare extension they had to put together a network of partners and organizations and people on the ground in communities in order to access this many practices in less than a year it wasn't a small task and so what what we did when we looked at that and we were looking at early and baseline data is we identified that they needed support around health technology using electronic medical records their ability to to really know what was there know it was possible and get the information that they needed so that it could be applied for quality improvement along with that there was this idea of practice capacity building and that's where the culture piece comes into play it's how do you keep small practices with three people from burning out trying to learn new skills how do you get them interested in energized and seeing how this benefits their patients developing motivation and resiliency and the third piece is really around making community connections trying to figure out what the resources are and this is a piece that comes directly from the agricultural model of having Extension agents having people who are embedded in local communities who know the practitioners who understand what the local issues are and are able to adapt their approach with those factors in mind what the model does is it sets up this bi-directional communication so you get it from the top down the policies get rolled out the evidence comes out of research centers but you also are able to tap into what's happening as far as local innovation what small practices are figuring out that works and like Arlene said you can create learning collaborative opportunities for these individual practitioners to interact with each other whether it's in person or we now have all of these virtual opportunities where you can asynchronously connect online trying to really capitalize on what's possible and the thing about evidence now that's been so phenomenal to watch is that all of these pieces that are talked about in different articles start to come together in this large-scale experiment and the last slide is really just to again acknowledge that this is an idea that is coming out of agriculture I'm from Portland Oregon and we're big on recycling and reusing things and if we don't have to rebuild it I think that that is probably a benefit the other thing about this is there's a possibility to create economies of scale this is not an inexpensive effort as others have said we invest a lot of money the government invests money and time in these initiatives and the idea that we aren't capitalizing on the work that's already been done is something I really hope is kept at the front of our minds and that putting an infrastructure like this in place may be a way to help facilitate that so that we are able to continue to grow and not just sort of circle over and over again so again I appreciate your time and inviting me thanks for keeping time thank you again I'll start off with a couple of questions Rocco I'm interested in the diffusion element you talked a lot in the paper and today about the lessons you have this unique statute that permits replication adoption based on actuarial determination that may have been part of also the impetus for fast evaluation so I just wonder if you could add based on the panel this morning some of the thoughts within I know you can't speak for your co-authors but some of the thoughts within the center reflecting on not just the innovations themselves but what you learned about diffusion mmm-hmm that's thank you well that's a great question so this is one of the things that we talked about inside that we didn't talk about a lot outside there was a theorist called Peter Rossi who developed the concept of the iron law of evaluation and one of the things we began to recognize in the work that we did is that as I mentioned in my remarks some participants did they did well some didn't do so well and others washed out in the middle when you fold that distribution over the net impact is zero so typically the impact that you're gonna see with any large type of health or social related program that's spread across multiple geographies is just based on distributional statistics a tendency for a net impact of of nothing so the the challenge that we had and we recognized this immediately was the goal was to identify who was doing well and who wasn't and in the collaborative sense you you have access to all of that information so we can begin to really understand and unpack where the opportunities are to improve because it is an iterative process it's why we had a rapid cycle evaluation group was built on that that principle of learning our way forward to try to fight against the net impact of knowing of no influence at all thank you Sarah it's notable to have the eastern plains of Colorado and Manhattan in a similar in the same project I'm thinking again if you could sort of reflect back on some of the elements of diffusion and in some of the taxonomy that Arlene presented how that difference plays out in terms of some of the dimensions obviously the ability to convene is different but trust methods how do those differences play out when you're looking at something that's that that is touching places as diverse as those right in a myriad of ways the one thing I think that Ark was incredibly smart about was putting in place a national evaluation not just because I get to be a part of that team but because what that does is it enables us to have a record of what was done in seven very distinct and different places so that we're able to to not only look across at what what comes out of all of this but also to really focus in on specific areas and be able to pull from a single region or across regions to say like what happened with solo practices that have three people in them what happened with rural practices what happened with practices that it either at the beginning or along the course of the project were incorporated into a health system how are those needs different so it was just incredible foresight I think to be able to put a mixed method study of this scale in place so that as we go forward and do you keep following us there's a lot to come we'll be able to look it at those variations three thanks we have time for a few questions as well for this group these are major initiatives that are complex and if you don't ask questions I will but yes let's Randee siebel cleveland the these presentations have been terrific both of the sessions so far I have something that I haven't heard though is positive deviance sort of the Jerry Sternin approach which is really sort of what we're doing in Cleveland and when you get as much data as we get about you metrics that were defined by the providers to improve care outcomes and cost that we use the data to look at the top decile top whatever and talk with them and try to disseminate it but and I haven't seen a lot of sort of a commitment to that kind of a model but I think it's sort of interesting I'll jump in that it sort of speaks to the previous point the the challenge is the positive deviants are lumped in with the negative deviants right so if if the goal is to look at this from a actuarial or a cost-benefit perspective to the trust fund that's a tough thing but your point is exactly right I mean that's what we were trying to do with the Innovation Center is identify those positive deviants and and build that until the work that we were doing either if it was with pioneer ACO so the comprehensive primary care initiative my colleague will Frank who oversaw the evaluation team was that was one of the things we talked about a lot is being able to harvest that learning from the myriad of models that we were testing and I would also say across different models right so there's the positive deviance within like a pioneer ACO but there's also positive deviance within a primary care initiative and because the systems are so related there are often lessons that can cut across different models as we think about the work moving forward yeah now collaborative if you look at there were four measures that they needed to improve on and some didn't actually need help which we're doing really well so I think learning where do you need to target but on the other hand is nobody was doing well on all of them right so and some were doing better on some than others so it's not like it's one organization that's the best it's how do you share that for the positive positive deviance across multiple practices and settings so let me close with a question about goal-setting raka the first example you gave was you open up and you got 7,000 applications but my general sense with what all of you have described are still the the top level goals seem to mostly be set centrally you may have willingness to have significant variance in methods or fidelity to a model to the extent one exists to what degree is goal-setting by the participants in the initiative a central part of successful diffusion I think they're absolutely critical and just from the time that I spent at the Innovation Center one of the first things that we would do with a group of participants I remember this distinctly with the Pioneer ACO is was come up with a goal because without that goal you never know if you're heading in the right direction or not and that was probably one of the most challenging situations is to get a group of stakeholders across multiple systems agree on a goal for a program at national scale as I still remember the goal in my head it was by year three of the Pioneer ACO as we wanted 100% of them to qualify for population-based payments based on their quality and cost performance and everything was engineered toward that objective the primary levers that we wanted to influence the changes that we were going to try how we're going to evaluate our work moving forward and then the second point that I think is absolutely critical is the the high level goal also has to be able to the vision that we set for the system so many of you know about the comprehensive primary care initiative and the first year the results came out and they were you know we weren't sure what we're not they were truly cost neutral if you added in the elevated care management fee there was some question about whether it meant statistical significance and so I can remember having a discussion with people thinking okay well it likes a map let's imagine it wasn't statistically significant does that mean we're gonna abandon comprehensive primary care now so there has to be this idea about the vision a clear goal in engineering everything around that and I would just add that I'm currently at a stage where I'm getting to interview practice facilitators the people who are on the ground in direct contact with practices and the first thing they'll tell you is there are the goals of the initiatives and there are the goals of practices and part of my job is to figure out where those can intersect and and to be able to meet a practice where it's at in this larger context they're having to adjust to payment value-based payment they're having to hit meaningful use a lot of things are coming coming at them and so listening for what those local goals are it's critical and I think that's part of why the extension model is compelling to me it really is a conversation that goes from from the ground up yeah I think you know I'm gonna step back and actually give a little different perspective because I think there's different kinds of collaborative with different purposes so clearly to scale you know performance or better care delivery that needs a clearly directed goal on the other hand there's collaborators like research collaborators where you're trying to develop methods how do you study complex patients with multiple you know chronic conditions where it might be more open-ended or the Medicaid medical directors is a good example where they we didn't define the agenda they defined it things that they needed to learn and share from each other on the other hand I think evidence now is really a model not only of like improving cardiovascular risk management but how do we learn how to make it work and I think more and more we need to do that parallel of studying what works in terms of implementation as we implement sort of like the CMS rapid cycle evaluation because we're doing this what we're doing in terms of implementing these collaborative is also we need evidence for that and how to do that best and most effectively and I think that's where we need to move the science it's kind of learn what as we implement and then share all these learnings so after the break we're gonna go deep into a number of examples of diffusion and what we learn from them but yeah we'll take 10 minutes before that please join me in thanking our panel if our first session was theory and our second session was how that theory plays out at the macro level this session starts to move us more directly into specific examples of innovations some really interesting stories and I'm very pleased to be able to turn the conversation this morning in that direction we'll hear from Rebecca Oni founder and CEO emeritus of health leads which is a is a nationally recognized leading organization at the intersection of social determinants population health and healthcare delivery we'll be hearing from Brian castle who is a palliative care research director at the Virginia Commonwealth University School of Medicine James Schuster chief medical officer for Medicaid special needs and behavioral services at the University of Pittsburgh Medical College insurance division and Maureen maverick who wrote the narrative matters piece in this month's issue of work who is board certified in family medicine and geriatrics it's my pleasure to turn it over first to Rebecca Thank You Alan not long ago an adolescent patient walked into a major Health System not far from here losing weight losing weight and just as the physicians were huddled up figuring out which blood tests and metabolic panels to run one of our health leads advocates asked out loud do you think the patient might be hungry and in fact the patient had been kicked out of his housing just a few weeks prior and just literally hadn't eaten in weeks he said he was so relieved that someone finally asked me and the article that we have written in this issue of Health Affairs is to some degree about an operational model for addressing patient social needs but really is much more so about the question of how do we change the understanding of what actually counts as healthcare why does ordering blood test count but asking a patient if he's hungry doesn't and I think for all of us engaged in health care transformation this is a critical question how do we actually transcend the status quo to create a new normal in this case my co-authors and I Chronicle health leads Odyssey in pursuit of this aim a healthcare system that addresses paste in social needs as a standard part of quality care I have to say that you know as many of you know case studies have many limitations but one of the benefits of obsessively focusing on a single project over a long period of time is that there are certain learnings that emerge only through contextual shifts and there were many of these that's what happened here we did a lot of really hard learning over 20 years as health leads evolved through five phases of diffusion which we show through the number of patients that we reached which is the blue line and the number of healthcare institutions that we reached the orange line phase 1 was really about testing and learning and honestly this phase was just a total mess we were experimenting with all different ways to try to impact the way that poverty shaped health outcomes and what's most important about this phase was less what we did but more how long it took it was 12 years and part of that was about our fixating on how to actually create a model but part of it was that the sector honestly just wasn't ready for this notion of Health in 2008 there were only 28 patient-centered medical homes in this country phase two is about model standardization by 2009 the Triple Aim had been introduced in a cos launch and we felt compelled to begin to standardize our model the idea is that when patients came into the clinic they'd be screened for unmet social needs the physician or other provider could then refer them to food or heat or other resources the same way they would have subspecialty referral the patient and the Advocate would then navigate together to access those resources in the existing landscape and finally the Advocate would close the loop with the provider to enable better critical a better clinical care phase three brought model replication in 2012 the US Supreme Court substantially upheld the ACA and within a couple of years we had had 1400 expansion requests from health systems as millions of newly insured patients presented not just with untreated chronic conditions but also with significant unmet social needs and we set out to replicate our model quickly with about 16 million dollars and funding from the Robert Wood Johnson Foundation and an additional 19 million dollars from other funders including the Physicians Foundation School Foundation and others almost immediately we realized that we had entirely the wrong plan with 70 million Medicaid patients in this country health needs goal of reaching 175 thousand patients over four years seemed laughably small and we quickly had to recognize that we had gotten it all wrong so in phase four we actually ditched the replication plan and instead began to provide tools data and best practices to health systems and payers to enable them to be able to address the social needs of patients and their own local context the number of health leads patients during this period of time that we reach directly plummeted but the number of health care institutions we reached increased exponentially from a couple of dozen to literally 2,300 institutions in 2017 and in Phase five which was the last one we began to catalyze broad adoption social determinants of health as you know is hot stuff today CMS has launched its first two models including social needs in care delivery and payment models informed by health Lee it's learning and we see today that VC firms are beginning to do due diligence on social care models commercial payers are grappling with the reality that up to 40% of patients who screen positive run met social needs have commercial insurance we see state medical societies like the North Carolina Medical Society beginning to prepare its members to address social needs in the context of value-based care so very quickly what do we learn regarding the diffusion of innovation in healthcare first a constant focus on a clear aim is essential and to stick with it over a long period of time the key here is the focus on the aim not on replicating the model this inoculates innovators against falling too much in love with their own idea whatever it is which in turn jeopardizes broad adoption philanthropy also has to understand that the only thing that's sacred is the aim and the values that enable you to achieve it everything else can and will change second investment model testing so much more than you would think because this ultimately enables responsiveness to market shifts diffusion pathways has have been talked about today are nonlinear especially in complex systems like healthcare and because the market and policy context can decelerate or accelerate diffusion a deep investment in the pilot phase is key to be able to seize the opportunities to go big when they arise to put a fine points on this as as Rocco Perla mentioned in an earlier paper CMS usually thinks about testing models for three to five years we were in the pilot phase for 12 for 12 years and finally innovators and their investors have to be willing to cede control of the model as fast as possible after the testing and standardization phases I'll be honest we had health aides never ever felt like we perfected the model but to achieve our aim we had to listen to our clinical partners like contra Costa health system or Kaiser Permanente who compelled us to evolve to a more flexible and more diffuse abow approach and for health leads to Steve control of our model we needed philanthropy to cede control of health leads holding us accountable for our aim rather than a rigid set of grant objectives I just want to say that despite all this learning this story is still a cliffhanger with the end unknown well addressing social needs be a fad or is the health system at last coming to terms with what actually drives health as patients know to be true as these data from recent focus groups in Charlotte North Carolina show two groups one of African American Democratic women and one of white Republican women were asked if you had hundred dollars to spend in investing and health in your community where would you spend it which remarkable is that they know you spend 134 money on health care two-thirds on everything else and despite the political differences their investments are identical 19 percent in affordable housing 25 percent in access to healthy food 14 percent in access to affordable childcare the question is our country may be fractured when it comes to health care but is it or could it be United when it comes to health thank you good morning thank you for having me to health affairs my name is Brian castle don't be speaking today about palliative care representing the palliative care leadership centers and the center to advance palliative care and my co-authors so for those of you who don't know specialist palliative care is a relatively new domain that's been carved out in the trajectory of serious illness care it adds a crucial layer of support to patients and families it is essentially an interdisciplinary effort that requires teams of specialists to work together effectively to prevent and treat suffering and pain to clarify prognosis of the patient and to clarify goals of care to communicate that effectively with the patient with the family with the care teams to address the whole biopsychosocial spiritual continuum of issues for the patient and the family and to bridge the gap between quote unquote curative care and quote unquote end-of-life care in our hospitals and elsewhere in the u.s. height of care is a distinct thing from hospice care for example there is no revenue stream specific to palliative care as there is to hospice care under the Medicare hospice benefit the field could essentially be said to have launched in the mid 1980s when several palliative journals were created when the subspecialty of palliative medicine was recognized in UK 18 years before it was recognized in the US and the first comprehensive program was founded by Declan Walsh at the Cleveland Clinic over the ensuing 30 years what we see is a very remarkable adoption of voluntary adoption of this high revenue I'm sorry high-value low revenue it's a crucial thing to get wrong sometimes we we say what we wish for and not what really is so it's a remarkable trajectory of adoption of a high-touch very personable high value but low revenue form of health care in the US you can see it increasing there from about one-quarter to three-quarters of US hospitals with 50 or more beds essentially tripling over those 15 years that are depicted in this slide the lower line depicts the number of hospital teams that were trained by through one method the height of care leadership centers which is a team to team training and educational program from the center to advanced palliative care that will describe in more detail so that two-thirds of hospitals with palliative care today have at some point sent a team through this piece ELC the PI declared leadership Center training to describe all of the purposeful work that the center to advance palliative care and others have done to advance the field of palliative care in the US it's useful to depict this using the Prochaska and DiClemente stages of change model and to talk about moving whatever level of entity you want to talk about whether it's a hospital or a health system or individual providers from pre contemplation to contemplation to determination to preparation to activation and maintenance etc and each of those crucial steps in growth and moving forward need to be addressed through dissemination through technical assistance through mentoring and through the building up of the field and connecting new adopters with those who have adopted previously in the field the one I want to talk about in depth that I think is lacking in a lot of dissemination work is the action step and this is where the palliative care leadership centers really have their work to help leaders who have committed to starting palliative care programs and their hospitals and health systems to do it well to do it quickly and to not fail the type of care leadership Center initiative was first funded as a lot of the innovation and dissemination work in the kind of care field was by the Robert Wood Johnson Foundation and in using the center to advance palliative care at Mount Sinai as the central hub creating a set of other hubs or intermediaries around the country that were centers of excellence that were recognized already as early adopters and innovators themselves geographically dispersed each of which had a characteristic that others would come to in addition to being geographically close for example cancer center based palliative care leadership centers or those that were affiliated closely with hospice or those in faith-based nonprofit health systems and the goal of these is really to get over the operational hurdles to make sure that the team is making its case clearly and effectively to hospital leaders and health system leaders and to take all of their aspirations and put them into concrete achievable goals and then mentoring them over a year to achieve them to use data effectively to make their case and then to measure and evaluate their progress and what more remains to be done and for themselves to become leaders in their own health systems so the leadership from Mount Sinai the leadership from the hubs and then creating leadership within these health systems that then further promulgates the palliative care principles and practices throughout their health systems which is not an easy task to overcome complacency and resistance in this area throughout hospitals in the u.s. so online preparation for didactic knowledge two to three day in-person training which is very intense and exhausting which ends with this concrete set of goals that then form the framework for the mentoring over the following year so two-thirds of hospitals now in the US have gone through that that have had palliative care palliative care leadership Center training I would like to say that we wish we had written this article after having listened to all the theories and models and other case studies that are described in the journal and the briefing it is such a unique laboratory for high value low revenue care in the US there's so much more that could be said and done about the diffusion of palliative care innovation across the u.s. so we welcome those of you who are here to contact us and to work further on this field to understand what has been done and what has worked and what remains to be done thank you thank you welcome and thank you for the opportunity to participate in the in the journal in this presentation so I'm James Schuster I'm with the University of Pittsburgh Medical Center Division of Insurance Services UPMC is a large integrated payer and provider system based in western Pennsylvania and as part of this study which was conducted through the UPMC Center for high-value health care which is a research and grants arm in the insurance division and community care which is a behavioral health managed care organization that manages services for about a million Medicaid members in Pennsylvania so I'm going to talk for a few minutes about the results of an intervention that we did and then for a bit also about the diffusion and innovation and certainly trying to improve services that our patients and members receives a key part we've been just as challenged as everybody else by the 10 to 20 year time frame for implementing evidence-based care so we were really pleased by this forum and work so we worked on implementing behavioral health homes and they're really a health home model for individuals with serious mental illness and we did that because as we all know individuals with serious mental illness die significantly younger than average individuals in the country and the primary causes of their death are related to cardiovascular disease COPD diabetes and other complications largely from smoking and obesity and actually many of the providers whom we helped to fund came to us and said behavior health providers came to us and said we worked with a lot of the individuals with serious mental illness a lot of our patients died young and we're not really sure how to engage and help them they often aren't very engaged with their primary care physician and we're not really sure what we should do or how we can best serve them so we there wasn't really a key comprehensive evidence-based model for how these providers could address this issue but we did try to pull information from several interventions that did have some evidence based either in this population or in other populations so the first there was some evidence certainly that placement of a nurse focused on physical health issues in a behavioral health setting helped engage people in primary care and enhance the rate of preventive services so we put that into our model the second is there's lots of evidence obviously about the impact of self-management tools in terms of helping people manage chronic illnesses third was there is significant evidence particularly in oncology but also other medical specialties about the value of health navigators or wellness coaches so we included that in the model as well and then the fourth piece is there's obviously many new innovations use registries to help people structure their intervention and and monitor their implementation over time so we also help the providers built our registry and then we reached out to the patient-centered outcomes Research Institute McCrory for funding to evaluate the intervention and we're successful and got lots of great support and advice from them during this during this intervention and assessment so what we did was we created two models to really test against each other our hope was that both models would be effective both models included wellness coaching and the use of a registry one model also included a wellness nurse in the provider setting working on physical health issues and the other intervention including included using self-management tool kits and other self-management strategies with individuals we used a cluster randomized design at 11 sites and the research participants were all enrolled in Medicaid 21 years or older diagnosed with a serious mental illness and we're receiving services primarily case management or peer support services at those providers we were able to pull data from a number of different data sources including a number of self-report measures that people completed through the study a number of secondary administrative sources both physical behavioral and pharmacy data and we also used a very structured learning collaborative model for implementation we heard some discussions earlier about the different types of learning collaborative models we use the learning collaborative model developed by the Institute for Healthcare Improvement and we use it with a number of other projects and our experience with this reflect our other experience which is that it's very important for the learning collaborative model to be very structured both in terms of what the goals are and what the expectations are in terms of the participants what we found in terms of outcomes was that there was a very significant increase in both arms in terms of patient activation we used the patient activation measure to assess that and increases in the patient activation measure are typically associated with improvements and utilization and improvement in costs we were pleased to see that there were certainly greater increase in activation for women and the provider supported part so the women enjoyed working with the nurses the men preferred to work on their own with self directed tools both arms had led to significant increase in frequency of visits with both primary care and other community-based physicians and there was an improvement in perceived mental health status and a small decline and perceived physical health status and we don't entirely know the reason for that but we hypothesize that it's at least partly because providers were now talking with individuals about their physical health status and physical health challenges we've now worked extensively to diffuse the in the innovation we have created wellness chain engaged wellness champions at different sites and used to train the trainer approach especially to disseminate the wellness coaching approach we actually have expanded now to 43 additional providers across the state and have used the similar learning collaborative model to support implementation and we've also used to pay for performance contracts to support implementation as well especially to support the nursing piece which is an incremental expense for the providers and we were recently fortunate in securing an additional dissemination and implementation award from picori which we were going to use to revise and implement the model in methadone treatment programs for individuals with opioid use disorders and in residential treatment programs for adolescents with behavioral challenges so we think the study findings can certainly inform national efforts to avoid or reduce mortality and individuals with serious mental illness we certainly can't intend to continue to monitor the impact of this work and we think also there's been a useful model for diffusing new models into behavioral health care settings and we think can apply to other settings as well for individuals with disabilities thank you Wow I am so honored to be here today thank you to the staff and editors of Health Affairs my name is maureen maverick and my article in this month's narrative matters of health affairs is rethinking the traditional doctor's visit it discusses a model of care that I was fortunate enough to pilot and participate in in a rural setting in a remote location in the Central Valley of California having previously led such a model with inner-city patient populations my narrative specifically discusses shared medical appointments block that we call it hunter's podemos are together we can this group which I ran with a health coach consisted of six to seven weekly sessions each session covered a specific topic such as emotions medication complications nutrition and exercise the rudiments of the appointment where patients would come in consent would be signed they would take their own blood pressures and weight and with a medical assistant standing by and then blue blood glucoses would be checked and while the health coach continued interactive discussion I would do brief exams heart lungs and feet while the patients were seated in the community room where the session took place as the session wrapped up I could also do medication refills but what made this appointment different was we ask patients to buddy with one another to provide support between appointments so in essence this was a model akin to a 12-step meeting where members shared experiences and support years ago I was intrigued by this model because I saw the confusion and the resignation on the faces of patients with chronic disease who left the traditional 15 to 20 minute visit I also saw the frustration on the faces of their clinicians even collisions early in their training the common thread was that both patient and clinician felt hopeless yet we didn't start out that way I saw a desire to partner and connect with patients to help improve health care in the residency application essays I read as a faculty attending in family medicine at UCLA it's this idealism and passion to innovate that we must preserve and nurture you see I've always believed that being a physician is a colleague it's a vocation and yet instead of feeling like innovators many of us feel caught up in a system that that drives burnout the non-traditional doctor's visit it's an opportunity to heal the hopelessness and the isolation that many patients with chronic disease of diabetes feel especially vulnerable populations impacted by social isolation economic challenges and poor health literacy my narrative tells the story of one such patient I called Mariana she's a woman in her late 50s who worked as a janitor and her buddy in the group I called Susie Mariana lived alone in a rural Hamlet she had no phone she couldn't read we had tried unsuccessfully to help Mariana reduce her blood glucose and despite being assigned a culturally sensitive health coach we were unable to budget that reading Marianas glucose was on average well above 300 and she already had retinal complications from her disease the fact is she was demoralized and truthfully the whole clinical team was – Mariana frequently missed appointments and she was an accessible via phone all right we did do some medication interventions with Mariana we added an insulin pen however if we discovered during her participation in the group that she wasn't using it correctly and so we helped with that but what the secret was to better health outcomes for Mariana was indeed the social connections she made in the group especially that with her buddy Susie after the shared medical block participation her blood glucose on average was reduced significantly from 300 to the low 200s not exactly a goal but it was it was moving in the right direction and through that connection to Susie Mariana who had been demoralized prior to honchos podemos was energized and smiling in fact they even wanted to do a testimonial and Susie in that testimonial also mentioned that she felt better physically and that helping Mariana had motivated her but you know that wasn't the whole story on the clinician side of the equation my team and me felt a sense of agency as we innovated and implemented juntos podemos when I saw the smiles on the faces of the nurse practitioner the health coach and the medical assistant as Mariana and Susy shared this success I know that on toast podemos was helping we providers of care to fill connection to our patients and each other when I picked up my New England Journal of Medicine in March of 2017 I was struck by two articles in the perspective section they discussed suicide and depression and medical students and clinicians ironically these searing pieces were immediately followed immediately followed by another article the case for shared medical appointments by commonly around us and our darcy who discussed four crucial components that need to be studied to make this innovative model of healthcare delivery a standard of care what's needed they said our rigorous scientific evidence supporting the value of shared medical appointments easy ways to pilot and to refine these shared appointment models before applying them in particular care settings regulatory changes or incentives to support the use of such models and patient and clinician education as I summarized in the last lines of my article by fostering interpersonal connections and relationships hunter's podemos resulted in more effective disease management and better health outcomes for Mariana it also restored to us her clinical team a sense of agency and autonomy so essential for preventing or reversing burnout in the past my patients hopelessness mirrored my own now their hope reflects my thank you I want to leave maximum time for you all given the breadth of topics covered in this but I Rebecca I have to ask one question I actually have a lot of questions but but your your your fidelity to aim instead of method I think was really fascinating realization what I'm my question is this you spent 12 years testing model then you shift model but you don't have 12 years to test the new model as in training institutions to do this as opposed to doing it yourself so how do you make the pivot not just to a different model but also dramatically shortening or maybe even forgoing the what what you had just said was a really important testing phase yeah I guess I would sort of shift the question because I think if we had been going from model a to model B we might have encountered the challenge that you mentioned but I think the shift was actually from a model to an approach and that was the distinction so the the notion was that you know in we thought we were standardizing and a and up sort of obsessively trying to improve the model to replicate it but it turned out what we're really doing was those activities in order to understand the constituent elements of the model so that we could then liberated and so it still required that like struggle the discipline the iteration the testing the data but you know the outcome at the end wasn't something then we would then you know sort of ship off the conveyor belt to replicate but actually to then put those those tools into the hands of folks who candidly like knew the context so much better than we did so you know our original model I think as folks probably know was promised on using college student advocates as that front line workforce in part because in the mid-90s when we started there were you know promontories and community health workers and care managers and case managers but you know by the time that the healthcare sector is really starting to heat up around driving towards health those work forces were in place and if we had continued to insist on a college student workforce we would have lost all the value and instead we were able to liberate that value by putting into the hands of you know other folks who just could much more deeply understand the work forces the patient populations the local resource landscapes and the you know the the parameters of clinical context that meant we could you know we could walk away from the model you know very quickly it's very helpful thank you I could ask more but it's good mic in the back very back and then we'll go to the gentlemen we've been in front of him thanks Ellen Mike Miller I'm a physician and health policy consultant they don't pick up on what Rebecca Justice was talking about because it worked with a lot of very large organizations trying to innovate and then diffuse that innovation except in their parlance it's usually spread best practices it's a very different kind of concept and it's the difference between between maybe health services research and management consulting but the last thing you said no one the other people could expand upon it about sort of abandoning the model and going to an approach because I've worked with people say well we're gonna do it here and then replicate it in all these other places and I have to try and convince them that you can't really replicate things because there's different cultures of resources it's like taking the model for a bayou house Louisiana trying to build it in the mountains of Colorado it just doesn't work so if you could talk about sort of that spreading of best practices outside of just the model concept thank you Brian I mean the paper talks about the role the leadership centers so maybe how does it how did that process reflect or or handle the notion of variability across the hospitals I I think the key point is that each hospital is different even within the VA we trained all hundred and fifty or so VA sites if you've seen one VA you've seen one VA center the the key was to take the essential practices and principles of palliative care and through the leadership centers teach the trainee team how to adapt and adopt the so not focusing so much on thou shalt but on what do you want to achieve how can palliative care principles and practices help you to do that how does that fit into your local health system and how can you use local data to supplement published research to make that case and really customize it I'll just say this is a great question so I think what did things that health leads did wrong in the beginning was that we would partner with health systems and you know the idea would be okay we're gonna pilot the model and then if it works it'll then get replicated within the system but you know you get just lost in the desert of like did it work you know did it work relative to what over what period of time with which patient population and in the sort of second half of the work we started to realize we have to actually start with a commitment to the aim is this Health System committed to recognizing up to 70% of what actually drives health outcomes social and environmental factors we can pilot to learn how to scale within the institution but that's different from piloting to then evaluate are we actually committed to the a.m. and so you know is that shift once institutions could commit to the aim all of the all of the the piloting and the diffusion was to drive learning towards the achievement of that a.m. rather than to decide whether we were committed to it in the first place gentlemen here thank you my questions for dr. Oni Rebecca you obviously been committed to addressing the social determinants of needs for patients for a long long time and now it's just become pretty popular in the healthcare policy arena what needs to be done to fully implement at the primary care level the social determinants of health care to improve access quality and outcome I think because if we're talking about containing cost of the federal and state level we've got to get the patients into the healthcare system early before they get into the more acute care system so how do you put the this issue on on steroids they really kick it into motion well first I like to say that Bob likes to haze me by calling me dr. Oney which I am NOT you know I think you know so speaking to the work that that we've done in North Carolina is a great example under the leadership of secretary Mandy Cohen there was the Secretary of Health and Human Services this is a great example of leadership committing at the outset as she did about a year ago and she began as Secretary to the integration of health care and social services and you know I think what's so powerful about that is you know her her the question she called was how do we look at every policy choice we make every operational choice we make in the state of North Carolina through a recognition of the broader set of drivers that impact health and you know I think what then happens is this doesn't get reduced to just you know mechanics in the clinic but every element whether you know it's the risk models the screening protocols the data that's collected statewide the workforces that are funded all of those things actually begin to be leveraged in order to drive towards this broader notion of health and you know the result of that is that yes clinical practices are asked to implement this were kind of at the frontlines of patient care but they do so in a in a policy context and in operating context that actually enables them to do so and I think we get lost if we think that we're going to you know transform the way care is delivered within the four walls of the practice without recognizing the the if the aim if the aim is not established around a broader notion of health it's it's extraordinarily unlikely that practice by practice we're going to be able to fundamentally change the way that care is delivered thanks as a payor I just want to add one additional comment which is I think value-based payment models can certainly help to drive the changes that you talked about we don't the reason you know you met you mentioned that people now talk about social determinants of care much more than they did five or ten years ago and it's not because we know lots more about social determinants of care it's because now people are starting to get paid on outcomes in terms of how their patients do over a period of time and the social determinants are important so I think continuing to build on the value-based payment models that we have and even thinking about additional tweaks to those so that for example there's no reason I you know we've done some work where we pay certain payers certain providers if the patients that they're treating employment rate improves or if their housing status improves so I think there are many ways to really build incentives for providers to address those issues to continue to make this further well we'll do the exception at the break so let's do one last question right here hi I'm Joanne Lynn from all terms program on improving elder care the new NPRM for Medicare Advantage procedures the chronic Act by allowing we're proposing to allow Medicare Advantage plans to begin to pay for health related non-medical services for all persons similarly situated which is not defined how do you think that that may change the game to move the health the health-related non-medical services into the eighty-five percent for Medicare Advantage plans are people going to be eager to take that up or and really help diffusion of some of these ideas or they're gonna sit back and sit on their hands and hope somebody else starts paying for meals on wheels that may be an imponderable at this Jasmine I tell you what I tell you what let me instead of talking about NPRM let's let me just broaden the question out to the payment role which James you mentioned in terms of value-based payment but payment you know you're a slip of the tongue about lucrative and and both of you I know your paper it not didn't come quite as much in the presentation but the role of philanthropy and supporting the models why don't we take this last question use an opportunity to sort of talk about supportive payment the problems of payment just get that in the mix because that is often the tool we we turn to and it has its pros and cons so who wants to take on that topic I can say briefly that the growing advance of alternative payment models including Medicare Advantage plans and AC OS and other things is leading to a second transformation and the palliative care field from Hospital based palliative care at a community based palliative care where the teams are working with those entities that are at some risk for not only outcomes but for cost and I I think that that is perfect for this stage of evolution in the field we're now we do have a pretty good workforce in the United States that is especially trained in palliative care and now ready to move upstream and prevent some of the problems we're seeing in hospitals where they should the patient's should not necessarily have had to go to get relief and good care I think that's certainly a good point the other piece that I think we probably should think about with regard to payment is that you know we're talking about social determinants of care and there are other funding streams in addition to health maybe not enough undoubtedly not enough but there are other funding streams through Human Services that address things like housing food and so on so I think many opportunities for us to look for partnerships and opportunities to braid funding to meet needs of individuals who have significant health problems and also lots of psychosocial stresses marine can you reflect on the role of payment in the transformation that you described with a an N of – maybe I did did roll out these shared medical appointments in inner-city Los Angeles and in rural California and there was very little cost except the Costco veggie tray that I would pick up and the the the the one hour a month in talking speaking to Rebecca's idea that you have to let go of the of the model I was able to get an hour a month in the Los Angeles County clinic I worked yet to do a shared medical appointment that was fine I didn't have any nursing support but I had medical students so we worked that out and then in the in the rural setting I was able to do it on a weekly basis so I think it's a question of having a physician champion or a medical champion and being flexible and and sticking to the sticking to a clear aim what are we looking for here and I listened to a quote talk by Vivek Murthy ethic who gave a wonderful podcast to the Commonwealth Fund in which he talked about external factors and internal factors of health the external being all these social determinants of housing food and what I was speaking to is the the internal social connections that help foster health and that you may still be living in a violent environment but if you have the friend and a resilience it helps so in terms of payment there is a the intangible payment of a champions desire and I think if we can cultivate that in our medical students and residents i think that'll go a long way and and with the inspiration of my colleagues here on the panel and the work they're doing to lead the way to further these models thank you I'll just quickly touch on this point around philanthropy because I think it gets lost in the shuffle you know I can assure you that there was no alternative payment model or health plan that was willing to be with us for 12 years as we really figure it out how do you integrate social needs into care delivery and I think we kind of radically underestimate what's required for innovation and the diffusion of innovation if we don't talk about the role of philanthropy and you know we were really lucky to be able to bring together a set of funders who were in fact willing to commit to a name and tolerate all the iterative loops that were required the shifting metrics in recognition that at the back end we could actually begin to bridge towards these these new payment models but those new peanut models it exists in the mid-90s when we began and so part of this is you know how do we view you know philanthropy is essentially the equivalent of venture capital that allows us to experiment and hone and learn so that we are ready when the payment models begin to line up terrific discussion terrific examples to learn from please join me in thanking our panelists we're going to bring up our next panel who will talk some about adoption and okay we're gonna move directly to three pay papers that look at the effects of various innovations is Shani Ganguly is a practicing primary care physician at the Brigham and Women's Hospital and instructor in medicine at the Harvard Medical School Jay or Lander associate chief for clinical and educational affairs medical services and acting chief section of general internal medicine at the VA Boston and vice chair for VA affairs at the Boston University School of Medicine Joseph Tannenbaum MD PhD candidate in the Department of population and quantitative Health Sciences at Case Western Reserve University School of Medicine I'll turn it over first to ishani thanks so much for having me I also want to thank my collaborators at Eve Mehrotra Geoff Souza and Michael McWilliams on this work so we're interested in thinking about the annual checkup this is the most common reason that patients go to see their primary care doctor despite ongoing debate about the value of these visits so in some ways this is a you know as old as time on the other hand Medicare just came out with a new version of this so in 2011 through the Affordable Care Act Medicare introduced its first version of an annual checkup that was free for beneficiaries and this was called the annual wellness visit and it was designed to address health risks in aging adults in particular through a number of different elements so you can see here are things like screening for home safety depression or doing a review of cancer screening talking about end-of-life care so what we found is that in previous work that these visit rates have have been pretty modest but rising steadily so you see in 2011 we've had about 8% of eligible beneficiaries across the country we're using annual wellness visits and that number rose to just about 19% in 2015 now the question is why is it been I've taken slow and we have heard over this morning about a number of reasons why that could be one possibility is that you know this visit has a lot of complex requirements as we've seen here and certain practices maybe better equipped to to address those requirements than others for example practices that can invest in workflows in physician extenders to to make these visits happen the same time practices may be motivated to offer these visits for a number of reasons so these visits are reimbursed at a higher rate than a typical problem-based visit so there's a revenue angle and then if you think specifically about some of the alternative payment models we discussed specifically accountable care organizations there and which are incentivized to keep costs within a certain budget for a population of patients and these visits may represent a way to sort of keep patients in the practice specifically improving stability of assignment for these patients so that they the practices can get credit for the the care that they're offering and the and the work they're doing to improve the health of these patients and you could also imagine that these visits may attract younger healthier patients and we have found in previous work that the patients who are getting B's for better or worse tend to be a bit younger and healthier and that's true in the wellness literature in general and so the question was sort of what's at play here so we decided to and try to understand what enabled and motivated practices to adopt these visits the first thing we found was that there was incredible variation in in practices adopting them so about half of practices provided zero visits in 2015 to their eligible beneficiaries another quarter of practices provided them to at least 25% of their eligible patients and recalled the the former group the non adopters and latter group adopters then we wondered and what are predictors of adopting these visits or visit use and what we found first was that practices caring for historically underserved populations rural populations racial and ethnic minorities dually enrolled patients with Medicaid and medically complex patients were less likely to offer the visit and what was very interesting about this was was there was twofold so firstly practices that cared for a disproportionate portion of these patients were less likely to offer the visit to anybody who walked in their door no matter that person status at the same time patients who fit the bill for this bill so patients on Medicaid for example we're less likely to get this visit no matter what practice they walked into so both of these things were at play next we looked at factors that we found a number of factors that predict a greater use of aw V's so practices that had a stronger primary care focus in other words they had a larger proportion of primary care physicians were more likely to offer these practices where there was a greater proportion of Medicare beneficiaries per doctor so in other words these individual dots were seeing a lot more patients who would be eligible had much higher rates and then practices that were part of a COS and practices that were involved in meaningful use and so therefore had the capability of using electronic health records we're much more likely to use these interestingly we found no association with the size of a practice so the size didn't really matter next we looked at some potential motivators for adopting these visits and first of all note that when we looked at visit trends you can see here that these zero later here you can see here that the the lighter bars on the Left represent the practices that adopted a wvs and you can see these the blue showing that those visits themselves the darker purple on the right represent the practices that did not adopt aw v's and you can see that essentially these visits did not add to the existing visits and it essentially it seemed to replace other problem based visits that practices were having and that could be because doctors were addressing these kinds of issues but just not calling them annual wellness visits or a number of reasons yet despite that practices that were that adopted annual wellness visits seem to have greater revenue and this is not to say that the these visits themselves caused this in fact we see a rise among adopters even before the visits were adopted but it does show suggest that protect that as part of the reason and suggests some important differences to think about in terms of paid the practices that are adopting primary care innovation versus not and these changes in revenue finally we looked at some of the potential motivators and here we found that practices that we use a difference-in-differences approach to essentially isolate the the of the annual on this visit adoption and we found that practices that had adopted the annual on this visit had on average slightly healthier population although there was no difference in in the age of the population and in addition these practices had greater stability of patient assignment in other words patients were more likely to stick with the practice for that period of time so in some we found wide variation in use of these visits across the country they were lower annual wellness visit rates in practices that were serving the the historically underserved but higher rates and practices that were involved in innovative models like a ce o–'s and who are using EHRs we found that there were some potential benefits of using these visits that so revenue potentially but also of seeing slightly healthier patient population and greater stickiness of these patients essentially in the practice and this brings up the question of how to the extent that these visits are useful and that's still a question that we were looking into you know how can we distribute those benefits more equitably to practices and therefore to patients and so one one way to think about that is is adapting these visits to the practices and the individuals that would most benefit from them so that's my time thank you I want to thank my wonderful team of collaborators and thank Health Affairs for this great opportunity to share our diffusions story that we observed and tried to understand primary care providers like myself often refer patients to specialists for help and patient care electronic consultation or a consult are a mechanism for provider to provide our communication typically within a shared electronic health record or web-based portal when used for primary care providers to communicate with specialists they've been shown to improve timeliness of care a decreased patient travel enhance efficiency of specialty clinic use and thereby improve access at reduced cost the VA is an ideal place to introduce this approach our unified electronic health record and salaried staff avoid issues of reimbursement for clinical effort in 2011 national VA leadership promoted the use of technology to improve specialty access to veterans due to sluggish uptake but a belief in their utility regional leadership in New England directed all eight of its VA Medical Center's to offer eeeek consults across all specialties that was back in 2013 once the II consult option is in our electronic record there are no restrictions on which clinicians can use it so we studied use the cross specialties in order to identify novel and creative practices we saw a large increase in use over the this time period but we're surprised to see that across all of New England now anesthesiologists were now completing more Econ sults than any other clinical service that surprised us because anesthesiology is not a specialty that we primary care providers usually consult as you can see in this slide prior to 2013 there was minimal use of e consults across anesthesia departments but by 2050 the four sites sites shown on this slide who performed the majority of surgeries in New England we're now completing 97% of all these anesthesia e consults but use clearly varied by site we read a sampling of the e consult to find that these were mostly surgical staff requesting pre-op assessments a specialty two specialty e consult this he also seemed to reflect the change in practice as commonly anesthesiologists saw patients in the days two weeks prior to a planned procedure we looked at factors such as size of the staff or number and complexity of cases to explain our finding but we couldn't identify a clear association we then interviewed staff from anesthesia departments and surgery surgical staff from each site from where and ask them about their preoperative processes and where Econ salts came in what we learned is the belief of the leaders of the anesthesia sections drove adoption Providence had a champion their head of anesthesia believed that nearly all the information he needed to provide a risk assessment and plan anesthesia care was present in the electronic record if they found a clinical concern using eConsult which would typically now occurred within a day or two of the surgeons planning of procedure there was time to sort things out Providence moved to nearly a hundred percent of their pre-op assessments by E consult including patients undergoing more complex procedures such as joint replacement and intra thoracic surgery and they eliminated their standing pre-op clinic they will see patients in person if the patients are the surgeons requested and sometimes call patients by phone to sort out a symptom or a concern they now report that e consults dramatically improved his staff workflow improve patient care and satisfaction an expedited ou are scheduling togas main providers heard about the approach from the togas chief our main colleagues as you can see here sort of a very rural population and they saw this approach as veteran centric innovation reducing the travel burden for patients that had ago many hours often just to attend the pre-op clinic they chose to adopt the consults for low-risk procedures such as cataract and hernia repair togas does a lot of e consults but that reflect reflecting the lower complexity of their surgical workload they now staff a pre-op clinic with an anesthesiologist just two days a week down from five and report improved scheduling and increased wait times Boston Massachusetts for I work mostly mimics the main Doc's in their enthusiasm for a consult as the principle tertiary care referral site for five New England states all except Connecticut Boston has the highest surgical volume and complexity of cases so with concurrence of their surgical colleagues they make an exception in complete a consult to accommodate some patients undergoing more complex surgeries who have long travel distances and whom the surgeons have deemed clinically appropriate they still staff a pre-op clinic five days a week but report having more time for the patients in the clinic reasons cited for limiting II consult use include the impersonal nature of the consult and perceived benefits of pre-op education on post-operative care West Haven Connecticut was our least enthusiastic group they believe that overall workload was not different but just shifted within the department and when he consults were requested and they expressed concern about potential inadequate preoperative education of patients due the perceived pressure to use he consults they set a target of twenty percent of all their elective surgeries would be done this way but they established the most restrictive criteria of all sites they still staff their clinic five days a week but do admit to fewer patient complaints and scheduling is a bit easier so our econ sults in the VA clearly have a lot of the characteristics we heard about on diffusion of technology but we attribute the diffusion of eConsult use in the pre-op process within VA New England to administrative leadership easily adaptable technology the opinion of a local champion and the perception of II consults on the impact of workflow patient convenience and quality care my name is Joseph Tannenbaum I'm an MD PhD candidate at Case Western in Cleveland Ohio and I want to echo my colleagues on the panel and thank Health Affairs for putting on this event and the opportunity to learn from so many folks from from across the country that are doing really innovative work my talk today is about the association of a regional health improvement collaborative or rick with rates of a mule Ettore care sensitive hospitalizations so a bit of background on the importance of the concept of ambulatory care sensitive hospitalizations particularly in a value-based payment environment conditions for which access to better primary care and more timely primary care may actually prevent hospitalizations are termed by arc in fact as ambulatory care sensitive conditions or acses and it's estimated that about 30 billion dollars is spent annually on a CSC related hospitalizations or acs h's and as a result acs HS are a widely accepted and utilized metric of primary care quality and in fact measuring a CSH rates drove some some of the innovative models that we've heard about today in particular accountable care organizations or a COS comprehensive primary care and now comprehensive primary care plus etc and particularly in the context of opportunities for shared savings and while a COS and CPC and now CB c plus emerged over the last decade as potentially transformative models of both healthcare delivery and health care payment simultaneously the the notion and the eye of regional health improvement collaboratives sort of co-evolved over a similar timeframe and so today I'm gonna be talking about one such regional health improvement collaborative that's called the better health partnership which is a which is a Rick that operates in Cuyahoga County and for those of you not familiar with your Ohio geography Cuyahoga County is Cleveland and the surrounding suburbs and so while a lot has been a lot of work has been done looking at a CO CPC etc and the potential impacts on a CSH rates very little has been done about the effect that Rick's might have in on similar outcomes and so a better health is a is a primary care driven model of regional health improvement collaborative and about 70 percent of all primary care that's delivered in Cuyahoga County is affiliated with better health at some level and I should mention that better health includes providers across the health care spectrum from large academic medical centers including Metro health and the Cleveland Clinic to safety net providers federally qualified health centers free clinics and and the Cleveland VA so really a wide spectrum as well as stakeholders and partners from inch from insurance to patient advocacy groups and so a better health one was founded decided to focus on three major conditions that affect population health diabetes heart failure and hypertension all of which contribute to arc to find ambulatory care sensitive hospitalizations and so the secret sauce of better health what better health tries to do and these are terms you've heard a lot about today in earlier sessions is to identify best practices using the positive deviance approach from the different members of the collaborative figure out who is delivering the highest quality care and then having yearly learning collaboratives at twice yearly learning collaborative excuse me to disseminate the ideas and best practices that have been sort of unearthed at the various clinical sites and perhaps even more importantly also using practice coaches who are sort of on the ground helping to redesign workflows and figure out how to deliver more effective care and and achieve better population health outcomes and so our question was was the implementation of better health associated with a significant change in the rate of ambulatory care sensitive hospitalizations and were there any potential cost save that might have been achieved by the broader health care system and so we used a difference in differences design where we the first difference asks was there any difference in the preventable hospitalizations rate for kya ho between Cuyahoga County and our comparator counties and those were the next five largest urban counties in Ohio which correspond to the cities of Toledo Dayton Akron Cincinnati and Columbus and then the second difference asked if there was a difference between the hospitalization rates for these conditions how did that evolve over time and so here what you can see this is the pretty better health era beginning in 2003 and you can see that hospitalization rates for these conditions was higher in Cuyahoga County relative to our comparator counties but from and there are a number of reasons for that I'd be happy to get into later that we we suspect is the driver of that but from an analytic perspective which you can see is that the rates evolved over time so the good news is hospitalization rates were generally going down before the implementation of better health both in Cuyahoga and in our comparator counties however those trends evolved similarly after better health was implemented what we see here is that the gap narrows considerably between Cuyahoga County and our comparator counties and up at the top you can see sort of an estimated trend line based on if you use just the data from the pre better health era what we would have projected hospitalizations to be and in our comparator counties they are much closer to expectation than we were in Cuyahoga County and a small point we actually estimated two different post periods because of some some differences in health care systems that that evolved in in about 2011 and onward so to put some numbers to that picture that I just showed you we actually estimated that about 5,700 hospitalizations for these conditions were averted after the implementation of better health in Cuyahoga County and if we look at the cost per hospitalization for each of these different types of conditions we estimate that about 40 million dollars was saved to the broader health care system by averting these hospitalizations and as sort of a small but very significant point here the annual budget of better health is about one and a half million dollars relative to up to the forty million dollars in potential savings to the broader healthcare system and so our conclusions are that better health was associated with a significant reduction in the rate of preventable hospitalizations and some cost savings were possibly achieved and taking sort of a step back to sort of tied together with some of the themes that we heard earlier today primary care focused and in specific provider led regional improvement collaboratives may be able to avert preventable hospitalizations for ambulatory care sensitive conditions and as a result we we believe that our results are actually a real cause for optimism that increased adoption of these collaborative based models that aim to bring together competitors to improve the the health of populations can really make a big difference particularly in the area of preventable hospitalizations and associated costs thank you very much can I ask a quick question of each of you and maybe a moment for the audience um Shani I can't help as I listen to your presentation to put it in the context of other sort of pay for performance initiatives particularly the literature growing that for example those service Patel's safety net hospitals being dis personally penalized by some of the value-based payment and and readmissions penalties and this sense of sort of the rich get richer because they have the infrastructure and they know that there's a new code here and they adopt and those who don't don't and fall behind hey I don't want to overstate that story so I just wonder if if that's my story is that your story do you have a sense of whether we're sort of trapped in this place where innovation is going to diffuse in a way that the leaders actually get further ahead and the laggers fall for the vine question I think you know in some sense you can think about these innovations in primary care payment as a broader effort to try to find all more money into the investment in primary care right and compared to other countries we under invest in that foundation of health care and so to the extent that aw V's inoue wellness visit payments are an investment in primary care writ large this is suggesting that that investment is going to practices that are already able to to reap the benefits of these payments and so I think it does speak to that narrative absolutely and then the second question that comes up is thinking about these visits in particular are there ways you know I as a primary care doctor I've performed these visits on a number of patients and I've seen firsthand that this that especially for patients with really complex social lives or complex issues at home financial issues etc this can feel like a distraction and so are there yet we still really need to think about preventive care in these populations and so the question arises are there better ways of formulating this visit making it less prescriptive or you know using alternative mechanisms to ensure that prime that preventive care is is met for for these populations so that we can still meet the goals of annual in this visit and alternative payment mechanisms to make sure that these practices are getting the support they need je given the earlier presentations the literature it's not surprising that diffusion would be in part dependent on a champion but it is interesting and it's what I was struck by is that the different champions had different reasons for championing so it's not a champion not champion is not a 1:0 thing I guess my question is what do we know and the lack of champions so what do we know about the creation of those champions was this just sort of random that you drop this innovation into some different settings and it just happens to be that person a views it this way person B views it that way or in the rollout is there anything that would guide people toward a certain narrative or away from another narrative well in the absence of heart obviously the the Providence site had a very low small staff only a little more than two full time equivalent anesthesiologists and a core group of nurse anesthetists so the challenges that that group had in covering the pre-op clinic and covering all scheduled surgeries was greater than some of the very large staffs in Boston or Connecticut so there was an inducement perhaps to want to believe that this improved patient care and scheduling and had no downside costs and yet there's also a group of academics I know within the facility which I work who want a whole hardly believed in one published paper that's a decade old prior to the true the current establishment of pre-op assessment where they believe that all their pre-op education has great post-op benefits and decreases length of stay but I would say that pre-op education could be given by the surgeons from the nurse practitioners from the inpatient nurses and it's not clear that the anesthesiologist needs to be the one delivering it so I think it's an example of where in the absence of data folks will try to believe in the improvements that they perceive and I think there's no right answer here as of yet and so they're opportunities to study it further just if this may be imperiled I am curious it's nice to see a positive outcome positive health outcome it's nice to see a positive return on investment or I wonder the degree to which that's actually the primary motivator for or a partial motivator for participants in the initiative what if you came out of a paper that said no no ROI here would that deter them or is it that they feel like they're doing the right thing I'm just anything in that sense but you know it's that that's a great question and having attended a number of these learning collaboratives and and visiting several of these clinics myself in the course of this work and others I can tell you that the enthusiasm is widespread the enthusiasm around the idea that coming together strengthens everybody involved is something that I think is a constant theme throughout the Cleveland healthcare area and I agree with you that it's nice that we got a positive result certainly for the purposes of a presentation but I know that I think I want to sort of take the answer in a little bit different direction which is to say that there have been a number of studies increasingly since we started this work about different collaboratives that have found mixed results and I think that something to highlight is that the the RIC model is not a one-size-fits-all you need to take into account the capacity of the region in which you're working to actually affect the changes that you're that you're trying to drive and so I think that you know to really highlight one point from this it's that you really need to be aware and cognizant of what's possible and move folks and folks forward in that direction encourage to hear that in it I reflect back on Maureen's presentation just in the prior panel the the importance of meaning and connection and a sense of purpose and that regardless of the five-year time cycle of an evaluation or even the complexities of of drawing clear conclusions about the effects that sense of coming together to do something I think is very powerful and we shouldn't side of that we can probably fit a question or two here as well and get the microphone great presentations my question is about equity and disparities so you know one of the things is that innovation and quality improvement can really widen disparities for the reason that a shiny brought up that the uptake is easier for those who have the resources to improve and I guess Jay I'm really interested like did you how did you address inequities did you measure them what can we do to build in equity as we innovate so we make sure that we're reaching everybody I mean all veterans have a similar access to care with the exception if they live further away and based on their service-connected illness they in this sense the I'm not sure where they are inequity would be the across the board there are challenges there are lots of publications that are now describing use of the II consults very broadly for some of the same reasons and I think the air across specialties some specialties data-driven specialties like hematology they're very popular they're popular in our facility and they're popular in other facilities whereas specialties that require more physical examination and the nuances of the history don't uptake them as much I think your question would be in patients who are subject to an e console perhaps with or without their permission are the outcome similar so then are we seeing differential use not based on what's better for the doctors not better for the patients and I don't think we have that information yet patients do seem like not to travel in the area of preoperative medicine they often don't understand why they have to tell their story to yet another clinician who's not doing their surgery and so that they seem to have appreciated that that part of the innovation and using e consults and it's from what I can tell from them Providence is their very few patients who ask to actually see the anesthesiologist now having said that anesthesia have to see every patient before surgery the difference in this process is they see them the day of surgery where they need to you know look at their throat examine their neck prior to an intubation or examine their spine and get informed consent so anesthesiologist are seeing every patient they're just not seeing them twice well in advance of the surgery and then right at the time of surgery so if I could just jump in there on to address that question um I think it's really important to think about disparities and inequities and that's something that better health really focused on from day one was how do we measure disparities in the population that we're serving and I can tell you anecdotally from looking at better Health's data and this isn't a forthcoming paper so watch out Alan and folks know that don't give it all the way here right but but but to make a very long story short disparities virtually disappeared across race across income groups across payor type in a lot of the outcomes that they were that better health focused on for hypertension heart failure and diabetes so I think that it's a really important to focus on disparities and inequities but there's a way to diffuse innovation in a way that that brings everybody along it's a really key point and one that requires a whole additional discussion quick question quick answers because I don't want to delay or yeah thank you very much for looking into the somewhat ritualized part of healthcare the annual physical exam I run the Marine Corps America every year since 2005 I call it the physical exam my question is simple when functional performance will become actually an integral part of this it'll cost so much about exercise and so forth but we never see that in the annual physical exams as with as much emphasis as it would deserve in in the population we're looking at in the aging population we do look at that annual wellness visit does include an assessment of functional fitness essentially it's not in the same population there's Marines and we're not looking at you know that your your I guess how long it takes you to run a mile but rather can you get up and go you know the time it takes to get up walk across the room and come back and that is in an element that's incorporated and speaks to in the sort of broader goals of the visit to focus on things that would be relevant to an aging population as opposed to listening to the heart listening to the lungs which are not required elements I can't speak to the broader trends towards including physical fitness though great please as I invite my final panelist up join me in thanking this terrific panel we're gonna close with a slightly more conversational approach you've heard a lot of papers but they're also tremendous number of themes that that we've gotten into and I want to see some of those out you'll be hearing from Sharon our arnold deputy director at the agency for Healthcare Research and quality and acting director of its Center for delivery organization in markets will Schrank who is our theme advisor for this issue our great our gratitude to him for all of that's involved in doing that chief medical officer for UPMC insurance services division fact with a faculty appointment in the Department of Internal Medicine Mike Squires vice president for innovation public policy at blue print healthcare IT care navigator Inc I will note and I don't draw any conclusions from this but we did ask CMMI to be here and they were not available to participate but they're obviously another part of the innovation ecosystem that has been referenced in the discussions earlier today so to effectuate a more conversational style I'm going to sit down see how successful that is in making it more conversational but again we've gone from sort of theory to practice and we've also talked a lot about models and I'd like to start with just a sense from each of you in some respects by way of your own introduction to the audience where do you fit in the ecosystem of innovation what is what is your role in promoting the diffusion of innovation Sharon you seem like a natural place to start so I'll turn to you well thank you very much for inviting me here and for this really important issue ARC's mission is to improve the quality and safety of health care and it does it by focusing on three pillars the first is producing evidence either through grants to independent researchers or through our own researchers and this is evidence about primarily kind of what works in making health care better so it's really – how to innovate we also develop tools and training to facilitate the spread of innovation and to make sure that researchers are poised to answer the right questions the questions that are of most importance to health systems and then finally we produce data and measures to track the performance so I feel like we're all over this and and our goal is to really facilitate the diffusion of innovation to improve the healthcare system well we're all very very grateful for the role that you play so I'm at the UPMC Insurance Services Division and as we know the the true innovation and dissemination happens in the marketplace so we need an environment that is conducive to testing and to learning and to spreading at UPMC I think we have a pretty unique environment to do that we have a large health plan with 3.3 million members we have a large health system 41 hospitals lots of doctors our health plan ensures about 40% of the patients that our health plan ensures are seen by UPMC doctors there's a sweet spot there where we are aligned between the payer and the provider and when you see efforts in the marketplace to disseminate to to test new models it's the fragmentation that often gets in the way who pays who ultimately why what makes a model sustainable but in an environment where you have a payer and a provider that are you know sort of responsible for the same bottom-line that are focused on improving the health of the same population that are deeply committed to reducing the healthcare costs and improving the care of the population that they serve together a lot of those barriers sort of diffuse away on top of that we've built a real infrastructure to support innovation a process that we call build the business innovation learning and dissemination group it's a goofy acronym I know but the goal is to pull together people with methodological and a deep commitment to rapid cycle learning and we all know that there's lots and lots of good ideas out there that do or don't get implemented do or don't get diffused based on how they're how what they'll take looks like how the implementation works it's not whether it's a good idea or a bad idea it's how it's actually a but it's actually about the operations and the goal of this build process is get really really deep into the operations process to learn fast to look at data rapidly to bring both the clinical operations the policy leaders the product leaders together to make rapid decisions about what is and isn't working mid-course Corrections and really drive a rapid learning exercise so I think though all of those pieces when put together offers us a unique opportunity to detest the kind of program that James Schuster described earlier with the seriously mentally ill behavioral health homes and to try to really partner more with external groups around solving the key problems the key aims that we all share thank you my Mike Squires I guess I represent the marketplace here this morning our company has really become laser focused on the world of care coordination and closing gaps in health care delivery and I've tried to point this and getting ready for today I tried to think what is all the disparate things that are going on in innovation and the complex 'ti and the diffusion and I came up with my my Triple Aim focus my triple dimension focus of innovation which were are I have to look at my notes here community automation and platform that to me are the three elements the three dimensions you need to look at to figure out how what innovation should be where it will work and how to accelerate that we got involved in we had high goals and starting out when the high tech Act passed what can we do to take away gaps that exist and we decided to do something unusual introduce speed dating to health IT John haladki called it the harmony for health IT we brought together in three regional sessions ten innovative entrepreneurs and 10 CIOs and three to meet them up through speed-dating and see what could come out of it and then we had our an advisory board who was talking and realized that they were less focused on the innovate the innovation programs were running and more interested in comparing notes with each other and we realized that several of them were from innovation centers provider based organizations and so we decided these people need to talk to each other they when they started an Innovation Center they would do some benchmarking going and talk around to each other and then they stayed focused on what their within themselves so we set up a series of five forms across the country starting with Mayo Clinic Johns Hopkins at Johns Hopkins innovation Institute on the west coast and then Mayo Clinic center for innovation and we did speed-dating among provider based innovation centers one getting them together to see what they could learn from each other and doing some brainstorming as well and we supported a Commonwealth Fund study in terms of providing some reach out to Innovation Centers to learn what Innovation Centers are doing and in this survey their top result of the focus of Innovation Centers which was kind of a surprised us this was in 2005 was care coordination that's our focus so we were pleased to learn that because we're trying to work in the same area and that's where we've gone to focus on those areas and it's really been working with things that have been actually created by arc and fundamentals working with one care Vermont which had a learning color which was built on a learning collaborative and took many of those fun to me fundamentals and decided they needed to in their communities really community of communities automate that that patient care and workflows and put it on a single platform to connect all these disparate groups within health service agencies so that's the short story perhaps too long a short story it's great you all are operating in very different environments and that was the goal for having you here Mike you started down this path but I'm gonna turn back to Sharon talk about us one or two successful models of diffusion again we sort of heard a lot about the theory we've heard some examples from your perspective where do you see success what are the key elements so I think we're particularly interested now in moving from kind of diffusion diffusing kind of one innovation I'm individually to thinking about how to set up the infrastructure in the healthcare system so that there's more of Hull of diffusion and it seamlessly gets incorporated into the healthcare system and so that's really where we're focused now so I'll talk about a couple of examples where we think we saw hints of this and where were further studying this area the first is the partnership for patients now many of you might think of this is the CMS innovation because they really funded a lot of this activity but we were a partner with them and we worked very closely with them to do some of the research behind some of the activities that were promoted in partnership for patients we developed the measurement structure and some of the tools that were used and so again this kind of infrastructure of tools that were not only focused on single interventions but could be adapted across a variety of interventions was really critical I think to the success of that initiative and that initiative was very successful in a four year period the project or the program saw a 20 21 percent reduction in hospital acquired conditions resulting in three million fewer adverse events and hospitals 125,000 lives saved and 28 billion in savings according to our estimates so pretty significant we heard about evidence now earlier about kind of creating this infrastructure for small and medium sized physician practices to improve heart health but it's not necessarily focused on heart health it's really focused on the infrastructure we've also got another project looking at health systems kind of the other end of the spectrum and trying to identify what makes health systems successful at implementing evidence with good outcomes and and just try and understand kind of what are the parameters of success well well I think I sort of answered this on the the first time just a couple highlights would be that the importance of alignment of payer and the provider and the effort to try to disseminate and learn rapid use of data rapid cycle assessment of performance and to build off Rebecca's comments the sort of the need to focus on the aim but be able to be flexible around what the model looks like I think is exactly right but it does you know I've just been reflecting at a higher level about hearing these great great presentations and thinking about for us or forum you know many of us who are in the in the field in the marketplace and trying to innovate and disseminate what a critical role the government has played in particular over the last seven years and really driving our opportunity to lead and driving our opportunity to test and learn and get better and improve I you know the arcs leadership is clear and I think you've it's been outlined very very very in broad strokes but with with a lot of great detail here I think your what you've done in terms of both creating taxonomy but also really focusing on specific programs and trying to in particular around patient-centered medical homes really disseminate what works and what doesn't work the role of picori in testing new models comparing new models developing better evidence really driving the science around how to disseminate and innovate and then as Rocco described CMS in the setting of the Innovation Center and the opportunity to to rapidly expand and scale on new payment models every one of those models had a learning and dissemination organization every one of those models had at its core focus on improvement and innovation and dissemination and there is more of a sense of inevitable inevitability now than I think there ever was I think 10 years ago folks in the marketplace would think about their innovations as proprietary it gave them a some sort of a competitive advantage I think now there is a clear expectation you're participating in something bigger and there's a an expectation essentially requirement to share that's happening I think it's palpable to me in the in the sort of circles that I spend time and I'm not sure that's true everywhere but I think a lot of that is due to what the government's done the government serving as a catalyst to really make this happen my cue your answer the first question was in essence this sense that people wanted to talk to each other he does that does that collaborative sense stick I don't want to get a trip over the question of where you've seen success and why but I do want to bring in this notion of of whether we move from a proprietary to a sharing environment for innovation that would be a big deal well even in our own case in New Jersey I would say before the high tech tax hospitals weren't really talking to each other about problems per se they're more like hiding their problems than sharing their problems hospital CEOs would get together in regular meetings and wouldn't talk about what their issues were so with the beginning of the high tech act people started actually they had to share and talk to each other about what that not only what their big successes were but what their problems were so that really has changed the whole you know environment I mean our our business wouldn't exist with what it's doing without what did without what CMS did government became a major cause of not only innovation but diffusion – and even what we were doing we had to be as much as we try to market things we had to be discovered people had to find out oh who can do these things so the first thing was with one care Vermont they had made an amazing toolkit which they have published online all their assessments their programs all the processes it was wonderful and I assumed that every ACO would have those well I was surprised that they didn't and we did a webinar with Pacific group for business with a with one care as well as some other people to describe what toolkits are and how people are using them but the key is now they we had to help one care move from using their care coordinators which weren't a single Hospital coordinators but diverse through primary care practices behavioral health organization social service agencies they had to talk to each other and they had these processes but they weren't real had a way of communicating so how could they do that one is when they did assessments how could they automate them they were they were using Excel spreadsheets and post it notes essentially which is what most care coordinators in the country used in the end they may have a mechanism in the HR where they look up information but when it comes to actually doing their work every day it's the Excel spreadsheet in the post-it notes how to automate that and not only automate it but potentially automate it into a shared care plan and so when it came to Massachusetts which is starting a new a CEO program now and their behavioral health people were charged with organizations we're coordinating for all of Health how could they do that it wasn't something that could come from an EHR yes they need EHR for information but how could they connect all those pieces and automate a shared care plan to be cut I should say automate assessments to turn them into shared care plans and so there are lots of challenges we push down to the lowest level but need some kind of innovation in order to actually make it work well and sharing you both describe environments rich with innovations going back to Jim's very first slide most innovations don't diffuse where's the decision point in deciding we ones are worth the effort if it's hard to diffuse and there's a lot bubbling up out there someone has to say this we're gonna put our bet on this one and we're not gonna put our bet on that one that seems like a very fundamental decision and I would love to hear thoughts about how those decisions get made in your two environments so I'll start it I think number one I think we need to look at innovation as part of a broader body of work I think we don't necessarily want to be in the position of every time there's a new finding going out and pushing to implement that before we understand how that fits and the body works so sometimes caution is appropriate and I you know I think that we need to make sure that we understand when it's appropriate and when it isn't but I think that from my perspective the innovation is really happening in the private sector and the federal government and academia can be it can be incubators but we need to transition in academia to anse to asking the right questions and developing innovations that have the ability to spread and are appropriate for the the private sector and I think that right now in academia folks are being promoted based upon the number of papers that are submitted to academic journals no accept it but but not really how they are you know how closely they're working with systems and how how creative they are coming up with innovations that can be disseminated and I think we need a mind shift on what is valued in terms of the health services research that gets offended I think it's one of the hardest questions because there are you know an infinite number of problems out there that we want to solve we have we if you're a pair you're you take care of a lot of vulnerable patients and we all know that there there's just so many different ways we could leverage our resources to help each and every patient in a way that would make their lives better we probably err on the side of going after too much rather than too little we're probably not quite as disciplined as we could be in that there's a deep sense of responsibility a deep sense of mission a real belief that what by leveraging more generous benefit designs in specific areas by addressing social determinants of health by sending community health workers to the home by leveraging technology to better engage patients that by creating more richer or medication-assisted therapy programs for patients with opiate problems by integrating behavioral and physical health homes that all of these Pro every one of these programs feels like it could work and it's hard to not go after all of them if you were to say how do you you know how do we sort of tease them out how do we say well we're gonna do 12 this year and not 18 I think it really comes down to where the mission seems most critical I don't think I can say quite clearly this is not a you know a dollars and cents decision this is a this is really a question about understanding our members understanding our population understanding the levers we have between the payer and the provider and how can we with a deep commitment to improving the health of the population we serve what's sort of the best ways that we think we can get there and you know some of these things are going to lose money but it's it's the right thing to do and Mike you mentioned an alignment around this notion of care coordination on the one hand that's a great source of alignment it does strike me however that that's a pretty broad term and so to say wow we've got all these centers out here and they all want to do this this could be really really different that's certainly excuse me that's certainly true so even when you look at any tool kid there can be a variety of things in there so which do you approach and how do you approach it and that's what always confused me about health care frankly because there are so many possibilities and so many things to do is what where do you start where do you put your where do you put your energy and I think that that's one of the and people will say well as that care management as a care coordination as a case management it was all these terms are somewhat interchangeable not really is specifically but in the broad sense yes so the question then became what do you need to do specifically so the more specific it could be the better we could potentially produce something that could help make that happen for people so what it's really gotten down to is if people actually use care coordinators whatever they are that are trying to make sure that they move people from one place to another that they have transportation to get there that they take into account social determinants of health that they just make sure that they're reminded to get to a medical appointment those are very basic blocking and tackling exercises but to do that how do you do it in an organized fashion when we did a survey of care coordinators and care coordinator managers what was their challenge just prioritizing what the day was I mean we talked to someone who went to an Excel spreadsheet every day their team would go through it who do we talk to what next so we had to find a way of organizing that process for care coordinators to deal with the specific patients and get those actions that needed to be done by dealing with the most difficult patients the more it's laid out I can't necessarily tell you the results I mean we'll have some results Commonwealth Fund is doing a case study of one care Vermont and what we did in that and the so a case study will be coming out in another month or two and there will be some results that will be shared there but one of the exciting examples this with MassHealth they have really laid what needs to be done for their behavioral health patients who need care that a countable care organization couldn't provide and they'll out laid out these are the actions that need to be taken and we're able to take that and actually put that in an organized fashion so they know what needs to be done and they're paid of eight or so qualifying activities that they have to do now will it work well we'll have to see but at least they have a very organized method for doing it that we could actually help them implement a system and we have now seven community partners which are they're a key part of this which is about a third of the of what's being done in Massachusetts to get together and make sure they do it at this point these are people are highly competitive from behavioral health organizations but now they're meeting together to figure out how to do this because they know they need to do it in order to serve their patients I'm gonna ask a final question I'm going to come down the line this way because the federal official has to go last will you you somewhat obliquely noted that resources that it's not about make you know it's not about what it costs always because it's about mission but resource constraints are real everywhere whether it's dollars or people or institutional support you've each learned a lot about innovation diffusion I'd like your thoughts if you were in a setting if you're setting had more resources for diffusion based on what you know about what makes diffusion successful where would you put those resources to maximize the likelihood that it would yield a higher rate of diffusion of the innovations that we need like I would say really it's the recognition aware a recognition awareness and money by generally state Medicaid agencies if not CMS where it's appropriate not only to assign opportunities and and goals and programs but to make sure there's some kind of technology that can really say or that which EHRs as they're setup today really can't support it they help but that's a nineties looking things it's not necessarily it's not the 21st century look and they're as great as they are they're amazing vastly disappeared to where they were before but it's not enough it won't it doesn't automate workflow it doesn't share information with parents and family members and cab drivers and other people that need to participate in this whole network of care so you need to look at not only is there technology that can be supportive of that but all to give to allow money for that but also to allow the individual organizations to decide on their own how to best do that MassHealth didn't they wanted add EHRs and then other people said who are in the organization's we need some kind of care coordination and in some cases EHRs couldn't do that so what can we find they discovered it MassHealth was smart enough to say go figure it out or they or the organization said we need to figure it out differently than what you're telling us and they gave them the room to be able to do that and that's really I think where a lot of this starts well I think I interpreted your question a little bit differently and that what if what resources would I need and as a you know someone working within a large insurance company trying to innovate disseminate the one resource that I think is missing most frequently across organizations like ours and I think we're trying to take action on this is the ability to gather qualitative information we all all of these ideas that we test are good ideas they're all good ideas and they're as I said before their success or failure comes down to how how they're implemented how they're engaged how they actually are applied what's working and not working on the ground in the real world and the claims data we look at sort of the the quality outcomes that we look at cannot capture the nuance of what's happening on the ground between a care manager and a patient or patient a member and an app or the the person who's trying to set up an appointment or the the transportation that gets that patient of the appointment and in the absence of that nuance the true day-to-day experiences that our members have that our providers have that our care managers have about how the care is actually delivered and where the true barriers are and where we can actually intervene to make a difference we are in the dark we end up figuring these things out it takes but it takes time and if we could if we had more resources to be on the ground to observe and to learn and to interview and to get that sort of qualitative information I think those cycles of improvement and innovation would be much more rapid and I think we'd get to better answers faster or in through Paula gist a shout out to our so I think we invest a lot of money in coming up with new innovations but not very much or not as much in understanding how to diffuse and so I think I'm you know I would put more money into understanding the how of the diffusion not only individual interventions but but how do you create organizations that look at their own data and identify where they need to improve and and experiment internally to take models externally and adapt them for their own use and so I think it's you know potentially using the extension model or others trying to support the healthcare delivery system to adapt and innovate and use their own data and improve that that's what I would like to spend the money that's great well as we close out our day I want to remind you that this issue and briefing were only possible because of the support of the Peterson center on health care the Blue Shield of California foundation Leon M and Harry B Helmsley Charitable Trusts and Ark it's been a rich day I hope you feel the same we have a lot to learn and a lot to learn about how we learn but we've also made tremendous progress so thank you for joining us today and please join me in thanking our pilot and we are adjourned

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *