Nobody argues with the idea that the next big advance in cancer care or HIV vaccine development requires a significant commitment to research. Nor would anyone argue that the quality of that research needs to be excellent or that it must yield reliable information and insight about what actually works and what doesn't. Would anyone invest in a global immunization campaign with a vaccine not proven effective in rigorously conducted trials? Of course not.
Tragically, when it comes to health care policy, the need for rigorous, long-term, high quality health systems research, is often over-looked, underappreciated, or pushed aside for expediency. Launching large-scale health policy initiatives without a clear understanding of precisely how care will be delivered more effectively to improve health outcomes is the equivalent of launching a massive immunization campaign not knowing whether the vaccine in the syringe is effective. More accurately – in the case of recent CMS, Center for Medicare and Medicaid Innovation initiatives focusing on payment reforms (like ACOs, Medicare shared savings, and bundled payments) – it's like starting the immunization campaign and telling each enlisted organization to come up with their own vaccine as they go. Never mind that most such organizations have relatively little capital, expertise, or experience in conducting such research or that a critical review of the results from their respective "labs", to date, show little high quality research conducted and scant evidence of effective interventions.
Payment Reforms Do Not Reduce the Need for Health Systems Research
The current massive allocation of resources and attention dedicated to fast-tracking payment reforms while providing comparatively less support for sustained health systems research appears to be based on 3 assumptions.
Assumption #1 – policy innovations (in this case, payment reforms) that begin as large-scale initiatives are more efficient because they obviate the need to commit additional time and effort to take interventions to scale or broadly disseminate best practices. Not true. Across the funded sites in the new payment reform programs, there will be a variety of models tried (codified to varying degrees) with a broad range of effectiveness. This will become apparent if enough resources and rigor are put into the evaluation of these programs. The task of figuring out what worked best and encouraging others to adapt effective models to different settings and contexts still remains. In other words, the challenge of disseminating innovations is merely deferred, not eliminated. In the short term, we might imagine that we are quickly moving the needle on a large scale – but that will not be true unless highly effective exemplars are reliably identified and their models widely adopted.
Assumption #2 – the expectation that large numbers of organizations can suddenly develop new delivery models that improve health outcomes and lower costs seems to rest on a naive assumption that this type of innovation (system redesign) is not complex and is largely dependent on organizational motivation. But there is little evidence to suggest that such innovations in delivery systems require less exploratory research, experimentation, analytical expertise or time than do biomedical or public health breakthroughs. Successful system redesign may require as much disciplined, iterative research as that required to develop an effective HIV vaccine or cancer treatment – in fact, it might require more. Underestimating the need for sustained systems research could pose a significant barrier to progress.
Assumption #3 – with cost saving "guarantees" to Medicare built into many of the new payment reforms, who cares whether participating organizations come up with better models of care delivery or not? – CMS will still save money for Medicare. This is shortsighted because health care organizations that try and fail to achieve a sustainable model of care will lose faith, withdraw from the effort, signal others to stay away, and mobilize political opposition to future reforms. In the worst-case scenario, trumpeting the success of models that turn out to be unsustainable erodes confidence that system innovations can lead to meaningful improvement.
Example: Without Good Evidence State Program In Jeopardy
A case in point is Community Care of North Carolina, a statewide approach to controlling Medicaid costs using a medical home care management model. The program offers providers considerable flexibility with respect to the specifics of implementation and is delivered differently by different provider networks across the state (i.e., high degree of local variation). Controversy has sprung up about the validity and continued relevance of earlier actuarial analyses used to evaluate the program's overall impact. In the absence of more direct and ongoing research methods, doubts about the program’s effectiveness have grown.
Quotes from a performance audit of North Carolina's Medicaid program issued January 2013 (click here for the full audit) highlight the need for a solid evidence base even in the case of one of the nation’s longest running and most often showcased large-scale deployment of the Patient Centered Medical Home model.
“While North Carolina Medicaid relies on several strategies to control consumption, the single strategy that is invested with creating the greatest cost savings is Community Care of North Carolina (CCNC). The State expected to save $90 million per year with CCNC during SFYs 2012 and 2013, but fell $39.5 million short of its goal in 2012. CCNC is a form of managed care that provides case management services in a medical home environment. It is assumed to provide savings in providing medical services to participants. More than a decade of data exists that would allow a study by medical researchers on whether the medical home model truly saves money and/or results in better medical outcomes. It would be a service to the nation as well as North Carolina to use this data to genuinely evaluate the questions associated with medical homes.”
One recommendation coming from the audit;
“The State of North Carolina should engage medical researchers to perform a scientifically valid study based upon actual data to determine whether the CCNC model saves money and improves health outcomes.”
CMS Halts Research Without Offering a Plan to Promote the Most Effective Model of Medicare Care Management
CMS recently notified Health Quality Partners (HQP) that it plans to terminate the long-running research HQP has been successfully conducting for over 11 years, in the Medicare Coordinated Care Demonstration (MCCD) despite having approved a new phase of research just 3 years ago, which was designed to validate the significant cost savings seen among higher-risk beneficiaries in prior subgroup analyses. The HQP community-based nurse care management program represents a new category of system redesign that fills the void between medical care and public health; an Advanced Preventive Service.
It has been shown to save lives, reduce hospitalizations, and lower cost among higher-risk beneficiaries (click on any of the following to see; Summary Slide by HQP, Third Report to Congress 2008, Fourth Report to Congress 2011, Health Affairs 2012, PLoS Medicine 2012). Given that the average life expectancy of Americans aged 65 is over 19 years, the durable longitudinal effectiveness (validated out to 4 years) of HQP’s program seems to be just what Medicare needs. HQP has requested CMS reconsider its decision and continue its support of HQP's research in order to further advance this critical knowledge base. In addition, it seems an opportune time for CMS to consider supporting pilots or demonstrations to test the replication and scalability of the program in other areas of the country wishing to implement this model.
We Need Both – Payment Reform to Spur Action and High Quality Health Systems Research
The issue is not pitting payment reforms against health systems research or trading one off for the other. The best hope for transforming the American health care system is a good balance and integration of these two important drivers of health system innovation. Achieving such a balance is a very complex challenge for the dedicated people at HHS, CMS, AHRQ, CDC, and other federal agencies. But such efforts, especially if undertaken in collaboration with states, is our best shot at creating a higher quality, person centered, preventive, and fiscally responsible health system.
Payment Reforms Do Not Reduce the Need for Health Systems Research
The current massive allocation of resources and attention dedicated to fast-tracking payment reforms while providing comparatively less support for sustained health systems research appears to be based on 3 assumptions.
Assumption #1 – policy innovations (in this case, payment reforms) that begin as large-scale initiatives are more efficient because they obviate the need to commit additional time and effort to take interventions to scale or broadly disseminate best practices. Not true. Across the funded sites in the new payment reform programs, there will be a variety of models tried (codified to varying degrees) with a broad range of effectiveness. This will become apparent if enough resources and rigor are put into the evaluation of these programs. The task of figuring out what worked best and encouraging others to adapt effective models to different settings and contexts still remains. In other words, the challenge of disseminating innovations is merely deferred, not eliminated. In the short term, we might imagine that we are quickly moving the needle on a large scale – but that will not be true unless highly effective exemplars are reliably identified and their models widely adopted.
Assumption #2 – the expectation that large numbers of organizations can suddenly develop new delivery models that improve health outcomes and lower costs seems to rest on a naive assumption that this type of innovation (system redesign) is not complex and is largely dependent on organizational motivation. But there is little evidence to suggest that such innovations in delivery systems require less exploratory research, experimentation, analytical expertise or time than do biomedical or public health breakthroughs. Successful system redesign may require as much disciplined, iterative research as that required to develop an effective HIV vaccine or cancer treatment – in fact, it might require more. Underestimating the need for sustained systems research could pose a significant barrier to progress.
Assumption #3 – with cost saving "guarantees" to Medicare built into many of the new payment reforms, who cares whether participating organizations come up with better models of care delivery or not? – CMS will still save money for Medicare. This is shortsighted because health care organizations that try and fail to achieve a sustainable model of care will lose faith, withdraw from the effort, signal others to stay away, and mobilize political opposition to future reforms. In the worst-case scenario, trumpeting the success of models that turn out to be unsustainable erodes confidence that system innovations can lead to meaningful improvement.
Example: Without Good Evidence State Program In Jeopardy
A case in point is Community Care of North Carolina, a statewide approach to controlling Medicaid costs using a medical home care management model. The program offers providers considerable flexibility with respect to the specifics of implementation and is delivered differently by different provider networks across the state (i.e., high degree of local variation). Controversy has sprung up about the validity and continued relevance of earlier actuarial analyses used to evaluate the program's overall impact. In the absence of more direct and ongoing research methods, doubts about the program’s effectiveness have grown.
Quotes from a performance audit of North Carolina's Medicaid program issued January 2013 (click here for the full audit) highlight the need for a solid evidence base even in the case of one of the nation’s longest running and most often showcased large-scale deployment of the Patient Centered Medical Home model.
“While North Carolina Medicaid relies on several strategies to control consumption, the single strategy that is invested with creating the greatest cost savings is Community Care of North Carolina (CCNC). The State expected to save $90 million per year with CCNC during SFYs 2012 and 2013, but fell $39.5 million short of its goal in 2012. CCNC is a form of managed care that provides case management services in a medical home environment. It is assumed to provide savings in providing medical services to participants. More than a decade of data exists that would allow a study by medical researchers on whether the medical home model truly saves money and/or results in better medical outcomes. It would be a service to the nation as well as North Carolina to use this data to genuinely evaluate the questions associated with medical homes.”
One recommendation coming from the audit;
“The State of North Carolina should engage medical researchers to perform a scientifically valid study based upon actual data to determine whether the CCNC model saves money and improves health outcomes.”
CMS Halts Research Without Offering a Plan to Promote the Most Effective Model of Medicare Care Management
CMS recently notified Health Quality Partners (HQP) that it plans to terminate the long-running research HQP has been successfully conducting for over 11 years, in the Medicare Coordinated Care Demonstration (MCCD) despite having approved a new phase of research just 3 years ago, which was designed to validate the significant cost savings seen among higher-risk beneficiaries in prior subgroup analyses. The HQP community-based nurse care management program represents a new category of system redesign that fills the void between medical care and public health; an Advanced Preventive Service.
It has been shown to save lives, reduce hospitalizations, and lower cost among higher-risk beneficiaries (click on any of the following to see; Summary Slide by HQP, Third Report to Congress 2008, Fourth Report to Congress 2011, Health Affairs 2012, PLoS Medicine 2012). Given that the average life expectancy of Americans aged 65 is over 19 years, the durable longitudinal effectiveness (validated out to 4 years) of HQP’s program seems to be just what Medicare needs. HQP has requested CMS reconsider its decision and continue its support of HQP's research in order to further advance this critical knowledge base. In addition, it seems an opportune time for CMS to consider supporting pilots or demonstrations to test the replication and scalability of the program in other areas of the country wishing to implement this model.
We Need Both – Payment Reform to Spur Action and High Quality Health Systems Research
The issue is not pitting payment reforms against health systems research or trading one off for the other. The best hope for transforming the American health care system is a good balance and integration of these two important drivers of health system innovation. Achieving such a balance is a very complex challenge for the dedicated people at HHS, CMS, AHRQ, CDC, and other federal agencies. But such efforts, especially if undertaken in collaboration with states, is our best shot at creating a higher quality, person centered, preventive, and fiscally responsible health system.