With the release of the final rule for Stage 3 of Meaningful Use (MU) of health information technology (IT), and modifications to Stage 2 of MU for 2015 – 2017, the Centers for Medicare and Medicaid Services (CMS) has taken the unusual step of asking for a second round of comments on Stage 3. Their rationale: their first request for comments was made prior to the passage of the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA).
This post draws from this second round of comments submitted in December 2015, on behalf of the American College of Physicians. We argue that, rather than tinkering around the edges, CMS should completely reconceptualize the program. A new MU program should focus on specialty-specific measures of quality and be based in normal clinical work patterns, rather than rewarding clinicians for meeting artificial “one size fits none” threshold requirements for specific functional uses of electronic health records (EHRs) and health IT.
On January 11, 2016, CMS Acting Administrator Andy Slavitt announced, “The meaningful use program as it has existed will now be effectively over and replaced with something better.”
The Advent Of Merit-Based Incentive Payment System Offers A Window For Rethinking MU
MACRA combines three existing programs—the Physician Quality Reporting System (PQRS), the Value-Based Payment Modifier (VBM), and MU (enacted in the Health Information Technology for Economic and Clinical Health (HITECH) Act), along with a new program—clinical practice improvement activities, into a single program, the Merit-Based Incentive Payment System (MIPS). Rule making is soon to be underway for this value-modified continuation of fee-for-service Medicare; thus, all components of MIPS are on the table for reconsideration.”
With MIPS on the near-term horizon, why not just drop MU? Several prominent thought-leaders have made this suggestion, and it’s not an unreasonable idea. One could make a strong case that, with a much clearer set of specific quality and cost objectives (PQRS and VBM), MU is not only no longer needed — it could only get in the way.
We share this concern. To stand a chance of successful participation in MIPS, clinicians need to begin the process now of constructively engaging with their technology and workflows. From what clinician survey data tell us, that will not happen as long as clinicians have a visceral, negative reaction to leveraging their electronic health records (EHRs) and other health IT.
Further, in our view, clinicians who excel in quality and cost savings, but fail MU because of missing a threshold of an EHR-use measure (such as ePrescribing 59.9 percent of the time), are more likely to react to this “gotcha” by leaving the Medicare program, rather than shifting their focus from quality and costs to compliance on EHR-use thresholds.
Just dropping MU could resolve all of these negatives. But that can’t be done without having Congress amend MACRA, as MACRA explicitly calls for a MU component within MIPS. Additionally, we believe there is still a need for a federal role within health IT, but ONLY if Stage 3 of MU looks very different from what has been proposed.
Others have called for CMS to delay final rulemaking on MU Stage 3. We believe this would be a mistake. Given that CMS is soon to be engaged in a rulemaking attempt to weave the existing PQRS, VBM, and MU programs into one, delaying rulemaking on the MU Stage 3 component would likely result in the existing, inadequate Stage 2 rules being cobbled into MIPS.
That said, we do believe that a stand-alone MU program should end in 2018 (if not before), and a very different and more appropriate MU program should emerge in 2019 for those clinicians participating in the MIPS program. Our recommendation is thus not a delay in finalizing Stage 3, but rather a transition from Stage 2 to implementation of MIPS in 2019, with a more appropriate and integrated MU component.
Part 2: Where Meaningful Use Went Off Course
At the outset, MU could not be judged on outcomes; neither the measures—eClinical Quality Measures (eCQMs)—nor clinicians were ready. Hence the construct of a dual measurement system—pay-for-reporting of quality measures and pay-for-performance of EHR-functional-use measures—was developed. EHR-functional-use measures are measures of how the EHR was used, such as what percent of patients seen in a given period were sent a secure electronic message.
While this construct held appeal, as it was based on activities that were measurable (EHR-functional-use), it was predicated on an idea that we believe was fundamentally wrong — that the potential value of EHRs and other health IT is the same for all specialties and in all settings of care. What followed was the development of a single set of EHR-functional-use measures across specialties and care settings. This is the polar opposite of eCQMs, which are more narrowly applicable, based on specialty and setting of care.
The eCQM ecosystem also includes a process for eCQM development, stewardship, and evolution, all of which is done by organizations with specific competencies and supported by evidence. eCQMs are expected to evolve over time, or even disappear, depending on evolving science. In contrast to this independent and evidence-based process for eCQMs, EHR-functional-use measures were developed via a political process, and largely supported by belief and/or self-referential logic. Thus, if the Health IT Policy Committee (HITPC) believed an EHR function was important, CMS proposed and then finalized a rule.
As only a few functions were left as professional attestations, CMS then had to define EHR-functional-use measures. In the interests of fairness to EHR vendors and clinicians, these quickly became highly prescriptive. What was conceptually a poor fit for many clinicians became an extremely poor fit; it was at this point that MU became widely derided as “one-size-fits-none.”
Further, clinicians regularly discovered that even when they performed an activity that satisfied the objective of an EHR-functional-use measure (such as providing patients with appropriate and specific educational resources), this activity did not count unless it was done in a highly specific way. This “gotcha” debacle arose from measures being over-defined and/or from the approach that EHR vendors had to take to document that a highly defined activity occurred.
With strictly defined EHR-functional-use measures in place, even highly functional EHRs had to be re-engineered. And those that couldn’t be easily re-engineered had MU EHR-functional-use measure checkboxes added, in some cases appearing as a parallel and duplicative process in a left or right column. EHR usability began to suffer, and satisfaction with EHRs, which had been steadily improving, now began to plummet.
To make things worse, CMS applied threshold requirements to EHR-functional-use measures. And where the functional-use measure did not fit into normal workflow, clinicians began to develop workarounds and gimmicks to satisfy thresholds. Since CMS considers a score above the threshold as validation that the measure is appropriate, whether the score results from gimmicks or clinically appropriate activity, EHR functional-use measures are blind to the very learning health care system that they were built to support. Unlike eCQMs, which will be withdrawn or modified if the clinical practices they support are no longer valid (or even harmful), CMS will only remove an EHR-functional-use measure if it is “topped out, redundant, or duplicative.”
Moreover, time wasted on deriving and using such gimmicks has taken its toll: Clinicians are angry, and even worse, where they should be engaging with health IT as a valuable tool, they are disengaged. From a perspective of readiness for value-based care, this is exactly the opposite of where they should be.
If MIPS was not on the near term horizon, one solution would be to redo EHR-functional-use measures in collaboration with medical professional and specialty societies, such that functional requirements made sense. And EHR-functional-use measures could then be subject to scientific study and emerging evidence, as eCQMs are. But with MIPS ready to start in just a few years and with the ongoing maturation of clinical quality and value measures, it makes more sense to conclude that EHR-functional-use measures as a vehicle for judging performance, once needed, are no longer necessary.
This is then the right time, as the PQRS and VBM programs are being re-conceptualized into the MIPS program, to course correct MU into something that will help to support the other components. And in our view, it doesn’t matter if one’s opinion is “MU, good riddance;” or “MU, time to declare victory and move on” — it is time to dramatically change the MU program.
Part 3: Guiding Principles For Meaningful Use Within The MIPS Program
While MACRA mandates an MU component within MIPS, it does not specify what that component should look like. We believe that CMS thus has a golden opportunity for an MU “do-over,” rather than merely making incremental changes in the current program. With a truer “north star” of evolving specialty-specific clinical quality measures combined with cost-of-care and relevant clinical practice improvement activities, there would no longer be the need for paying (or penalizing) clinicians based on their threshold performance on a “one-size-fits-none” set of EHR-functional-use measures.
We believe the relevant question for CMS then becomes, “What would otherwise be missing from technology or workflow if EHR technology and use were only incented by quality and resource use measures (as per the current PQRS and VBM programs), as well as by clinical practice improvement activities?” This is where a new, more relevant MU program should be focused.
MU should follow its own advice and avoid requiring duplicative activities. Patient-centered quality and value measures should be the primary drivers of how EHRs and other health IT are used. A new MU within MIPS should aim only to fill in key gaps and strive to incent optimization of value from health IT (based on specialty and setting of care). MU must permit and even encourage flexibility and innovation.
MU measures should not be burdensome and should be built into existing or emerging workflows, such that as care is provided process or activity measurements can be auto-generated. Processes and/or activities would then be determined by specialty-specific best practices or even self-determined, as are quality improvement or value improvement projects.
When clinicians within MIPS need to satisfy the quality, resource use, and clinical practice improvement categories, as well as MU, their MU activities should be in support of these MIPS activities and not an additional set of unrelated requirements.
In defining new, more appropriate MU measures and activities, CMS and the Office of the National Coordinator for Health IT (ONC) should set a new tone by collaborating with physicians and other clinicians to determine the key tasks that clinicians and staff will need to perform to improve care. This further flips the existing paradigm from “health IT as an end point to health IT as enabling infrastructure” — a means to an end.
While some tasks could be performed more efficiently with better technology and standards, physicians should have opportunities to work on these tasks now, with currently available technology. There should not be requirements for mature standards and certification criteria before practices can begin work on activities such as smoothing out the wrinkles in care coordination and chronic care management.
MU measures should not be defined only where standards and certification criteria exist. MU measures and activities should challenge practices to work on solutions in advance of standards and certification criteria so that the learnings from the activities can better inform future standards and certification criteria.
If this approach is taken now, it may not be too late to turn around clinician anger and disengagement. Clinicians engaged with meaningful and helpful health IT will be better prepared for the future of health care delivery.
Meaningful Use Within MIPS
1. Eliminate All EHR-Functional-Use Measure Thresholds
We believe there is no place for the continuation of EHR-functional-use measure thresholds in this last stage of MU. As noted above, there is no EHR-functional-use process that is or should be the same for all specialties and settings of care. And where functional use measures are not consistent with what physicians naturally do, physicians see them as annoying distractions and develop workarounds and/or gimmicks merely to achieve the thresholds.
EHR-functional-use data are most useful when they reflect actual workflow, not contrived attempts to achieve a performance threshold. Learning is enhanced when reported data include naturally occurring variance and are not restricted to CMS’s prescriptive definition of threshold achievement. Furthermore, due to the threshold requirements, CMS only gets data from the relatively few successful attesters — we learn absolutely nothing from those who fail to attest because they have not met the thresholds.
If thresholds were not required, we would have data from far more practices. These data could tell us volumes about what works and what doesn’t, and the possible causes of variation. With health IT, as with clinical medicine, we can learn a great deal from the near misses.
2. Judiciously Continue EHR-Functional-Use Measurement Reporting (Without Thresholds) To Develop New Learnings On Health IT-Enabled Care And Improve EHR Usability
Where key questions as to health IT processes persist, data collection and analysis should continue. This approach to EHR-functional-use measurement must avoid the prior pitfall of narrow and/or overly prescriptive definitions that have been the cause of compliance-driven and/or duplicative clinical workflows, poor EHR usability, and distraction from the development of more usable and useful software.
Less prescriptive EHR-functional-use measures would free up EHR developer time, allowing them to be both innovative and responsive to their customers (physicians), as well as permitting physicians to be innovative and more responsive to their customers, patients. It would additionally permit physicians to use the next several years to develop and test “advanced clinical processes” so that they are better prepared for clinical practice and practice improvement, under either MIPS or an Alternative Payment Model.
We believe this approach is responsive to HITECH and MACRA. Workflow processes that CMS, ONC, professional/specialty societies, or researchers believe hold the potential for improving care can be engineered into EHRs and auto-reported. CMS and ONC could then collaborate on using these process data to learn, rather than to grade.
3. Demonstrate Continuing Education Within The Domains Of Health IT
While it is now common for health system leaders to be physicians and other clinicians with advanced training in clinical informatics (Chief Medical Information Officers and Chief Nursing Information Officers), medical and graduate medical education does not yet consistently impart basic and specialty-specific education on health IT. More than anything else, physicians now need health IT guidance, best practices, implementation strategies, and new ideas. One new meaningful activity should involve continuing medical education (CME) on health IT and data management topics.
To that end, we propose that CMS and ONC work with the physician professional and specialty societies to create or endorse relevant continuing education, and to jointly determine the amount and frequency of such health IT-related education. We suggest the following as examples of relevant modules for continuing education:
- Health IT to improve quality of care for patients, and where appropriate, to improve quality for select populations;
- Safe use of health IT, with a focus on reducing errors of both omission and commission;
- Health IT to improve value; and,
- Health IT to engage patients and families
4. Advance Meaningful And Practical Interoperability
Interoperability is meaningful when it supports the liquidity of relevant narrative and structured sets of contextual data, so that the data can support clinical needs. Interoperability is practical when it is meaningful and it further supports the needs of the participants in data exchange. Interoperability is neither meaningful nor practical when the content of every exchange is predetermined and mandated by regulations.
Many thought leaders have speculated that health information exchanges have had difficulty staying in business not because of technological challenges, but rather due to the lack of a sustainable business case. The introduction of MIPS and APMs should present a real business opportunity for relevant data exchange and thus meaningful and practical interoperability.
5. Support Bi-Directional And Less Burdensome Reporting To Public Health
We appreciate the need for clinicians and public health authorities to share relevant data. We believe that this has been seen as burdensome only because it has been presented historically as a one-sided requirement on clinicians. Also, we believe that all data recipients have an obligation to supply clinically valuable information resulting from the reporting activity back to the reporting clinicians. Reporting clinicians should receive the benefit of bi-directional data exchange for their reporting activities.
The requirement becomes yet more burdensome when the obligation to report is not necessarily satisfied via a standard report into a single utility. As currently proposed for Stage 3, clinicians may have to create and support multiple interfaces to public health entities; each public health entity may require custom changes to reports and/or duplicative entry into unique forms.
These practices must stop. We believe that significant cost and effort could be saved through the development of a standard Application Programming Interface (API) for all public health, quality, and registry reporting.
6. Develop And Use Flexible Measures Of Patient Engagement That Respect Patient And Family Needs And Preferences
We recommend that measures of patient engagement shift from a singular set of prescriptive EHR-functional-use measures to attestations of patient engagement activities that reflect the setting of care, context, and patient needs and preferences. Clinicians should have the option to develop a case report describing a patient engagement problem and the actions the practice took, including use of health IT, to resolve the problem.
7. Encourage Clinician Engagement And Innovation By Allowing For New Types Of Activities To Replace Existing Measures
The MIPS program requires that CMS assess clinician performance along a linear scale (from 0-100), and this variable scoring is meant for all scorable activities within MIPS, including MU. This opens the door even further for a truly flexible MU program within MIPS. And while MACRA includes practice improvement as a component of MIPS, we feel that health IT-related practice improvement projects should count as meaningful use of health IT. Here are several examples:
- Precision Medicine / Learning Health System: participation in practice-based research or other observational study efforts.
- Clinical Informatics: support of iterative improvement in practical informatics via use of an “EHR Feedback” application or via participation in EHR user groups.
- Local Quality, Safety, or Value Improvement Projects Leveraging Health IT: several specialty societies have similar requirements for board certification (but without specification of IT use).
- Patient Safety and Near-Miss Reporting: done locally, as part of a professional or specialty society effort, or more formally with a patient safety organization.
- Development of eCQMs that Support Quality Improvement: done locally, as part of one’s engagement with a Qualified Clinical Data Registry, or as part of a quality improvement community of interest (such as the Million Hearts™ Initiative).
The views presented do not necessarily represent official policy of the American College of Physicians, or of MedStar Health. We thank Shari M. Erickson, Vice President, Governmental Affairs & Medical Practice, American College of Physicians, and Brooke Rockwern, Analyst, Regulatory Affairs, American College of Physicians, for their contributions.