MACRA began with a sensible promise: stop threatening physicians with automatic Medicare cuts, move beyond payment for volume, and reward better care. Congress passed the Medicare Access and CHIP Reauthorization Act of 2015 with strong bipartisan support, finally eliminating the Sustainable Growth Rate formula. That achievement was real. The disaster was the machinery built around it.
The Merit-based Incentive Payment System, or MIPS, turned “value” into a complicated contest involving measure selection, data submission, scoring rules, exceptions, delayed feedback, and consultant-friendly alphabet soup. It often identifies who can document compliance most efficiently rather than who keeps patients healthier. Medicare still needs payment reform, but physicians should not need a minor degree in federal spreadsheetology to participate.
What MACRA was supposed to fix
Before MACRA, the Sustainable Growth Rate repeatedly triggered scheduled Medicare cuts that Congress blocked with temporary “doc fixes.” MACRA repealed that formula and created the Quality Payment Program, offering two broad routes: MIPS or qualifying participation in Advanced Alternative Payment Models. CMS describes the framework as a way to reward high-quality, cost-efficient care rather than volume alone.
The goal was reasonable. Fee-for-service can reward more visits, tests, and procedures even when prevention, coordination, or a well-timed phone call would serve the patient better. But MIPS did not truly redesign payment. It bolted a complicated performance score onto the existing fee schedule and hoped the dashboard would transform medicine.
Why MIPS has become a policy disaster
1. It rewards reporting skill more reliably than clinical excellence
Traditional MIPS scores clinicians across quality, cost, improvement activities, and promoting interoperability. For the 2026 performance year, the threshold remains 75 points. Clinicians can face a negative adjustment as large as 9%, while positive adjustments are scaled to preserve budget neutrality. In other words, the threat is clear, but the reward depends partly on how many other clinicians lose money.
The system also encourages strategic measure selection. Research has found that practices may choose measures on which they already perform well, allowing impressive scores without clear gains in patient experience. GAO reported that at least 93% of providers earned positive adjustments in the early years, yet the largest positive adjustment was only 1.88%. A grading system that gives almost everyone an A and then hands out pocket change is not a transformation strategy.
2. The administrative cost can outweigh the practical value
A JAMA Health Forum study estimated that physicians and staff spent more than 50 physician hours per year and roughly $12,800 per physician participating in MIPS in 2019. Medical-group testimony continues to cite those findings while describing MIPS as a major regulatory burden. That money and time come from appointment capacity, staff attention, technology budgets, and the patience of people already wrestling with prior authorization and overflowing inboxes.
CMS has tried to simplify the program through MIPS Value Pathways, specialty groupings, exceptions, and technical assistance. Yet simplification that requires another pathway, registration process, and instruction manual can feel like reducing clutter by buying a larger filing cabinet. CMS plans to sunset traditional MIPS through future rulemaking, but the replacement still relies heavily on the same measurement machinery.
3. Small, solo, and rural practices face a steeper hill
Large systems can spread compliance work across analysts, informatics teams, quality departments, and hundreds of clinicians. Small practices cannot. AMA reporting on 2025 payment adjustments said 14% of all MIPS-eligible clinicians were penalized, compared with 29% in small practices, 49% in solo practices, and 18% in rural practices. Those gaps suggest that administrative capacity influences the score.
If payment policy pushes independent practices toward consolidation merely because they cannot afford the reporting infrastructure, Medicare may gain cleaner dashboards while communities lose local clinicians. Research has also found that practices embedded in larger health systems are more likely to participate intensively in alternative payment models, making the road to “value” easier for organizations that already possess capital and scale.
4. Social risk can be mistaken for poor performance
Patients with unstable housing, transportation barriers, food insecurity, limited caregiver support, or multiple chronic conditions often need more resources and may have worse measured outcomes despite excellent care. Studies have associated higher patient social risk with lower MIPS scores, while psychiatrists have received lower scores and more penalties than other outpatient physicians. A value program that inadequately adjusts for complexity can encourage clinicians to avoid difficult patients or underserved locations.
5. Feedback arrives too late
MIPS separates performance from payment by a long delay. By the time a score affects reimbursement, staff, workflows, and patient panels may have changed. Delayed feedback is useful for an autopsy, not for steering a living practice. Current reform proposals therefore emphasize providing several quarters of claims data and performance reports during the same year, when clinicians can still improve care.
6. MACRA never solved the base-payment problem
For 2026, CMS finalized separate conversion factors for qualifying Advanced APM participants and other clinicians, reflecting statutory updates of 0.75% and 0.25%, a one-year 2.5% increase enacted by Congress, and technical adjustments. The temporary increase improves 2026 payment, but it does not establish a durable inflation-based update. Medicare cannot demand continuing investment in staff, access, and technology while treating rising practice costs as an occasional surprise.
What should replace MIPS
1. Establish a stable, inflation-aware base payment
Congress should separate adequate baseline payment from quality incentives. Annual physician updates should reflect a credible measure of practice-cost inflation, with particular attention to labor-intensive services such as primary care, geriatrics, psychiatry, and chronic disease management. A practice should not have to win a quality contest merely to keep up with wages, rent, malpractice coverage, and health information technology.
2. Use a small, claims-based accountability set
Medicare should retire the giant pick-your-own-measure menu and use a limited national set of reliable outcomes and utilization signals calculated primarily from claims and existing data. Examples include avoidable admissions, preventable emergency department use, readmissions, medication safety, timely follow-up after hospitalization, and total cost trends under clinically credible attribution rules.
Claims are imperfect, so chart-based information will still be needed for selected measures. The principle is simple: minimize clinician-generated reporting and reserve it for data that add real clinical meaning. If Medicare already has the information, practices should not be required to type it again while wearing a quality-improvement hat.
3. Reward improvement, not just high absolute scores
A safety-net practice should earn meaningful rewards for measurable improvement even if it has not yet reached a wealthy-region benchmark. Payment formulas should combine attainment, improvement, and equity, supported by strong clinical and social risk adjustment, minimum case thresholds, and appropriate peer comparisons. The goal is to make complex patients attractive to excellent clinicians, not financially radioactive.
4. Build realistic on-ramps to alternative payment models
Advanced APMs should remain available, but participation should not require an immediate leap into substantial downside risk. Medicare can offer staged models: prospective care-management payments, episode bundles, population-based payments, and shared savings before carefully limited losses. Primary care practices might receive risk-adjusted monthly payments for access and coordination. Specialists could use episode payments designed with their societies and patient groups.
CMS generally requires Advanced APMs to connect payment to quality and cost while placing meaningful risk on participants. Many practices, however, lack the capital, data systems, patient volume, or local partners to absorb that risk. Medicare should help build readiness with upfront infrastructure support rather than distributing a PDF and wishing everyone luck.
5. Provide timely, understandable data
Practices should receive monthly or quarterly claims feeds, clear attribution lists, benchmark explanations, and drill-down tools showing what drives their results. Data should arrive soon enough to change referral patterns, discharge follow-up, medication management, and care-coordination outreach. Clinicians must also be able to challenge attribution errors before payment is affected.
6. Make incentives predictable and large enough to matter
Quality incentives should be prospectively defined, transparently funded, and large enough to justify investment. Budget neutrality makes the MIPS upside a moving target. The AMA’s proposed Data-Driven Performance Payment System is one attempt to emphasize claims-based measurement, faster feedback, and fewer reporting requirements. Its details require debate, but its diagnosisthat MIPS should be replaced rather than endlessly decoratedis sound.
A practical transition plan
In year one, Congress should freeze MIPS penalties, enact an inflation-linked fee update, require quarterly feedback, and protect small and rural practices. During years two and three, CMS should test a short core measure set in primary care and selected specialties, with voluntary participation and upfront care-management support. Independent evaluators should examine outcomes, administrative hours, equity, access, and consolidation.
By year four, Congress should retire MIPS and offer three understandable lanes: sustainable fee-for-service with light claims-based accountability, voluntary episode or population models, and advanced models with larger rewards and carefully calibrated risk. Three clear choices would beat dozens of pathways, exceptions, weights, categories, and annual surprises.
Experience from the front lines: what the current system feels like
Consider a composite primary care practice based on patterns documented in surveys, studies, and physician-group testimony. It has four clinicians, an office manager, several medical assistants, and a patient panel heavy with diabetes, heart failure, chronic lung disease, and transportation problems. Nobody was hired as a MIPS specialist. Nevertheless, the practice must identify its status, choose measures, confirm data completeness, monitor EHR mappings, document improvement activities, investigate exclusions, and later verify a score tied to future payment.
In January, the office manager discovers that last year’s workflow is not automatically this year’s workflow. The EHR vendor offers a quality module, but setup and support cost extra. One measure looks sensible until its denominator produces patients the doctors barely recognize. Another is easy to score well on but has little connection to the practice’s biggest challenge: keeping fragile patients out of the emergency department. The rational choice is the easier measure. The clinically satisfying choice is care coordination. The program nudges the team toward the former while giving speeches about the latter.
By spring, a physician spends part of an evening reviewing lists with duplicate and outdated attribution. A nurse calls patients after hospital discharge because the work genuinely prevents harm, but documenting the call in exactly the right field becomes almost as important as making it. The office manager attends a webinar explaining another acronym. Everyone laughs when the presenter says the changes reduce burden. It is not joyful laughter.
Compare that office with a large multispecialty system. The system employs analysts who optimize measure selection, validate files, monitor dashboards, and appeal errors. Its clinicians may provide excellent care, but they also benefit from an administrative exoskeleton. When the large group earns a higher score, MIPS cannot cleanly distinguish better medicine from better machinery.
The small practice then postpones hiring a care manager because compliance consumed the available margin. One partner considers selling to a hospital. Another temporarily closes the panel to new Medicare patients. No dramatic collapse occurs. There is simply one fewer appointment here, one delayed outreach call there, and one more independent office absorbed into a larger organization. Policy failure often arrives not with a siren, but with an unavailable new-patient slot.
Now imagine the same practice under a better model. Medicare sends a risk-adjusted monthly primary care payment, quarterly claims data, and a short list of transparent goals. The practice sees that post-discharge follow-up is weak and several emergency visits involve medication confusion. It hires a part-time nurse care manager, contacts patients promptly after discharge, and reconciles medications. Improvement earns a predictable payment, and the data arrives soon enough to guide the next quarter.
That experience would feel less like test preparation and more like quality improvement. It would not eliminate every debate about attribution, risk adjustment, or measurement. Health care is too complicated for a perfect scorecard. But it would align payment with work patients recognize: easier access, safer transitions, fewer avoidable crises, and clinicians who have time to listen. That is the practical standard any MACRA replacement should meet.
Conclusion: keep the goal, replace the machinery
MACRA solved one historic problem by repealing the Sustainable Growth Rate formula, but MIPS has not delivered a credible or efficient path to value. It is expensive, biased toward organizations with greater reporting capacity, vulnerable to measure gaming, slow to provide feedback, and weakly connected to outcomes that matter to patients.
The answer is not a return to unlimited fee-for-service. It is a simpler architecture: inflation-aware base rates, a small set of reliable measures generated mostly from existing data, meaningful rewards for improvement, strong adjustment for complexity, timely feedback, and voluntary alternative models with realistic on-ramps. Medicare should pay for better care without forcing clinicians through an annual scavenger hunt.
Note: This policy commentary reflects publicly available information current through July 7, 2026. Medicare payment rules can change through legislation and annual CMS rulemaking.
