Written and maintained by CASRAI Editorial Board
Last updated
The Hospital Readmissions Reduction Program (HRRP) reduces Medicare inpatient payments to hospitals whose risk-standardised 30-day readmission performance on six specific condition and procedure cohorts exceeds what their peer group achieves. It was created by section 1886(q) of the Social Security Act, added by the Affordable Care Act, and applies to discharges beginning 1 October 2012 — fiscal year 2013. Its mechanics live at 42 CFR 412.150, 412.152 and 412.154.
It is also one of the most heavily criticised payment policies in Medicare, with a substantial peer-reviewed literature arguing it under-delivered on its own objective and may have carried unintended clinical consequences. Both halves belong on the same page: a hospital quality team needs the arithmetic to know what its penalty is, and the critique to know what the number does and does not mean. This guide sets out the mechanics from CMS’s own primary sources, then presents the documented criticisms attributed to the specific published work that made them, without endorsing or dismissing any of them.
The Six Measures
CMS states on its own HRRP programme page (last modified 16 March 2026) that the programme includes the following condition- or procedure-specific 30-day risk-standardised unplanned readmission measures:
- Acute myocardial infarction (AMI)
- Chronic obstructive pulmonary disease (COPD)
- Heart failure (HF)
- Pneumonia
- Coronary artery bypass graft (CABG) surgery
- Elective primary total hip arthroplasty and/or total knee arthroplasty (THA/TKA)
The regulation does not name these six; it defines an applicable condition generically as one the Secretary selects among conditions and procedures whose readmissions are high volume or high expenditure and for which endorsed measures exist with exclusions for readmissions unrelated to the prior discharge — such as planned readmissions or transfers to another applicable hospital — or among other conditions the Secretary determines appropriate. Measure selection therefore happens through Inpatient Prospective Payment System rulemaking, not by regulation text. Check the current IPPS final rule rather than assuming the cohort list is static.
Readmission itself is defined narrowly and mechanically at 42 CFR 412.152: the admission of an individual to the same or another applicable hospital within 30 days of discharge from an applicable hospital. Where the patient goes back does not matter; whether it was the same institution does not matter.
The Excess Readmission Ratio
The excess readmission ratio (ERR) is the unit of measurement, and its definition contains a detail that changes how the whole programme behaves. Per 42 CFR 412.152, it is a hospital-specific ratio for each applicable condition and applicable period: risk-adjusted readmissions based on actual readmissions, divided by risk-adjusted expected readmissions — but not less than 1.0.
That floor is asymmetric by design. A hospital that performs better than expected on a condition gets an ERR of 1.0 for it, exactly the same as a hospital that performs precisely as expected. There is no credit below 1.0 and therefore no reward for outperformance; the programme is a penalty mechanism, not a two-sided incentive. Excelling on one condition cannot offset a poor ERR on another.
The applicable period is the data window. Per the regulation: for FY 2022 it was the 3-year period 1 July 2017 to 30 June 2020; beginning with FY 2023 the 3-year period advances by one year each programme year; and beginning with the FY 2027 programme year the applicable period becomes a 2-year period advancing by one year annually, unless the Secretary specifies otherwise. That shortening — added by the FY 2026 IPPS rulemaking recorded at 90 FR 37201 — is a material change for anyone modelling future penalties, because it halves the smoothing effect of the third year.
Peer Grouping by Dual-Eligible Proportion
The programme did not always compare hospitals to a stratified peer group. CMS’s own account: the 21st Century Cures Act directs CMS to assess a hospital’s performance relative to other hospitals with a similar proportion of beneficiaries dually eligible for Medicare and full Medicaid benefits, beginning in FY 2019. The legislation also requires that estimated payments under the peer-grouping methodology equal those estimated under the non-peer-grouping methodology used from FY 2013 to FY 2018, to maintain budget neutrality — which is what the regulation’s neutrality modifier does: a multiplicative factor that equates total Medicare savings under the stratified methodology to the previous non-stratified one.
Two definitions carry the stratification. Proportion of dual-eligibles is the number of dual-eligible patients among all Medicare fee-for-service and Medicare Advantage stays during the applicable period. Dual-eligible, for adjustment factors from FY 2021 onward, is a beneficiary identified as having full benefit status in both Medicare and Medicaid in data sourced from State MMA files for the month of discharge — with an exception for beneficiaries who die in the month of discharge, who are identified from the previous month’s data.
Because the payment formula compares a hospital’s ERR to the peer-group median ERR rather than to a single national benchmark, the peer group a hospital lands in materially affects its penalty. The number of strata CMS uses and how their boundaries are drawn are set through IPPS rulemaking rather than in the regulation text, so confirm the current stratification against the applicable final rule before modelling.
How the Payment Adjustment Is Computed
The arithmetic runs through three defined quantities, all at 42 CFR 412.152 and 412.154.
- Aggregate payments for excess readmissions — for each applicable condition, the product of four terms, summed across conditions: the base operating DRG payment amount for that condition; the number of admissions for that condition; the hospital’s ERR minus the peer-group median ERR; and the neutrality modifier.
- Aggregate payments for all discharges — the sum of base operating DRG payment amounts for all discharges for all conditions in the applicable period.
- Readmissions adjustment factor — the greater of (a) 1 minus the ratio of the first quantity to the second, or (b) the floor adjustment factor.
The floor adjustment factor is set in the regulation: 0.99 for FY 2013, 0.98 for FY 2014, and 0.97 for FY 2015 and every subsequent fiscal year. That is the statutory maximum penalty in the form it is actually applied — CMS states it plainly as a payment reduction capped at 3 percent.
The reduction applies to the base operating DRG payment amount for each discharge during the fiscal year, and that base is defined restrictively. It is the wage-adjusted DRG operating payment plus any applicable new technology add-on payments, determined without regard to Hospital Value-Based Purchasing adjustments, and excluding indirect medical education payments, disproportionate share payments, outlier payments and low-volume adjustments. A sole community hospital paid on its hospital-specific rate has the difference between that rate and the Federal rate left untouched by the adjustment; for a Medicare-dependent small rural hospital, that difference is included.
What a Hospital Can and Cannot Contest
CMS provides each applicable hospital with a confidential Hospital-Specific Report and discharge-level information used to calculate its excess readmission ratios, and the hospital has 30 days from receipt to review and submit corrections to those ratios. The limits on that review are as important as the right itself: the regulation states that the administrative claims data used to calculate an applicable hospital’s excess readmission ratios… are not subject to review and correction, and CMS’s own guidance is that hospitals cannot submit corrections to the underlying claims data or add new claims to the data extract during the period. The review is for discrepancies in the calculation, not in the inputs.
Beyond that, 42 CFR 412.154(e) forecloses appeal entirely on six items: there is no administrative or judicial review of the determination of base operating DRG payment amounts, the methodology for determining the adjustment factor (including the ERR and both aggregate-payment quantities), the applicable period, the neutrality modifier, the proportion of dual-eligibles, or the applicable conditions.
One relief valve exists. Under 412.154(d), CMS may grant an extraordinary circumstance exception (ECE) for events beyond a hospital’s control — the regulation names natural or man-made disasters such as a hurricane, tornado, earthquake, terrorist attack or bombing — that affected its ability to comply with reporting requirements. A hospital must request one within 60 calendar days of the date the circumstance occurred, via QualityNet or a successor site. CMS may also grant an ECE to hospitals that did not request one, where a systemic problem with a CMS data collection system or a circumstance affecting an entire region or locale is identified.
The Documented Criticisms
What follows attributes each critique to the specific published work that made it. These are the arguments as their authors made them; where the literature disagrees with itself, that is stated rather than resolved.
Did it reduce readmissions by as much as intended?
Gupta and Fonarow, reviewing the programme in the European Journal of Heart Failure (2018;20(8):1169–1174, DOI 10.1002/ejhf.1212) under the title “The Hospital Readmissions Reduction Program — Learning From Failure of a Healthcare Policy”, argue that the policy “had limited success in achieving its primary objective of reducing readmissions as the achieved reduction in heart failure readmissions was much smaller (~9%) than anticipated (~25%)”, and that some of the observed reduction in risk-standardised readmission rates is “attributable to the artifact of administrative upcoding post-HRRP rather than an actual decline in readmissions.” They also raise, as concerns present from the passage of the law, gaming behaviours including increases in observation stays, delaying readmissions beyond discharge day 30, and inappropriate emergency-department triage strategies.
The observation-stay substitution concern
This is the most-cited worry about the programme, and the primary study of it does not support it. Zuckerman, Sheingold, Orav, Ruhter and Epstein — writing from the HHS Office of the Assistant Secretary for Planning and Evaluation — examined monthly hospital-level readmission and observation-service rates among elderly Medicare beneficiaries from October 2007 through May 2015 across 3,387 hospitals (New England Journal of Medicine 2016;374(16):1543–1551, DOI 10.1056/NEJMsa1513024).
They found readmission rates for targeted conditions declined from 21.5% to 17.8% between 2007 and 2015, and for non-targeted conditions from 15.3% to 13.1%. Observation-unit stays for targeted conditions rose from 2.6% in 2007 to 4.7% in 2015, and for non-targeted conditions from 2.5% to 4.2% — so both trends are real. But the substitution hypothesis requires them to be linked within hospitals, and they were not: the authors report “no significant association between changes in observation-unit stays and readmissions after implementation of the ACA”, concluding that they “did not find evidence that changes in observation-unit stays accounted for the decrease in readmissions.”
The tension between this finding and the gaming concerns catalogued by Gupta and Fonarow is genuine and unresolved in the literature. It is also why the inpatient-versus-observation status decision has its own body of rules worth understanding independently — see the two-midnight rule and condition code 44.
The mortality signal
Wadhera, Joynt Maddox, Wasfy, Haneuse, Shen and Yeh studied 8.3 million Medicare fee-for-service hospitalisations for heart failure, AMI and pneumonia across four periods from April 2005 to March 2015 (JAMA 2018;320(24):2542–2552, DOI 10.1001/jama.2018.19232). They report that HRRP announcement and implementation were each significantly associated with an increase in 30-day post-discharge mortality for heart failure and pneumonia, but not for AMI — where announcement was in fact associated with a decline. The increase for HF and pneumonia was, in their analysis, mainly among patients who were not readmitted but died within 30 days of discharge.
The authors’ own caveats are load-bearing and are frequently dropped when this study is cited. They note that HRRP implementation was not significantly associated with an increase in mortality within 45 days of admission relative to pre-HRRP trends, and conclude that “given the study design and the lack of significant association of the HRRP with mortality within 45 days of admission, further research is needed to understand whether the increase in 30-day postdischarge mortality is a result of the policy.” This is an observational association with an explicitly unsettled causal interpretation, not a demonstrated harm.
The socioeconomic-adjustment debate
The longest-running structural criticism is that the ERR’s risk adjustment accounts for clinical risk but not for the social and economic circumstances that drive a substantial share of readmission risk — so hospitals serving poorer populations would be penalised for their patient mix rather than their care. Unlike the other critiques, this one produced a legislative response: the 21st Century Cures Act’s requirement, described above, that CMS compare each hospital to others with a similar proportion of dual-eligible beneficiaries, effective FY 2019.
Whether peer grouping by dual-eligibility proportion adequately captures socioeconomic risk is a separate and still-open question. Dual-eligibility is a binary programme-enrolment status, not a measure of income, housing stability, health literacy or access to follow-up care, and the budget-neutrality requirement means stratification redistributes penalties among hospitals rather than reducing them in total. Anyone assessing the fairness argument should treat the Cures Act change as a partial, specific response to a broader claim, not as a resolution of it.
What This Means for a Hospital Quality Team
Three practical consequences fall out of the mechanics rather than the debate. First, because the ERR floors at 1.0, effort is only rewarded on conditions where the hospital is currently above expected — there is no return on pushing a good cohort further. Second, because the penalty is applied to every base operating DRG payment for the year, not just to the six cohorts, a small adjustment factor moves a large amount of money, and the exposure scales with total inpatient volume rather than with readmission volume. Third, because the applicable period is a multi-year rolling window — narrowing from three years to two from FY 2027 — improvement made this year appears in the penalty calculation years later, and any internal dashboard that treats current-month readmission rates as a proxy for current penalty exposure will mislead.
Readmission-reduction work itself sits outside this programme’s design: HRRP measures outcomes and adjusts payment, and says nothing about method. The interventions with the clearest evidence base — medication reconciliation at transitions, and structured improvement cycles such as the PDSA model — are worth evaluating on their own merits rather than as penalty-avoidance tactics.
Frequently Asked Questions
What is the maximum HRRP penalty?
Three percent. The regulation sets the floor adjustment factor at 0.97 for FY 2015 and all subsequent fiscal years, and CMS describes this as the payment reduction being capped at 3 percent. Earlier years were capped lower: 1 percent for FY 2013 and 2 percent for FY 2014.
Which conditions count toward HRRP?
Six: acute myocardial infarction, COPD, heart failure, pneumonia, CABG surgery, and elective primary total hip and/or total knee arthroplasty. The measures are 30-day risk-standardised unplanned readmission measures — planned readmissions and transfers to another applicable hospital are excluded by measure design.
Does a readmission to a different hospital count?
Yes. The regulation defines a readmission as admission to the same or another applicable hospital within 30 days of discharge, and the penalty attaches to the discharging hospital.
Can a hospital appeal its HRRP penalty?
Only narrowly. There is a 30-day review-and-correction period after receipt of the Hospital-Specific Report, limited to discrepancies in the calculation — the underlying claims data cannot be corrected and new claims cannot be added. Beyond that, 42 CFR 412.154(e) bars administrative and judicial review of the methodology, the applicable period, the neutrality modifier, the proportion of dual-eligibles and the applicable conditions. An extraordinary circumstance exception is available for disasters and similar events, requested within 60 calendar days.
Does the programme adjust for how poor a hospital’s patients are?
Indirectly and partially. Since FY 2019, under the 21st Century Cures Act, CMS compares each hospital to a peer group of hospitals with a similar proportion of dual-eligible beneficiaries, and the formula measures the hospital’s ERR against its peer-group median. The change is budget-neutral by statute, so it redistributes penalties rather than reducing them, and dual-eligibility status is a coarse proxy for socioeconomic risk.
Did HRRP work?
That is genuinely contested rather than unknown. Readmission rates for targeted conditions fell substantially over the period studied (Zuckerman et al., 21.5% to 17.8% from 2007 to 2015), but non-targeted conditions fell too; Gupta and Fonarow argue the heart-failure reduction was far below what was anticipated and partly an upcoding artifact; and Wadhera et al. report an association with increased 30-day post-discharge mortality for heart failure and pneumonia whose causal status the authors themselves describe as unresolved. Read the primary papers rather than any summary of them, including this one.
This guide summarises 42 CFR 412.150, 412.152 and 412.154 as codified on 1 August 2026, and CMS’s own Hospital Readmissions Reduction Program page as last modified 16 March 2026. Measure selection, peer-group construction and applicable periods are set through annual IPPS rulemaking and change; verify against the current final rule before relying on any of this operationally. The critiques are attributed to their published sources and are presented as those authors framed them, not as CASRAI’s assessment.
Related: see the Patient Safety & Infection Prevention hub, and the CLABSI guide for how a different CMS-linked hospital measure is constructed from a surveillance definition rather than from claims.
