The Standardized Readmission Ratio (SRR) for a dialysis facility is the ratio of the number of observed index discharges from acute care hospitals to that facility that resulted in an unplanned readmission to an acute care hospital within 4-30 days of discharge, to the expected number of readmissions given the discharging hospital’s and patient’s characteristics, and based on a national event rate. Note that the measure applies exclusively to Medicare covered ESRD chronic dialysis patients enrolled in either Fee-for-Service (FFS) or Medicare Advantage (MA) programs.
This measure was previously endorsed under CBE ID 2496 (2015-2020). We are now submitting it with improvements for endorsement under a new CBE ID 5110.
Measure Specs
General Information
Unplanned readmission rates are an important indicator of patient morbidity and quality of life. On average, dialysis patients are admitted to the hospital nearly 1.5 times a year and hospitalizations account for approximately 35% of total Medicare expenditures for dialysis patients [2]. In both 2021 and 2022, between 31-33% of dialysis patient discharges from an all-cause hospitalization were followed by an unplanned readmission within 30 days [2]. Measures of the frequency of unplanned readmissions, such as SRR, help efforts to control escalating medical costs, play an important role in providing cost-effective health care, and support coordination of care across inpatient and outpatient settings. Preventive interventions such as fluid weight management, management of mineral and bone disease, anemia management as well as post-discharge processes of care (medication reconciliation) by dialysis facilities, and coordination of care with other providers in the pre- and post-discharge periods (communication with the dialysis provider; medication reconciliation) have the potential to prevent hospital readmissions for ESRD dialysis patients. Preventing hospital readmissions is regarded as a shared responsibility that can be impacted by both dialysis providers and hospitals.
Finally, findings across all performance years of the Center for Medicare and Medicaid Innovation’s Comprehensive ESRD Care Initiative suggest care coordination may reduce readmission risk [1]. The findings of this controlled study showed an overall decrease of 2% in the percentage of Medicare beneficiaries with at least one readmission, among those aligned to an ESRD Seamless Care Organization, relative to a matched comparison group of facilities.
References:
[1] Marrufo et al., Fifth Annual Evaluation Report of the Comprehensive ESRD Care Initiative. Prepared for the Centers for Medicare and Medicaid Services. Jan 2022. https://www.cms.gov/priorities/innovation/data-and-reports/2022/cec-annrpt-py5. Accessed Aug 8, 2025.
[2] United States Renal Data System. 2024 USRDS annual data report: Epidemiology of kidney disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2024
Data are derived from the EQRS patient-specific clinical and administrative data, including ESRD patient list, CMS-2728 Medical Evidence Form, CMS-2746 Death Notification Form, and patient admission and discharge data, from all Medicare certified dialysis facilities, the Medicare Enrollment Database (EDB), and Medicare claims data.
In addition, the database includes transplant data from the Scientific Registry of Transplant Recipients (SRTR), data from the Nursing Home Minimum Dataset, and the provider and survey and certification data from the Internet Quality Improvement and Evaluation System (iQIES) data.
Information on hospitalizations is obtained from Medicare inpatient and skilled nursing claims Standard Analysis Files (SAFs), and past-year comorbidity data are obtained from multiple claim types (inpatient, home health, hospice (Part A only), skilled nursing facility claims).
Fee-for-service (FFS) Medicare Part A (inpatient) and Part B (outpatient and physician supply) claims for dialysis patients are included in the current database; additionally, the database now incorporates Part C Medicare Advantage (MA) data for the MA enrollees. This database ensures that hospital, outpatient dialysis, and other billable services under Medicare – whether FFS or MA – are captured.
Numerator
The number of hospital discharges among eligible Medicare ESRD dialysis patients during the reporting period that are followed by an unplanned hospital readmission within 4-30 days of discharge.
Index discharges are restricted to Medicare-covered hospitalizations for inpatient care at short-term acute care hospitals and critical access hospitals. Discharges from SNFs, long-term care hospitals (LTCHs), rehabilitation hospitals and PPS-exempt cancer hospitals - as well as those from separate dedicated units for hospice, rehabilitation and psychiatric care - are excluded.
Potential readmissions are:
- Medicare-covered hospitalizations for inpatient care at short-term acute care hospitals and critical access hospitals.
- Classified as either a planned or unplanned admission according to the planned readmission algorithm (see below for further discussion).
Note that hospitalizations where the patient dies on the date of discharge are included for consideration as potential readmissions (in other words, the outcome of the readmission itself is not relevant to this measure).
The numerator for a given facility is the total number of index hospital discharges during the reporting period that are followed by unplanned readmissions within 4-30 days of discharge and that are not preceded by a “planned” readmission or other competing event that also occurred within 4-30 days of discharge. If the first event during days 4-30 after discharge is an unplanned hospitalization, then the index discharge is classified as having a readmission. If the first event during days 4-30 is a planned hospitalization or other competing event, then the index discharge is classified as not having a readmission. Competing events include admissions to rehabilitation or psychiatric hospitals, death, transplant, loss to follow-up, withdrawal from dialysis, and recovery of renal function. A readmission is considered “planned” under three scenarios:
i) The patient undergoes a procedure that is always considered planned (e.g., kidney transplant) or has a primary diagnosis that always indicates the hospitalization is planned (e.g., maintenance chemotherapy).
ii) The patient undergoes a procedure that MAY be considered planned if it is not accompanied by an acute diagnosis. For example, a hospitalization involving a heart valve procedure accompanied by a primary diagnosis of diabetes would be considered planned, whereas a hospitalization involving a heart valve procedure accompanied by a primary diagnosis of acute myocardial infarction (AMI) would be considered unplanned.
iii) The readmission was to a rehabilitation, long-term care, or psychiatric facility.
This definition follows from the algorithm developed by Yale New Haven Health Services Corporation/Center for Outcomes Research & Evaluation (YNHHSC/CORE) for The Centers for Medicare and Medicaid Services 2018 All-Cause Hospital Wide Measure Updates and Specifications Report Hospital Level 30-Day Risk-Standardized Readmission Measure – Version 7.0. https://www.qualitynet.org/files/5d0d375a764be766b010141f?filename=2018_Rdmsn_Updates%26Specs_Rpts.zip
Denominator
The expected number of index discharges during the reporting period that are followed by an unplanned readmission within 4-30 days in each facility. The expected number is derived from a model that accounts for patient characteristics, the dialysis facility to which the patient is discharged, and the discharging acute care or critical access hospital involved.
General Inclusion Criteria for Dialysis Patients
Patients are included in the measure only after they have had ESRD for greater than 90 days. This minimum 90-day period assures that patients are eligible for Medicare, either as their primary or secondary insurer, and that follow-up is complete. Thus, the measure excludes hospitalizations during the first 90 days of ESRD as well as patients who die or recover kidney function during that time period.
In order to assure completeness of information on hospitalizations for all patients included in the analysis, we restrict to Medicare patients who are either enrolled in Medicare Advantage or who reach a certain threshold of Medicare outpatient dialysis and inpatient claims. Specifically, months within a given dialysis patient-period are used for SRR calculation when the patient is enrolled in Medicare Advantage or meets the criterion of being within two months after a month with either: (a) $1200+ of Medicare-paid dialysis claims OR (b) at least one Medicare inpatient claim.
Index discharges are restricted to Medicare-covered hospitalizations for inpatient care at short-term acute care hospitals and critical access hospitals. Discharges from SNFs, long-term care hospitals (LTCHs), rehabilitation hospitals and PPS-exempt cancer hospitals - as well as those from separate dedicated units for hospice, rehabilitation and psychiatric care - are excluded. Index discharges are attributed to the dialysis provider to which the patient is discharged at the end of the hospital stay. In other words, the facility to which the patient is discharged is held responsible for any unplanned readmissions occurring within 4-30 days of the index discharge, regardless of whether the patient is still being treated at the facility associated with the index discharge at the time of readmission.
Exclusions
An Index Discharge is excluded when it:
- Occurs at a non-acute care hospital
- Ends in death
- Is against medical advice
- Includes a primary diagnosis for certain types of cancer, mental health or rehab prosthesis
- Includes a revenue center code indicating rehabilitation
- Occurs after a patient’s 12th admission in the calendar year
- Is from a PPS-exempt cancer hospital
- Includes a patient not on dialysis and under care of a dialysis facility at discharge
- Is followed within three days by any hospitalization (at acute care, long-term care, rehabilitation, or psychiatric hospital or unit), death, transplant, loss to follow-up, withdrawal from dialysis, or recovery of renal function
- Is associated with an inpatient stay of 365 days or longer
- Discharged against medical advice: We determine discharge status from the inpatient claim.
- Certain diagnoses: The primary diagnosis at discharge is available on the inpatient claim; we group these diagnoses into more general categories using AHRQ’s Clinical Classification Software (CCS; see
- http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp for descriptions of each CCS). The excluded CCSs are shown below.
- Cancer: 42, 19, 45, 44, 17, 38, 39, 14, 40, 35, 16, 13, 29, 15, 18, 12, 11, 27, 33, 32, 24, 43, 25, 36, 21, 41, 20, 23, 26, 28, 34, 37, 22, 31, 30
- Psychiatric: 657, 659, 651, 670, 654, 650, 658, 652, 656, 655, 662
- Rehab for prosthesis: 254
- Presence of one or more of the following revenue center codes: 0024, 0118, 0128, 0138, 0148, 0158
- Number of admissions: We remove any records for a patient after his/her 12th discharge in the calendar year.
- PPS-exempt cancer hospitals: The following hospitals are listed as PPS-exempt cancer hospitals in the Federal Register (http://www.gpo.gov/fdsys/pkg/FR-2011-07-18/html/2011-16949.htm): 050146, 050660, 100079, 100271, 220162, 330154, 330354, 360242, 390196, 450076, 500138
- Any index discharge with an inpatient readmission of any type, a death, a transplant, loss to follow-up, withdrawal from dialysis, or recovery of renal function occurring within the first 0-3 days following the index discharge.
Measure Calculation
The numerator is the observed number of hospital discharges that are followed by an unplanned readmission within 4-30 days. The denominator is the expected number of index discharges that are followed by an unplanned readmission within 4-30 days in each facility adjusted for the patient mix and the discharging hospital characteristics. The measure for a given facility is calculated by dividing the numerator by the denominator. See 1.18a for SRR_Flowchart_04-2026_508.pdf attachment
N/A
There is not a minimum sample size needed to calculate the performance score. Public reporting of this measure on DFCC is restricted to facilities with at least 11 index discharges to ensure stable estimates and for the measure to comply with restrictions on reporting of potentially identifiable patient information related to small cell size.
Supplemental Attachment
Point of Contact
Not applicable
Wilfred Agbenyikey
Baltimore, MD
United States
Jonathan Segal
UM-KECC
Ann Arbor, MI
United States
Importance
Evidence
Several studies and commentaries strongly suggest pre- and post-discharge interventions within the purview of dialysis providers may reduce the risk of unplanned readmissions within the ESRD chronic dialysis population [3, 7, 13, 14, 23]. Plantinga et al found that interventions in the immediate post-discharge period were associated with reduced readmission risk among hemodialysis patients [23]. They also suggest that post-discharge processes of care may help identify certain patients at higher risk for readmission, creating opportunities for dialysis providers to initiate interventions to reduce readmissions. However, early readmissions raised the risk of subsequent mortality [23]. Chan and colleagues found that certain post-discharge assessments and changes in treatment at the dialysis facility may be associated with a reduced risk of readmission [8]. Assessments included hemoglobin testing and modification of EPO dose; mineral and bone disease testing and modification of vitamin D; and, importantly, modification of dry weight after discharge. The risk of unplanned hospital readmission was reduced when these assessments were completed within the first seven days post-hospital discharge. In a commentary, the Chan 2009 study and several others are cited as examples of the potential for care coordination to reduce readmissions among ESRD dialysis patients [25]. The findings from Chan and colleagues are further supported by results from a recent study comparing principal diagnosis of index hospitalizations and their associated readmissions [8, 17]. Tables included in the paper’s supplementary materials clearly demonstrate that a significant portion of readmission principal discharge diagnoses are for dialysis-related conditions. For example, regardless of the index hospitalization cause (i.e. infectious, endocrine, cardiovascular, GI, dermatologic, renal, etc), the top principal discharge diagnosis lists for related readmissions prominently included diagnoses typically associated with fluid overload and failure of fluid management in dialysis patients (fluid overload, hypertension, CHF, etc). These results, in combination with more recent work [25], suggest that adjustment of patient target weight by the dialysis facility in the early post-hospitalization discharge period (to adjust for the frequent weight loss and/or in-hospital re-assignment of a lower post-dialysis target weight) is a likely mechanism for a substantial minority of unplanned readmissions in the US chronic dialysis population.
Studies in the non-dialysis setting have cited post-interventions or a combination of pre-and post-discharge interventions as drivers for reducing unplanned readmissions [1, 2, 3, 5, 6, 7, 9, 10, 11, 12, 15, 16, 19, 20, 22]. However, a recent study and related commentary challenge the reported magnitude of reductions in hospital-wide readmissions since 2010, as part of the publicly reported Hospital Wide Readmission (HWR) measure for the Hospital Readmission Reduction Program (HRRP) [21, 24]. They suggest the potential driver of these reductions is in part attributed to a change in diagnosis coding policy for inpatient claims that took effect in October 2012. While it is not yet settled whether the reductions were primarily or only nominally driven by the ability of hospitals to report more condition diagnoses, resulting in more robust comorbidity risk adjustment in the measure, the concern has generated attention about whether reported improvements in readmission rates is a result of the HWR and by extension better care delivery by hospitals. These concerns are not considered germane to drivers of readmission reduction based on the dialysis facility readmission measure. The SRR was implemented by CMS in 2015, after the 2012 coding changes took effect. Therefore, trends in dialysis patient 30-day readmissions only reflect the period since the claims-based diagnoses coding changes, and observed reductions since that time are not considered an artifact of the 2012 inpatient diagnosis coding changes.
Finally, findings across all performance years of the Center for Medicare and Medicaid Innovation’s Comprehensive ESRD Care Initiative suggest care coordination may reduce readmission risk [18]. The findings of this controlled study showed an overall decrease of 2% in the percentage of Medicare beneficiaries with at least one readmission, among those aligned to an ESRD Seamless Care Organization, relative to a matched comparison group of facilities.
References:
[1] Ahmed A, Thornton P, Perry GJ, Allman RM, DeLong JF. Impact of atrial fibrillation on mortality and readmission in older adults hospitalized with heart failure. Eur J Heart Fail. 2004;6(4):421–426.
[2] Anderson MA, Clarke MM, Helms LB, Foreman MD. Hospital readmission from home health care before and after prospective payment. J Nurs Scholarsh. 2005;37(1):73–79.
[3] Assimon, M.M.; Wang, L.; Flythe, J.E. Failed Target Weight Achievement Associates with Short-Term Hospital Encounters among Individuals Receiving Maintenance Hemodialysis. 1,3 Journal of the American Society of Nephrology 29(8):p 2178-2188, August 2018. DOI: 10.1681/ASN.2018010004
[4] Azevedo A, Pimenta J, Dias P, Bettencourt P, Ferreira A, Cerqueira-Gomes M. Effect of a heart failure clinic on survival and hospital readmission in patients discharged from acute hospital care. Eur J Heart Fail. 2002 Jun;4(3):353–359.
[5] Balaban RB, Weissman JS, Samuel PA, Woolhandler S. Redefining and redesigning hospital discharge to enhance patient care: a randomized controlled study. J Gen Intern Med. 2008;23(8):1228–1233.
[6] Bostrom J, Caldwell J, McGuire K, Everson D. Telephone follow-up after discharge from the hospital: Does it make a difference? Appl Nurs Res. 1996;9:47–52.
[7] Braun E, Baidusi A, Alroy G, Azzam ZS. Telephone follow-up improves patients satisfaction following hospital discharge. Eur J Internal Med. 2009;20:221–225.
[8] Chan K, Lazarus M, Wingard R, et al. “Association between repeat hospitalization and early intervention in dialysis patients following hospital discharge.” Kidney International (2009) 76:331-41.
[9] Coleman E, Parry C, Chalmers S, et al. The care transitions intervention. Arch Internal Med. 2006;166:1822–1828.
[10] Creason H. Congestive heart failure telemanagement clinic. Lippencotts Case Management: Managing the Process of Patient Care. 2001 Jul-Aug;6(4):146-56.
[11] Dudas V, Bookwalter T, Kerr KM et al. The impact of follow-up telephone calls to patients after hospitalization. American Journal of Medicine. 2001; 111(9B):26S-30S
[12] Dunn JM, Elliot TB, Lavy JA et al. Outpatient clinic review after arterial reconstruction: is it necessary? Annals of the Royal College of Surgeons of England. 1994 Sep;76(5):304-6.
[13] Flythe, J.E.; Assimon, M.M.; Overman, R.A. Target weight achievement and ultrafiltration rate thresholds: potential patient implications. Published: 02 June 2017. Volume 18, article number 185, (2017).
[14] Flythe, J.E.; Katsanos, S. L.; Hu, Yichun; Kshirsagar, Abhijit V.; Falk, Ronald J.; Moore, Carlton R. Predictors of 30-Day Hospital Readmission among Maintenance Hemodialysis Patients: A Hospital’s Perspective. Clinical Journal of the American Society of Nephrology 11(6):p 1005-1014, June 2016.
[15] Jack B, Chetty V, Anthony D, et al. “A reengineered hospital discharge program to decrease rehospitalizaton.” Annals of Internal Medicine (2009) 150:178-88.
[16] Koehler BE, Richter KM, Youngblood L et al. Reduction of 30-day postdischarge hospital readmission or emergency department (ED) visit rates in high-risk elderly medical patients through delivery of a targeted care bundle. Journal of Hospital Medicine. 2009 Apr;4(4):211-8.
[17] Lin, E.; Bhattacharya, J.; Chertow, G.M. Prior Hospitalization Burden and the Relatedness of 30-Day Readmissions in Patients Receiving Hemodialysis. Journal of the American Society of Nephrology 30(2):p 323-335, February 2019.
[18] Marrufo et al., Fifth Annual Evaluation Report of the Comprehensive ESRD Care Initiative. Prepared for the Centers for Medicare and Medicaid Services. Jan 2022. https://www.cms.gov/priorities/innovation/data-and-reports/2022/cec-annrpt-py5. Accessed Aug 8, 2025.
[19] McDonald, MD. The hospitalist movement: wise or wishful thinking? Nurse management. 2001 Mar;32(3):30-1.
[20] Naylor M, Brooten D, Jones R et al. Comprehensive discharge planning for the hospitalized elderly. A randomized clinical trial. Annals of Internal Medicine. 1994 Jun 15;120(12):999-1006.
[21] Ody, C.; Msall, L.; Dafny, L.S.; Grabowski, D.C.; Cutler, D.M. Health Affairs. Vol. 38, No. 1. Decreases In Readmissions Credited To Medicare’s Program To Reduce Hospital Readmissions Have Been Overstated.
[22] Parry C, Min SH, Chugh A et al. Further application of the care transitions intervention: results of a randomized controlled trial conducted in a fee-for-service setting. Home Health Care Services Quarterly. 2009;28(2-3):84-99.
[23] Plantinga LC, King L, Patzer RE, Lea JP, Burkart JM, Hockenberry JM, Jaar BG. Early hospital readmission among hemodialysis patients in the United States is associated with subsequent mortality. Kidney Int. 2017 Oct;92(4):934-941. doi: 10.1016/j.kint.2017.03.025. Epub 2017 May 20.
[24] Wadhera, R.K.; Joynt Maddox, K.E. Hospital revisits within 30 days after discharge for medical conditions targeted by the Hospital Readmissions Reduction Program in the United States. August 2019. BMJ 2019;366:l4563.
[25] Wish, J.B. The Role of 30-Day Readmission as a Measure of Quality. Clinical Journal of the American Society of Nephrology 9(3):p 440-442, March 2014.
[26] Plantinga LC, Masud T, Lea JP, Burkart JM, O'Donnell CM, Jaar BG. Post-hospitalization dialysis facility processes of care and hospital readmissions among hemodialysis patients: a retrospective cohort study. BMC Nephrol. 2018 Jul 31;19(1):186.
Measure Impact
Not applicable. This measure was previously endorsed under CBE ID 2496 (2015-2020). We are now submitting it with improvements for endorsement under a new CBE ID 5110. The measure has been implemented in two federal programs for many years.
Not applicable. This measure was previously endorsed under CBE ID 2496 (2015-2020). We are now submitting it with improvements for endorsement under a new CBE ID 5110. The measure has been implemented in two federal programs for many years.
There are several studies indicating that patients with kidney failure who require dialysis value an assessment of hospitalization rates at the dialysis facility level. This would extend to re-hospitalizations.
In a study of 81 dialysis patients and 45 caregivers that assessed what outcomes were most important, respondents gave high priority to clinical outcomes such as infection that are associated with hospitalization [1]. In this study, outcomes such as hospitalization were given higher importance due to the disruption and inconvenience of daily living. Similarly, in a study of over 4,000 HD patients, reduction in hospital stays was among the highest priority outcomes when asked to rate the importance of 23 patient-relevant outcomes [2]. Finally, the ESRD Networks that are charged with helping dialysis facilities improve quality of care have reduction in hospitalizations as part of the statement of work. These Networks have Patient Advisory Committees that provide input and peer to peer communication to help reduce hospitalizations.
References:
[1] Manera KE, Johnson DW, Craig JC, Shen JI, Ruiz L, Wang AY, Yip T, Fung SKS, Tong M, Lee A, Cho Y, Viecelli AK, Sautenet B, Teixeira-Pinto A, Brown EA, Brunier G, Dong J, Dunning T, Mehrotra R, Naicker S, Pecoits-Filho R, Perl J, Wilkie M, Tong A. Patient and Caregiver Priorities for Outcomes in Peritoneal Dialysis: Multinational Nominal Group Technique Study. Clin J Am Soc Nephrol. 2019 Jan 7;14(1):74-83. doi: 10.2215/CJN.05380518. Epub 2018 Dec 20. PMID: 30573659; PMCID: PMC6364541.
[2] Janssen IM, Scheibler F, Gerhardus A. Importance of hemodialysis-related outcomes: comparison of ratings by a self-help group, clinicians, and health technology assessment authors with those by a large reference group of patients. Patient Prefer Adherence. 2016 Dec 13;10:2491-2500. doi: 10.2147/PPA.S122319. PMID: 28008235; PMCID: PMC5171198.
Performance Gap
Data for Table 1 are from the data described in 1.25 for the year 2023. The total number of dialysis facilities included in the performance score calculation was 7,710, and the total number of index discharges was 460,594.
See 2.4a for SRR_Table 1_Final_508.pdf attachment for table and text
Care Gaps
Closing Care Gaps
This domain is optional for the Spring 2026 cycle.
Feasibility
Feasibility
All the data incorporated into our database come from structured data. Data collection for this measure is accomplished via data sources including EQRS, a web-based and electronic batch submission platform maintained and operated by CMS contractors, Medicare Claims, and other supplemental data sources (see Section 1.25 Data Source Details). Publicly reported measures like this one are reviewed on a regular basis by dialysis facility providers and rare instances of inaccurate or missing data are present (based on comments received during facility previews).
As the data required for this measure is already part of routine data collection, no additional costs or burden are anticipated.
Public reporting of this measure on DFCC would be restricted to facilities with at least 11 eligible index discharges to comply with restrictions on reporting of potentially identifiable patient information related to small cell size.
No feasibility challenges have been identified that resulted in a change to the measure. We have expanded our inclusion of Medicare Advantage patients in identifying index discharges, readmissions, and comorbidities. All other measure specifications have remained unchanged since the previous endorsement cycle. The feasibility profile is not affected by the changes made.
Proprietary Information
Scientific Acceptability
Testing Data
EQRS data from January-December 2023 and Medicare claims data from January-December 2023. See section 1.25 for more detail about data sources.
EQRS data from January-December 2023 and Medicare claims data from January-December 2023. See section 1.25 for more detail about data sources.
None
In 2023 there were 7,252 facilities included in the measure. Facilities had a median caseload of around 29 eligible patients and median eligible index discharges of 55.
Number of facilities and median facility size, 2023
| Year | Total Facilities | Total Index Discharges | Median Index Discharges Per Facility | Median Patients Per Facility |
| 2023 | 7,252 | 457,965 | 55 | 29 |
See 7.1 Supplemental Attachment section for SRR_5.1.4_Final_508.pdf, which contains the text and tables for this question
Reliability
A key metric for SRR is the inter-unit reliability (IUR), which quantifies the proportion of total variation in a measure that is attributable to true differences between facilities, rather than to random variation. By definition, IUR ranges from 0 to 1, with higher values indicating that most of the observed variation in the quality measure reflects actual differences in facility performance—thereby implying higher precision in comparing facilities.
However, due to the ratio form of SRR, directly estimating the within-facility variance is not straightforward. We use a bootstrap-based approach to estimate this component of variability.
Let T1,…,TN represent the SRR values for N facilities. For each facility i with ni observations, we draw B bootstrap samples with replacement from its patients (we found B=100 to be sufficient based on numerical experiments). For each sample, we compute the corresponding bootstrapped SRRs, denoted of T*i,1, … T*i,B. We then compute the sample variance of these bootstrapped SRRs for each facility, denoted Si*2.
An estimate of the within-facility variance of SRR, namely, σ t,w2, is given by the bootstrap variance:
St,w2= ΣNi=1[(ni-1) S i*2]/ ΣNi=1(ni-1).
Calling on formulas from the one-way analysis of variance, an estimate of the overall variance of Ti is
St2= ΣNi=1[ni (Ti -Ť)2]/ [n’(N-1)],
where
Ť = Sni Ti / Sni
is the weighted mean of the observed SRR and
n’ = (Sni - Sni2/Sni)/(N-1)
is approximately the average facility size (number of observations per facility). Note that St2 is the total variation of SRR and is an estimate of σ b2 + σ t,w2, where σ b2is the between-facility variance, the true signal reflecting the differences across facilities. Thus, the estimated IUR, which is defined by
IUR = σ b2 /( σ b2 + σ t,w2),
can be estimated with (St2- St,w2)/St2.
Note: SRR calculations were restricted to facilities with at least 11 index discharges to ensure stable estimates and comply with restrictions on reporting of potentially identifiable patient information related to small cell size.
Data Element Reliability
EQRS Data: Some data for this measure comes from the End Stage Renal Disease Quality Reporting System (EQRS), a CMS-owned data system that collects data directly from all Medicare-certified dialysis facilities. EQRS has processes in place [1] to ensure the reliability and validity of the patient level data used for a broad array of measure calculations, including this measure. Briefly, CMS performs a random selection of 300 eligible dialysis facilities each year. Ten patient records are randomly selected from a single quarter each year from each of the facilities selected to participate. The most recent reported review included EQRS entries from April 1, 2025 to June 30, 2025. Experienced nurse reviewers assessed the data obtained from the medical records on each of 60 data elements selected from EQRS for the reporting month.
- Medicare Claims Data: CMS claims data are routinely used for quality measures that UM-KECC and other measure developers have crafted and have long been considered reliable. CMS routinely assesses the accuracy of claims codes through auditing programs as part of an effort to ensure appropriate billing by providers in both the Fee-for-Service and Medicare Advantage programs [2]. In addition, CMS conducts data analysis to identify potential problem areas and detect fraud, and audits important data fields used in our measures, including diagnosis and procedure codes and other elements that are consequential to payment.
References:
1. End Stage Renal Disease Facility Data Validation. CMS QualityNet. https://qualitynet.cms.gov/files/68d58e4e9fb3148bd3307ea6?filename=2025_EQRS_ExecSummary.pdf (accessed 5/12/2026)
2. Center for Medicare and Medicaid Services, FFS and Part C audits: https://www.cms.gov/data-research/monitoring-programs/medicare-fee-service-compliance-programs/medicare-fee-service-recovery-audit-program
https://www.cms.gov/medicare/audits-compliance/part-c-d/program-audits
The IUR for SRR in 2023 is 0.336, which means that one third of the variation can be attributed to between-facility variation. The SRR measure IUR is similar to previous cycles, and has been endorsed and re-endorsed for the last several cycles. Please see the SRR_Table 2a_Table 2b_IUR Info_Final_508.pdf attachment in Section 5.2.3a for additional information but to summarize:
- Dialysis facilities are extremely small compared to other health care entities (e.g. hospitals, nursing homes) such that risk adjusted measures do not have a large enough facility size to achieve an IUR of 0.6
- Determining if a facility is “worse than expected” uses statistical hypothesis testing to mitigate the risk of inappropriately flagging small facilities. Specifically, smaller facilities need to have an SRR farther from the median to be flagged compared to larger facilities.
- Star Ratings for dialysis facilities combine information across multiple measures to reduce random noise so that even a measure with a low IUR can contribute to raising the overall reliability of the combined measure set.
- The Quality Incentive Program (QIP) uses a small-facility adjuster (generally applied to facilities with 25 or fewer eligible patients), which helps mitigate the low IURs that would otherwise contribute to payment reductions.
- The number of preventable events, even for facilities in the lower IUR decile groups, is generally >3, suggesting the potential for improvement at a given facility.
Data element reliability
- EQRS: Per the executive summary [1], the rate of correct matches between data extracted from medical records and the same data fields in EQRS was 97.0% for all data elements. A total of 1.7% of entries in either EQRS (.1%) or Medical Records (1.6%) contained missing information. The rate of discrepant comparisons (incorrect matches between data elements in the medical record and EQRS) was 1.3% in CY2024. Of note, this overall error rate has been steadily declining over the past 5 years for a rate of 4.9% in CY2020. Of the 60 data elements examined, error rates generally ranged from 0 - 2.6%
References:
1. End Stage Renal Disease Facility Data Validation. CMS QualityNet. https://qualitynet.cms.gov/files/68d58e4e9fb3148bd3307ea6?filename=2025_EQRS_ExecSummary.pdf (accessed 5/12/2026)
In terms of CMS Claims: Please see section 5.2.2
When stratified by facility size, we find that, as expected, larger facilities have greater IUR.
See Pages 2-5 of SRR_Table 2a_Table 2b_IUR Info_Final_508.pdf (attached to 5.2.3a) for additional information about SRR IUR.
Data Element Reliability
- The data element reliability is also quite strong with 97% of data elements in EQRS correctly matching the same elements in the medical records. Missing data and data errors in EQRS are very rare. Of note, EQRS data are a primary source for the federal ESRD Quality Incentive Program, a value-based purchasing program. This measure has been reported in the ESRD QIP for many years. In addition, both EQRS and Medicare claims are used in the federally funded ESRD registry database (United States Renal Data System) that includes all patients who are on dialysis in the US. As such, they are tested for reliability for use in these federal programs and are considered highly reliable based on that testing.
See 5.2.3a for SRR_Table 2a_Table 2b_IUR Info_Final_508.pdf, which contains the table and text for this question
See 5.2.3a for SRR_Table 2a_Table 2b_IUR Info_Final_508.pdf, which contains the table and text for this question
Validity
To validate SRR, we first stratified facilities into the ‘better than expected’, ’as expected’, and ‘worse than expected’ categories of SRR. Next, we calculated mean performance scores for several quality measures that are expected to be clinically associated with 30-day readmissions: Standardized Mortality Ratio (SMR), Standardized Hospitalization Ratio (SHR), Standardized Transfusion Ratio (STrR), Standardized Fistula Rate (SFR), and Long-term Catheter (LTC). We then compared mean performance scores across the three strata of ‘better than expected’, ‘as expected’, and ‘worse than expected’ categories for SRR.
We expect better mean performance on the above quality measures for facilities classified as ‘better than/as expected’ for SRR compared to facilities classified as ‘worse than expected’, with the exception of LTC, where lower mean performance is expected. Compared to facilities that perform ‘worse than expected’, facilities that perform ‘better than/as expected’ on SRR are likely to have more successful care coordination and other processes of care in place that may help patients avoid a readmission visit in the vulnerable period following a recent discharge.
Standardized Mortality Ratio (SMR): We expect to observe a lower mean standardized mortality ratio for facilities in the ‘better than/as expected’ categories for SRR compared to facilities classified as ‘worse than expected.’ Patients who require acute inpatient medical care represent an at-risk population for mortality since they likely have greater acute medical needs or complications from chronic comorbid conditions that put them at higher risk for death.
Standardized Hospitalization Ratio (SHR): We expect to observe a lower mean standardized hospitalization ratio for facilities in the ‘better than/as expected’ categories for SRR compared to facilities classified as ‘worse than expected.’ We expect this trend to be fairly strong with SHR since readmissions are also hospital admissions. Additionally, both hospitalization and readmission are a reflection of hospital utilization and increased comorbidity burden.
STrR: We expect to observe a lower mean standardized transfusion event ratio for facilities in the ‘better than/as expected’ categories for SRR compared to facilities classified as ‘worse than expected.’ Facilities that have a lower STrR likely have processes of care in place to support robust anemia management and other care processes compared to facilities with a higher STrR.
Vascular Access: Standardized Fistula Rate (SFR) – We expect to observe a higher mean standardized fistula rate for facilities in the ‘better than/as expected’ category for SRR compared to facilities classified as ‘worse than expected.’ Successfully creating an AVF is generally seen as representing a robust process to coordinate care outside of the dialysis facility, and potentially reduces the likelihood of adverse events, like infection that can increase the risk of patient hospitalization and hospital readmission. Facilities that do a better job at care coordination reduce the likelihood that patients will experience re-hospitalization as measured by SRR.
- Vascular Access: Long-term catheter rate (catheter in use >=3 continuous months) – We expect to observe a lower mean catheter rate for facilities in the ‘better than/as expected’ categories for SRR compared to facilities classified as ‘worse than expected.’ Long-term catheters put patients at increased risk for infection and other complications. Additionally, a high long-term catheter rate also indicates a higher patient comorbidity burden at the facility level such that sicker patients who have a long-term catheter may also be more likely to be hospitalized and re-admitted after initial hospitalization.
Data element validity
- EQRS: Some data for this measure comes from the End Stage Renal Disease Quality Reporting System (EQRS), a CMS-owned data system that collects data directly from all Medicare-certified dialysis facilities. EQRS has processes in place [1] to ensure the reliability and validity of the patient level data used for a broad array of measure calculations, including this measure. Briefly, CMS performs a random selection of 300 eligible dialysis facilities each year. Ten patient records are randomly selected from a single quarter each year from each of the facilities selected to participate. The most recent reported review included EQRS entries from April 1, 2025 to June 30, 2025. Experienced nurse reviewers assessed the data obtained from the medical records on each of 60 data elements selected from EQRS, including the dialysis modality type for the reporting month.
- Medicare Claims: UM-KECC is unable to perform validity testing on claims-based measures. However, the data elements used to specify the measure are structured, used for reimbursement, and audited. Claims data have long been considered to be valid and reliable for use in quality measurement. In addition, there are examples of CBE-endorsed measures used for public reporting that have been validated with models using chart-abstracted data for risk adjustment [2]. In general, the risk-standardized rates using the claims-based approach had a high level of agreement with results based on models using medical records, supporting the use of claim-based models for quality measurement used in public reporting.
References:
1. End Stage Renal Disease Facility Data Validation. CMS QualityNet. https://qualitynet.cms.gov/files/68d58e4e9fb3148bd3307ea6?filename=2025_EQRS_ExecSummary.pdf (accessed 5/12/2026)
2. Bratzler DW, Normand SL, Wang Y, O'Donnell WJ, Metersky M, Han LF, Rapp MT, Krumholz HM. An administrative claims model for profiling hospital 30-day mortality rates for pneumonia patients. PLoS One. 2011 Apr 12;6(4):e17401. doi: 10.1371/journal.pone.0017401. PMID: 21532758; PMCID: PMC3075250.
See 5.3.4a for SRR_5.3.4_Final_508.pdf, which contains the text and table for this question
On average the standardized mortality ratio was 6% higher than the national average for facilities that were ‘worse than expected,’ and 7% lower than the national average (SMR = 0.93) for facilities that were ‘better than’ expected for SRR.
Facilities classified as ‘better than expected’ for SRR performed, on average, 21% better than the national average on hospitalization rates (SHR = 0.79) while those classified as ‘worse than expected’ performed, on average, 32% worse than the national average (SHR = 1.32).
On average, the standardized transfusion event ratio was 30% higher than the national average for facilities classified as ‘worse than expected’ while the ‘better than expected’ classification group of facilities were 24% lower than the national average. This suggests that facilities which have lower numbers of transfusion events likely have better processes of care in place to support robust anemia management and other care processes, thus reducing the need for re-hospitalization.
Overall, the average SFR was 58.7% in facilities classified as ‘better than expected’ and 54.43% in facilities classified as ‘worse than expected.’ The results reinforce the observation that patients with AVFs have lower risk of infection and potential need for acute care or hospitalization compared to patients with other access types, such as long-term catheter. Higher facility standardized fistula rates suggests facilities may be doing a better job at care coordination, reducing many acute care needs necessitating re-hospitalization. While the difference in fistula rates was small between facilities, this may reflect the fact that national trends in AVF rates have generally plateaued across many US dialysis facilities.
The mean LTC in facilities classified as ‘better than expected’ was 18.07% compared to facilities classified as ‘worse than expected’ (20.45%), suggesting that facilities that have lower rates of patients dialyzing with a catheter likely have reduced infection risk and other patient comorbidity burden which, in turn, reduce the risk of readmission after initial hospitalization.
Taken together these results provide validation support for SRR. Performance on key quality measures that were expected to be related to hospital readmission was also related to facility flagging in the respective ‘better than expected’ or ‘worse than expected’ categories.
The data element validity is also quite strong with 97% of data elements in EQRS correctly matching the same elements in the medical records. Missing data and data errors in EQRS are very rare. Of note, EQRS data are a primary source for the federal ESRD Quality Incentive Program, a value-based purchasing program. This measure has been reported in the ESRD QIP for many years. In addition, both EQRS and Medicare claims are used in the federally funded ESRD registry database (United States Renal Data System) that includes all patients who are on dialysis in the US. As such, they are tested for validity for use in these federal programs and are considered highly valid based on that testing.
Risk Adjustment
To estimate the probability of 30-day unplanned readmission, we use a three-stage model, the first of which is a fixed-effects logistic regression model. In this step, facility-hospital combinations are included as fixed effects, adjusting for a set of patient-level characteristics. The results of this step are estimates of the regression coefficients of patient-level characteristics in the logistic regression model. These estimates avoid issues of bias that arise through estimation of regression coefficients in a model with random effects. In particular, these estimates are unbiased regardless of correlations between hospital effects or facility effects and patient-case mix. These estimated regression coefficients are then used as an offset variable in the second stage model.
The next stage is a double random-effects logistic regression model. In this stage of the model, both dialysis facilities and hospitals are represented as random effects, and the sum of regression adjustments multiplied by estimated parameters obtained in the first stage is included as the offset variable. From this model, we obtain the estimated standard deviation of the random effects of hospitals [13].
The third stage of the model is a mixed-effects logistic regression model, in which dialysis facilities are modeled as fixed effects and hospitals are modeled as random effects, with the standard deviation specified as equal to its estimate from the second-stage model and the estimated parameters obtained in the first stage providing an offset. The expected number of readmissions for each facility is estimated as the sum of the probabilities of readmission of all index discharges in this facility and assuming the national norm (i.e., the median) for the facility effect. This model accounts for a given facility’s case mix using the same set of patient-level characteristics as those in the first model.
The model and methods are described in some additional detail below:
To estimate the probability of 30-day unplanned readmission following an index discharge, we use a three-stage approach. The main model, which produces the estimates used to calculate SRR, takes the form:
log {pijk/(1-pijk)}= ri + αj + βT Zijk, (1)
Where pijk represents the probability of an unplanned readmission for the kth discharge among patients who are discharged from jth hospital to the ith facility, and Zijk represents the set of patient-level characteristics. Here, ri is the fixed effect for facility and αj is the random effect for hospital j. It is assumed that the αjs arise as independent normal variables (i.e., αj ~ N(0, σ2)).
We then use the estimates from this model to calculate each facility’s SRR:
SRRi = Oi / Ei = Oi / ƩjϵH(i) Ʃk Ṕijk, (2)
where, for the ith facility, Oi is the number of observed unplanned readmissions, Ei is the expected number of unplanned readmissions for discharges, H(i) is the collection of indices of hospitals from which patients are discharged, and Ṕijk is the predicted probability of unplanned readmission under the national norm for each discharge. Specifically, Ṕijk takes the form
Ṕijk = exp( řM+ᾰj + ḂT Zijk )/{1+exp(řM+ᾰj + ḂT Zijk)}, (3)
which estimates the probability that a discharge from hospital j of an individual in facility i with characteristics Zijk would result in an unplanned readmission if the facility effect corresponded to the median of national facility effects, denoted by řM. Here, ᾰj and Ḃ are estimates from model (1). The sum of these probabilities is the expected number of unplanned readmissions Ei at facility i; e.g., the number of readmissions that would have been expected in facility i had they progressed to the readmissions at the same rate as the national population of dialysis patients.
Patient-Level Risk Adjustors
As mentioned previously, the model accounts for a set of patient-level characteristics:
- Sex
- Age
- Years on dialysis
- Medicare Advantage status at discharge
- Nursing home status in past year at discharge
- None (0 days)
- Short term (<90 days)
- Long term (>=90 days)
- Diabetes as cause of ESRD
- Interaction of age and diabetes as cause of ESRD
- BMI at incidence of ESRD*
- < 18.5
- 18.5 - 24.9
- 25-29.9
- 30+
*missing included with the 30+ group
- Length (days) of index hospitalization
- Past-year comorbidities: We identify all unique ICD-10 diagnosis codes from each patient’s prior year of Medicare inpatient claims. We group these diagnosis codes by diagnosis area using the v2019.1 Agency for Healthcare Research and Quality (AHRQ) Clinical Classifications Software (CCS) diagnosis categories. The CCS diagnosis categories used in calculation of the SRR are:
- CCS 6: Hepatitis
- CCS 10: Immunizations and screening for infectious disease
- CCS 42: Secondary malignancies
- CCS 50: Diabetes mellitus with complications
- CCS 51: Other endocrine disorders
- CCS 52: Nutritional deficiencies
- CCS 55: Fluid and electrolyte disorders
- CCS 59: Deficiency and other anemia
- CCS 64: Other hematologic conditions
- CCS 95: Other nervous system disorders
- CCS 96: Heart valve disorders
CCS 97: Peri-; endo-; and myocarditis; cardiomyopathy (except that caused by
tuberculosis or sexually transmitted disease)
- CCS 100: Acute myocardial infarction
- CCS 101: Coronary atherosclerosis and other heart disease
- CCS 102: Nonspecific chest pain
- CCS 106: Cardiac dysrhythmias
- CCS 107: Cardiac arrest and ventricular fibrillation
- CCS 108: Congestive heart failure; nonhypertensive:
- CCS 117: Other circulatory disease
- CCS 118: Phlebitis; thrombophlebitis and thromboembolism
- CCS 120: Hemorrhoids
- CCS 121: Other diseases of veins and lymphatics
CCS 122: Pneumonia (except that caused by tuberculosis or sexually transmitted
disease)
- CCS 127: Chronic obstructive pulmonary disease and bronchiectasis
- CCS 130: Pleurisy; pneumothorax; pulmonary collapse
- CCS 131: Respiratory failure; insufficiency; arrest (adult)
- CCS 133: Other lower respiratory disease
- CCS 134: Other upper respiratory disease
- CCS 135: Intestinal infection
- CCS 138: Esophageal disorders
- CCS 140: Gastritis and duodenitis
- CCS 141: Other disorders of stomach and duodenum
- CCS 151: Other liver diseases
- CCS 152: Pancreatic disorders (not diabetes)
- CCS 153: Gastrointestinal hemorrhage
- CCS 154: Noninfectious gastroenteritis
- CCS 155: Other gastrointestinal disorders
- CCS 158: Chronic kidney disease
- CCS 159: Urinary tract infections
- CCS 197: Skin and subcutaneous tissue infections
- CCS 198: Other inflammatory condition of skin
- CCS 199: Chronic ulcer of skin
- CCS 201: Infective arthritis and osteomyelitis (except that caused by tuberculosis or sexually transmitted disease)
- CCS 237: Complication of device; implant or graft
- CCS 244: Other injuries and conditions due to external causes
- CCS 251: Abdominal pain
- CCS 253: Allergic reactions
- CCS 255: Administrative/social admission
- CCS 259: Residual codes; unclassified
- CCS 651: Anxiety disorders
- CCS 659: Schizophrenia and other psychotic disorders
- CCS 660: Alcohol-related disorders
- CCS 661: Substance-related disorders
Discharged with high-risk condition: We define a high-risk diagnosis as any diagnosis area that was rare in our population but had a 30-day readmission rate of at least 40%. We did not include high-risk diagnosis groups related to cancer or mental health. We group these conditions using the Agency for Healthcare Research and Quality (AHRQ) Clinical Classifications Software (CCS). The CCS areas identified as high-risk are:
- CCS 5: HIV infection
- CCS 6: Hepatitis
- CCS 56: Cystic fibrosis
- CCS 57: Immunity disorders
- CCS 61: Sickle cell anemia
- CCS 190: Fetal distress and abnormal forces of labor
- CCS 151: Other liver diseases
- CCS 182: Hemorrhage during pregnancy; abruptio placenta; placenta previa
- CCS 186: Diabetes or abnormal glucose tolerance complicating pregnancy; childbirth;
- CCS 210: Systemic lupus erythematosus and connective tissue disorders
- CCS 243: Poisoning by nonmedicinal substances
or the puerperium
The list of 53 past-year comorbidity variables were selected from 233 indicators of AHRQ CCS diagnosis categories with prevalence greater than 0.1% using a score-test based sample splitting forward selection approach. In particular, the data sample is randomly split into two halves. The first half is used for fitting a first-stage fixed effects logistic regression model to select a set of comorbidity variables via a forward selection scheme using single variable score tests with 0.01 p-value cutoff and adjusting for patient-level characteristics such as age splines, sex, BMI, etc. The second half is then used to fit another first-stage model adjusting for patient-level risk factors as well as those selected variables using the first-half data sample. Single variable score tests are performed after model fitting to obtain p-values for selected variables. A common p-value of 1 is assigned to unselected variables using the first-half data sample. The steps above are repeated 50 times to generate 50 sets of p-values for all 233 variables. The 50 p-values of each variable are aggregated following Bühlmann and van de Geer and the 53 prevalent comorbidities with aggregated p-values less than 0.01 are selected.
Finally, the relationship between patient level SDS, socioeconomic disadvantage and health care utilization such as hospitalization is well-established in the general population and has received considerable attention over the years [2-6]. The likelihood of hospitalization is related to socioeconomic disadvantage through differences in health status, insurance coverage, and access to quality primary care [8, 9]. Further, individual and market or area-level measures of deprivation have been shown to contribute independently to preventable hospitalizations [20].
Health care outcomes and utilization are associated with area-level income and residential segregation, but particularly so for racial minorities [24, 25]. This suggests the interplay of patient level (race) and area level SES factors related to lower income, neighborhood poverty, segregation, levels of educational attainment, and unemployment levels that jointly influence key health outcomes related to morbidity [1, 24, 25].
Within the dialysis population area-level SES are associated with poor outcomes [7]; while patient level factors such as race are predictive of differences in certain clinical outcomes by race [26, 28]. In a study of first year hemodialysis patients, patients of Hispanic ethnicity had lowest all-cause hospital length of stay compared to whites, while patients of black race had intermediate all-cause hospital admissions that was lower relative to whites but higher than Hispanic patient, with differences observed across certain age groups [28]. Moreover, the study authors found that infection-related hospitalizations were significantly higher for black and Hispanic patients compared to non-Hispanic whites. These associations could indicate certain facility level practices related to effective infection control and prevention may unevenly impact patients of black race and Hispanic ethnicity [28].
Insurance status is also related to health outcomes but this has not been studied extensively within the dialysis population as it relates to hospitalization, though the association has been documented in studies of the general dual Medicare and Medicaid population. Dual eligibility typically confers greater comorbidity burden and access to care barriers which in turn drives higher hospital utilization [15, 19, 27].
Given these observed linkages we tested these patient- and area-level SDS/SES variables based on the conceptual relationships as described above and demonstrated in the literature, as well as the availability of data for the analyses. In total, we tested the following variables:
Patient level:
- Sex
- Race
- Ethnicity
- Medicare dual eligible
- ZIP code level – Area Deprivation Index (ADI) from Census data (2009-2013). Based on patient zip-code. We use the publicly available Area Deprivation Index (ADI) originally developed by Singh and colleagues at the University of Wisconsin. We applied the updated ADI based on 2009-2013 census data [23]. The ADI reflects a full set of SES characteristics, including measures of income, education, and employment status, measured at the ZIP code level.
References:
[1] Agency for Healthcare Research and Quality, Rockville, MD. Internet Citation: Chapter 3: Creation of New Race-Ethnicity Codes and SES Indicators for Medicare Beneficiaries - Chapter 3. January 2008. Publication # 08-0029-EF. http://archive.ahrq.gov/research/findings/final-reports/medicareindicators/medicareindicators3.html
[2] Agency for Healthcare Research and Quality (AHRQ). 2010 National Health Care Disparities Report. Washington, DC: AHRQ; 2011).
[3] Agency for Healthcare Research and Quality (AHRQ). 2011 National Health Care Disparities Report. Washington, DC: AHRQ; 2012).
[4] Agency for Healthcare Research and Quality (AHRQ). 2012 National Health Care Disparities Report. Washington, DC: AHRQ; Reports: 2013).
[5] Agency for Healthcare Research and Quality (AHRQ). 2013 National Health Care Disparities Report. Washington, DC: AHRQ; Reports: 2014).
[6] Agency for Healthcare Research and Quality (AHRQ). 2014 National Health Care Disparities Report. Washington, DC: AHRQ; 2015).
[7] Almachraki, F., Tuffli, M., Lee, P., Desmarais, M., Shih, H.C., Nissenson, A., and Krishnan, M. Population Health Management. Volume 19, Number 1, 2016.
[8] Basu, J., Thumula, V., and Mobley, L.R. (2012, July-September). Changes In Preventable Hospitalization Patterns Among Adults. A Small Area Analysis of U.S. States. Journal of Ambulatory Care Management 35(3), pp.3280-3290.
[9] Blustein, J., Hanson, K., and Shea, S. Preventable Hospitalizations and Socioeconomic Status. Health Affairs 17, no.2 (1998):177-189).
[10] Bühlmann, P. and van de Geer, S. (2011). Statistics for High-Dimensional Data: Methods, Theory and Applications. Springer.
[11] Cox DR: Regression models and life tables (with discussion). JRStat Soc [SerB]34: 187–220, 1972.
[12] Curtin R., Oberley E., Sacksteder P., and Friedman A. Differences Between Employed and Nonemployed Dialysis Patients. AJKD Vol 27:4. (April) 1996. 533-540.
[13] Diggle, P.J., Heagerty, P., Liang, K-Y., Zeger, S.L. Analysis of Longitudinal Data. 2 New York: Oxford Univ. Press; 2002.
[14] He, K., Kalbfleisch, J.D., Li, Y., Li, Y. “Evaluating readmission rates in dialysis facilities with or without adjustment for hospital effects.” Unpublished manuscript. 2012.
[15] Jiang, H., Wier, L., Potter, D.E.B., Burgess, J. AHRQ Statistical Brief #96 Potentially Preventable Hospitalizations among Medicare-Medicaid Dual Eligibles, September 2010.
[16] Kalbfleisch J.D., Prentice R.L.: The statistical analysis of failure time data, Hoboken, New Jersey, John Wiley & Sons, Inc., 2002.
[17] Lawless, J.F. and Nadeau, C. Some Simple Robust Methods for the Analysis of Recurrent Events Technometrics. Vol. 37, No. 2 (May, 1995), pp. 158-168.
[18] Liu, D., Schaubel, D.E. and Kalbfleisch, J.D. (2012). Computationally efficient marginal models for clustered recurrent event data. Biometrics, 68, 637-647.
[19] Moon S., Shin J. BMC Public Health. 2006 Apr 5;6:88. Health Care Utilization Among Medicare-Medicaid Dual Eligibles: A Count Data Analysis.
[20] Moy E., Chang E., and Barrett M. Potentially Preventable Hospitalizations — United States, 2001–2009. CDC Morbidity and Mortality Weekly Report (MMWR). Supplements November 22, 2013 / 62(03);139-143.
[21] Rodriguez R., Sen S., Mehta K., Moody-Ayers S., Bacchetti P., and O’Hare A. Geography Matters: Relationships among Urban Residential Segregation, Dialysis Facilities, and Patient Outcomes. Ann Intern Med. 2007; 146:493-501.
[22] Singh, G.K. Area Deprivation and Widening Inequalities in US Mortality, 1969–1998. Am J Public Health. 2003; 93(7):1137–1143.
[23] University of Wisconsin School of Medicine Public Health. 2015 Area Deprivation Index v2.0. Downloaded from https://www.neighborhoodatlas.medicine.wisc.edu10/31/2018.
[24] Williams, D. “Race, Socioeconomic Status, and Health: The Added Effects of Racism and Discrimination. Annals of the New York Academy of Sciences. Volume 896, Issue 1, Article first published online: 6 Feb 2006.
[25] Williams, D. and Collins, C. Racial Residential Segregation: A Fundamental Cause of Racial Disparities in Health. Public Health Reports / September–October 2001. Volume 116. 404-416.
[26] Whittle, J.C., Whelton, P.K., Seidler, A.J., and Klag, M.J. Does Racial Variation In Risk Factors Explain Black-White Differences In The Incidence Of Hypertensive End-Stage Renal Disease? Arch Intern Med. 1991 Jul;151(7):1359-64.
[27] Wright, B., Potter, A., and Trivedi, A. Federally Qualified Health Center Use Among Dual Eligibles: Rates of Hospitalizations And Emergency Department Visits Health Affairs, 34, no.7 (2015):1147-1155.
[28] Yan, G., Norris, K., Greene, T., Yu, A., Ma, J., Yu, W., and Cheung, A. Race/Ethnicity, Age, and Risk of Hospital Admission and Length of Stay during the First Year of Maintenance Hemodialysis. Clin J Am Soc Nephrol 9: epub (June), 2014.
See 5.4.3a for SRR_5.4.3_Final_508.pdf, which contains the table and text for this question
See 5.4.4a for SRR 5.4.4_Final_508.pdf, which contains the table and text for this question
The model's ability to distinguish between patients who will and will not have a hospital readmission within 4-30 days was measured using the Area Under the Receiver Operating Characteristic (AUC) curve. The predicted AUC value is 0.683, which indicates the model has fair discriminatory power. This means the model is effective at differentiating between patients with higher and lower risk of a hospital readmission. Specifically, if a patient who was readmitted to the hospital and a patient who was not are randomly selected, the model will correctly identify which patient was readmitted 68.3% of the time.
See 5.4.5a for calibration and discrimination testing results, found in SRR_5.4.5a_Final_508.pdf.
While the inclusion of some patient SES characteristics such as Hispanic, Medicare Dual Eligible, and Asian and Black race are significant, other patient SES characteristics are not (Area Deprivation Index, American Indian or Alaskan Native, and Other). Race, ethnicity, dual eligible status, and area deprivation are not included in the final risk adjusted model. Other studies have reported associations between patient-level race, ethnicity, dual eligible status, neighborhood deprivation and acute care utilization, however it is unclear whether these differences are due to underlying biological or other patient factors, or represent disparities in care. Adjusting for these social risk factors could have the unintended consequence of creating or reinforcing disparities and limiting access to care. The primary goal should be to implement quality measures that result in the highest quality of patient care and equitable access for all patients.
Use & Usability
Use
Dialysis Facility Care Compare (DFCC) helps patients find detailed information about Medicare-certified dialysis facilities. They can compare the services and the quality of care that facilities provide.
United States. All Medicare-certified dialysis facilities who are eligible for the measure and have at least 11 index discharges are included in the measure calculation for the program. For the October 2024 Dialysis Facility Compare refresh, SRR results were reported for 458,111 index discharges in 7,250 U.S. dialysis facilities.
All Medicare-certified dialysis facilities who are eligible for the measure and have at least 11 index discharges are included in the measure calculation for the program.
The ESRD QIP will reduce payments to ESRD facilities that do not meet or exceed certain performance standards. The measure was added to the program for PY2017.
United States.
Patients/index discharges included: All patients/index discharges who meet the requirements to be included in the measure from included facilities. Patient counts could not be included here as they were not available in this program’s public use files.
For the most recent QIP report that is publicly available (PY 2026), 7,207 facilities received a measure score.
All Medicare-certified dialysis facilities that are eligible for the measure and have at least 11 patient years at risk (due to public reporting requirements).
The Dialysis Facility Reports (DFRs) are provided as a resource for characterizing selected aspects of clinical experience at this facility relative to other caregivers in this state, End Stage Renal Disease (ESRD) Network, and across the United States. Since these data could be useful in quality improvement and assurance activities, each state’s surveying agency may utilize the DFRs as a resource during their survey and certification process. Measures included in the DFRs are updated annually and available to dialysis facilities to review and submit comments prior to their release to State Survey Agencies and Regional Offices in September of each year.
United States. All Medicare-certified dialysis facilities who are eligible for the measure and have at least 11 index discharges are included in the measure calculation for the program. For the FY 2025 Dialysis Facility Reports, SRR results were reported for 458,111 index discharges in 7,250 U.S. dialysis facilities.
Facility level, Dialysis Facilities
This measure is best suited for an accountability program that focuses on End Stage Renal Disease (ESRD) patients. Specifically, ESRD patients with Medicare coverage (either traditional Medicare or a Medicare Advantage Plan) would be the target population. Programs that focus on the dialysis facility as the accountable entity are ideal, as opposed to programs that focus on the Nephrologist or provider. As such, this is an outpatient measure with limited adjustments for social risk factors. However, additional adjustments could be made at the program level based on the needs or design of the program.
Usability
There are a number of actions that dialysis facility providers can take to help manage high risk patients and avoid preventable readmissions. Examples include:
- Optimize fluid management: Evaluation of the target weight after hospital discharge is important to avoid fluid related re-admissions. Encouraging patients to complete the full duration of their treatments along with not missing treatments (or rescheduling missed treatments) can be helpful in achieving optimal target weight.
- Medication reconciliation: this is especially important to avoid medication related adverse events.
- Education: Deliver education to patients about when to obtain hospital care vs. care at facility or by other providers and who to contact if questions arise between treatments.
- Coordination of care: facility staff can help ensure that follow up appointments are arranged post-discharge
- Anemia Management: appropriate use of erythrocyte stimulating agents can avoid the need for hospitalization for red blood cell transfusion.
These processes are directly tied to the required credentials and training facilities must have per the Conditions for Coverage and to be certified and in good standing with Medicare.
For DFCC, feedback can be provided any time through contacting the dialysisdata.org helpdesk. Preview periods allow for specific times for facilities to review and comment on measure calculations and provide an opportunity to request a list of patients included in the measure calculation.
Comments received during DFCC preview periods tend to be technical in nature, asking for clarification on how the SRR is calculated for particular facilities, including questions about patient assignment and application of risk adjustment criteria.
Below we have explained our response to the common questions we noted above.
Several comments questioned the use of both SHR and SRR which could doubly penalize facilities since a readmission would count in both the SHR and SRR measures. While the SHR and SRR may both count the same hospitalization event, we believe this is appropriate because it places additional emphasis on the importance of avoiding hospitalizations and, separately, re-hospitalization for dialysis patients. Doing so can help reduce this major cost driver as well as promote better patient health-related quality of life. In addition, while the SRR and SHR are moderately correlated with one another, it is possible for a facility to score relatively well on one measure, and relatively poorly on the other. We also believe that the measures capture distinct aspects of the quality of care provided by a dialysis facility. The SRR assesses the coordination of care during transitions as dialysis patients are discharged from an acute care hospital into the care of a dialysis facility, and the SHR evaluates the facility’s overall performance in reducing hospitalizations.
Based on enrollment information from the Medicare Enrollment Database (EDB), the percentage of ESRD dialysis beneficiaries enrolled in Medicare Advantage (MA) has steadily increased over time. From 12% in 2010, the proportion rose to 22% by 2020. Prior to 2020, there was an annual increase of approximately 1%. However, since 2021, the annual increase has been more than 5%.
The growth in ESRD beneficiaries joining MA plans carries significant implications for the metrics used to assess dialysis facility performance. Contrary to the data from Fee-For-Service (FFS) Medicare beneficiaries, MA outpatient encounters and administrative records have not been readily available for the purposes of analyzing facility quality, except for internal CMS use in risk adjustment and performance assessment.
See 7.1 Supplemental Attachment section for SRR_6.2.4_Final_508.pdf, which contains the table and text for this question
None
None
Comments
Staff Preliminary Assessment
CBE #5110 Staff Assessment
Importance
Strengths:
- A clear logic model is provided, depicting the relationships between inputs (e.g., Quality Improvement Staff and facility-specific Policies and Procedures that reflect requirements in CMS’ Conditions for Participation in the Medicare ESRD Chronic Dialysis Program) activities (e.g., identification of high-risk patients and root cause analysis for common emergency department (ED) visits, delivery of education to patients on when to visit the emergency department and who they can call if questions arise between visits), and desired outcomes (e.g., Reduce overall average of unplanned readmissions that occur between 4-30 days post discharge (SRR). This model demonstrates how the measure's implementation will lead to the anticipated outcomes.
- The problem this measure addresses presents a significant burden to dialysis patients, with between 31-33% of dialysis patient discharges from an all-cause hospitalization being followed by an unplanned readmission within 30 days in 2021 and 2022.
- Data from dialysis treatment facilities from 2023 show a performance gap, with decile ranges from 0.44 in the lowest decile to 1.53 in the highest decile, indicating variation in measure performance.
Limitations:
- Section 2.5 Health Care Quality Landscape: There is no description of other existing measures or programs, and no description of a search conducted to identify other existing measures or programs is provided.
- The literature review includes many studies and reports that are more than 5 years old. The submission could be strengthened by including more recent and higher quality empirical studies and more clearly linking each claim in the evidence review to the sources in the reference list.
- The literature supporting meaningfulness to patients was more than five years old. This submission could be strengthened by collecting patient input using other means, like technical expert panels (TEPs).
Rationale:
- This new measure is rated as 'Not Met but Addressable' for importance due to 1) dated evidence not addressing anticipated measure impact and 2) lack of information regarding potential competing measures.
Closing Care Gaps
The developer did not address this optional domain.
Feasibility Assessment
Strengths:
- All required data elements are routinely generated during care delivery, and required elements are available from digital or electronic sources.
- The developer stated there are no costs or burden were associated with data collection and data entry, validation, and analysis.
- The developer described how all required data elements can be collected without risk to patient confidentiality, specifically DFCC would be restricted to facilities with at least 11 eligible index discharges to comply with restrictions on reporting of potentially identifiable patient information related to small cell size.
- There are no fees, licensing, or other requirements to use any aspect of the measure (e.g., value/code set, risk model, programming code, algorithm).
Limitations:
- None
Rationale:
- This new measure meets all criteria for 'Met' for feasibility due to its well-documented feasibility assessment, clear and implementable data collection strategy, and transparent handling of patient confidentiality, burden, licensing, and fees. These factors collectively ensure that the measure can be implemented effectively and sustainably in a real-world healthcare setting.
Scientific Acceptability
Strengths:
- The developer performed the required reliability testing for this new measure by presenting existing evidence from an annual data validation report prepared for CMS and details of auditing performed on CMS claims data to support data element reliability. This report shows results for data entries from April 1, 2025, through June 30, 2025.
- According to the report, the data element used by this measure had an error rate less than 2.6%.
- Data sources used for reliability analysis are adequately described and include a CMS-owned data system which collects data directly from dialysis facilities.
Limitations:
- Note that accountable entity-level reliability testing is not required for initial endorsement, and is not considered in the rating.
Rationale:
- This new measure is rated as 'Met' for reliability because the developer provided the required evidence for this measure to demonstrate sufficient reliability at the data element-level.
Strengths:
- While not considered in the rating for validity because such testing is optional for initial endorsement, the developer also performed accountable entity level testing using data from the required timeframe.
- The developer conducted statistical risk adjustment, based on a conceptual model, selecting risk factors that are present at the start of care, have a significant association with the outcome, and vary in prevalence across measured entities. The model has acceptable calibration.
Limitations:
- The developer did not provide sufficient evidence for data element validity for all critical data elements (i.e., the numerator, denominator, exclusions), which is required for new measures. The developer cited data element validity testing performed on EQRS data, but the source they provided does not specify which data elements used in this measure were included in this testing. The developer also cited an article from 2011 that evaluated the claims-based approach for measuring 30-day mortality rates among patients with pneumonia, a different outcome and condition from the focus of this measure.
The developer indicated in the numerator statement that the numerator definition was based on the algorithm developed by Yale for the all-cause hospital wide readmission (HWR) measure and cited the 2018 measure update report. Additional information, for example, regarding how many of the data elements used in this measure were included in the HWR measure algorithm, should be provided.
Finally, The data dictionary appears incomplete. The submission would be strengthened by submitting a complete data dictionary. - The developer reported a c-statistic of 0.683, indicating moderate model discrimination.
Rationale:
- This maintenance measure is rated as ‘Not Met But Addressable’ for validity; data element validity testing results partially support an inference of validity, suggesting that the measure somewhat accurately reflects performance on quality and can distinguish good from poor performance to a limited extent.
- The risk adjustment methods used are appropriate and demonstrate variation in the prevalence of risk factors across measured entities. The model performance is acceptable.
Use and Usability
Strengths:
- The measure is currently used in the Dialysis Facility Care Compare (CMS), Dialysis Facility Reports (CMS) and ESRD QIP (CMS).
Attributes of a suitable program for this measure are described, and these include ESRD patients with Medicare coverage, programs that focus on dialysis facilities at the accountable entity level.
The developer provided a summary of how accountable entities can use the measure results to improve performance. Specifically, optimize fluid management, encouraging patients to complete the full duration of their treatments, medication reconciliation, patient education, coordination of care, anemia management.
Giving feedback is available at anytime via the dialysisdata.org helpdesk. The developer noted they have received questions on the use of both SHR and SRR as there are concerns this could doubly penalize facilities since a readmission would count in both the SHR and SRR measures. The developer notes that they believe this is appropriate based on measure focus, the possibility of different facility scores.
The developer reported no unexpected findings or unintended consequences.
Limitations:
- As a new measure, the developer is not required to provide performance trend data. However, the data that was shared shows that readmission rates have largely remained steady from 2014 to 2023. This limitation does not affect the rating
Rationale:
- This maintenance measure is rated ‘Met’ for use and usability because it is actively used in at least one accountability application, with a systematic feedback approach that allows for continuous updates based on stakeholder feedback.
Committee Independent Review
(No subject)
Scientific Acceptability
the staff assessment is that because this is being presented as a new measure, although an improvement on a prior measure, entity level reliability testing is not required. However, the entity level reliability measures are far below what has been considered acceptable, and this is not truly a new measure. This is not acceptable.
The developers report variations in the outcomes and some factors likely to be associated with performance for facilities characterized as above and below expected, but I am not seeing the actual number of facilities classified as above and below presented in the documentation. How large are these groups?
#5110 Standardized Readmission Ratio for Dialysis Facilities
Importance
Clearly stated evidence of the issue and importance to measure.
Closing Care Gaps
N/A
Feasibility Assessment
Measure uses existing data.
Scientific Acceptability
Sufficient and appropriate reliability testing for a new measure.
Agree with staff preliminary assessment that rationale or evidence is missing on the conceptual model for risk adjustment. This could also be strengthened by highlighting how this builds on the previously cited measure with data from prior measure.
Use and Usability
Already in use.
Summary
Well structured and specified measure with some addressable gaps. Namely, providing a conceptual framework and rationale for selected risk adjustment variables.
#5110 - SRR for dialysis facilities
Importance
I find this measure, as presented, lacking current research and patient feedback. Would be helpful to see more recent patient feedback.
Measure is using data from 2021-2022 - period when all care was significantly impacted by COVID-related shifts in access to care.
Closing Care Gaps
optional/not required
Feasibility Assessment
Scientific Acceptability
I find evidence for validity to be insufficient. Some of the materials/articles cited are not related to the subject matter/outcomes of #5110. Also noted lack of clarity with the data dictionary, that should be expanded/completed.
Use and Usability
Summary
While meeting feasibility and usability, would like to see more recent/applicable research and recent patient feedback. Data dictionary should be revised/completed.
The measure assumes dialysis…
Importance
While unplanned readmissions remain clinically important and costly, the measure submission places substantial emphasis on the overall burden of readmissions but provides more limited evidence that dialysis facility-specific quality gaps persist, that dialysis facilities are the primary drivers of performance variation, or that improvements in SRR consistently translate into improved patient-centered outcomes or cost reduction.
Closing Care Gaps
See notes below in scientific acceptability.
Feasibility Assessment
Scientific Acceptability
Same concern as CBE 3566. It is unclear is whether smaller facilities can be fairly compared using SRR. Because the measure is based on observed versus expected readmissions, facilities with fewer eligible discharges may experience greater score instability. The re-endorsement package should clearly demonstrate that performance estimates remain reliable and interpretable across the full range of facility volumes.
In addition, the measure relies on claims-based risk adjustment using patient and hospital characteristics. However, scientific acceptability would be strengthened by additional evidence demonstrating that residual differences in SRR are not driven primarily by unmeasured patient complexity or social risk. Additional evidence could include sensitivity analyses incorporating social risk variables, subgroup analyses by race, ethnicity, dual-eligibility status, rurality, and socioeconomic status. In addition, evaluation of potential confounding patient attributes including frailty and functional status could be driving readmissions, data elements not included in the risk model.
It is unclear whether the evidence submitted for CBE #5110 sufficiently demonstrates that facility-level differences in SRR scores represent true differences in dialysis facility quality rather than differences in patient risk, healthcare-system factors, or random variation. The re-endorsement materials primarily demonstrate measure calculation, implementation history, and statistical modeling, but provide more limited evidence supporting attribution, risk-adjustment adequacy, and the ability of the measure to consistently distinguish meaningful differences in quality across facilities, particularly among lower-volume providers.
A major validity concern is whether dialysis facilities should be held accountable for readmissions that may be influenced by hospitals, primary care providers, specialists, social services, and patient factors. The current rationale describes potential opportunities for dialysis facilities to influence readmissions but provides limited empirical evidence quantifying the facility's contribution to variation in readmission risk.
Use and Usability
Summary
The measure assumes dialysis facilities substantially influence readmission risk through: care coordination, medication reconciliation, post-discharge follow-up, and ongoing dialysis management.
However, hospital readmissions are influenced by many additional factors:
Hospital discharge quality, specialist access, primary care access, social determinants, caregiver support, patient confounding attributes, and patient adherence. Risk model is inadequate to account for these confounding variables.
(No subject)
Importance
Need updated research and information on potential cross-over measures
Closing Care Gaps
Optional domain
Feasibility Assessment
Does not appear to have any additional reporting burden
Scientific Acceptability
additional data needed for confirming data reliability
Use and Usability
(No subject)
Importance
The developer clearly demonstrates that hospital readmissions are a major issue for patients receiving dialysis. ESRD patients experience very high rates of hospitalization, and approximately one-third of hospital discharges are followed by an unplanned readmission within 30 days. Readmissions are associated with substantial costs, increased morbidity, and poorer quality of life, making this an important outcome for both patients and Medicare.
The literature cited by the developer, although dated, supports a connection between dialysis-facility practices—such as post-discharge follow-up, medication reconciliation, fluid management, and care coordination—and lower readmission risk. The measure addresses a well-established quality gap and focuses on an outcome that patients and caregivers consistently identify as important.
Closing Care Gaps
Not addressed
Feasibility Assessment
No concerns given how measure is calculated and lack of burden on providers
Scientific Acceptability
The data element reliability evidence is strong enough. The measure relies on EQRS and Medicare claims data, both of which are used extensively in federal quality programs. The developer reports that 97% of EQRS data elements matched medical records during validation review, with very low rates of missing or discrepant data. In addition, the claims data used for the measure are structured, used for reimbursement, and routinely audited by
My primary concern is the accountable entity-level reliability. The reported inter-unit reliability (IUR) is 0.336, which is well below the 0.60 threshold. The developer provides a reasonable explanation that dialysis facilities are relatively small, making high reliability difficult to achieve, and notes that CMS uses additional safeguards such as minimum reporting thresholds, statistical testing, Star Ratings aggregation, and small-facility adjustments to mitigate the impact of measurement noise. Still, nearly all deciles fall well below this range and I am unsure how to ignore such a large issue.
While the evidence for the data element validity is not the strongest, I do think in general the construct and how the measure use the data are not too great of a concern. I do not think the developer needs to do too much to strengthen this argument.
The clinical rationale for the measure is persuasive. The submission cites evidence that dialysis-facility processes such as fluid management, medication reconciliation, post-discharge assessment, and care coordination can influence readmission risk. The validity testing is supported by a conceptual model linking these facility actions to lower readmissions, and the risk-adjustment model demonstrates acceptable discrimination with an AUC of 0.683.
Use and Usability
No concerns, but I will say due to low reliability it does make it more challenging for facilities to know if their performance is due to noise or a real signal.
(No subject)
Importance
Changes as recommended by public comment to use ICD10 grouper vs ICD9 needs to be addressed in the measure specification.
Closing Care Gaps
not addressed
Feasibility Assessment
Scientific Acceptability
Use and Usability
Not met but addressable concerns
Importance
I agree with Staff Assessment This new measure is rated as 'Not Met but Addressable' for importance due to 1) dated evidence not addressing anticipated measure impact and 2) lack of information regarding potential competing measures.
Closing Care Gaps
As mentioned in the Staff Assessment, the developer did not address this optional domain.
Feasibility Assessment
I agree with the Staff Assessment, noting all required data elements are routinely generated during care delivery, and required elements are available from digital or electronic sources.
Scientific Acceptability
In particular the low error rate from CMS audit data supports this rating.
I agree with the Staff Assessment that the developer did not but should provide sufficient evidence for data element validity for all critical data elements (i.e., the numerator, denominator, exclusions), which is required for new measures.
Use and Usability
I agree with the Staff Assessment, as summarized that this maintenance measure is rated ‘Met’ for use and usability because it is actively used in at least one accountability application, with a systematic feedback approach that allows for continuous updates based on stakeholder feedback.
Summary
I feel there are several not met but addressable aspects to this measure so I endorse with modifications.
Public Comments
ASN Comments
While ASN appreciates the updates made to measure CBE 5510: Standardized Readmission Ratio for Dialysis Facilities (SRR), the society remains concerned that the revised SRR specifications do not meaningfully address the substantive methodological issues raised by the society in previous comment letters[i].
Specifically, ASN continues to urge CMS to transition from a ratio-based metric to a true risk-standardized rate measure. As currently specified, the SRR remains an opaque "observed-over-expected" ratio. This structure lacks the transparency necessary for patients and care partners to make well-informed healthcare decisions, while leaving small dialysis facilities highly vulnerable to extreme score volatility driven by random statistical variation.
[i] 240826ASNESRDPPSQIPCommentsFINAL.pdf
Need to update comorbidity adjustment method
"Methodology for prevalent comorbidity selection: We began the selection process with the 283 AHRQ CCS groupers for calendar year 2015."
This approach is antiquated. The Clinical Classification Software (CCS) is a dead grouper that was designed for ICD-9-CM and is no longer supported or recommended by AHRQ for ANY purpose. The current diagnosis grouper is the Clinical Classification Software Refined (CCSR), which was purpose-built for ICD-10-CM. Unlike CCS, CCSR is updated annually and therefore properly handles the 5,000 or more new ICD-10-CM codes that have been introduced over the past decade. For example, 1504 new codes were introduced in 2023 alone, and these codes cannot be handled by a dead grouper such as CCS. The CCSRs are designed for backward compatibility to October 2015, and must therefore be used for any purpose where CCS would have been used prior to that date. Please refer to these websites or consult with the AHRQ HCUP Research Tools team for further information.
https://www.icd10data.com/ICD10CM/Codes/Changes/New_Codes?year=2025
https://hcup-us.ahrq.gov/tools_software.jsp
https://hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp
RPA Comments on 5110
The Renal Physicians Association (RPA) is the professional organization of nephrologists whose goals are to ensure optimal care under the highest standards of medical practice for patients with kidney disease and related disorders. RPA acts as the national representative for physicians engaged in the study and management of patients with kidney disease. RPA appreciates the opportunity to submit comments on the kidney related quality measures currently under review by the Partnership for Quality Measures.
While the RPA supports the goal of reducing unplanned hospital readmissions for patients receiving dialysis, the measure is overly broad and has not demonstrated that it will close care gaps. The measure does not capture whether the patient was seen at the facility prior to readmission. Additionally, by including the vast majority of conditions rather than limiting to those related to dialysis care, the capacity of the facility to intervene may be diluted. As demonstrated in the data provided by the measure developers in SRR_6.2.4_Final_508.pdf, the impact on readmission rates is virtually unchanged between 2014-2023, when the nearly identical measure 2496 was in use. Therefore, RPA does not support the endorsement of CBEID 5110).