The Standardized Mortality Ratio (SMR) is defined as the ratio of the number of deaths that occur for Medicare ESRD dialysis patients (both Fee For Service and Medicare Advantage) treated at a particular facility to the number of deaths that would be expected given the characteristics of the dialysis facility’s patients and the national event rate for dialysis facilities. This measure is calculated as a ratio but can also be expressed as a rate.
When used for public reporting, the measure calculation will be restricted to facilities with at least three expected deaths in the reporting year. This restriction is required to ensure patients cannot be identified due to small cell size.
Measure Specs
General Information
For individuals with chronic kidney failure (end stage renal disease), regular dialysis treatments are life-sustaining when performed properly. However, the demanding technical environment creates numerous opportunities for life-threatening complications, including infections related to vascular access care and water treatment, air embolism, cardiovascular events related to stress of treatment, rapid electrolyte shifts, and acute volume depletion. In addition, inadequate treatment of chronic volume excess and ESRD-related mineral and bone disease can increase long-term cardiovascular risk associated with vascular calcification and critical organ arterial supply, as well as, contribute to left ventricular hypertrophy, all associated with the markedly increased risk of long-term cerebrovascular and cardiovascular complications and death from these dialysis-related complications. Quality measures that evaluate the large facility-level differences in patient mortality in dialysis facilities provide essential information for consumers about effectiveness of this life-sustaining therapy and facility-level differences in prevention of life-threatening complications of the treatment. Dialysis facilities also benefit from the information included in the SMR, which provides an overall benchmark of their success in preventing avoidable death for their patients. Preventing avoidable deaths also avoids some of the costs of treating the life-threatening complications (indirectly), and may allow more patients to survive long enough to receive a kidney transplant, further benefitting the patient and society.
Mortality rates among ESRD patients on chronic dialysis decreased in the US between 2012 to 2019, and mortality rates among ESRD patients increased during the COVID-19 pandemic from 2020-2022. Individuals receiving dialysis had a far lower number of expected remaining years of life relative to age-matched individuals in the general population [1]. In addition, mortality among ESRD dialysis patients varies across dialysis facilities, even after adjustment for patients’ characteristics. An adjusted facility-level mortality, which accounts for differences in patients’ characteristics, is one of several important health outcomes used by providers, health consumers, and insurers to evaluate the quality of care provided in dialysis facilities.
Reference:
[1] 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 measure now incorporates Part C Medicare Advantage (MA) data for the MA enrollees. This ensures that hospital, outpatient dialysis, and other billable services under Medicare – whether FFS or MA – are captured.
Numerator
Number of deaths among Medicare ESRD dialysis patients at the facility during the time period.
Information on death is obtained from several sources which include the CMS ESRD Program Medical Management Information System, the Death Notification Form (CMS Form 2746), and the Social Security Death Master File. The number of deaths that occurred among eligible dialysis patients during the time period is calculated. This count includes only Medicare patients, as detailed below.
Denominator
The denominator is the number of deaths that would be expected among Medicare ESRD dialysis patients at the facility during the time period, given the national average mortality rate and the patient mix at the facility.
Assignment of Patients to Facilities
The ESRD Quality Reporting System (EQRS), including CMS Medical Evidence Form (Form CMS-2728) and Death Notification Form (Form CMS-2746) is the primary basis for placing patients at dialysis facilities. Outpatient dialysis claims are used as an additional source when needed. We create a complete history of the status, location, and dialysis treatment modality of an ESRD patient from the date of the first ESRD service until the patient dies or the data collection cutoff date is reached. A new record is created each time a patient changes facilities or dialysis treatment modality; therefore, each record represents a time period associated with a specific modality and dialysis facility. Information regarding first ESRD service date, death and transplant is obtained from additional sources including the CMS Enrollment Database (EDB), transplant data from the Organ Procurement and Transplant Network (OPTN), and the Social Security Death Master File.
We detail patient inclusion criteria, facility assignment and how to count days at risk, all of which are required for the risk adjustment model. As patients can receive dialysis treatment at more than one facility in a given year, we assign each patient day to a facility (or no facility, in some cases) based on a set of conventions below.
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 deaths during the first 90 days of ESRD as well as patients who recover kidney function during that time period.
In order to exclude patients who only received temporary dialysis therapy, we assign patients to a facility only after they have been on dialysis there for at least 60 days. This 60-day period is used both for patients who started ESRD for the first time and for those who returned to dialysis after a transplant. That is, deaths and survival during the first 60 days of dialysis at a facility do not affect the SMR of that facility.
Identifying Facility Treatment Histories for Each Patient
For each patient, we identify the dialysis provider at each point in time. Starting with day 91 after onset of ESRD, we attribute patients to facilities according to the following rules. A patient is attributed to a facility once the patient has been treated there for at least 60 days. When a patient transfers from one facility to another, the patient continues to be attributed to the original facility for 60 days and then is attributed to the destination facility from day 61. In particular, a patient is attributed to their current facility on day 91 of ESRD if that facility has treated them for at least 60 days. If on day 91, the facility has not treated a patient for at least 60 days, we wait until the patient reaches day 60 of continuous treatment at that facility before attributing the patient to that facility. When a patient is not treated in a single facility for a span of 60 days (for instance, if there were two switches within 60 days of each other), we do not attribute that patient to any facility. Patients are removed from a facility’s analysis upon receiving a transplant. Patients who withdrew from dialysis or recovered renal function remain assigned to their treatment facility for 60 days after withdrawal or recovery.
If a period of one year passes with neither Medicare dialysis claims nor EQRS information to indicate that a patient was receiving dialysis treatment and if there is no earlier evidence of transfer, recovery, or death, we consider the patient lost to follow-up and do not include that patient in the analysis. If evidence of dialysis reappears, the patient is entered into analysis after 60 days of continuous therapy at a single facility. All EQRS records noting continuing dialysis are extended until the appearance of any evidence of recovery, transfer, or death. Lost to follow-up periods are not created in these cases since the instructions for EQRS only require checking patient data for continued accuracy, but do not have a requirement for updating if there are not any changes.
Days at Risk for Each Patient-Record
After patient treatment histories are defined as described above, periods of follow-up time since ESRD onset are created for each patient. In order to adjust for duration of ESRD appropriately, we define six time intervals with cut points at 6 months, 1 year, 2 years, 3 years, and 5 years. A new time period begins each time the patient is determined to be at a different facility, has a change in Medicare eligibility, has a change in Medicare Advantage status, at the start of each calendar year, or when crossing any of the above cut points.
The number of days at risk in each of the six time-intervals listed above is used to calculate the expected number of deaths for the patient during that period. The SMR for a facility is the ratio of the total number of observed deaths to the total number of expected deaths during all time periods at the facility.
Exclusions
Exclusions that are implicit in the denominator definition include time at risk while a patient has had ESRD for 90 days or less. In addition, the measure does not include deaths from non-prescription drug overdose or accidents unrelated to treatment as indicated on CMS form 2746, since these deaths are unlikely to be related to care at the dialysis facility.
See 1.15a Denominator Details, above
Measure Calculation
See SMR Flowchart_Final_Oct 2025._508 PDF, attached to 1.18a
N/A
There is not a minimum sample size needed to calculate the performance score. Public reporting of this measure on Dialysis Facility Care Compare (DFCC) would be restricted to facilities with at least three expected deaths for the measure to comply with restrictions on reporting of potentially patient identifiable information related to small cell size.
Supplemental Attachment
Measure Record
Point of Contact
N/A
Wilfred Agbenyikey
Baltimore, MD
United States
Jonathan Segal
UM-KECC
Ann Arbor, MI
United States
Importance
Evidence
Note: This Evidence Summary was modified from the 2019-2020 Evidence Form submitted to the CBE. We performed an updated PUBMED Literature search using multiple search strategies for the years 2017-present. The search results were initially reviewed by one investigator to remove duplicates and off-topic citations. The resulting set of citations was independently reviewed by two investigators with clinical dialysis training and experience. Of the 174 references reviewed, 23 were identified with consensus agreement. We have included selected recent references and updated the Evidence Summary with these additional citations in the summary and reference list below.
All-cause mortality for ESRD patients on chronic dialysis far exceeds age matched controls in the general and Medicare populations [1]. Mortality rates across dialysis facilities vary, even after controlling for multiple patient characteristics and comorbidities [2, 41]. Selection of dialysis modality, sometimes the result of dialysis facility practices, likely influences mortality [3]. Furthermore, mortality is associated with certain conditions resulting from kidney failure and chronic dialysis care, including uremic toxin accumulation, volume overload/hypertension and its treatment, bone/mineral disease, and infections related to dialysis access, have been described in detail [4-6, 40, 44].
Specific dialysis practices have been identified for several of these ESRD-related conditions that can improve patient survival and morbidity, including provision of adequate small solute clearance [7], control of total body volume while guarding against rapid ultrafiltration [8-11, 50-52], control of electrolytes, particularly potassium [46, 47], and appropriate management of mineral and bone disorders [12-14, 43, 45, 53-59]. In addition, improved infection prevention efforts by dialysis providers can result in reduced infection-related hospitalization and mortality [15-20].
Additional studies have bolstered the importance of fluid management in improving patient survival [24, 26, 37]. Rescheduling missed dialysis treatments [21], as well as providing longer treatment times at dialysis initiation [33], while being mindful to preserve residual kidney function [30] all have the potential to reduce patient mortality. Nutrition counseling, and how the interdisciplinary team manages potassium [38, 43, 45], phosphorus [31], and encourages healthy eating habits with fruits/vegetables [39] also impact patient outcomes. Sustained efforts at influenza vaccinations can impact mortality [32]. Lastly, in the midst of a national opioid epidemic, dialysis patients are at particularly increased risk of adverse outcomes related to mis-use of analgesic and sedative drugs. Careful attention is needed to avoid excess mortality [25, 42].
Although the basic technology of hemodialysis has not changed dramatically in the last two decades, overall mortality of individuals on chronic dialysis has improved, both in absolute and relative terms [48, 49]. The trend towards reduced mortality is temporally correlated with the introduction of public reporting of the Standardized Mortality Ratio (SMR) in the early 2000’s. Much of the observed reduction in mortality was likely driven by changes in dialysis facility practice patterns (e.g. less aggressive erythropoietic stimulating agent [ESA], increased use of home dialysis modalities, broader attention to the risks of aggressive fluid removal during dialysis and more holistic approaches to overall volume management, and introduction of additional medical options for treatment of mineral and bone disease). Despite these modest gains, mortality is far too common for patients receiving chronic dialysis and continued public reporting of this important health outcome is necessary to incentivize additional improvement over the coming years.
References
[1]. United States Renal Data System. 2018 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, 2018.
[2]. Kalbfleisch J, Wolfe R, Bell S, Sun R, Messana J, Shearon T, Ashby V, Padilla R, Zhang M, Turenne M, Pearson J, Dahlerus C, Li Y. Risk Adjustment and the Assessment of Disparities in Dialysis Mortality Outcomes. J Am Soc Nephrol. 2015; Nov;26(11):2641-5.
[3]. Weinhandl ED, Nieman KM, Gilbertson DT, Collins AJ. Hospitalization in daily home hemodialysis and matched thrice-weekly in-center hemodialysis patients. Am J Kidney Dis. 2015 Jan;65(1):98-108.
[4]. Himmelfarb J, Ikizler T. Hemodialysis N Engl J. 2010 Nov; 363:1833–1845.
[5]. Kliger AS. Maintaining Safety in the Dialysis Facility. Clin J Am Soc Nephrol. 2015 Apr 7;10(4):688-95.
[6]. Hung AM, Hakim RM. Dialysate and Serum Potassium in Hemodialysis. Am J Kidney Dis. 2015 Jul;66(1):125-32.
[7]. Port FK, Ashby VB, Dhingra RK, Roys EC, Wolfe RA: Dialysis dose and body mass index are strongly associated with survival in hemodialysis patients. J Am Soc Nephrol 13:1061-1066, 2002.
[8]. Saran R, Bragg-Gresham JL, Levin NW, Twardowski ZJ, Wizemann V, Saito A, Kimata N, Gillespie BW, Combe C, Bommer J, Akiba T, Mapes DL, Young EW, Port FK. Longer Treatment Time and Slower Ultrafiltration in Hemodialysis: Associations With Reduced Mortality in the DOPPS. Kidney Int. 2006 Apr;69(7):1222-8.
[9]. FHN Trial Group, Chertow GM, Levin NW, Beck GJ, Depner TA, Eggers PW, Gassman JJ, Gorodetskaya I, Greene T, James S, Larive B, Lindsay RM, Mehta RL, Miller B, Ornt DB, Rajagopalan S, Rastogi A, Rocco MV, Schiller B, Sergeyeva O, Schulman G, Ting GO, Unruh ML, Star RA, Kliger AS. In-center hemodialysis six times per week versus three times per week. N Engl J Med. 2010 Dec 9;363(24):2287-300.
[10]. Flythe JE, Curhan GC, Brunelli SM. Disentangling the Ultrafiltration Rate–Mortality Association: The Respective Roles of Session Length and Weight Gain. Clin J Am Soc Nephrol. 2013 Jul;8(7):1151-61.
[11]. Weiner DE, Brunelli SM, Hunt A, Schiller B, Glassrock R, Maddux FW, Johnson D, Parker T, Nissenson A. Improving clinical outcomes among hemodialysis patients: a proposal for a "volume first" approach from the chief medical officers of US dialysis providers. Am J Kidney Dis. 2014 Nov;64(5):685-95.
[12]. Block GA, Kilpatrick RD, Lowe KA, Wang W, Danese MD. CKD-mineral and bone disorder and risk of death and cardiovascular hospitalization in patients on hemodialysis. Clin J Am Soc Nephrol. 2013 Dec;8(12):2132-40.
[13]. Pun PH, Horton JR, Middleton JP. Dialysate calcium concentration and the risk of sudden cardiac arrest in hemodialysis patients. Clin J Am Soc Nephrol. 2013 May;8(5):797-803.
[14]. Ishani A, Liu J, Wetmore JB, Lowe KA, Do T, Bradbury BD, Block GA, Collins AJ. Clinical outcomes after parathyroidectomy in a nationwide cohort of patients on hemodialysis. Clin J Am Soc Nephrol. 2015 Jan 7;10(1):90-7.
[15]. Gilbertson DT, Unruh M, McBean AM, Kausz AT, Snyder JJ, Collins AJ. Influenza vaccine delivery and effectiveness in end-stage renal disease. Kidney Int. 2003 Feb;63(2):738-43.
[16]. Rosenblum A, Wang W, Ball LK, Latham C, Maddux FW, Lacson E Jr. Hemodialysis catheter care strategies: a cluster-randomized quality improvement initiative. Am J Kidney Dis. 2014 Feb;63(2):259-67.
[17]. Patel PR, Kallen AJ. Bloodstream infection prevention in ESRD: forging a pathway for success. Am J Kidney Dis. 2014 Feb;63(2):180-2.
[18]. Dalrymple LS, Mu Y, Romano PS, Nguyen DV, Chertow GM, Delgado C, Grimes B, Kaysen GA, Johansen KL. Outcomes of infection-related hospitalization in Medicare beneficiaries receiving in-center hemodialysis. Am J Kidney Dis. 2015 May;65(5):754-62.
[19]. Dalrymple LS, Mu Y, Nguyen DV, Romano PS, Chertow GM, Grimes B, Kaysen GA, Johansen KL. Risk Factors for Infection-Related Hospitalization in In-Center Hemodialysis. Clin J Am Soc Nephrol. 2015 Dec 7;10(12):2170-80.
[20]. Gilbertson DT, Wetmore JB. Infections Requiring Hospitalization in Patients on Hemodialysis. Clin J Am Soc Nephrol. 2015 Dec 7;10(12):2101-3.
[21] Dena E. Cohen, Kathryn S. Gray, Carey Colson, David B. Van Wyck, Francesca Tentori, and Steven M. Brunelli. Impact of Rescheduling a Missed Hemodialysis Treatment on Clinical Outcomes. Kidney Medicine. Volume 2, Issue 1, January–February 2020, Pages 12-19.
[22] Abdulkareem Agunbiade , Abhijit Dasgupta , Michael M Ward. Racial/Ethnic Differences in Dialysis Discontinuation and Survival After Hospitalization for Serious Conditions Among Patients on Maintenance Dialysis. J Am Soc Nephrol, 31 (1), 149-160 Jan 2020.
[23] Fozia Ajmal, Janice C Probst, John M Brooks, James W Hardin, Zaina Qureshi, Tazeen H Jafar . Freestanding Dialysis Facility Quality Incentive Program Scores and Mortality Among Incident Dialysis Patients in the United States. Am J Kidney Dis, 75 (2), 177-186 Feb 2020.
[24] Magdalene M Assimon, Julia B Wenger, Lily Wang, Jennifer E Flythe. Ultrafiltration Rate and Mortality in Maintenance Hemodialysis Patients. Am J Kidney Dis, 68 (6), 911-922 Dec 2016.
[25] Kimmel PL, Fwu CW, Abbott KC, Eggers AW, Kline PP, Eggers PW. J Am Soc Nephrol. 2017 Dec;28(12):3658-3670. doi: 10.1681/ASN.2017010098. Epub 2017 Sep 21. Opioid Prescription, Morbidity, and Mortality in United States Dialysis Patients.
[26] Zoccali C, Moissl U, Chazot C, Mallamaci F, Tripepi G, Arkossy O, Wabel P, Stuard S. J Am Soc Nephrol. 2017 Aug;28(8):2491-2497. doi: 10.1681/ASN.2016121341. Epub 2017 May 4. Chronic Fluid Overload and Mortality in ESRD.
[27] Ku E, Yang W, McCulloch CE, Feldman HI, Go AS, Lash J, Bansal N, He J, Horwitz E, Ricardo AC, Shafi T, Sondheimer J, Townsend RR, Waikar SS, Hsu CY; Am J Kidney Dis. 2019 Nov 12:S0272-6386(19)30974-6. doi: 10.1053/j.ajkd.2019.08.011. Online ahead of print. Race and Mortality in CKD and Dialysis: Findings From the Chronic Renal Insufficiency Cohort (CRIC) Study. CRIC Study Investigators. Collaborators: Appel LJ, Kusek JW, Rao PS, Rahman M.
[28] Am J Nephrol. 2019;49(3):241-253. doi: 10.1159/000497446. Epub 2019 Feb 28. Temporal Trends in Incident Mortality in Dialysis Patients: Focus on Sex and Racial Disparities. Shah S, Leonard AC, Meganathan K, Christianson AL, Thakar CV.
[29] Bowman B, Zheng S, Yang A, Schiller B, Morfín JA, Seek M, Lockridge RS. Am J Kidney Dis. 2018 Aug;72(2):278-283. doi: 10.1053/j.ajkd.2018.01.035. Epub 2018 Mar 3. Improving Incident ESRD Care Via a Transitional Care Unit.
[30] Li T, Wilcox CS, Lipkowitz MS, Gordon-Cappitelli J, Dragoi S. Am J Nephrol. 2019;50(6):411-421. doi: 10.1159/000503805. Epub 2019 Oct 18. Rationale and Strategies for Preserving Residual Kidney Function in Dialysis Patients.
[31] Hou Y, Li X, Sun L, Qu Z, Jiang L, Du Y. Clin Chim Acta. 2017 Nov;474:108-113. doi: 10.1016/j.cca.2017.09.005. Epub 2017 Sep 10. Phosphorus and mortality risk in end-stage renal disease: A meta-analysis.
[32] Gilbertson DT, Rothman KJ, Chertow GM, Bradbury BD, Brookhart MA, Liu J, Winkelmayer WC, Stürmer T, Monda KL, Herzog CA, Ashfaq A, Collins AJ, Wetmore JB.J Am Soc Nephrol. 2019. Feb;30(2):346-353. doi: 10.1681/ASN.2018060581. Epub 2019 Jan 24.Excess Deaths Attributable to Influenza-Like Illness in the ESRD Population.
[33] Swaminathan S, Mor V, Mehrotra R, Trivedi AN. Am J Kidney Dis. 2017 Jul;70(1):69-75. doi: 10.1053/j.ajkd.2016.11.017. Epub 2017 Feb 21. Initial Session Duration and Mortality Among Incident Hemodialysis Patients.
[34] Schold JD, Flechner SM(, Poggio ED, Augustine JJ, Goldfarb DA, Sedor JR, Buccini LD .Am J Kidney Dis. 2018 Jul;72(1):19-29. doi: 10.1053/j.ajkd.2017.12.014. Epub 2018 Mar 7. Residential Area Life Expectancy: Association With Outcomes and Processes of Care for Patients With ESRD in the United States. Comment in Am J Kidney Dis. 2018 Jul;72(1):4-6.
[35] BMC Nephrol. 2019 Jul 29;20(1):285. doi: 10.1186/s12882-019-1473-0. Long-term outcomes among Medicare patients readmitted in the first year of hemodialysis: a retrospective cohort study. Ross KH, Jaar BG, Lea JP, Masud T, Patzer RE, Plantinga LC.
[36] Clin Nephrol. 2016 Nov;86 (2016)(11):262-269. doi: 10.5414/CN108816.Data completeness as an unmeasured confounder in dialysis facility performance comparison with 1-year follow-up. Liu J, Krishnan M, Zhou J, Nieman KM, Peng Y, Gilbertson DT.
[37] Am J Kidney Dis. 2017 Mar;69(3):367-379. doi: 10.1053/j.ajkd.2016.08.030. Epub 2016 Nov 17. Interdialytic Weight Gain: Trends, Predictors, and Associated Outcomes in the International Dialysis Outcomes and Practice Patterns Study (DOPPS). Wong MM, McCullough KP, Bieber BA, Bommer J, Hecking M, Levin NW, McClellan WM, Pisoni RL, Saran R, Tentori F, Tomo T, Port FK, Robinson BM.
[38] Am J Kidney Dis. 2017 Jul;70(1):21-29. doi: 10.1053/j.ajkd.2016.10.024. Epub 2017 Jan 19. Serum Potassium and Short-term Clinical Outcomes Among Hemodialysis Patients: Impact of the Long Interdialytic Interval. Brunelli SM, Du Mond C, Oestreicher N, Rakov V, Spiegel DM. Comment in Am J Kidney Dis. 2017 Jul;70(1):4-7.
[39] Clin J Am Soc Nephrol. 2019 Feb 7;14(2):250-260. doi: 10.2215/CJN.08580718. Epub 2019 Jan 31. Fruit and Vegetable Intake and Mortality in Adults undergoing Maintenance Hemodialysis. Saglimbene VM, Wong G, Ruospo M, Palmer SC, Garcia-Larsen V, Natale P, Teixeira-Pinto A, Campbell KL, Carrero JJ, Stenvinkel P, Gargano L, Murgo AM, Johnson DW, Tonelli M, Gelfman R, Celia E, Ecder T, Bernat AG, Del Castillo D, Timofte D, Török M, Bednarek-Skublewska A, Duława J, Stroumza P, Hoischen S, Hansis M, Fabricius E, Felaco P, Wollheim C, Hegbrant J, Craig JC, Strippoli GFM.
[40] Beta-Blocker Use and Risk of Mortality in Heart Failure Patients Initiating Maintenance Dialysis. Zhou H, Sim JJ, Shi J, Shaw SF, Lee MS, Neyer JR, Kovesdy CP, Kalantar-Zadeh K, Jacobsen SJ. Am J Kidney Dis. 2021 May;77(5):704-712. doi: 10.1053/j.ajkd.2020.07.023. Epub 2020 Nov 15.
[41] Risk factors for mortality in elderly haemodialysis patients: a systematic review and meta-analysis. Song YH, Cai GY, Xiao YF, Chen XM. BMC Nephrol. 2020 Aug 31;21(1):377. doi: 10.1186/s12882-020-02026-x.
[42] Benzodiazepines, Co-dispensed Opioids, and Mortality among Patients Initiating Long-Term In-Center Hemodialysis. Muzaale AD, Daubresse M, Bae S, Chu NM, Lentine KL, Segev DL, McAdams-DeMarco M. Clin J Am Soc Nephrol. 2020 Jun 8;15(6):794-804. doi: 10.2215/CJN.13341019. Epub 2020 May 26.
[43] Relationship between serum calcium or phosphate levels and mortality stratified by parathyroid hormone level: an analysis from the MBD-5D study. Asada S, Yokoyama K, Miyakoshi C, Fukuma S, Endo Y, Wada M, Nomura T, Onishi Y, Fukagawa M, Fukuhara S, Akizawa T. Clin Exp Nephrol. 2020 Jul;24(7):630-637. doi: 10.1007/s10157-020-01879-8. Epub 2020 Mar 31.
[44] Association of all-cause mortality with pre-dialysis systolic blood pressure and its peri-dialytic change in chronic hemodialysis patients. Zhang H, Preciado P, Wang Y, Meyring-Wosten A, Raimann JG, Kooman JP, van der Sande FM, Usvyat LA, Maddux D, Maddux FW, Kotanko P. Nephrol Dial Transplant. 2020 Sep 1;35(9):1602-1608. doi: 10.1093/ndt/gfz289.
[45] Cinacalcet Treatment Significantly Improves All-Cause and Cardiovascular Survival in Dialysis Patients: Results from a Meta-Analysis. Zu Y, Lu X, Song J, Yu L, Li H, Wang S. Kidney Blood Press Res. 2019;44(6):1327-1338. doi: 10.1159/000504139. Epub 2019 Nov 20.
[46] Post-dialysis Hypokalemia and All-Cause Mortality in Patients Undergoing Maintenance Hemodialysis. Ohnishi T, Kimachi M, Fukuma S, Akizawa T, Fukuhara S. Clin J Am Soc Nephrol. 2019 Jun 7;14(6):873-881. doi: 10.2215/CJN.07950718. Epub 2019 May 2.
[47] Racial and Ethnic Differences in Mortality Associated with Serum Potassium in a Large Hemodialysis Cohort. Kim T, Rhee CM, Streja E, Soohoo M, Obi Y, Chou JA, Tortorici AR, Ravel VA, Kovesdy CP, Kalantar-Zadeh K. Am J Nephrol. 2017;45(6):509-521. doi: 10.1159/000475997. Epub 2017 May 20.
[48] Temporal Trends in Incident Mortality in Dialysis Patients: Focus on Sex and Racial Disparities. Shah S, Leonard AC, Meganathan K, Christianson AL, Thakar CV. Am J Nephrol. 2019;49(3):241-253. doi: 10.1159/000497446. Epub 2019 Feb 28.
[49] Changes in Excess Mortality from End Stage Renal Disease in the United States from 1995 to 2013. Foster BJ, Mitsnefes MM, Dahhou M, Zhang X, Laskin BL. Clin J Am Soc Nephrol. 2018 Jan 6;13(1):91-99. doi: 10.2215/CJN.04330417. Epub 2017 Dec 14.
[50] Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Jennifer E. Flythe, Stephen E. Kimmel and Steven M. Brunelli. Kidney International (2011) 79, 250–257; doi:10.1038/ki.2010.383
[51] Ultrafiltration Rate and Mortality in Maintenance Hemodialysis Patients. Magdalene M. Assimon, Julia B. Wenger, Lily Wang, and Jennifer E. Flythe. Am J Kidney Dis. 2016;68(6):911-922.
[52] A Body Size–Adjusted Maximum Ultrafiltration Rate Warning Level Is Not Equitable for Larger Patients. John T. Daugirdas. Clin J Am Soc Nephrol 16: 2021.
[53] Changes in 3-month mineral and bone disorder patterns were associated with all-cause mortality in prevalent hemodialysis patients with secondary hyperparathyroidism. Chihiro Kato, Naohiko Fujii, Chisato Miyakoshi, Shinji Asada , Yoshihiro Onishi, Shingo Fukuma, Takanobu Nomura1, Michihito Wada1, Masafumi Fukagawa8, Shunichi Fukuhara3 and Tadao Akizawa9 BMC Nephrol 21: 432, 2020.
[54] Abnormal Mineral Metabolism and Mortality in Hemodialysis Patients With Secondary Hyperparathyroidism: Evidence From Marginal Structural Models Used to Adjust for Time-Dependent Confounding. Masafumi Fukagawa, Ryo Kido, Hirotaka Komaba, Yoshihiro Onishi, Takuhiro Yamaguchi, Takeshi Hasegawa, Noriaki Kurita, Shingo Fukuma, Tadao Akizawa, and Shunichi Fukuhara. Am J Kidney Dis 63(6):979-87, 2014
[55] Facility-level CKD-MBD composite score and risk of adverse clinical outcomes among patients on hemodialysis. Geoffrey A. Block, Akeem A. Yusuf, Mark D. Danese, Heidi S. Wirtz, Yan Hu, Thy P. Do, Kerry Cooper, David T. Gilbertson, Brian D. Bradbury and Allan J. Collins. BMC Nephrology (2016) 17:166.
[56] Impact of longer term phosphorus control on cardiovascular mortality in hemodialysis patients using an area under the curve approach: results from the DOPPS. Marcelo Barreto Lopes, Angelo Karaboyas , Brian Bieber, Ronald L. Pisoni, Sebastian Walpen, Masafumi Fukagawa , Anders Christensson, Pieter Evenepoel, Marisa Pegoraro, Bruce M. Robinson and Roberto Pecoits-Filho. Nephrol Dial Transplant 35:1794-1801, 2020.
[57] Association of single and serial measures of serum phosphorus with adverse outcomes in patients on peritoneal dialysis: results from the international PDOPPS. Marcelo Barreto Lopes, Angelo Karaboyas, Junhui Zhao, David W. Johnson, Talerngsak Kanjanabuch, Martin Wilkie, Kosaku Nitta, Hideki Kawanishi, Jeffrey Perl and Ronald L. Pisoni. Nephrol Dial Transplant 38: 193-202, 2023.
[58] Phosphorus Binders and Survival on Hemodialysis. Tamara Isakova, Orlando M. Gutierrez, Yuchiao Chang, Anand Shah, Hector Tamez, Kelsey Smith, Ravi Thadhani, and Myles Wolf. J Am Soc Nephrol 20(2):388-96, 2009.
[59] Use of phosphate-binding agents is associated with a lower risk of mortality. Jorge B. Cannata-Andıa, Jose L. Fernandez-Martın, Francesco Locatelli, Gerard London, Jose L. Gorriz, Jurgen Floege, Markus Ketteler, Anıbal Ferreira, Adrian Covic, Boleslaw Rutkowski, Dimitrios Memmos, Willem-Jan Bos, Vladimir Teplan, Judit Nagy, Christian Tielemans, Dierik Verbeelen, David Goldsmith, Reinhard Kramar, Pierre-Yves Martin, Rudolf P. Wuthrich, Drasko Pavlovic, Miha Benedik, Jose Emilio Sanchez, Pablo Martinez-Camblor, Manuel Naves-Dıaz, Juan J. Carrero and Carmine Zoccali. Kidney Int 84:998-1008, 2013.
Measure Impact
There are several studies indicating that patients with kidney failure who require dialysis value an assessment of mortality rates at the dialysis facility level [1,2]. Interestingly, patients tend to place an even higher value on issues related to quality of life compared to longevity. In comparison with patients, dialysis providers and administrators place an even higher value on a mortality relative to other measures [2].
References:
[1] Weiner DE, Delgado C, Flythe JE, Forfang DL, Manley T, McGonigal LJ, McNamara E, Murphy H, Roach JL, Watnick SG, Weinhandl E, Willis K, Berns JS; KDOQI Patient-Centered Quality Measures for Dialysis Care Workshop Participants. Patient-Centered Quality Measures for Dialysis Care: A Report of a Kidney Disease Outcomes Quality Initiative (KDOQI) Scientific Workshop Sponsored by the National Kidney Foundation. Am J Kidney Dis. 2024 May;83(5):636-647.
[2] Parra, E., Arenas, M.D., Fernandez-Reyes Luis, M. et al. Evaluation of dialysis centres: values and criteria of the stakeholders. BMC Health Serv Res 20, 297 (2020)
Performance Gap
The average SMR remained stable across years and during the 2020-2023 period. The average SMR varied from 0.93 to 1.05. However, within any given year, there was a substantial gap in performance as SMR varied widely across facilities, with the 10th decile being as low as 0.50 and the 90th decile being as high as 1.57.
Distribution of SMRs of all facilities by year (2020-2023):
2020: Facilities = 6,521, Mean SMR = 0.98, Standard Deviation = 0.41, 10th =0.50, 25th = 0.70, 50th = .95, 75th = 1.22, 90th = 1.52
2021: Facilities = 7,074, Mean SMR = 1.05, Standard Deviation =0 .41, 10th = 0.58, 25th = 0.77, 50th = 1.00, 75th = 1.27, 90th = 1.57
2022: Facilities =7,091, Mean SMR = 0.98, Standard Deviation = 0.38, 10th = 0.54, 25th = 0.73, 50th = .95, 75th = 1.20, 90th = 1.47
2023: Facilities = 7,026, Mean SMR = 0.93, Standard Deviation = 0.37, 10th = 0.50, 25th =0.68, 50th = .90, 75th = 1.14, 90th = 1.39
Across the 4-year SMR (2020-2023): Facilities = 7,824, Mean SMR = 0.99, Standard Deviation = 0.27, 10th = 0.68, 25th = 0.81, 50th = 0.97, 75th = 1.14, 90th = 1.32
See SMR_2.4 Table 1_Revised Nov 2025_508 PDF, attached to 2.4a, for Table 1 data and caption.
Care Gaps
Closing Care Gaps
This field is optional for Fall 2025.
Feasibility
Feasibility
All the data incorporated into this measure 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 Dialysis Facility Care Compare (DFCC) would be restricted to facilities with at least three expected deaths for the measure to comply with restrictions on reporting of potentially patient identifiable information related to small cell size.
No changes were made.
Proprietary Information
Scientific Acceptability
Testing Data
Data from 2020-2023 were used to calculate SMR. Please refer to Section 1.25 Data Source Details for information on data sources.
2020-2023
None
See SMR_5.1.3_Final_Oct 2025_508 PDF, attached in Section 7.1 Supplemental Attachment, for full response to this question
Please see SMR_5.1.4_Final_Oct 2025_508 PDF, which is attached in Section 7.1 Supplemental Attachment, for full response to this question
Reliability
The reliability of the Standardized Mortality Ratio (SMR) was assessed using data among ESRD dialysis patients during 2020-2023. If the measure were a simple average across individuals in the facility, the usual approach for determining measure reliability would be a one-way analysis of variance (ANOVA), in which the between and within facility variation in the measure is determined. The inter-unit reliability (IUR) measures the proportion of the total variation of a measure that is attributable to the between-facility variation. The SMR, however, is not a simple average and we instead estimate the IUR using a bootstrap approach, which uses a resampling scheme to estimate the within facility variation that cannot be directly estimated by ANOVA. A small IUR (near 0) reveals that most of the variation of the measures between facilities is driven by random noise, indicating the measure would not be a good characterization of the differences among facilities, whereas a large IUR (near 1) indicates that most of the variation between facilities is due to the real difference between facilities.
Here we describe our approach to calculating IUR. Let T1,…,TN be the SMR for these facilities. Within each facility, select at random and with replacement B (say 100) bootstrap samples. That is, if the ith facility has ni subjects, randomly draw with replacement ni subjects from those in the same facility, find their corresponding SMRi and repeat the process B times. Thus, for the ith facility, we have bootstrapped SMRs of T*i,1, … T*i,B Let Si*2 be the sample variance of this bootstrap sample. From this it can be seen that
St,w2= ΣNi=1[(ni-1) S i*2]/ ΣNi=1(ni-1).
is a bootstrap estimate of the within-facility variance in the SMR, namely, σ t,w2. 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 SMR and
n’ = (Sni - Sni2/Sni)/(N-1)
is approximately the average facility size (number of patients per facility). Note that St2 is the total variation of SMR and is an estimate of σ b2 + σ t,w2, where σ b2 is 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.
The SMR calculation only included facilities with at least 3 expected deaths for each year.
The overall IUR for the four-year SMR (2020-2023) is 0.47, which means that a little less than half of the variation in the 4-year SMR can be attributed to the between-facility variation. The SMR measure IUR is similar to previous cycles, and has been endorsed and re-endorsed for the last several cycles. Please see SMR_5.2.3a Table 2 and IUR Reliability_Revised Nov 2025_508 PDF attached to 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 SMR 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.
This value of IUR indicates a moderate degree of reliability. When stratified by facility size, we find that, as expected, larger facilities have greater IUR.
See SMR_5.2.3a Table 2 and IUR Reliability_Revised Nov 2025_508 PDF, attached to Section 5.2.3a, for Table 2 data and caption.
Validity
We have assessed the validity of the measure through various comparisons of this measure with other quality performance measures in use, using Spearman correlations.
Negative Relationships
- Vascular Access: Standardized Fistula Rate (SFR) – We expect a negative association between SFR and SMR. 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 mortality. Higher rates of the facility level SFR will be negatively associated with mortality as measured by SMR.
- Kt/V ≥ 1.2: We expect a negative association between the facility percentage of patients with Kt/V>= 1.2 and SMR. Facilities that have a high proportion of patients with adequate small solute clearance may also have processes of care in place that would likely avoid adverse outcomes. In addition, patients who are unable to achieve a Kt/V of 1.2 may be morbidly obese, use a catheter for vascular access, or be non-adherent to treatment recommendations such that they may be at higher risk for mortality. Higher rates of the facility level percentage of patients with adequate dialysis (facility percentage Kt/V> 1.2) will be negatively associated with SMR.
Positive Relationships
- Vascular Access: Long-term catheter rate (catheter in use >=3 continuous months) – We expect a positive association between the long-term catheter rate and SMR. 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 be at higher risk of mortality. Higher long-term catheter rates will be positively associated with SMR.
- Standardized Hospitalization Ratio (SHR): We expect a positive association between SHR and SMR. Patients who require acute medical care in the hospital 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 Readmission Ratio (SRR): We expect a positive association between SRR and SMR. Both hospitalization and readmission are a reflection of hospital utilization and increased comorbidity burden. Additionally, patients readmitted after a recent discharge indicates they still require acute medical attention or experience other post-discharge complications placing them at higher risk for mortality.
- Standardized Transfusion Ratio (STrR): We expect a positive association between STrR and SMR. Patients with severe anemia may require hospitalization and blood transfusion, placing them at risk for other adverse events and potentially higher risk for mortality.
Please see SMR_5.3.4a_Final_Oct 2025_508 attachment in 5.3.4a for full response to this question
SMR is correlated with each of the quality performance measures in the expected direction. All correlations are statistically significant. As expected, the SMR is positively correlated for each individual year with the SHR-Admissions, SRR-Readmissions, and the STrR. The SMR is negatively correlated with the percent of hemodialysis patients with Kt/V>=1.2, in the direction expected indicting lower SMRs are associated with a higher percentage of patients receiving adequate dialysis dose. The SMR is negatively correlated with the percentage of patients in the facility with an AV Fistula as measured by SFR indicating lower standardized mortality rates are associated with a higher standardized fistula rate. On the other hand, the SMR is positively correlated with long-term catheter rates indicating that higher values of SMR are associated with higher rates of long-term catheters.
Risk Adjustment
The methods for development of the risk factor models have been published and documented previously (Wolfe 1992; Wolfe 2001). The final risk adjustment is based on a Cox or relative risk model. In this model, covariates are taken to act multiplicatively on the death rate and the adjustment model is fitted with facility defining strata so as to provide valid estimates even if the distribution of adjustment variables differs across facilities. Relevant references are Cox (1972) and Kalbfleisch and Prentice (2002). All analyses are performed using SAS.
The denominator of SMR for a facility is the expected number of deaths from the patient-records meeting the inclusion criteria, based on the number of days attributed to that facility (the assignment rule will be detailed later), if the facility conforms to the national norm. Specifically, the expectation is calculated using a two-stage model. At Stage 1, we fit a Cox model [1] stratified by facility and adjusted for patient age, race, ethnicity, sex, diabetes, duration of ESRD, nursing home status, patient comorbidities at incidence, prevalent comorbidities, body mass index (BMI) at incidence, Medicare Advantage status, and calendar year. This stratified model allows each facility to have a distinct baseline survival function while retaining the same regression coefficients of all the adjusters across all the facilities. Stratification by facility avoids estimating facility effects directly and also reduces computational burden. A linear predictor using the estimates of regression coefficients will be computed for each patient and will be used as the offset term in the Stage 2 modeling. At Stage 2, we fit an unstratified Cox model, which includes the offset term from Stage 1 model as well as the race-specific age-adjusted state population death rates. The baseline hazard or survival function of this model has national norm interpretations. With the fitted model at Stage 2, we compute the expected probability of death for each patient based on the aforementioned adjusters and the number of days assigned to a facility. The denominator of SMR for a facility is then the summation of expected probabilities of death from all the patients assigned to that facility.
The patient characteristics included in the stage 1 model as covariates are:
- Age: Age is included as a piecewise continuous variable with different coefficients based on whether the patient is 0-13 years old, 14-60 years old, or 61+ years old.
- Sex
- Race: White, Black, Asian/PI, Native American or other
- Ethnicity: Hispanic, non-Hispanic or unknown
- Diabetes as cause of ESRD
- Duration of ESRD:
- Less than one year
- 1-2 years
- 2-3 years
- 3+ years
- Nursing home status in previous 365 days:
- None (0 days)
- Short term (0-89 days)
- Long term >=90 days)
- BMI at ESRD incidence:
- BMI < 18.5
- .5 ≤ BMI < 25
- 25≤ BMI < 30
- BMI ≥30
- Comorbidities at ESRD incidence:
- Atherosclerotic heart disease
- Cardiac disease
- Diabetes other than as primary cause of ESRD (all types including diabetic retinopathy)
- Congestive heart failure
- Inability to ambulate
- Chronic obstructive pulmonary disease
- Inability to transfer
- Malignant neoplasm, cancer
- Peripheral vascular disease
- Cerebrovascular disease, CVA, TIA
- Tobacco use (current smoker)
- Alcohol dependence
- Drug dependence
- No Medical Evidence (CMS-2728) Form
- At least one of the comorbidities listed
- A set of prevalent comorbidities based on Medicare inpatient claims (individual comorbidities categorized into 90 groups – see below)
- Includes an adjustment for Less than 6 Medicare covered months in prior calendar year
- Calendar year
- Medicare Advantage coverage
Beside main effects, two-way interaction terms between age, race, ethnicity, sex, duration of ESRD and diabetes as cause of ESRD are also included:
- Age and Race: Black
- Ethnicity and Race: Non-White
- Diabetes as cause of ESRD and Race
- Diabetes as cause of ESRD and Duration of ESRD
- Duration of ESRD: less than or equal to 1 year and Race
- Sex and Race: Black
Below we discuss how factors were considered for inclusion in the statistical risk model.
Risk adjustment factors were selected for testing based on several considerations, specifically clinical criteria, expert input, factors identified in the literature as associated with mortality, and data availability. We began with a large set of patient demographics, comorbidities (at ESRD incidence and prevalent), anthropometrics, and other characteristics. Facility characteristics were also considered. Risk factors were evaluated for appropriateness of the adjustment. For instance, it is important not to adjust for factors that reflect the results of treatment. Factors considered appropriate and supported in the literature were then investigated with statistical models, including interactions between sets of adjusters, to determine if they were empirically related to mortality. Risk factors were also evaluated for face validity as potential predictors of mortality.
Consideration of prevalent comorbidities as risk adjusters, in addition to incident comorbidities, is in part a response to stakeholder interest to adjust for more current (prevalent) comorbidities to reflect the current health status of dialysis patients, and conditions associated with mortality. CMS contracted with UM-KECC to convene a Technical Expert Panel (TEP) to consider the addition of prevalent comorbidities in the SMR and SHR risk adjustment models. The summary report for the TEP can be found here: https://dialysisdata.org/content/esrd-measures. Specific objectives of this TEP and a detailed description of the evaluation process and criteria for identifying appropriate comorbidities for adjustment are provided above.
This process resulted in the TEP recommending a list of 210 individual ICD-9 diagnosis codes for inclusion as risk adjustors. The TEP further recommended that: (1) comorbidities for inclusion as risk-adjusters in a particular year should be present in Medicare claims in the preceding calendar year; and (2) determination of a prevalent comorbidity required at least two outpatient claims or one inpatient claim. With the expansion of diagnostic codes that accompanied the transition from ICD-9 to ICD-10 in 2015, the original list of 210 comorbidities grew to over 1000 ICD-10 codes. The 210 individual ICD-9 codes were collapsed into 91 clinical groups using the AHRQ CCS categories as the framework for grouping the selected prevalent comorbidities. Using a crosswalk, the ICD-10 codes were then mapped to the 91 clinical comorbidity groups that are included in the SMR risk adjustment model (comorbidity groups are listed in the model results table in the section below). The decision to group the comorbidities was to achieve greater model parsimony.
Ascertainment of prevalent comorbidities is based on both outpatient (OP, SN, HH, HS, and PS claim types) and inpatient (IN claim types) Medicare claims, including those from Part C.
A patient is considered to have a particular prevalent comorbid condition if one of the ICD-10 codes for that condition (see SMR_DataDictionary_Final_Oct 2025.xlsx for list of codes) appears on a claim for the patient in the prior year. If no such claim is found, the patient is considered to not have the condition. If a patient has less than 6 months of Medicare coverage in the prior year, we consider the prevalent comorbidity information to be missing. This requirement is intended to allow us to distinguish between a patient who does not have a particular comorbidity from one who does not have claims during enough of the year to determine whether the condition is present or not.
Finally, SDS/SES factors were evaluated based on appropriateness (whether related to disparities in care), empirical association with the outcome, and support in published literature.
References:
[1] Cox, D.R. (1972) Regression Models and Life Tables (with Discussion). J. Royal statistical Society, Series B, 34, 187-220.
[2] Kalbfleisch, J.D. and Prentice, R. L. The Statistical Analysis of Failure Time Data. Wiley, New York, 2002.
Please see SMR_5.4.3_Final_Oct 2025_508, attached to 5.4.3a, for full response to this question
Please see SMR_5.4.4_Final_Oct 2025_508, attached to Section 7.1 Supplemental Attachment, for full response to this question
To assess model performance, we evaluated discrimination using the C-statistics. The C-statistics quantifies the model’s ability to discriminate between outcomes based on the included risk factors. Specifically, the SMR model is a time-to-event model, for which the C-statistics measures the concordance between the observed mortality rates and the model-based predicted rates.
The C-statistic for SMR is 0.68, which indicates moderate model discrimination, reflecting the model’s ability to distinguish high-risk from low-risk subjects.
Note: this text is also uploaded as an attachment to 5.4.5a since that is a required field.
In addition to clinical factors, we evaluated patient- and area-level SDS/SES social risk factors as risk adjusters. These were in addition to the current inclusion of race, ethnicity, and sex included in the currently endorsed and implemented SMR as described in the 2016 submission.
The relationships among individual SDS factors, socioeconomic disadvantage and mortality is well-established in the general population [18] [23] [24]. Further, individual- and market- or area-level measures of deprivation have been shown to contribute independently to higher mortality [19].
The relationship between race and mortality, Hispanic ethnicity and mortality, as well as both race and area-level SES factors and mortality in the dialysis population, is also well documented [1] [5] [7] [8] [10] [11] [12] [15] [16] [20] [28] [29]. However, the direction of the relationship between race and mortality is inverted relative to the general population, with lower observed mortality in blacks on chronic dialysis compared to whites, although the relationship is mediated by sociodemographic and clinical factors [13] [14] [3].
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:
- Employment status 6 months prior to ESRD
- 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 [22]. The ADI reflects a full set of SES characteristics, including measures of income, education, and employment status, measured at the ZIP code level.
Our fully risk-adjusted model includes all but two of the SDS factors listed above. The Medicare Dual Eligibility covariate was noted to have a very small hazard ratio estimate and was not significantly associated with mortality when other SDS and clinical risk factors were included. The ADI variable was significantly associated with mortality, but the parameter estimate was extremely small. Each 10-point increase in ADI is associated with only a 1% increase in the risk of mortality. In addition, otherwise fully risk adjusted models that either included or excluded the dual eligible and ADI covariates resulted in very similar dialysis facility-level flagging. Given these results we chose to exclude these two socioeconomic covariates to contribute to model parsimony. We did consider whether stratified reporting could be used, but the extremely limited contribution of these two variables did not justify that approach, in our opinion. In addition, the small numbers of patients treated in most U.S. facilities, results in suppression of stratified results for a relatively large percentage of facilities, based on the small cell size rules utilized in federal public reporting and other federally supported programs.
References:
[1] Burrows N, Cho P, Bullard KM, Narva A, and Eggers P. Survival on Dialysis Among American Indians and Alaska Natives With Diabetes in the United States, 1995–2010. American Journal of Public Health. Supplement 3, 2014, Vol 104, No. S3. S490.
[2] CDC National Vital Statistics Reports, Vol. 61, No. 6, October 10, 2012, Table A
[3] Cowie C, Port F, Rust K, Harris M: Differences In Survival Between Black And White Patients With Diabetic End-Stage Renal Disease. Diabetes Care 17: 681–687, 1994
[4] Cox, D.R. (1972) Regression Models and Life Tables (with Discussion). J. Royal statistical Society, Series B, 34, 187-220.
[5] Crews D, Sozio S, Liu Y, Coresh J, and Powe N. Inflammation and the Paradox of Racial Differences in Dialysis Survival. J Am Soc Nephrol 22: 2279–2286, 2011.
[6] Curtin R, Oberley E, Sacksteder P, and Friedman A. Differences Between Employed and Nonemployed Dialysis Patients. AJKD Vol 27:4. (April) 1996. 533-540.
[7] Eisenstein E, Sun J, Anstrom K, Stafford J, Szczech L, Muhlbaier L, Mark D. Do Income Level and Race Influence Survival in Patients Receiving Hemodialysis? The American Journal of Medicine (2009) 122, 170-180.
[8] Johns T, Estrella M, Crews D, Appel L, Anderson C, Ephraim P, Cook C, and Boulware L. Neighborhood Socioeconomic Status, Race, and Mortality in Young Adult Dialysis Patients. Am Soc Nephrol 25: epub, 2014.
[9] Kalbfleisch, J.D. and Prentice, R. L. The Statistical Analysis of Failure Time Data. Wiley, New York, 2002.
[10] Kalbfleisch J, Wolfe R, Bell S, Sun R, Messana J, Shearon T, Ashby V, Padilla R, Zhang M, Turenne M, Pearson J, Dahlerus C, and Li Y. Risk Adjustment and the Assessment of Disparities in Dialysis Mortality Outcomes. J Am Soc Nephrol 26: 2641–2645, 2015
[11] Kimmel P, Fwu CW, and Eggers P. Segregation, Income Disparities, and Survival in Hemodialysis Patients. JASN. February 2013 vol. 24 no. 2 293-301.
[12] Kucirka L, Grams M, Lessler J, Hall E, James J, Massie A, Montgomery R, and Segev D. Age and Racial Disparities in Dialysis Survival. JAMA. 2011 August 10; 306(6): 620–626. doi:10.1001/jama.2011.1127
[13] Norris K, Mehrotra R, Nissenson A. Racial Differences in Mortality and ESRD. American Journal of Kidney Diseases, Volume 52, Issue 2, August 2008, Pages 205–208.
[14] Powe, NR. Reverse Race And Ethnic Disparities In Survival Increase With Severity Of Chronic Kidney Disease: What Does This Mean? Clin J Am Soc Nephrol 1: 905–906, 2006;
[15] Ricks R, Molnar M, Kovesdy C, Kopple J, Norris K, Mehrotra R, Nissenson A, Arah O, Greenland S, and Kalantar-Zadeh K. Racial and Ethnic Differences in the Association of Body Mass Index and Survival in Maintenance Hemodialysis Patients. Am J Kidney Dis. 2011 October ; 58(4): 574–582.
[16] 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.
[17] Singh, G. Area Deprivation and Widening Inequalities In US Mortality, 1969–1998. Am J Public Health. 2003;93(7):1137–1143
[18] Singh G and Siahpush M. Widening Socioeconomic Inequalities In US Life Expectancy, 1980–2000. Int. J. Epidemiol. (August 2006) 35 (4): 969-979
[19] Smith G, Hart C, Watt G, Hole D, Hawthorne V. Individual Social Class, Area-Based Deprivation, Cardiovascular Disease Risk Factors, And Mortality: the Renfrew and Paisley Study. J Epidemiol Community Health 1998; 52:399-405
[20] Streja E, Kovesdy C, Molnar M, Norris K, Greenland S, Nissenson A, Kopple J, and Kalantar-Zadeh K. Role of Nutritional Status and Inflammation in Higher Survival of African American and Hispanic Hemodialysis Patients. Am J Kidney Dis. 2011 June ; 57(6): 883–893.
[21] University of Wisconsin School of Medicine Public Health. 2015 Area Deprivation Index v2.0. Downloaded from https://www.neighborhoodatlas.medicine.wisc.edu. Accessed 10/31/2018.
[22] University of Wisconsin School of Medicine Public Health. 2013 Area Deprivation Index v1.5. Downloaded from https://www.neighborhoodatlas.medicine.wisc.edu/ October 31, 2018.
[23] 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 February 2006.
[24] 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.
[25] Wolfe RA et al: New Dialysis Facility Mortality Statistics (Smrs) Adjust For More Patient Characteristics. J Am Soc Nephrol 2001; 12; A1802
[26] Wolfe R et al. Using USRDS Generated Mortality Tables To Compare Local ESRD Mortality Rates To National Rates. Kidney Int 1992; 42: 991-96
[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, Xin W, Ma J, Yu A, Greene T, Yu W, and Cheung A. Facility Size, Race and Ethnicity, and Mortality for In-Center Hemodialysis. J Am Soc Nephrol 24: 2062–2070, 2013. doi: 10.1681.
[29] Yan G, Norris K, Yu A, Ma J, Greene T, Yu W, and Cheung A. The Relationship of Age, Race, and Ethnicity with Survival in Dialysis Patients. Clin J Am Soc Nephrol 8: 953–961, 2013. doi: 10.2215.
Use & Usability
Use
Dialysis Facility Compare helps patients find detailed information about Medicare-certified dialysis facilities. Patients 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 3 expected deaths are included in the measure calculation for the program. For the October 2024 Dialysis Facility Compare refresh, 7,324 U.S. dialysis facilities serving a total of 2,090,013 patients treated in the 4-year period from 2020-2023 had SMR results reported.
Facility level, Dialysis Facilities
Usability
There are a number of actions that dialysis facility providers can take to help improve patient mortality. Examples include:
- Optimize dialysis adequacy: Ensuring that adequate small solute clearance is achieved by measuring Kt/V regularly and making appropriate dialysis prescription adjustments. In addition, encouraging patients to complete the full duration of their treatments along with not missing treatments is also important.
- Managing cardiovascular risk factors: Controlling blood pressure and optimizing fluid management to avoid chronic volume overload is central to reducing cardiovascular morbidity. Attention to ultrafiltration rates during treatment can also impact cardiovascular outcomes.
- Infection prevention: Monitoring and reducing blood-stream infections, particularly those that are dialysis catheter related is a cornerstone. In addition, promoting and administering vaccinations with influenza, pneumococcal, and hepatitis B.
- Nutrition and Metabolic support: Managing mineral and bone disorder (MBD) with control of hyperphosphatemia, avoidance of hypercalcemia, and treatment of hyperparathyroidism can all impact patient mortality.
- Improve vascular access by avoiding long-term catheters when possible.
- Enhance care coordination: Reconcile medications when patients return from hospital.
- Encourage kidney transplantation which has improved patient survival compared to long-term dialysis.
For DFCC, feedback can be provided any time through contacting the dialysisdata.org helpdesk. Preview periods allow for specific times for facilities review and comment on measure calculations, and provide an opportunity to request a patient list.
Comments received during DFC preview periods tend to be technical in nature, asking for clarification on how the SMR is calculated for particular facilities, including questions about patient assignment and application of exclusion and risk adjustment criteria.
The revisions made to the measure specifications during this maintenance review were not directly in response to specific feedback received during public reporting (which, as described above, was more general in nature).
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.
Mortality rates from 2020-2023 may be difficult to interpret because of the COVID-19 pandemic effects, but mortality rates are lower in 2023 (reference year) as evidenced by the hazard ratios for calendar year from the SMR model. The risk of mortality for 2020 was 5% higher compared to 2023 (p-value<0.0001). The risks of mortality in 2021 and 2022 were 12% and 6% higher, respectively, compared to 2023 (p-value <0.0001 for each year).
2020: Coefficient = 0.05, Hazard Ratio= 1.05, P-value = <0.0001
2021: Coefficient = 0.11, Hazard Ratio= 1.12, P-value = <0.0001
2022: Coefficient = 0.05, Hazard Ratio= 1.06, P-value = <0.0001
2023: Reference Category
None
Comments
Staff Preliminary Assessment
CBE #0369 Staff Preliminary Assessment
Importance
Strengths
- A clear logic model is provided, depicting the relationships between inputs (e.g., dialysis facility staff and quality improvement team), activities (e.g., strict adherence to infection prevention and control standards, identifying high-risk patients for interventions), and desired outcomes (e.g., improvement in dialysis access infections, reduction in emergency department visits, and reduction in chronic dialysis patient mortality rates). This model demonstrates how the measure's implementation will lead to the anticipated outcomes.
The problem this measure addresses represents a significant issue, mortality among end stage renal disease (ESRD) patients on chronic dialysis. By highlighting facility-level differences in patient outcomes, the measure helps ensure that issues such as infections, cardiovascular instability, electrolyte disturbances, and inadequate management of chronic volume and mineral disorders are addressed, ultimately improving patient survival and quality of life.
The measure is supported by a comprehensive literature review, including systematic reviews/meta-analyses (e.g., Song, 2020), randomized trials (Rosenblum, 2014), and observational cohort studies (e.g., Magdalene, 2016), demonstrating a clear net benefit in terms of reducing mortality for ESRD patients on chronic dialysis.
Data from 7,824 dialysis facilities from 2020-2023 show a performance gap, with decile ranges from 0.58 to 1.51, indicating variation in measure performance across the target population.
Description of patient input supports the conclusion that the measured outcome is meaningful with at least moderate certainty. Patient input was demonstrated through two studies (2024 and 2020) that indicated dialysis patients with kidney failure value a facility-level assessment of morality rates.
Limitations
- Evidence is predominately observational, showing associations between facility practices and mortality.
Rationale
- This maintenance measure meets all criteria for 'Met' for importance due to the significance of the problem it addresses, its robust evidence base, a documented performance gap, and well-articulated logic model, making it essential for addressing mortality rates of ESRD patients on chronic dialysis.
There is at least moderate confidence that the business case is adequate, i.e., the anticipated impacts of the measure on patient outcomes justify use of the measure.
Closing Care Gaps
The developer did not address this optional domain.
Feasibility Assessment
Strengths
- All required data elements are routinely generated and available in structured fields within administrative data, claims data, or registries (e.g., transplant data).
The developer indicated there have been no changes to the measure specification.
The developer described that there are no additional costs and burden associated with data collection and data entry, validation, and analysis as the data are routinely collected. Based on comments they received during facility previews, there are only rare instances of inaccurate or missing data.
The developer described how all required data elements can be collected without risk to patient confidentiality, including restricting public reporting to facilities with at least three expected deaths to avoid small cell size issues.
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 identified.
Rationale
- This maintenance 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 health care setting.
Scientific Acceptability
Strengths
- The developer performed the required reliability testing for this maintenance measure, namely, they conducted accountable entity-level (“measure score”) reliability testing at the level for which the measure is specified. Data sources used for reliability analysis are adequately described and include data from over 2.1 million ESRD dialysis patients at 7,840 facilities during the four-year period of 2020-2023.
Limitations
- The developer conducted signal-to-noise reliability testing at the accountable entity-level. Less than 20% of accountable entities were above the expected threshold of 0.6. The developer did provide some interpretation and rationale for these results. They acknowledge that the low reliability is driven by the typical small size of dialysis facilities, but that a lower standard for reliability may be warranted in this situation due to the impactfulness of this measure. They state that the impact of small facility size is reduced by combination with other measures for Star Ratings and that QIP uses a small facility adjuster to limit effects on payment reductions.
Rationale
- This maintenance measure is rated as ‘Not Met’ for reliability because the reliability testing results significantly fall below the established thresholds, indicating major issues with the consistency and accuracy of the results across different settings and populations. Reliability will be even lower for periods of performance less than four years. The developer did provide an interpretation and some rationale for these results for consideration.
Strengths
- The developer performed the required validity testing for this maintenance measure, namely, they provided accountable entity-level (“measure score”) validity testing at the level for which the measure is specified. Data sources used for validity analysis are adequately described and include administrative data and Medicare claims (Parts A, B, and C), and additional data from the Scientific Registry of Transplant Recipients (SRTR), the Nursing Home Minimum Dataset, and the Internet Quality Improvement and Evaluation System (iQIES) from the period 2020-2023. The number of entities included in the analysis ranged from 7,979 (2023) to 8,108 (2021), and entities annually served a median 60 to 62 patients each.
The developer conducted Spearman's rank-order correlations at the accountable entity-level between the measure and 6 other measures that the evidence review indicates are related to the measure focus, hypothesizing the direction of the relationship and rationale, but not the expected magnitude, for each: Vascular Access: Standardized Fistula Rate (SFR) (negative correlation); Kt/V ≥ 1.2 (negative); Vascular Access: Long-term catheter rate (LTC) (positive correlation); Standardized Hospitalization Ratio (SHR) (positive); Standardized Readmission Ratio (SRR) (positive); and, Standardized Transfusion Ratio (STrR) (positive). The results showed statistically significant correlations coefficients all in the expected directions, with negative correlations for SFR (rho = -0.11) and Kt/V>=1.2 (rho = -0.12),and positive correlations for LTC (rho = 0.08), SHR (rho = 0.17), SRR (rho = 0.07), and STrR (rho = 0.15).
A well-developed logic model and thorough literature review suggest adequate “ruling in” of mechanisms that can explain the measure focus.
The developer conducted statistical risk adjustment, based on a conceptual model, selecting risk factors that are present at the start of care and have a significant correlation with the outcome.
Limitations
- While the results of entity-level validity testing supported the developer's hypotheses regarding expected direction of effects, the coefficients indicated weak to negligible relationships. The developer did not provide hypotheses for expected magnitudes of these relationships, or a rationale for why the coefficients are so small. For example, was it expected that SMR's relationship with SHR and STrR would be the strongest of this group, and if so, why? The submission could be strengthened by a discussion of these results.
The measure's low reliability does not support an inference of validity, because it may indicate that observed relationships are not real.
The variable distributions provided by the developer are presented in pooled form, therefore they do not demonstrate differences in facility-level prevalence of risk factors that would warrant adjustment for fair comparisons. The developer reported a c-statistic of 0.68, indicating moderate model discrimination, but did not provide performance metrics reflecting model calibration.
Rationale
- This maintenance measure is rated as ‘Not Met But Addressable’ for validity because the validity testing results partially support an inference of validity for the measure, suggesting that the measure somewhat accurately reflects performance on quality and/or resource use and can distinguish good from poor performance to a limited extent.
The risk adjustment methods used are appropriate and demonstrate that the risk factors contribute to unique variation in the outcome. The model discrimination is acceptable, but the developer did not provide model calibration testing results or evidence of variation in the prevalence of risk factors across measured entities.
Use and Usability
Strengths
- The measure is currently used in CMS Dialysis Facility Care Compare.
The developer provided a summary of how accountable entities can use the measure results to improve performance. Specifically, managing cardiovascular risk factors, infection prevention activities, nutrition and metabolic support, and optimizing dialysis adequacy. These possible actions are reflected in the measure’s logic model.
Interested parties can submit ongoing feedback through an online helpdesk. A preview period is also held to allow dialysis facilities to review, comment, and ask questions about measure calculations before public reporting. The developer noted that dialysis facilities ask technical questions related to patient assignment and risk adjustment criteria.
The developer reported changes in performance from calendar years 2020-2023. Compared to 2023, mortality risk was 5% higher in 2020, 12% higher in 2021, and 6% higher in 2022. These results indicate measurable improvement over time, supporting the usability of this measure.
The developer reported no unexpected findings.
Limitations
- The developer notes that mortality rates from 2020-2023 may be difficult to interpret due to the effects of the COVID-19 pandemic.
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. The measure also demonstrates a positive trend in performance results, affirming its ongoing usability. The developer reported no unexpected findings.
Committee Independent Review
Support
Importance
Literature appears to be primarily older, although the developers do describe a more recent review.
Closing Care Gaps
Not addressed/required
Feasibility Assessment
Leverages extant data with no additional collection burden.
Scientific Acceptability
Testing results did not meet generally accepted thresholds, but mitigation strategies and rationale for lower results were presented. This contextual information should be reflected in public reporting (in plain language) for awareness,
Agree with staff assessment. More information on the hypotheses underlying the testing would be useful.
Use and Usability
Would be helpful to have examples of any specific/successful interventions individual facilities have undertaken to address mortality and/or QI activities stemming from/using measure results.
Summary
Would be interested in learning more about how the three age groups were determined for the purposes of risk adjustment.
I placed comments in above…
Importance
Agree with staff review
Closing Care Gaps
Agree with staff review
Feasibility Assessment
Please provide clarification on the use of Medicare Advantage claims- per the specs- the numerator only includes Medicare Part A, not Med Advantage-, If I am reading correctly, there is an adjustment added for Medicare Advantage patients. However, per the additional supplemental information that was provided, I am reading this to say that during this 2025 measure cycle, they are recommending inclusion of Medicare Adv as they now have access to this data. If so, will there be updates to the measure specifications to show this inclusion?
Scientific Acceptability
Agree with Staff assessment
Agree with staff assessment
Use and Usability
Agree with staff assessment
Summary
I placed comments in above areas to ask for some clarification on this measure. Also, I agree with the ASN comments, regarding excluding the patient if it were their choice to withdraw from dialysis and I appreciate the ASN comment on modification for a risk-standardized rate instead of a ratio.
Measure 0369 Summary
Importance
Agree with staff assessment: strong logic model with evidence based relationships supported between processes and outcomes, updated literature review, variation across facilities (performance gap).
In addition, American Society of Nephrology supported importance in public comments.
Literature review for consumers/patients mixed about importance. While there were references supporting importance of mortality to patients, there also were references indicating that quality of life may be more important to them than mortality rates.
Closing Care Gaps
Optional and not addressed
Feasibility Assessment
Agree with staff assessment. Data for this measure available in current data sets.
Scientific Acceptability
Measure of reliability (IUR) does not reach acceptable level. Attributed to small sample size at dialysis facilities which will continue to be an issue.
Question for measure developers - can this be improved?
Agree with staff review - direction of expected relationships was consistently as hypothesized but magnitude low.
Risk adjustment methods reviewed by TEP.
Question to measure developers - explanation for low correlations?
Use and Usability
Currently used in public reporting
While important to have an overall measure of success at prevention of potential causes of mortality, this indicator does not provide information about the multiple potential causes at play in increased mortality rates, e.g. infection, volume depletion, electrolytes etc. Usability is strong only to the extent that each of these potential causes also are tracked and measured.
Summary
Important aggregate indicator for dialysis patients. Strong feasibility. Reliability and validity measures low. Wonder about the overall benefit of aggregate measures like mortality if major contributing causes not tracked.
Patient Partner review of 0369 SMR
Importance
As a patient partner, I believe this is met as the measure is evidence based (been in use since 2008) and is making gains in healthcare quality (mortality rates) and costs. For example, preventing avoidable deaths also avoids some of the costs of treating the life-threatening complications.
Closing Care Gaps
Not required
Feasibility Assessment
The data required for this measure is already part of routine data collection, no additional costs or burden are anticipated.
Scientific Acceptability
As a patient partner, I am not a technical expert in this area. As such, I agree with the staff assessment.
As a patient partner, I am not a technical expert in this area. As such, I agree with the staff assessment.
Use and Usability
In use via CMS Dialysis Facility Compare which helps patients find detailed information about Medicare-certified dialysis facilities
Summary
I support this measure but would like to discuss the public comments.
Patients who withdrew from dialysis remain assigned to their treatment facility for 60 days after withdrawal – it seems these patients should be excluded in the rate as withdrawal from dialysis would lead to death.
Edit: I looked up the demographics and it appears that these patients may already be accounted for in the forecasted number of deaths due to demographics: The demographics of dialysis withdrawal have been studied at length. Patient characteristics associated with withdrawal are older age, female, white race, longer duration of dialysis, higher educational level, living alone, severe pain, and comorbidity (with chronic or progressive diseases)
agree with exception proposed by ASN
Importance
agree
Closing Care Gaps
agree
Feasibility Assessment
agree
Scientific Acceptability
agree
agree - except would appreciate discussion of ASN comment about use of rate vs relying upon o/e (?) ratio
Use and Usability
agree with comment about excluding patients who withdraw from dialysis
Summary
would exclude patients who withdraw from ongoing dialysis; would also appreciate discussion of rationale for using o/e ratio vs rate
discontinue this measure
Importance
From the measure: "“Although the basic technology of hemodialysis has not changed dramatically in the last two decades, overall mortality of individuals on chronic dialysis has improved, both in absolute and relative terms [48, 49]. The trend towards reduced mortality is temporally correlated with the introduction of public reporting of the Standardized Mortality Ratio (SMR) in the early 2000’s.”
Correlation is not evidence of the importance of this measure. As the measure documents, there have been a variety of other factors that may have contributed to decreased mortality. The largest of these is the increase of modalities other than in center hemodialysis.
This is a very general measure and has not effect on whether other patient care improvements will be undertaken by a particular center. For this reason, the usability of the measure is zero.
Feasibility Assessment
Scientific Acceptability
Use and Usability
The resulting statistical information provides no guidance to dialysis units for improvement.
So what potential use does this measure have for patients? Would new patients rely on this information to make decisions about their care i.e. where to dialyze? They would be naive if they did and if they did, there is no guarantee they would receive better care.
Mortality depends on so many factors as this measure states. So the usability from a patient's perspective is not there either.
Summary
This measure provides no value to patients or dialysis facilities. That it is easy to collect and statistically analyze is not a reason to continue it.
#0369 summary
Importance
Agree with staff prelim assessment; logic model is appropriate and a care/practice gap exists. The measure steward reports that providers and administrators place a higher value on this metric than patients, who place higher value on QOL.
Closing Care Gaps
Not addressed
Feasibility Assessment
Agree with staff prelim assessment. Data is captured in structured and standardized way; no burden for centers
Scientific Acceptability
I defer to the Staff prelim assessment. I do understand the low IUR (0.47), however the stewards report that this IUR is similar to the value presented in past endorsement and re-endorsement cycles. Were there requests or expectations for additional data during those review cycles?
The correlations for known groups analysis is logical and results were as predicted, however the staff preliminary assessment makes a good point in relation to magnitude of the relationships
Use and Usability
This metric is currently used in Care Compare. Per the logic model, there are structures and processes to evaluate for quality improvement if an org is not meeting benchmarks.
I do agree with ASN comments about excluding people who choose to withdraw from care
Summary
Payers, regulatory agencies, administrators, and quality/safety teams always want a mortality metric when assessing care quality. To that end, I support the continuation of this metric, but I do think the ASN recommendations for exclusions should be thoughtfully considered
Nothing more to add
Importance
Agree with staff assessment
Closing Care Gaps
Not required
Feasibility Assessment
Agree with staff assessment.
Scientific Acceptability
Agree with staff assessment.
Agree with staff assessment.
Use and Usability
Agree with staff assessment
Summary
Nothing more to add.
Await committee discussion
Importance
The measure is important for making significant gains in health care quality or cost where there is variation in or overall less-than-optimal performance.
Closing Care Gaps
Not addressed. The measure can identify and reduce gaps in care for ESRD patients
Feasibility Assessment
Data are readily available OR could be captured without undue burden AND can be implemented for performance measurement.
Scientific Acceptability
The measure, as specified, can produce consistent (reliable) results about the quality of care
The measure, as specified, can produce credible (valid) results about the quality of care
Use and Usability
Measure results can be used for both accountability and performance improvement to achieve the goal of high-quality efficient health care for individuals or populations.
Summary
Await committee discussion
Important unintended consequences raised in public comments
Importance
Importance is substantiated. Excellent logic model. Concur with staff assessment.
Closing Care Gaps
Not required/not submitted.
Feasibility Assessment
Concur with staff assessment.
Scientific Acceptability
agree with staff assessment
agree with staff assessment
Use and Usability
agree with staff assessment
Summary
Important measure. The ASN comments suggest an exclusion to avoid unintended consequences--dialysis centers discouraging patient choice to terminate dialysis. This seems like a reasonable exclusion and should be discussed. Also would like to discuss ratio vs. rate as basis for measure as recommended by ASN.
meaningful but careful
Importance
Appreciate the importance
Closing Care Gaps
optional
Feasibility Assessment
Support
Scientific Acceptability
agree with staff assessment
agree with staff assessment
Use and Usability
support
Summary
While it may be significant, careful integration as a measure would be necessary when dealing with mortality.
Public Comments
ASN Comments on Standard Mortality Ratio for Dialysis Facilities
ASN continues to support the Standard Mortality Ratio for Dialysis Facilities but emphasizes that death due to patient choice to withdraw from dialysis should be explicitly excluded from the measure. Mortality-based metrics should not result in a facility working to dissuade individuals from making an informed choice to withdraw from dialysis should the benefits of dialysis no longer exceed the burden of dialysis. In addition, this measure should be modified to be a true risk-standardized rate as opposed to a ratio. This modification allows accurate assessment of improvement as well as accurate benchmarking, elements that are critical to the quality program and to individual dialysis facilities as they seek to improve the care provided to individual patients.