[Skip to Content]
[Skip to Content Landing]
Figure 1.
Weighted Kaplan-Meier Curves for Cardiovascular Events
Weighted Kaplan-Meier Curves for Cardiovascular Events
Figure 2.
Results of Subgroup Analyses for the Outcome Cardiovascular Events Primary As-Treated Analysis
Results of Subgroup Analyses for the Outcome Cardiovascular Events Primary As-Treated Analysis

Error bars indicate 95% CI. CV indicates cardiovascular.

Table 1.  
New Users of Phosphate Binders With Therapy Start Between Day 1 and 180 Days After Start of Hemodialysis
New Users of Phosphate Binders With Therapy Start Between Day 1 and 180 Days After Start of Hemodialysis
Table 2.  
Incidence Rates of the Cardiovascular Events and Cardiovascular Mortality (Primary) and All-Cause Mortality (Secondary) Outcomes
Incidence Rates of the Cardiovascular Events and Cardiovascular Mortality (Primary) and All-Cause Mortality (Secondary) Outcomes
Table 3.  
Main Results Comparing Patients Receiving Sevelamer vs Calcium Acetatea
Main Results Comparing Patients Receiving Sevelamer vs Calcium Acetatea
Supplement.

eMethods. Technical Details on Statistical Models

eFigure 1. Schematic Depiction of Study Design

eFigure 2. Cohort Assembly Flow Chart

eFigure 3. Monthly Mean Weights of the Morginal Structural Model for Cardiovascular Events

eFigure 4. Monthly Mean Weights of the Morginal Structural Model for All-Cause Mortality

eTable 1. Full Set of Baseline Characteristics Before and After Propensity Score Weighting

eTable 2. Censoring Reasons and Follow-up Time Before and After PS Weighting for Outcome Fatal or Non-Fatal Cardiovascular Event

eTable 3. Censoring Reasons and Mean Follow-up Before and After PS Weighting for Outcome All-Cause Mortality

eTable 4. Censoring Reasons and Mean Follow-up Including Data 2008-2013

eFigure 5. Results of Subgroup Analyses for the Outcome All-Cause Mortality (Primary As Treated Analysis)

eFigure 6. Weighted Kaplan Meier Curves for All-Cause Mortality in the As-Treated and Intention-to-Treat Follow-up (Bottom) Model in the Primary (left) and Secondary Study Population

eTable 5. Serum Phosphorus and Albumin-Corrected Calcium Values Over Time

eTable 6. Post-hoc Sensitivity Analysis in Patients With and Without Nephrology Care Prior to Start of HD (As Treated Follow-up)

eFigure 7. Kaplan Meier Curves (As-Treated) for All-Cause Mortality Comparing Sevelamer to Calcium Acetate in Patients Without Nephrology Care Prior to HD Start

eTable 7. Cohort Creation Flor Chart for New Prevalent Users of Phosphate Binders Used in eTable 8

eTable 8. Baseline Covariates in Prevalent Phosphate Binder Users (Excluded) vs Incident Phosphate Binder Users (Included) Before PS Weighting

1.
St Peter  WL, Wazny  LD, Weinhandl  E, Cardone  KE, Hudson  JQ.  A review of phosphate binders in chronic kidney disease: incremental progress or just higher costs?  Drugs. 2017;77(11):1155-1186. doi:10.1007/s40265-017-0758-5PubMedGoogle ScholarCrossref
2.
Spoendlin  J, Schneeweiss  S, Tsacogianis  T,  et al.  Association of Medicare’s bundled payment reform with changes in use of vitamin D among patients receiving maintenance hemodialysis: an interrupted time-series Analysis.  Am J Kidney Dis. 2018;72(2):178-187. doi:10.1053/j.ajkd.2018.03.027PubMedGoogle ScholarCrossref
3.
Kidney International Supplements. KDIGO 2017 Clinical Practice Guideline for the Diagnosis, Evaluation, Prevention and Treatment of Chronic Kidney Disease Mineral and Bone Disorder (CKD-MBD). https://kdigo.org/wp-content/uploads/2017/02/2017-KDIGO-CKD-MBD-GL-Update.pdf. Published July 2017. Accessed March 4, 2019.
4.
St Peter  WL, Wazny  LD, Weinhandl  ED.  Phosphate-binder use in US dialysis patients: prevalence, costs, evidence, and policies.  Am J Kidney Dis. 2018;71(2):246-253. doi:10.1053/j.ajkd.2017.09.007PubMedGoogle ScholarCrossref
5.
Block  GA, Raggi  P, Bellasi  A, Kooienga  L, Spiegel  DM.  Mortality effect of coronary calcification and phosphate binder choice in incident hemodialysis patients.  Kidney Int. 2007;71(5):438-441. doi:10.1038/sj.ki.5002059PubMedGoogle ScholarCrossref
6.
Di Iorio  B, Molony  D, Bell  C,  et al; INDEPENDENT Study Investigators.  Sevelamer versus calcium carbonate in incident hemodialysis patients: results of an open-label 24-month randomized clinical trial.  Am J Kidney Dis. 2013;62(4):771-778. doi:10.1053/j.ajkd.2013.03.023PubMedGoogle ScholarCrossref
7.
Suki  WN, Zabaneh  R, Cangiano  JL,  et al.  Effects of sevelamer and calcium-based phosphate binders on mortality in hemodialysis patients.  Kidney Int. 2007;72(9):1130-1137. doi:10.1038/sj.ki.5002466PubMedGoogle ScholarCrossref
8.
Patel  L, Bernard  LM, Elder  GJ.  Sevelamer versus calcium-based binders for treatment of hyperphosphatemia in CKD: A meta-analysis of randomized controlled trials.  Clin J Am Soc Nephrol. 2016;11(2):232-244. doi:10.2215/CJN.06800615PubMedGoogle ScholarCrossref
9.
Yusuf  AA, Weinhandl  ED, St Peter  WL.  Comparative effectiveness of calcium acetate and sevelamer on clinical outcomes in elderly hemodialysis patients enrolled in Medicare part D.  Am J Kidney Dis. 2014;64(1):95-103. doi:10.1053/j.ajkd.2013.11.015PubMedGoogle ScholarCrossref
10.
Borzecki  AM, Lee  A, Wang  SW, Brenner  L, Kazis  LE.  Survival in end stage renal disease: calcium carbonate vs. sevelamer.  J Clin Pharm Ther. 2007;32(6):617-624. doi:10.1111/j.1365-2710.2007.00871.xPubMedGoogle ScholarCrossref
11.
United States Renal Data System. USRDS annual data report 2016, chapter 3: clinical indicators and preventive care. https://www.usrds.org/2016/view/v2_03.aspx. Accessed October 3, 2018.
12.
United States Renal Data System. https://www.usrds.org/. Accessed October 3, 2018.
13.
United States Renal Data System. Volume 2: ESRD analytic methods. https://www.usrds.org/2017/view/v2_00_appx.aspx#_Toc501030698. Accessed October 3, 2018.
14.
Kiyota  Y, Schneeweiss  S, Glynn  RJ, Cannuscio  CC, Avorn  J, Solomon  DH.  Accuracy of Medicare claims-based diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records.  Am Heart J. 2004;148(1):99-104. doi:10.1016/j.ahj.2004.02.013PubMedGoogle ScholarCrossref
15.
Kumamaru  H, Judd  SE, Curtis  JR,  et al.  Validity of claims-based stroke algorithms in contemporary Medicare data: reasons for geographic and racial differences in stroke (REGARDS) study linked with Medicare claims.  Circ Cardiovasc Qual Outcomes. 2014;7(4):611-619. doi:10.1161/CIRCOUTCOMES.113.000743PubMedGoogle ScholarCrossref
16.
Rocco  MV, Yan  G, Gassman  J,  et al; Hemodialysis Study Group; The National Institutes of Health-funded Hemodialysis. Health Care Financing Administration.  Comparison of causes of death using HEMO Study and HCFA end-stage renal disease death notification classification systems.  Am J Kidney Dis. 2002;39(1):146-153. doi:10.1053/ajkd.2002.29905PubMedGoogle ScholarCrossref
17.
Foley  RN, Collins  AJ.  The USRDS: what you need to know about what it can and can’t tell us about ESRD.  Clin J Am Soc Nephrol. 2013;8(5):845-851. doi:10.2215/CJN.06840712PubMedGoogle ScholarCrossref
18.
Desai  RJ, Rothman  KJ, Bateman  BT, Hernandez-Diaz  S, Huybrechts  KF.  A propensity score based fine stratification approach for confounding adjustment when exposure is infrequent.  Epidemiology. 2017;28(2):249-257. doi:10.1097/EDE.0000000000000595PubMedGoogle ScholarCrossref
19.
Austin  PC.  Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research.  Commun Stat Simul Comput. 2009;38(6):1228-1234. doi:10.1080/03610910902859574Google ScholarCrossref
20.
Franklin  JM, Rassen  JA, Ackermann  D, Bartels  DB, Schneeweiss  S.  Metrics for covariate balance in cohort studies of causal effects.  Stat Med. 2014;33(10):1685-1699. doi:10.1002/sim.6058PubMedGoogle ScholarCrossref
21.
Danaei  G, García Rodríguez  LA, Cantero  OF, Logan  RW, Hernán  MA.  Electronic medical records can be used to emulate target trials of sustained treatment strategies.  J Clin Epidemiol. 2018;96(96):12-22. doi:10.1016/j.jclinepi.2017.11.021PubMedGoogle ScholarCrossref
22.
Danaei  G, Rodríguez  LA, Cantero  OF, Logan  R, Hernán  MA.  Observational data for comparative effectiveness research: an emulation of randomised trials of statins and primary prevention of coronary heart disease.  Stat Methods Med Res. 2013;22(1):70-96. doi:10.1177/0962280211403603PubMedGoogle ScholarCrossref
23.
Hernán  MA, Brumback  B, Robins  JM.  Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men.  Epidemiology. 2000;11(5):561-570. doi:10.1097/00001648-200009000-00012PubMedGoogle ScholarCrossref
24.
Desai  RJ, Solomon  DH, Schneeweiss  S, Danaei  G, Liao  KP, Kim  SC.  Tumor necrosis factor alpha inhibitor use and the risk of incident hypertension in patients with rheumatoid arthritis.  Epidemiology. 2016;27(3):414-422. doi:10.1097/EDE.0000000000000446PubMedGoogle ScholarCrossref
25.
Goodman  WG, Goldin  J, Kuizon  BD,  et al.  Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis.  N Engl J Med. 2000;342(20):1478-1483. doi:10.1056/NEJM200005183422003PubMedGoogle ScholarCrossref
26.
Mansournia  MA, Etminan  M, Danaei  G, Kaufman  JS, Collins  G.  Handling time varying confounding in observational research.  BMJ. 2017;359:j4587. doi:10.1136/bmj.j4587PubMedGoogle ScholarCrossref
27.
Di Iorio  BR, Cucciniello  E, Bellizzi  V.  Vascular calcification and QT interval in incident hemodialysis patients.  J Nephrol. 2009;22(6):694-698.PubMedGoogle Scholar
28.
Niu  J, Shah  MK, Perez  JJ,  et al.  Dialysis modality and incident atrial fibrillation in older patients with ESRD.  Am J Kidney Dis. 2019;73(3):324-331.PubMedGoogle Scholar
29.
Block  GA, Spiegel  DM, Ehrlich  J,  et al.  Effects of sevelamer and calcium on coronary artery calcification in patients new to hemodialysis.  Kidney Int. 2005;68(4):1815-1824. doi:10.1111/j.1523-1755.2005.00600.xPubMedGoogle ScholarCrossref
30.
Raggi  P, Boulay  A, Chasan-Taber  S,  et al.  Cardiac calcification in adult hemodialysis patients: a link between end-stage renal disease and cardiovascular disease?  J Am Coll Cardiol. 2002;39(4):695-701. doi:10.1016/S0735-1097(01)01781-8PubMedGoogle ScholarCrossref
31.
Sato  H, Nagasawa  T, Saito  A, Miyazaki  M.  Risk of cardiovascular mortality predicted by the serum calcium level and calcification score at the initiation of dialysis.  Clin Exp Nephrol. 2018;22(4):957-966. doi:10.1007/s10157-017-1527-1PubMedGoogle ScholarCrossref
32.
Kaltenboeck  A, Bach  PB.  Value-based pricing for drugs: theme and variations.  JAMA. 2018;319(21):2165-2166. doi:10.1001/jama.2018.4871PubMedGoogle ScholarCrossref
33.
Nazareth  T, Ko  JJ, Sasane  R,  et al.  Outcomes-based contracting experience: research findings from US. and European stakeholders.  J Manag Care Spec Pharm. 2017;23(10):1018-1026.PubMedGoogle Scholar
34.
Palmer  SC, Gardner  S, Tonelli  M,  et al.  Phosphate-binding agents in adults with CKD: a network meta-analysis of randomized trials.  Am J Kidney Dis. 2016;68(5):691-702. doi:10.1053/j.ajkd.2016.05.015PubMedGoogle ScholarCrossref
35.
Bousher  A, Al-Makki  A, Sutton  J, Shepler  B.  A review of sucroferric oxyhydroxide for the treatment of hyperphosphatemia in patients receiving dialysis.  Clin Ther. 2014;36(12):2082-2093. doi:10.1016/j.clinthera.2014.09.021PubMedGoogle ScholarCrossref
36.
DOPPS Practice Monitor. Mineral bone disorder, national sample for use of iron based phosphate binders in the last 3 months. https://www.dopps.org/dpm/DPMSlideBrowser.aspx. Published 2018. Accessed August 2, 2018.
37.
Johns  TS, Estrella  MM, Crews  DC,  et al.  Neighborhood socioeconomic status, race, and mortality in young adult dialysis patients.  J Am Soc Nephrol. 2014;25(11):2649-2657. doi:10.1681/ASN.2013111207PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    EXPAND ALL
    Debate Not Which, But “If”
    Bharat Sachdeva, Associate Professor | LSU Heath SHREVEPORT
    I read with interest this study comparing the benefit of one drug vs another to bind phosphate and lower mortality.

    There is yet not one RCT that has shown that lowering phosphate, no matter how, reduces CKD or ESRD mortality.

    Diet exposure is the key. If one hot dog has 100 mg of phosphate you will need 4 calcium acetate or 5 sevelamer pills at a cost of ($15-$25) to bind all the phosphate in that hot dog. 

    Phosphate binders have given physicians a fall sense of security and an excuse to patients to increase phosphate
    intake knowing they can manage their levels with a binder. But the fact remains that binders will bind few mg of the total 1000+ mg that we ingest every day.

    It’s time we rethought our focus of managing phosphate in CKD patients toward a primary reduction of intake, as there is no evidence so far that phosphate binders will help them on meaningful measures of clinical outcomes.
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Original Investigation
    May 6, 2019

    Cardiovascular Outcomes of Calcium-Free vs Calcium-Based Phosphate Binders in Patients 65 Years or Older With End-stage Renal Disease Requiring Hemodialysis

    Author Affiliations
    • 1Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
    • 2Renal Section, Renal Division, Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
    • 3New England Geriatrics Research Education and Clinical Center, Veterans Affairs Boston Healthcare System, Boston, Massachusetts
    • 4Harvard Medical School, Boston, Massachusetts
    JAMA Intern Med. 2019;179(6):741-749. doi:10.1001/jamainternmed.2019.0045
    Key Points

    Question  Is the calcium-free phosphate binder sevelamer carbonate associated with superior cardiovascular end points compared with calcium acetate in patients 65 years or older with end-stage renal disease who are undergoing dialysis?

    Findings  In this cohort study of data from 2639 patients 65 years or older with end-stage renal disease in the United States Renal Data System, a similar risk of cardiovascular events and death between sevelamer and calcium acetate initiators was noted after adjusting for 78 potential confounders, including serum calcium and phosphorous levels.

    Meaning  This null result suggests that any potential increased safety of sevelamer compared with calcium-based phosphate binders on cardiovascular events observed in previous small trials with nonrepresentative populations may not translate into routine clinical practice; this observation questions the high cost incurred to national budgets by use of sevelamer and calls for well-designed randomized clinical trials.

    Abstract

    Importance  Guidelines restricting use of calcium-based phosphate binders in all patients with end-stage renal disease owing to their potential contribution to increased cardiovascular risk shifted prescribing from calcium acetate toward the costlier sevelamer carbonate products.

    Objective  To compare cardiovascular events and mortality between patients with end-stage renal disease (ESRD) undergoing hemodialysis receiving sevelamer vs calcium acetate in real-world practice.

    Design, Setting, and Participants  An observational cohort study was conducted using the United States Renal Data System linked to Medicare claims data (May 1, 2012, to December 31, 2013). Data analysis was performed from October 2017 to September 2018. Participants included patients 65 years or older with ESRD within 180 days after starting hemodialysis (sevelamer, 2647; calcium acetate, 2074).

    Exposures  New use of sevelamer (calcium-free phosphate binder) vs calcium acetate (calcium-based phosphate binder).

    Main Outcomes and Measures  Hazard ratios (HRs) with 95% CIs were estimated for fatal or nonfatal cardiovascular events (myocardial infarction or ischemic stroke: primary outcome) and all-cause mortality (secondary outcome) using Cox proportional hazards regression with fine stratification on the propensity score to control for potential confounders, including phosphorus and calcium levels.

    Results  After propensity score weighting, 2639 patients initiating sevelamer treatment (1184 men [44.9%]; mean [SD] age, 75.6 [6.9] years) and 2065 patients initiating calcium acetate treatment (930 men [45.0%]; mean [SD] age, 75.5 [7.1] years) were included in the analysis. Crude incidence rates (IRs) for cardiovascular events of 458 per 1000 person-years for sevelamer and 464 per 1000 person-years for calcium acetate were observed. After propensity score fine-stratification weighting, HRs of 0.96 (95% CI, 0.84-1.10) for cardiovascular events were observed. Results were consistent within subgroups of age (<75 y: primary outcome, HR, 1.02; 95% CI, 0.85-1.24; vs ≥75 years: primary outcome, HR, 0.83; 95% CI, 0.69-1.01) and sex (primary outcome in men: HR, 1.02; 95% CI, 0.83-1.26).

    Conclusions and Relevance  The results of the study do not suggest increased cardiovascular safety of sevelamer in the routine clinical practice of patients with ESRD compared with calcium acetate; this study’s findings suggest that well-designed, long-term, randomized clinical trials are needed.

    Introduction

    Hyperphosphatemia is present in most patients with end-stage renal disease (ESRD) and has been associated with increased cardiovascular mortality.1 Phosphate binders (calcium based and calcium free) are the mainstay pharmacologic treatment to lower phosphorus levels in patients with ESRD. In 2013, more than 75% of patients undergoing hemodialysis (HD) in the United States with Medicare Part D benefits filled 1 or more prescriptions for a phosphate binder, most frequently calcium-free sevelamer carbonate (50.1%) and calcium acetate (33.7%).2

    The 2017 update of the Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guideline suggests restricting use of calcium-based phosphate binders (CBPBs) in patients with ESRD irrespective of baseline calcium levels.3 This treatment change will further the shift of prescribing away from calcium acetate ($678 per user-year) toward the costlier agent, sevelamer ($4924 per user-year for sevelamer carbonate), which accounted for 83% of the total cost of phosphate binders to Medicare in 2014 (>$1.5 billion).2,4 The basis of this recommendation was the results of 3 open-label randomized clinical trials (RCTs) comparing cardiovascular survival between sevelamer and CBPBs, which were graded as methodologically weak (evidence 2B) by the KDIGO group.3 The main limitations of the trials are limited sample size leading to estimates with low precision, underrepresentation of frail patients at greatest risk of cardiovascular events, inconsistent findings, and high loss to follow-up with subsequent incomplete outcome data and risk of bias.3,5-8 Two prior observational studies compared treatment with calcium-free vs calcium-based phosphate binders but could not account for potential bias by differences in phosphorus and calcium levels.9,10

    We aimed to compare cardiovascular events and all-cause mortality between sevelamer and calcium acetate in a population of Medicare-insured patients 65 years or older with ESRD receiving newly initiated HD. Serum phosphorus and corrected calcium levels at baseline and over time were included in the analysis.

    Methods
    Data Source

    We conducted an observational cohort study using data from the United States Renal Data System (USRDS) from May 1, 2012, through December 31, 2013. This study period was chosen because the selected monthly laboratory data including serum phosphorus and calcium levels have been available in the USRDS since May 2012.11,12 Data analysis was performed from October 2017 to September 2018. The USRDS collects information on all patients with ESRD in the United States, including demographics, preexisting comorbidities, and longitudinal data for death and transplants. The USRDS is linked to all Medicare Parts A and B claims, including information on diagnoses and procedures from hospitalizations and outpatient visits, as well as all Medicare Part D claims containing information on prescription drugs. We obtained data use agreements by the USRDS and this study was approved with waiver of informed consent by the Brigham and Women’s Hospital institutional review board.

    Study Population

    The population of interest was patients 65 years or older who initiated treatment with sevelamer carbonate or hydrochloride or calcium acetate within 180 days of initiating HD between October 1, 2012, and December 31, 2013. The 6-month period before treatment start was defined as the baseline period. We further required continuous HD treatment, no prior use of phosphate-binding agents (including lanthanum), and at least 1 recorded value for serum phosphorus and corrected calcium levels during the baseline period. Patients with reports of kidney transplant or alcohol abuse before the index date were excluded.

    Drug Exposure

    Exposure was defined as the initiation of sevelamer or calcium acetate, whereby each patient contributed only 1 treatment episode. Patients beginning treatment with more than 1 phosphate binder (including lanthanum) on the same date were excluded. We followed up patients in an as-treated approach from the day after treatment initiation until the occurrence of an outcome. Patients were censored in case of a treatment switch (on the day of a prescription of a different phosphate binder), discontinuation (no prescription refill for the index drug within 60 days after the estimated last day of supply), disenrollment from any part of Medicare, kidney transplant, end of HD treatment, noncardiovascular-associated death (primary analysis), or at the end of the study period (eFigure 1 in the Supplement).

    In a secondary analysis, we followed up patients in an intention-to-treat (ITT) approach. Patients were monitored in their initially assigned treatment group irrespective of treatment switch or discontinuation.

    Study Outcomes

    The primary outcome was occurrence of a fatal or nonfatal cardiovascular event defined as either cardiovascular death,13 myocardial infarction (positive predictive value [PPV], 93.7% overall, 88.1% in the presence of an myocardial infarction code within 1 year prior14) or ischemic stroke (PPV, 99.1%15). The secondary outcome was all-cause mortality (99% sensitivity).16

    Patient Characteristics

    We assessed 78 patient characteristics during the baseline period (eTable 1 in the Supplement),17 including demographics, ESRD- and HD-related covariates, proxies for socioeconomic status (eg, employment status, low-income subsidy status) and frailty (eg, inability to ambulate, status of institutionalization, comorbidity index), cardiovascular disease, comedication, and health care use covariates. We further assessed monthly values for serum phosphorus and corrected calcium levels between start of HD and the index date, as well as during follow-up.

    Statistical Analysis

    We calculated a propensity score (PS) (ie, the probability of receiving sevelamer vs calcium acetate) for each patient including all covariates listed in eTable 1 in the Supplement. We trimmed the nonoverlapping portions of the PS distributions to exclude noncomparable patients and created 50 PS strata based on the PS distribution in the sevelamer group.18 We weighted calcium acetate users proportional to the distribution of sevelamer users in the PS stratum within which they fell. Covariate balance before and after weighting was shown using standardized differences.19 We quantified the crude and weighted C statistic as a measure of overall covariate balance (perfect balance, 0.5).20 We plotted weighted Kaplan-Meier curves and used weighted Cox proportional hazards regression models to estimate adjusted hazard ratios (HRs) with 95% CIs comparing sevelamer with calcium acetate. The proportional hazards assumption was tested with an interaction term between exposure and time of follow-up in the primary as-treated analysis.

    Subgroup Analyses

    We performed prespecified analyses for both outcomes in a subgroup excluding patients with a recorded myocardial infarction or stroke during the baseline period and by sex and age (<75 vs ≥75 years). We implemented a post hoc subgroup analysis in patients with and without nephrology care before the start of HD.

    Adjusting for Phosphorus and Calcium Levels Over Time

    High serum phosphorus and calcium levels may lead physicians to switch or discontinue an ongoing phosphate binder therapy, particularly in patients at high cardiovascular risk. Censoring patients based on such treatment changes driven by change in prognosis introduces the possibility of informative censoring. To minimize this possibility, we conducted a secondary analysis using marginal structural models with stabilized inverse probability treatment weights to account for changing phosphorus and calcium levels during follow-up.21 We calculated monthly weights accounting for serum phosphorus and calcium levels. Missing monthly laboratory values were imputed by carrying the last observation forward (missing values during follow-up months: phosphorus level, 16.5%; calcium level, 18.2%; no missingness at baseline by design). Weights were truncated on the 1st and 99th percentiles to reduce influence by outliers. The eMethods in the Supplement provides technical details of this model.22-24

    Study Population 2

    Because follow-up was limited to a maximum of 1.5 years in the primary study population, we established study population 2, including patients with phosphate binder treatment initiated between January 1, 2008, and December 31, 2013, based on the same inclusion and exclusion criteria, but not taking into account baseline phosphorus or calcium values (available since 2012). Besides the absence of phosphorus and calcium level values, statistical analyses were conducted identical to the main analyses. Given the longer follow-up, we conducted an additional analysis where we introduced a lag period of 180 days after treatment start, excluding patients whose follow-up ended before completion of the lag period. Follow-up started on day 181 after treatment was initiated, accounting for the slow progression of coronary artery calcification and the high mortality rate shortly after initiating HD.25 The larger sample size further allowed a subgroup analysis excluding patients with dysrhythmia (61% of study population) during baseline in accordance with 2 previous RCTs.5,6 We considered an α level of .05 as significant, and testing was 2-tailed. All analyses were conducted using SAS, version 9.4 (SAS Institute Inc).

    Results
    Cohort Characteristics

    We identified 2647 patients (mean [SD] age, 75.6 [6.9] years) receiving newly initiated sevelamer treatment and 2074 patients (mean [SD] age, 75.5 [7.1] years) receiving newly initiated calcium acetate treatment (cohort creation flowchart in eFigure 2 in the Supplement). Imbalances mainly resulted from higher baseline calcium levels in patients (mean [SD], 9.3 [0.6] mg/dL; to convert to millimoles per liter, multiply by 0.25) with newly initiated sevelamer compared with those with newly initiated calcium acetate (9.0 [0.6] mg/dL) (absolute standardized difference, 39.1%) and dialysis provider (8% more sevelamer use in DaVita compared with Fresenius Kidney Care clinics). After PS weighting, comparability was achieved (C = 0.50) (Table 1; eTable 1 in the Supplement provides a complete list of covariates before and after weighting).

    Comparative Safety

    Crude incidence rates (IRs) for fatal or nonfatal cardiovascular events were 458 per 1000 person-years for sevelamer and 464 per 1000 person-years for calcium acetate, and were 458 per 1000 person-years vs 486 per 1000 person-years after PS weighting. Crude IRs for all-cause mortality were 208.4 per 1000 person-years for sevelamer and 216.8 per 1000 person-years for calcium acetate, again, largely unchanged after PS weighting (208.4 vs 216.8 per 1000 person-years) (Table 2; censoring reasons in eTable 2 and eTable 3 in the Supplement). Kaplan-Meier plots for the as-treated and ITT follow-up model are shown in Figure 1 (CV events) and in eFigure 6 in the Supplement (mortality). The Cox proportional hazards assumption was maintained (P = .47), and we observed a weighted hazard ratio (HR) of 0.96 (95% CI, 0.84-1.10) for fatal or nonfatal cardiovascular events. The weighted HR for all-cause mortality was 0.96 (95% CI, 0.80-1.17). The secondary ITT analyses revealed null results similar to those of the primary as-treated analysis (Table 3).

    Subgroup and Sensitivity Analyses

    Men and patients younger than 75 years had slightly higher HRs for fatal and nonfatal cardiovascular events (Figure 2) (men: HR, 1.02; 95% CI, 0.83-1.26; age: HR, 1.02; 95% CI, 0.85-1.24) and all-cause mortality (men: HR, 0.89; 95% CI, 0.69-1.16; age: HR, 0.79; 95% CI, 0.58-1.07) (eFigure 5 in the Supplement), but overall results were similar across subgroups. Excluding patients with a recorded myocardial infarction or ischemic stroke during the baseline period did not meaningfully change the results. In addition, findings were unchanged in patients with nephrology care prior to the start of HD. Among patients without pre-HD nephrology care, the adjusted HR was 0.61 (95% CI, 0.42-0.87) for all cause-mortality and 0.85 (95% CI, 0.65-1.12) for cardiovascular events. However, results were attenuated to a null result after implementation of a 60-day lag period, suggesting that this finding may be explained by residual confounding (eTable 6 in the Supplement and Kaplan-Meier curves for mortality without lag period in eFigures 7 and eFigure 8 in the Supplement).

    Adjusting for Phosphorus and Calcium Levels Over Time

    Serum phosphorus levels decreased in patients initiating sevelamer (−0.21 [1.73] mg/dL) and calcium acetate (−0.12 [1.74] mg/dL) during follow up (eTable 5 in the Supplement). However, calcium levels remained unchanged during treatment with sevelamer (0.01 [0.64] mg/dL) but increased by 0.20 (0.68) mg/dL (P < .001) during treatment with calcium acetate. Stratification of laboratory values measured in the last month of follow-up by censoring reasons revealed significant differences between patients receiving sevelamer and calcium acetate who were censored for treatment switch or discontinuation, indicating potential selection bias.26 However, the inverse probability treatment weights analysis, which adjusted for time-varying selection bias, revealed unchanged adjusted HRs of 0.97 (95% CI, 0.71-1.32) for cardiovascular events and 0.96 (95% CI, 0.77-1.19) for all-cause mortality (Table 3). The overall mean (SD) marginal structural model weights after truncation were 1.00 (0.023) for the cardiovascular event analysis and 1.00 (0.023) for the mortality analysis (eFigure 3 and eFigure 4 in the Supplement show monthly mean of weights).

    Study Population 2: 2008-2013

    In study population 2, we identified 16 447 patients receiving sevelamer and 15 780 receiving calcium acetate between 2008 and 2013 (mean follow-up and censoring reasons reported in eTable 4 in the Supplement). After PS weighting, we observed a null result in the as-treated as well as the ITT analysis, with HRs of 0.94 (95% CI, 0.86-1.02) and 0.98 (95% CI, 0.95-1.01). Kaplan-Meier plots were overlapping during the entire study period of 5 years (Figure 2B and D for CV events; eFigure 6 in the Supplement for mortality). Introduction of a lag period of 180 days did not meaningfully change these results for primary outcome of fatal or nonfatal CV event (HR, 0.94; 95% CI, 0.86-1.02) and secondary outcome of all-cause mortality (HR, 1.00; 95% CI, 0.91-1.09) (Table 3).

    Discussion

    The results of this observational cohort study do not suggest that sevelamer is associated with superior cardiovascular safety or survival compared with calcium acetate in a routine care setting of patients 65 years or older with ESRD requiring maintenance HD in the United States.

    The 2017 update of the KDIGO guidelines recommending more restricted CBPB use was based on weak evidence (grade 2B)3 consisting of 3 open-label RCTs comparing clinical outcomes between sevelamer and CBPBs with inconsistent results. The Renagel in New Dialysis (RIND) Study (United States, N = 127)5 and INDEPENDENT (Reduce Cardiovascular Calcifications to Reduce QT Interval in Dialysis) study (Italy, N = 466)6,27 included patients with incident HD and both reported an approximately 4-fold survival benefit (secondary end point) for sevelamer compared with CBPBs. However, both trials were limited by a low event count in the sevelamer group (11 in RIND and 28 in INDEPENDENT), leading to potential for chance findings. Furthermore, exclusion of patients with atrial fibrillation/flutter (RIND5) and cardiac arrhythmia (INDEPENDENT27), which are common,28 limits their generalizability to patients with typical ESRD who have a high comorbidity burden. In addition, both trials reported insufficient balance after randomization for key covariates, such as baseline phosphorus levels6 and atherosclerotic heart disease.29 The third trial, Dialysis Clinical Outcomes Revisited (DCOR), recruited 2103 patients with prevalent HD in the United States and reported no overall mortality benefit with sevelamer but a modest benefit in the subgroup of patients 65 years or older (HR, 0.77; 95% CI, 0.61-0.96).5-7 However, 46% of the patients were lost to follow-up, introducing concerns related to healthy survivor bias.7 In addition, all 3 trials applied an open-label design, which may have influenced investigator-chosen comedication or drug dosage following randomization.

    The survival benefit with sevelamer became apparent after 1 to 2 years in the DCOR and RIND trials.6 Our primary study population was limited in size and duration of follow-up (maximum, 1.5 years) and could not rule out a 20% survival benefit of sevelamer compared with calcium acetate. However, study population 2, including patients between 2008 and 2013, revealed unchanged association estimates, with narrow 95% CIs and overlapping Kaplan-Meier curves during the entire 5 years of follow-up.

    A prior observational cohort study using USRDS data (2006-2011) reported a small but statistically significant reduction in all-cause mortality in patients 65 years or older with incident HD and newly started sevelamer vs calcium acetate therapy (HR, 0.94; 95% CI, 0.91-0.98), but no difference in CV mortality and all-cause or CV hospitalization.9 However, the authors could not account for baseline differences in serum phosphorus and calcium levels. Another observational cohort study compared sevelamer with CBPBs among veterans with new HD treatment and reported a survival benefit for sevelamer (HR, 0.65; 95% CI, 0.52-0.80). However, that study was limited by unmeasured confounding because only 7 potential confounders were accounted for, not including baseline calcium and phosphorus levels,10 which are known risk factors for cardiovascular morbidity and mortality.30,31

    Our real-world study provides a more nuanced analysis accounting for a comprehensive range of confounders, including serum phosphorus and calcium levels. The inconsistency with previous RCTs calls into question the updated KDIGO recommendations.3 Although the observational nature of this study does not allow a definitive causal conclusion, it suggests that the observed greater benefit of sevelamer vs calcium acetate in small, highly selected populations may not translate to patients receiving routine care. In light of the limitations of the methods of previous RCTs and the high cost to national budgets incurred by treatment with sevelamer, our results call for well-designed, long-term RCTs comparing relevant clinical outcomes between different phosphate binders. Considering the time and costs that such trials would entail, innovative, outcome-based payment models for sevelamer may provide a more actionable solution in the near future.32 The wealth of data routinely collected by the USRDS may facilitate tracking of real-world outcomes and could make for an informative case study for implementation of such payment models.33

    Strengths and Limitations

    This study has several other strengths. First, the fact that results were consistent between the as-treated and ITT follow-up models suggests that nonadherence and informative censoring are unlikely to explain our null result. Next, our cohort was sampled from the entire US population of patients older than 65 years with Medicare benefits undergoing HD and, hence, our results have generalizability. In addition, the good covariate balance on proxies for frailty, socioeconomic status, comorbidities, and comedication even before PS weighting demonstrates the strength of a new-user active comparator design and suggests that unmeasured confounding may be limited.

    This study has some limitations. First, exclusion of patients with phosphate-binder use before initiation of HD may have led to overselection of patients with insufficient pre-HD nephrology care. We compared patient characteristics between those with pre-HD phosphate-binder use (excluded, eTable 7 in the Supplement) and new users after HD initiation (included), and observed a higher proportion of patients without pre-HD nephrology care, catheter access at HD start, and low-income subsidy status in our study population (eTable 8 in the Supplement). Thus, our results may be most representative for socioeconomically disadvantaged patients and may not be extrapolated to those with prevalent HD. However, generalizability has no value in the absence of internal validity, which is challenging to attain with a prevalent user design owing to risk of bias from depletion of susceptible patients. Second, USRDS data are released with a lag of 3 years, and we were thus not able to include the iron-based phosphate binders sucroferric oxyhydroxide (US Food and Drug Administration [FDA] approval in 2013) and ferric citrate (FDA approval in 2014) into our analyses.34,35 However, in 2018, less than 10% of patients in the United States undergoing HD were receiving these drugs.36 Third, phosphate-binder dose changes are frequent, but we did not undertake a formal dose-response analysis. Future research shedding light on the association between phosphate-binder doses and risk of CV events is recommended.

    Fourth, serum levels of parathyroid hormone were not captured in USRDS data during our study period. Fifth, we observed a significantly lower risk of mortality in patients initiating sevelamer therapy who did not have pre-HD nephrology care; however, this association was primarily driven by early events and, hence, likely to be attributable to unmeasured confounding by socioeconomic factors that are associated with early mortality during HD treatment.37 However, such confounding would have biased our results toward a protective effect of sevelamer. Sixth, calcium carbonate is sold as an over-the-counter product in the United States and its use was not captured in our study.

    Conclusions

    We observed no significant differences in the risk of cardiovascular events and all-cause mortality when comparing patients with new use of sevelamer vs calcium acetate among older patients with ESRD undergoing newly initiated HD in the United States. These results suggest the need for long-term, double-blind RCTs to establish the effectiveness of sevelamer in reducing the risk of cardiovascular events and mortality in this population.

    Back to top
    Article Information

    Accepted for Publication: January 6, 2019.

    Corresponding Author: Rishi J. Desai, MS, PhD, Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, 1620 Tremont St, Ste 3030-R, Boston, MA 02120 (rdesai@bwh.harvard.edu).

    Published Online: May 6, 2019. doi:10.1001/jamainternmed.2019.0045

    Author Contributions: Drs Spoendlin and Desai had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Spoendlin, Kim, Schneeweiss, Desai.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Spoendlin, Desai.

    Critical revision of the manuscript for important intellectual content: Paik, Tsacogianis, Kim, Schneeweiss, Desai.

    Statistical analysis: Spoendlin, Tsacogianis, Desai.

    Obtained funding: Spoendlin, Schneeweiss.

    Administrative, technical, or material support: Paik.

    Supervision: Kim, Schneeweiss, Desai.

    Conflict of Interest Disclosures: Dr Spoendlin reported receiving grants from Swiss National Science Foundation during the conduct of the study. Dr Kim reported receiving grants from Pfizer and grants from Bristol-Myers Squibb outside the submitted work. Dr Schneeweiss reported receiving personal fees from WHISCON LLC, personal fees and other support from Aetion, Inc, and grants from Bohringer-Ingelheim, Bayer, Vertex outside the submitted work. Dr Desai reported receiving grants from Merck, grants from Bayer, and grants from Vertex outside the submitted work. No other disclosures were reported.

    Disclaimer: The data reported here have been supplied by the United States Renal Data System. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US government.

    Additional Contributions: Michael Bodmer, MD (Internal Medicine, Cantonal Hospital Zug, Zug, Switzerland) helped to derive the study question. There was no financial compensation.

    References
    1.
    St Peter  WL, Wazny  LD, Weinhandl  E, Cardone  KE, Hudson  JQ.  A review of phosphate binders in chronic kidney disease: incremental progress or just higher costs?  Drugs. 2017;77(11):1155-1186. doi:10.1007/s40265-017-0758-5PubMedGoogle ScholarCrossref
    2.
    Spoendlin  J, Schneeweiss  S, Tsacogianis  T,  et al.  Association of Medicare’s bundled payment reform with changes in use of vitamin D among patients receiving maintenance hemodialysis: an interrupted time-series Analysis.  Am J Kidney Dis. 2018;72(2):178-187. doi:10.1053/j.ajkd.2018.03.027PubMedGoogle ScholarCrossref
    3.
    Kidney International Supplements. KDIGO 2017 Clinical Practice Guideline for the Diagnosis, Evaluation, Prevention and Treatment of Chronic Kidney Disease Mineral and Bone Disorder (CKD-MBD). https://kdigo.org/wp-content/uploads/2017/02/2017-KDIGO-CKD-MBD-GL-Update.pdf. Published July 2017. Accessed March 4, 2019.
    4.
    St Peter  WL, Wazny  LD, Weinhandl  ED.  Phosphate-binder use in US dialysis patients: prevalence, costs, evidence, and policies.  Am J Kidney Dis. 2018;71(2):246-253. doi:10.1053/j.ajkd.2017.09.007PubMedGoogle ScholarCrossref
    5.
    Block  GA, Raggi  P, Bellasi  A, Kooienga  L, Spiegel  DM.  Mortality effect of coronary calcification and phosphate binder choice in incident hemodialysis patients.  Kidney Int. 2007;71(5):438-441. doi:10.1038/sj.ki.5002059PubMedGoogle ScholarCrossref
    6.
    Di Iorio  B, Molony  D, Bell  C,  et al; INDEPENDENT Study Investigators.  Sevelamer versus calcium carbonate in incident hemodialysis patients: results of an open-label 24-month randomized clinical trial.  Am J Kidney Dis. 2013;62(4):771-778. doi:10.1053/j.ajkd.2013.03.023PubMedGoogle ScholarCrossref
    7.
    Suki  WN, Zabaneh  R, Cangiano  JL,  et al.  Effects of sevelamer and calcium-based phosphate binders on mortality in hemodialysis patients.  Kidney Int. 2007;72(9):1130-1137. doi:10.1038/sj.ki.5002466PubMedGoogle ScholarCrossref
    8.
    Patel  L, Bernard  LM, Elder  GJ.  Sevelamer versus calcium-based binders for treatment of hyperphosphatemia in CKD: A meta-analysis of randomized controlled trials.  Clin J Am Soc Nephrol. 2016;11(2):232-244. doi:10.2215/CJN.06800615PubMedGoogle ScholarCrossref
    9.
    Yusuf  AA, Weinhandl  ED, St Peter  WL.  Comparative effectiveness of calcium acetate and sevelamer on clinical outcomes in elderly hemodialysis patients enrolled in Medicare part D.  Am J Kidney Dis. 2014;64(1):95-103. doi:10.1053/j.ajkd.2013.11.015PubMedGoogle ScholarCrossref
    10.
    Borzecki  AM, Lee  A, Wang  SW, Brenner  L, Kazis  LE.  Survival in end stage renal disease: calcium carbonate vs. sevelamer.  J Clin Pharm Ther. 2007;32(6):617-624. doi:10.1111/j.1365-2710.2007.00871.xPubMedGoogle ScholarCrossref
    11.
    United States Renal Data System. USRDS annual data report 2016, chapter 3: clinical indicators and preventive care. https://www.usrds.org/2016/view/v2_03.aspx. Accessed October 3, 2018.
    12.
    United States Renal Data System. https://www.usrds.org/. Accessed October 3, 2018.
    13.
    United States Renal Data System. Volume 2: ESRD analytic methods. https://www.usrds.org/2017/view/v2_00_appx.aspx#_Toc501030698. Accessed October 3, 2018.
    14.
    Kiyota  Y, Schneeweiss  S, Glynn  RJ, Cannuscio  CC, Avorn  J, Solomon  DH.  Accuracy of Medicare claims-based diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records.  Am Heart J. 2004;148(1):99-104. doi:10.1016/j.ahj.2004.02.013PubMedGoogle ScholarCrossref
    15.
    Kumamaru  H, Judd  SE, Curtis  JR,  et al.  Validity of claims-based stroke algorithms in contemporary Medicare data: reasons for geographic and racial differences in stroke (REGARDS) study linked with Medicare claims.  Circ Cardiovasc Qual Outcomes. 2014;7(4):611-619. doi:10.1161/CIRCOUTCOMES.113.000743PubMedGoogle ScholarCrossref
    16.
    Rocco  MV, Yan  G, Gassman  J,  et al; Hemodialysis Study Group; The National Institutes of Health-funded Hemodialysis. Health Care Financing Administration.  Comparison of causes of death using HEMO Study and HCFA end-stage renal disease death notification classification systems.  Am J Kidney Dis. 2002;39(1):146-153. doi:10.1053/ajkd.2002.29905PubMedGoogle ScholarCrossref
    17.
    Foley  RN, Collins  AJ.  The USRDS: what you need to know about what it can and can’t tell us about ESRD.  Clin J Am Soc Nephrol. 2013;8(5):845-851. doi:10.2215/CJN.06840712PubMedGoogle ScholarCrossref
    18.
    Desai  RJ, Rothman  KJ, Bateman  BT, Hernandez-Diaz  S, Huybrechts  KF.  A propensity score based fine stratification approach for confounding adjustment when exposure is infrequent.  Epidemiology. 2017;28(2):249-257. doi:10.1097/EDE.0000000000000595PubMedGoogle ScholarCrossref
    19.
    Austin  PC.  Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research.  Commun Stat Simul Comput. 2009;38(6):1228-1234. doi:10.1080/03610910902859574Google ScholarCrossref
    20.
    Franklin  JM, Rassen  JA, Ackermann  D, Bartels  DB, Schneeweiss  S.  Metrics for covariate balance in cohort studies of causal effects.  Stat Med. 2014;33(10):1685-1699. doi:10.1002/sim.6058PubMedGoogle ScholarCrossref
    21.
    Danaei  G, García Rodríguez  LA, Cantero  OF, Logan  RW, Hernán  MA.  Electronic medical records can be used to emulate target trials of sustained treatment strategies.  J Clin Epidemiol. 2018;96(96):12-22. doi:10.1016/j.jclinepi.2017.11.021PubMedGoogle ScholarCrossref
    22.
    Danaei  G, Rodríguez  LA, Cantero  OF, Logan  R, Hernán  MA.  Observational data for comparative effectiveness research: an emulation of randomised trials of statins and primary prevention of coronary heart disease.  Stat Methods Med Res. 2013;22(1):70-96. doi:10.1177/0962280211403603PubMedGoogle ScholarCrossref
    23.
    Hernán  MA, Brumback  B, Robins  JM.  Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men.  Epidemiology. 2000;11(5):561-570. doi:10.1097/00001648-200009000-00012PubMedGoogle ScholarCrossref
    24.
    Desai  RJ, Solomon  DH, Schneeweiss  S, Danaei  G, Liao  KP, Kim  SC.  Tumor necrosis factor alpha inhibitor use and the risk of incident hypertension in patients with rheumatoid arthritis.  Epidemiology. 2016;27(3):414-422. doi:10.1097/EDE.0000000000000446PubMedGoogle ScholarCrossref
    25.
    Goodman  WG, Goldin  J, Kuizon  BD,  et al.  Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis.  N Engl J Med. 2000;342(20):1478-1483. doi:10.1056/NEJM200005183422003PubMedGoogle ScholarCrossref
    26.
    Mansournia  MA, Etminan  M, Danaei  G, Kaufman  JS, Collins  G.  Handling time varying confounding in observational research.  BMJ. 2017;359:j4587. doi:10.1136/bmj.j4587PubMedGoogle ScholarCrossref
    27.
    Di Iorio  BR, Cucciniello  E, Bellizzi  V.  Vascular calcification and QT interval in incident hemodialysis patients.  J Nephrol. 2009;22(6):694-698.PubMedGoogle Scholar
    28.
    Niu  J, Shah  MK, Perez  JJ,  et al.  Dialysis modality and incident atrial fibrillation in older patients with ESRD.  Am J Kidney Dis. 2019;73(3):324-331.PubMedGoogle Scholar
    29.
    Block  GA, Spiegel  DM, Ehrlich  J,  et al.  Effects of sevelamer and calcium on coronary artery calcification in patients new to hemodialysis.  Kidney Int. 2005;68(4):1815-1824. doi:10.1111/j.1523-1755.2005.00600.xPubMedGoogle ScholarCrossref
    30.
    Raggi  P, Boulay  A, Chasan-Taber  S,  et al.  Cardiac calcification in adult hemodialysis patients: a link between end-stage renal disease and cardiovascular disease?  J Am Coll Cardiol. 2002;39(4):695-701. doi:10.1016/S0735-1097(01)01781-8PubMedGoogle ScholarCrossref
    31.
    Sato  H, Nagasawa  T, Saito  A, Miyazaki  M.  Risk of cardiovascular mortality predicted by the serum calcium level and calcification score at the initiation of dialysis.  Clin Exp Nephrol. 2018;22(4):957-966. doi:10.1007/s10157-017-1527-1PubMedGoogle ScholarCrossref
    32.
    Kaltenboeck  A, Bach  PB.  Value-based pricing for drugs: theme and variations.  JAMA. 2018;319(21):2165-2166. doi:10.1001/jama.2018.4871PubMedGoogle ScholarCrossref
    33.
    Nazareth  T, Ko  JJ, Sasane  R,  et al.  Outcomes-based contracting experience: research findings from US. and European stakeholders.  J Manag Care Spec Pharm. 2017;23(10):1018-1026.PubMedGoogle Scholar
    34.
    Palmer  SC, Gardner  S, Tonelli  M,  et al.  Phosphate-binding agents in adults with CKD: a network meta-analysis of randomized trials.  Am J Kidney Dis. 2016;68(5):691-702. doi:10.1053/j.ajkd.2016.05.015PubMedGoogle ScholarCrossref
    35.
    Bousher  A, Al-Makki  A, Sutton  J, Shepler  B.  A review of sucroferric oxyhydroxide for the treatment of hyperphosphatemia in patients receiving dialysis.  Clin Ther. 2014;36(12):2082-2093. doi:10.1016/j.clinthera.2014.09.021PubMedGoogle ScholarCrossref
    36.
    DOPPS Practice Monitor. Mineral bone disorder, national sample for use of iron based phosphate binders in the last 3 months. https://www.dopps.org/dpm/DPMSlideBrowser.aspx. Published 2018. Accessed August 2, 2018.
    37.
    Johns  TS, Estrella  MM, Crews  DC,  et al.  Neighborhood socioeconomic status, race, and mortality in young adult dialysis patients.  J Am Soc Nephrol. 2014;25(11):2649-2657. doi:10.1681/ASN.2013111207PubMedGoogle ScholarCrossref
    ×