Urgent-start dialysis in patients referred early to a nephrologist—the CKD-REIN prospective cohort study

Abstract

Background

The lack of a well-designed prospective study of the determinants of urgent dialysis start led us to investigate its individual- and provider-related factors in patients seeing nephrologists.

Methods

The Chronic Kidney Disease Renal Epidemiology and Information Network (CKD-REIN) is a prospective cohort study that included 3033 patients with CKD [mean age 67 years, 65% men, mean estimated glomerular filtration rate (eGFR) 32 mL/min/1.73 m²] from 40 nationally representative nephrology clinics from 2013 to 2016 who were followed annually through 2020. Urgent-start dialysis was defined as that ‘initiated imminently or <48 hours after presentation to correct life-threatening manifestations’ according to the Kidney Disease: Improving Global Outcomes 2018 definition.

Results

Over a 4-year (interquartile range 3.0–4.8) median follow-up, 541 patients initiated dialysis with a known start status and 86 (16%) were identified with urgent starts. The 5-year risks for the competing events of urgent and non-urgent dialysis start, pre-emptive transplantation and death were 4, 17, 3 and 15%, respectively. Fluid overload, electrolytic disorders, acute kidney injury and post-surgery kidney function worsening were the reasons most frequently reported for urgent-start dialysis. Adjusted odds ratios for urgent start were significantly higher in patients living alone {2.14 [95% confidence interval (CI) 1.08–4.25] or with low health literacy [2.22 (95% CI 1.28–3.84)], heart failure [2.60 (95% CI 1.47–4.57)] or hyperpolypharmacy [taking >10 drugs; 2.14 (95% CI 1.17–3.90)], but not with age or lower eGFR at initiation. They were lower in patients with planned dialysis modality [0.46 (95% CI 0.19–1.10)] and more nephrologist visits in the 12 months before dialysis [0.81 (95% CI 0.70–0.94)] for each visit.

Conclusions

This study highlights several patient- and provider-level factors that are important to address to reduce the burden of urgent-start dialysis.

Open in new tabDownload slide

INTRODUCTION

The transition from chronic kidney disease (CKD) to kidney failure requiring kidney replacement therapy (KRT) is a critical period with major effects on mortality, morbidity, quality of life and health resource utilization [1–3]. Conditions of dialysis initiation, whether planned or unplanned, with a timely or urgent start, may strongly affect patients’ outcomes [4]. Pre-dialysis care and education programmes have been associated with fewer urgent starts, less catheter use and better survival [5–7]. Nevertheless, despite these programmes’ growing availability, urgent start remains common and associated with high early mortality [5, 8–10]. In France, it has been remarkably stable at ~30% of incident dialysis over the past 15 years [11]. Until recently, the lack of a consensus definition for ‘urgent-start dialysis’ made research on its risk factors difficult [12]. In 2018, a Kidney Disease: Improving Global Outcomes (KDIGO) Consensus Conference [4] clarification of this definition as ‘dialysis which must be initiated imminently or in less than 48 hours after presentation to correct life-threatening manifestations’ opened up new research perspectives on this serious and unsolved issue.

Late nephrologist referral is a well-documented determinant of urgent-start dialysis [13–16]. A few studies using a variety of definitions have nonetheless shown that this risk may be common even with early nephrology care [13, 17–19]. They identified older age, high body mass index (BMI), cardiovascular disease, acute kidney injury (AKI), non-attendance at dialysis information sessions and surgical delays for vascular access as risk factors for starting dialysis urgently or without planning, defined as initiation as an inpatient or with a central venous catheter [12, 14, 18, 20]. Their retrospective design, however, may have induced information bias. Moreover, social factors such as education, health literacy or living conditions and patterns of nephrology care related to dialysis start conditions have not been adequately investigated, although they have been shown to affect modality choice [21, 22], CKD progression [23, 24] and survival [25]. A well-designed prospective study was needed to determine the relative impact of patient-, disease- and provider-related factors associated with urgent-start dialysis.

We used data from the Chronic Kidney Disease Renal Epidemiology and Information Network (CKD-REIN) cohort study [26] to estimate the risk of urgent-start dialysis and examine its determinants and early mortality in patients under nephrology care. We also compared reasons and modalities of dialysis initiation between patients starting urgently and non-urgently, as well as their early mortality.

MATERIALS AND METHODS

Study design and participants

The CKD-REIN is a prospective cohort study in 40 French nephrology clinics, nationally representative geographically and by legal status (public, private non-profit and private for-profit). From July 2013 to March 2016, 3033 adult patients ≥18 years of age with a proven CKD diagnosis, an estimated glomerular filtration rate (eGFR) ˂60 mL/min/1.73 m² and no prior chronic dialysis or kidney transplantation were enrolled during a routine nephrology visit. All patients signed an informed consent and were followed up annually thereafter. A detailed study protocol and cohort profile have previously been published [26, 27]. The ‘Institut National de la Santé et de la Recherche Médicale’ (INSERM) Institutional Review Board approved the study protocol (IRB00003888) on 18 June 2013. The study is registered at ClinicalTrials.gov (NCT03381950).

Information

Baseline data collection instruments included patient- and provider-level questionnaires coordinated with the international CKD Outcomes and Practice Patterns Study [28]. Sociodemographic characteristics such as education, health literacy and living conditions were recorded from patient interviews or self-administered questionnaires. Health literacy was assessed by asking patients ‘How often do you need help to understand written documents from your doctor or your pharmacist? Never, rarely, often, always?’ [29]. Clinical data and medications were collected from medical records. Longitudinal data including the number of nephrologist visits, hospitalizations and their causes and laboratory measures were collected at 1-year intervals from medical records. We used the Chronic Kidney Disease Epidemiology Collaboration equation for eGFR [30]. Data about the information and education patients received regarding KRT options and planning were also collected longitudinally. Hospitalizations at dialysis starts were categorized as planned or unplanned. Planned hospitalizations were defined as hospitalizations scheduled to initiate dialysis. Provider-level data included clinic legal status and the availability of the patients’ nephrologists, assessed by the delay for outpatient appointments.

Transition to KRT events and death

Deaths and KRT events—dialysis or pre-emptive transplantation—were tracked from study enrolment through January 2020. Deaths were identified from medical records or reported by family members at the annual follow-up. Transitions to KRT events were reported by patients or identified from their medical records or by quarterly record linkage to the national REIN registry [31]. Reasons to initiate dialysis were also collected from medical records, including volume management, low kidney function, uraemic symptoms, malnutrition, electrolyte problems, post-surgical worsening of kidney failure and AKI.

Identification of urgent-start dialysis

We assessed all patients who began maintenance dialysis to identify those who experienced an urgent start as defined by the KDIGO guidelines [4]. Dialysis starts were classified as urgent when this was reported in patients’ hospitalization or dialysis initiation reports or from the REIN registry data at dialysis initiation when these were missing [31]. The REIN registry’s definition of urgent start is similar to that of KDIGO except that dialysis is to be initiated in ˂24 h for the REIN instead of 48 h after presentation. Using both data sources, we were able to ascertain dialysis start status in 541 of 581 patients (93%) who initiated dialysis during follow-up; 72 of 541 (13%) were identified only from the REIN registry.

Statistical analyses

We first determined the baseline characteristics of all participants, as percentages, means and standard deviations (SDs) or medians and interquartile ranges (IQRs) as appropriate. We then estimated the cumulative incidence for the competing risks [32, 33] of urgent-start dialysis, non-urgent-start dialysis, pre-emptive kidney transplantation and death in the overall population. To reduce bias due to prevalent cases, we also estimated these risks among incident CKD Stage 5 patients, defined as an eGFR <15 mL/min/1.73 m². To estimate the time to reach an eGFR <15 mL/min/1.73 m², we used the intercept and slope predicted by a linear regression for the 2910 patients with an eGFR ≥15 mL/min/1.73 m² at baseline. In each model we used all eGFR values available since enrolment, i.e. a median of 11 (IQR 7–16) values per patient. After the exclusion of 57 patients with a single eGFR measurement and 19 with a predicted intercept of <15 mL/min/1.73 m², this analysis included 2834 patients. Observations were censored at the time of the most recent information in any of the analyses.

Using chi-squared, Student or Wilcoxon–Mann–Whitney tests as appropriate, we studied the crude associations of urgent-start dialysis with patient-, disease- and provider-related factors among the 541 patients with known dialysis start status, regardless of their CKD stage baseline. The patient-related factors studied included sociodemographic variables, BMI, any missed nephrologist visits in the year before dialysis or reported dialysis refusal. Disease-related factors included the duration of nephrology care at inclusion, comorbidities, hyperpolypharmacy (taking ≥10 drug classes) [34], albuminuria, hospitalization for AKI from 1 to 12 months before starting dialysis and haemoglobin level and eGFR at this start. Provider-related factors included the number of visits in the 12 months before dialysis, nephrology appointment delay and KRT education and modality planning during follow-up. We used multiple logistic regression to estimate adjusted odds ratios (aORs) of urgent start associated with age and gender, as well as with any factors with a P-value <0.10 in the crude analysis unless they were strongly correlated with another such factor. Robust variance estimates were used to take the facility clustering effect into account. For variables with missing data, a multiple imputation procedure was applied (multiple imputation with chained equations, 50 iterations, 20 datasets) including all covariates considered in the logistic regression model as well as two others considered relevant (dialysis modality and access) [35]. We fitted the model on each imputed dataset and pooled the estimates into a single set of estimates to report the aOR of each variable. Finally, we used chi-squared to compare reasons for dialysis initiation, dialysis start conditions and 3-month mortality between urgent- and non-urgent-start patients. Data were analysed with R software version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Participants’ baseline characteristics

In the overall patient population, the median eGFR at baseline was 32 (IQR 23–41) mL/min/1.73 m² and 123 patients had CKD Stage 5 (Table 1). The mean age was 67 years and 65% were men. More than half of the patients had cardiovascular disease, 43% had diabetes and 43% were exposed to hyperpolypharmacy.

Cumulative incidence of urgent-start dialysis before and after reaching CKD Stage 5

Over a median follow-up of 4 years (IQR 3.0–4.8),581 patients started maintenance dialysis (86 urgently), 84 had a pre-emptive transplantation and 402 died before starting any KRT. In the overall population, the 5-year risk of urgent- and non-urgent-start dialysis, pre-emptive transplantation and death were 4, 17, 3 and 15%, respectively. Among the 2834 patients with CKD Stages 2–4 at inclusion, the linear model identified 687 patients who reached eGFR <15 mL/min/1.73 m² during follow-up. Their 5-year risk to reach eGFR <15 was 28%, while their risks for urgent- and non-urgent-start dialysis and death were 1, 2 and 13%, respectively (Figure 1A). Among the 687 incident Stage 5 patients, the median follow-up was 44 weeks (IQR 24–73) and the 2-year risks for urgent- and non-urgent-start dialysis and death were 7, 54 and 10%, respectively (Figure 1B). Of note, 45 of the 687 patients had no further eGFR measurements after they reached eGFR <15 mL/min/1.73 m²; 21 of them had either died or started KRT. The other 642 had a mean of 6.1 ± 5.0 measurements, with a mean eGFR of 11.9 ± 2.5 mL/min/1.73 m².

Determinants of urgent-start dialysis

Among the 541 patients who initiated dialysis and had data available for their dialysis start status, 86 (16%) were identified with urgent starts. In crude analyses, patient-related factors significantly associated with urgent-start dialysis included older age, living alone and low health literacy (Table 2). Patients who started urgently tended to have missed a nephrology visit in the year before dialysis and reported refusing dialysis more often than those who did not, but these differences were not statistically significant. They also had more cardiovascular comorbidities and were more often exposed to hyperpolypharmacy. However, urgent-start dialysis was not significantly associated with gender, BMI, baseline albuminuria or either haemoglobin or eGFR at dialysis start. Significant provider-related factors included fewer nephrology visits during the year before dialysis initiation and fewer KRT planning and education programme sessions during follow-up.

In the multivariate logistic regression, the aOR for urgent-start dialysis was significantly higher in patients living alone and in those with low health literacy, heart failure or hyperpolypharmacy (Table3). In contrast, it was significantly lower for those with a higher number of nephrology visits and on the borderline of significance for planned KRT modality. The latter was strongly correlated with ‘attending an education programme’, which accordingly was not included in the model. Of note, the associations of urgent-start dialysis with age, diabetes and cardiovascular disease (excluding heart failure) were no longer significant after adjustment.

Modalities of dialysis initiation and early mortality

Fluid overload, hydroelectrolytic disorders, AKI and post-surgery kidney function worsening were more frequently reported as reasons to initiate dialysis in urgent- than non-urgent-start patients (Figure 2). A third of the non-urgent starts began in the hospital, and almost half of them were unplanned, while 31% of the urgent starts used a functional dialysis access (arteriovenous fistula or prosthetic graft or peritoneal catheter) (Table 4). There was only one urgent-start peritoneal dialysis. Early mortality was >4 times higher among urgent- than non-urgent-start patients.

DISCUSSION

This study confirms that a number of patients with CKD experience urgent start dialysis despite early referral to nephrology care. Its prospective design enabled us to estimate the relative weight of urgent-start dialysis compared with timely start KRT and death in this setting. A novelty of the study is that it identified both immediate trigger factors of urgent start and risk factors present ahead of the event that are potentially modifiable. It especially highlights the importance of social factors, hyperpolypharmacy, as well as some nephrology practices in the risk for urgent start, independent of clinical factors, notably heart failure.

To our knowledge, this study is the first study to estimate the competing risks of urgent-start dialysis versus timely KRT start or death in patients under nephrology care. While the absolute risks for these events strongly depend on age and eGFR at study entry, their relative weight may reflect patients’ profiles and their experience in the transition to kidney failure as managed in each clinical setting. In the overall CKD-REIN population (mean age 67 years; mean eGFR 32 mL/min/1.73 m²), the estimated 5-year risk of urgent dialysis start—4%—was significant compared with the 17% risk of non-urgent dialysis start and the 15% risk of death. Expressed as a percentage of all dialysis starts, the observed 16% of urgent starts was lower, as expected in this selected population of patients attending nephrology clinics, than the 30% reported by the REIN registry for the overall dialysis population, which includes many patients referred late to specialists [11, 15]. However, this percentage is not comparable with those of other studies that used different definitions than ours, including dialysis start with an intravenous catheter or in-hospital [13, 17–19]. Interestingly, our study shows that close to 30% of the patients who started urgently had a functional dialysis access at initiation and ˃75% had a planned KRT, versus 74% and 88%, respectively, for those with non-urgent starts. On the other hand, a third of the non-urgent starts initiated dialysis in the hospital, half of them unplanned. These findings demonstrate the importance of the clarification made by the KDIGO between urgent versus non-urgent start and planned versus unplanned dialysis modality; their causes may be different and require specific investigations [4].

Patients seen by nephrologists were recently shown to have more complex illnesses than those seen by other subspecialists, due to their large number of comorbidities and prescribed medications and their high mortality risk [36]. This complexity has been defined as ‘an interaction between the personal, social and clinical aspects of the patient’s experience that complicates patient care’. Transition to dialysis is a critical period where the combination of these factors may significantly amplify the risk of adverse events. As we reported previously, heart failure was a major risk factor for urgent start in this cohort [12, 17, 20]. Heart failure management in advanced CKD is challenging and volume management may require that dialysis start urgently [37].

We also identified hyperpolypharmacy as another major risk factor. Polypharmacy (≥5 drug classes) and hyperpolypharmacy (≥10 drug classes) are highly prevalent in CKD patients and associated with adverse drug effects, hospitalizations and mortality [38–41]. Their association with urgent-start dialysis has not previously been described. Whether this association is causal, resulting from over-, under- or otherwise inappropriate prescription of specific drug classes or their misuse by patients is uncertain. Nevertheless, these findings should encourage physicians to reassess medication prescriptions regularly in CKD patients and consider dose adjustment or deprescription whenever possible [42–44]. Of note, in contrast to other studies, we did not find that urgent-start dialysis was significantly associated with BMI, diabetes or cardiovascular disease other than heart failure after adjustment for other patient characteristics [12, 20].

Among the sociodemographic variables, older age is one the risk factors most frequently associated with urgent-start dialysis in crude analyses, while in studies that perform multivariate analyses, it is often non-significant, as it is in ours [12]. Socio-economic status, social isolation and health literacy have often been associated with poor outcomes in CKD, but to our knowledge, these factors have not previously been investigated in relation to dialysis start status [45–47]. In this study, a lower education level was not significantly associated with urgent-start dialysis. In contrast, we observed strong associations between an urgent start and both living alone and low health literacy, independent of clinical and provider-related factors. Health literacy is rarely assessed in routine nephrology practice but should be given greater attention in view of the complexity of kidney failure management [36].

Pre-dialysis educational programmes and timely modality planning have long been established as major determinants of better outcomes in the transition to dialysis [6, 7, 12, 15, 17–19]. Both were also associated with lower risks of urgent-start dialysis in crude analyses, but because they are strongly correlated, it was not possible to assess the independent impact of each on dialysis start status. However, although statistically significant, the observed difference in the percentage of patients who attended KRT education (89% and 81% in non-urgent- and urgent-start patients, respectively) may not be considered clinically meaningful. Beyond these two well-known factors, this study emphasized the importance of closer follow-up during the last year before dialysis. In multivariate analyses, each additional nephrology visit was significantly associated with 20% lower odds of urgent start independent of multiple potential confounders. A similar finding was also reported recently in the French incident dialysis population by the national REIN registry [15]. This result did not seem to be related to waiting time for an appointment or to missed appointments; those rates, although slightly higher in urgent- than non-urgent-start patients, did not differ significantly between them. Although 10% of the patients who started urgently had chosen peritoneal dialysis, our finding that only one was able to use it at initiation highlights the barriers to urgent-start peritoneal dialysis [4]. Similarly, that 4% of urgent-start patients had chosen conservative care may reflect the difficulty in exercising this option in the face of acute complications and the inability of supportive care to manage symptoms effectively [48].

Major strengths of our study include the representativeness of the nephrology clinic sample and the thorough identification of KRT events through annual contact with patients combined with record linkage with the national REIN registry. The use of a standardized definition for urgent-start dialysis and its cross-validation with the REIN registry data significantly decreased the risk of misclassification for the main study outcome. Finally, the extensive and prospective data collection made it possible to investigate a number of novel determinants and reduce information bias and potential confounding.

Our study also has limitations. First, while missing KRT events are unlikely, dialysis start status was missing for 40 (7%) dialysis events. Moreover, 13% of the dialysis start statuses we studied were only identified from the REIN registry, which uses an urgent-start definition similar to that of KDIGO, except for the shorter time to initiate dialysis after presentation, which is only 24 h in REIN rather than 48 h. Both these limitations may have resulted in slightly underestimating the risk of urgent start and reduced study power for multivariate analyses. In addition, although we define urgent-start dialysis accurately using CKD-REIN and REIN data and hospital reports, some subjectivity remains in the interpretation of ‘life-threatening manifestations’ requiring dialysis initiation in ˂48 h [4]. Third, we did not consider non-linear GFR trajectories to predict time to eGFR <15 mL/min/1.73 m², which could have affected our risk estimate for this event, as this type of trajectory can vary over time [49, 50]. Nonetheless, this had no impact on our main analyses of the determinants of urgent-start dialysis. Fourth, patients consenting to take part in research, even observational studies, are often different from those declining or not invited. In the CKD-REIN, however, we have previously shown that although study participants were slightly younger than overall nephrology clinic–eligible patients who were not invited to participate, their CKD stage and diabetes prevalence were similar at baseline [27]. Finally, despite its large sample size, this study lacked power to analyse risk factors for urgent start by subgroups of causes.

Our findings have several implications for patients and nephrology practices. They provide new insights on the competing risks of urgent-start dialysis and death for those who reached CKD Stage 5 and call attention to this critical aspect of the transition period to KRT, previously poorly documented. They also reveal the importance of social factors, often overlooked in this context. Patients living alone or with a poor understanding of written documents from their doctor should be considered at high risk and receive closer monitoring in advanced-stage CKD, especially if they have heart failure and/or are prescribed a large number of medications. Finally, they identified several areas of potential improvement in clinical practices, including increasing intensity of nephrology care and attendance at education programmes in advanced CKD, developing urgent-start peritoneal dialysis programmes and improving conservative care management.

In conclusion, urgent-start dialysis remains a challenging issue, even among patients with early nephrology care. This study identified a number of potentially modifiable risk factors, the consideration of which may substantially reduce the incidence of urgent-start dialysis. Whether the current implementation in France of a care bundle for patients with advanced CKD, including several nephrologist, advanced practice nurse and dietician visits combined with KRT education, will improve dialysis transition and reduce the urgent-start rate will need to be evaluated.

ACKNOWLEDGEMENTS

We acknowledge the CKD-REIN study coordination staff for their efforts in setting up the CKD-REIN cohort: Marie Metzger, Elodie Speyer, Céline Lange, Reine Ketchemin, Natalia Alencar de Pinho and all the clinical research associates. We thank Jo Ann Cahn for editing the English version. All legal authorizations were obtained, including those from the Comité consultatif sur le traitement de l'information en matière de recherche dans le domaine de la santé (CCTIRS No12.360), the Commission nationale de l’informatique et des libertés (CNIL N°DR-2012-469) and the Kremlin-Bicêtre Comité de protection des personnes (CPP No IDRCB 2012-A00902-41). CKD-REIN biological collection is registered in the management application of the COnservation D'Eléments du COrps Humain (CODECOCH No 2012-1624). The Inserm Institutional Review Board approved the study protocol (IRB00003888). ClinicalTrials.gov Identifier: NCT03381950. The CKD-REIN Study Collaborators include members of the steering committee and coordinating centre and site investigators. Steering committee and study coordination: Natalia Alencar de Pinho, Carole Ayav, Serge Briançon, Dorothée Cannet, Christian Combe, Denis Fouque, Luc Frimat, Yves-Edouard Herpe, Christian Jacquelinet, Maurice Laville, Ziad A Massy, Christophe Pascal, Bruce M Robinson, Bénédicte Stengel, Céline Lange, Karine Legrand, Sophie Liabeuf, Marie Metzger and Elodie Speyer. CKD-REIN site investigators: Thierry Hannedouche, Bruno Moulin, Sébastien Mailliez, Gaétan Lebrun, Eric Magnant, Gabriel Choukroun, Benjamin Deroure, Adeline Lacraz, Guy Lambrey, Jean Philippe Bourdenx, Marie Essig, Thierry Lobbedez, Raymond Azar, Hacène Sekhri, Mustafa Smati, Mohamed Jamali, Alexandre Klein, Michel Delahousse, Christian Combe, Séverine Martin, Isabelle Landru, Eric Thervet, Ziad A. Massy, Philippe Lang, Xavier Belenfant, Pablo Urena, Carlos Vela, Luc Frimat, Dominique Chauveau, Viktor Panescu, Christian Noel, François Glowacki, Maxime Hoffmann, Maryvonne Hourmant, Dominique Besnier, Angelo Testa, François Kuentz, Philippe Zaoui, Charles Chazot, Laurent Juillard, Stéphane Burtey, Adrien Keller, Nassim Kamar, Denis Fouque and Maurice Laville.

FUNDING

CKD-REIN is funded by the Agence Nationale de la Recherche through the 2010 ‘Cohortes-Investissements d’Avenir’ programme (ANR-IA-COH-2012/3731) and by the 2010 national ‘Programme Hospitalier de Recherche Clinique’. The CKD-REIN is also supported through a public–private partnership with Amgen, Fresenius Medical Care and GlaxoSmithKline since 2012, Lilly France since 2013 and Otsuka Pharmaceutical since 2015, Baxter and Merck Sharp & Dohme-Chibret (MSD France) from 2012 to 2017, Sanofi-Genzyme from 2012 to 2015 and Vifor Fresenius and AstraZeneca since 2018. Inserm Transfert set up and has managed this partnership since 2011. The funding sources played no role in the study design, conduct and reporting.

AUTHORS’ CONTRIBUTIONS

V.F., N.A.P. and B.S. designed the present study. C.C., L.F., M.L., C.J., D.F., S.L., R.P.F., Z.A.M. and B.S. designed the CKD-REIN cohort. C.C., D.F., L.F., C.J., M.L., Z.A.M., C.L. and B.S. were responsible for data acquisition. V.F., N.A.P., M.M., A.H., J.B. and B.S. were responsible for data analysis and/or interpretation. V.F., N.A.P., J.B. were responsible for statistical analysis. N.A.P., J.B. and B.S. were responsible for supervision or mentorship.

CONFLICT OF INTEREST STATEMENT

B.S. reports grants for the CKD-REIN cohort study from Amgen, Baxter, Fresenius Medical Care, GlaxoSmithKline, Merck Sharp and Dohme-Chibret, Sanofi-Genzyme, Lilly, Otsuka and Vifor Fresenius, as well as speaker honoraria at the French Society of Diabetology from Lilly and at the French-speaking Society of Nephrology, Dialysis and Transplantation from MSD, unrelated to the content of this manuscript. Z.A.M. reports grants for CKD-REIN and other research projects from Amgen, Baxter, Fresenius Medical Care, GlaxoSmithKline, Merck Sharp and Dohme-Chibret, Sanofi-Genzyme, Lilly, Otsuka and the French government, as well as fees and grants to charities from Astellas, Baxter, Daichii and Sanofi-Genzyme; these sources of funding are not necessarily related to the content of the present manuscript. D.F. reports grants for other research projects from Fresenius Kabi, Sanofi. R.P.-F. reports research grants from Fresenius Medical Care, National Council for Scientific and Technological Development, grants (paid to employer) from AstraZeneca, Boehringer-Lilly, Novo Nordisk, Akebia and Bayer for participation in advisory boards and educational activities. C.C. reports grants from Novartis and Sanofi and honoraria from Amgen, Fresenius and Travere, not necessarily related to the content of the present manuscript. The other authors declare that they have no competing interests. The results presented in this article have not been published previously in whole or part.

REFERENCES

Author notes