This year, the #NephJC crew is bringing you certain highlights from Kidney Week with daily blogs. Our focus will be mostly on the late breaking science, and any other simultaneous publications that we come across. There is a lot more at Kidney Week of course - but this is high level clinical science that we thought our readers would be interested in.
Day 2 Studies:
PISCES: Fish oil in Dialysis; Lok et al NEJM 2025
LIBERATE-D: Dialysing less in AKI to promote recovery; Liu et al JAMA 2025
Lower IS in older Kidney transplants: Sanders et al JASN 2025
Discordance between Cystatin C and Creatinine, a meta-analysis; Estrella et al JAMA 2025
Flozination effects by diabetes status and level of albuminuria- meta-analysis; Staplin, et al, JAMA, 2025
Flozination Kidney outcomes by GFR and Albuminuria; Neuen et al JAMA 2025
Global Burden of Disease, CKD edition: GBD Consortium Lancet 1025
CHW support for Hispanic/Latino HD individuals; Cervantes et al JAMA IM 2025 and patient experience; Rizzolo et al JAMA NO 2025; Cervantes et al KI Reports 2025
Note: all studies do not have a PMID yet, so we have embedded the journal url in the title - or click the links above.
Something fishy, or finally a catch? The PISCES trial on Omega-3s in hemodialysis
N Engl J Med. 2025 Nov 7. doi: 10.1056/NEJMoa251303. Online ahead of print.
Fish-Oil Supplementation and Cardiovascular Events in Patients Receiving Hemodialysis
Charmaine E. Lok, Michael Farkouh, Brenda R. Hemmelgarn, Louise M. Moist, Kevan R. Polkinghorne, George Tomlinson, Paul Tam, Marcello Tonelli, and Jacob A. Udell, for the PISCES Investigators
PMID: 41201837
Why was the study done?
For decades, dialysis cardioprotection has been a story of disappointment: statins, ACE inhibitors, and antiplatelets all swam upstream and died on the rocks of uremia. The PISCES trial (Lok CE, et al, NEJM, 2025) finally tested a hypothesis that had been circling since the FISH study (Lok CE, et al, JAMA, 2012): could long-chain omega-3 fatty acids (eicosapentaenoic acid -EPA, and docosahexaenoic- DHA) reduce the relentless cardiovascular mortality in hemodialysis? Prior observational and small interventional work showed low plasma omega-3 levels in dialysis patients and suggested inverse correlations with sudden cardiac death and major cardiovascular events (Friedman AN, et al, Am J Nephrol, 2012; Madsen T, et al. J Ren Nutr, 2011). Mechanistic data supported effects on membrane stabilization, anti-inflammatory signaling, and arrhythmia suppression- an appealing angle in a population where sudden cardiac death remains the principal cause.
How was it done and What did it show?
In PISCES, 1,228 patients from 26 Canadian and Australian centers were randomized to high-dose fish oil (4 g/day of combined EPA+DHA) or corn oil placebo for 3.5 years. The main age was about 64 years, with roughly two-thirds male participants. Diabetes (~55%) and hypertension (~85%) dominated the comorbidities profiles, while 30% had prior coronary artery disease and 19% HF (supplementary table S3). Importantly, two-thirds of patients had no prior CV event, positioning the trial closer to primary prevention than most nephrology CV studies. 40% of the patients were White, 15-16% each Asian and Southeast Asian, and 12-14% Black. Dialysis vintage was modest (2.5 years). The results were unexpectedly striking. The survival curves for the primary endpoint begin to separate almost immediately- a surreal sight in nephrology.
Figure 1. The mean cumulative number of serious cardiovascular events per participant in each trial group, from Lok CE, et al, NEJM, 2025
Rates of serious cardiovascular events (defined as fatal and nonfatal myocardial infarction, stroke, peripheral vascular disease requiring amputation, and cardiac death) were nearly halved (0.31 vs 0.61 per 1000 patient-days, HR 0.57, 95% CI 0.47-0.70). Event reductions were consistent across components: cardiac death (HR 0.55, 0.40-0.75), myocardial infarction (0.56, 0.40-0.80), stroke (0.37, 0.18- 0.76), and peripheral vascular disease (0.57, 0.38-0.86).
Table 2. Primary and secondary endpoints, from Lok CE, et al, NEJM, 2025
What does it all mean?
This was not a perfect catch. Patients with implantable cardioverter-defibrillators were excluded, trimming the highest-risk arrhythmic subgroup. The pragmatic design allowed real-world variability but sacrificed control over diet, inflammation, and concomitant therapies. Reliance on self-reported adherence over 3.5 years invites some drift despite biochemical checks. And while the endpoint was adjudicated blindly, the composite’s breadth complicates mechanistic attribution- was the benefit driven by antiarrhythmic, vascular, or anti-inflammatory effects?
Nevertheless, PISCES delivers something rare in dialysis trials: a clean, statistically robust signal favoring an accessible, low-toxicity therapy. Unlike in REDUCE-IT (Bhatt et al NEJM 2018), which demonstrated an impressive 25% reduction in MACE and 20% reduction in CV death in ~ 8000 patients with hypertriglyceridemia, the placebo was not mineral oil - it was corn oil and unlikely to be toxic. PISCES was also funded by peer-reviewed agencies and not the drug company. If confirmed, this would mark the first intervention to lower cardiovascular events in end-stage kidney disease in decades. Until replication or refutation, we find ourselves in an unusual place- hooked with a positive trial in dialysis. Watch out for a more detailed dive in an upcoming NephJC.
LIBERATE-D trial: Should we dialyze less to heal more?
JAMA Published Online: November 7, 2025, doi: 10.1001/jama.2025.21530
A Conservative Dialysis Strategy and Kidney Function Recovery in Dialysis-Requiring Acute Kidney Injury
The Liberation From Acute Dialysis (LIBERATE-D) Randomized Clinical Trial
Kathleen D. Liu, Edward D. Siew; Delphine S. Tuot, Anitha Vijayan; Gonzalo Matzumura Umemoto; Bethany C. Birkelo; Benjamin J. Lee; Y. Diana Kwong; Ian E. McCoy; Kevin Delucchi; Hanjing Zhuo; Chi-yuan Hsu
PMID 41201895
Why was the study done?
Dialysis-requiring acute kidney injury (AKI-D) has always been more art than algorithm. In contrast to maintenance kidney failure, where thrice-weekly schedules became ritual, AKI-D management hinges on judgment: when to start, when to stop, and how much to do in between. The physiological balance is delicate: each dialysis session clears solutes but risks hypotension, myocardial stunning, and loss of residual renal function. Clinicians often default to chronic-dialysis patterns out of habit rather than evidence. LIBERATE-D (Liu KD, et al, JAMA, 2025) directly challenges that reflex. It tested whether a conservative, indication-based dialysis approach- dialyzing only when clearly needed- could hasten kidney recovery compared with conventional thrice weekly therapy. The trial builds on observational signals that higher dialysis frequency may delay the recovery (Hsu RK, et al, JASN, 2013; Siew ED, Kidney Int. 2015) but brings prospective data to debate that has remained largely philosophical.
How was it done and What did it show?
LIBERATE-D enrolled 221 hospitalized adults with AKI-D across four US centers (110 conservative, 111 conventional). Participants were middle-aged (mean 56 years), 67% male, and 60% White, with a median of 9 days of prior dialysis. Baseline eGFR (i.e. prior to developing AKI) averaged 65 mL/min/1.73m2, and both groups were well balanced (supplement table 1). The conservative strategy reduced treatment intensity (median 1.8 vs 3.1 sessions per week; difference -1.4, 95% CI -1.8 to -1.0) and lowered intradialytic hypotension (69 vs 97 events; eTable 6). Kidney recovery before discharge occurred in 64% vs 50% (difference 13.8%, 95% CI 0.8-26.9, p=.04). In adjusted models, the effect attenuated (OR 1.56, 95% CI 0.86, P=0.15). Recovery curves began separating early, consistent with biologically plausible benefit from reduced hemodynamic stress.
Figure 2. Time to first kidney function recovery before hospital discharge, from Liu KD, et al, JAMA, 2025
What does it all mean?
The study’s modest size limits precision; confidence intervals include both modest benefit and no effect. Its open-label design introduces potential clinician bias in defining indications. The enrolled cohort excluded the most unstable or ventilated patients, tempering generalizability. Long-term renal outcomes and mortality beyond 90 days remain unknown.
LIBERATE-D reframes AKI-D not as a protocol to execute but as a process to interpret. The findings suggest that restraint, guided by physiology, may serve recovery better than routine. Future multicenter trials with larger samples, objective ultrafiltration triggers, and follow-up beyond discharge could confirm whether less dialysis truly means more kidney survival. For now, Table 3 (+ supplement eTable 3) illustrates a simple truth: in the art of AKI dialysis, sometimes the most therapeutic act is knowing when not to dialyze.
Table 3. Participants who experienced adverse events by treatment group, from Liu KD, et al, JAMA, 2025
Can we lighten the load? Immunosuppression tailoring in older kidney transplant recipients
J Am Soc Nephrol. 2025 Nov 7. doi: 10.1681/ASN.0000000924. Online ahead of print.
Immunosuppression in Older Kidney Transplant Recipients: A Randomized Controlled Trial
Jan-Stephan F Sanders , Silke E de Boer, Jip Jonker , Frederike J Bemelman, Michiel G H Betjes, Aiko P J de Vries , Luuk Hilbrands , Marc Hilhorst , Dirk R J Kuypers, Priya Vart 1, Arjan D van Zuilen, Dennis A Hesselink, Stefan P Berger
PMID: 41201871
Why was the study done?
Older kidney transplant recipients present a unique biological paradox: lower immunologic vigor but higher vulnerability to infection, malignancy, and drug toxicity. Despite this, most immunosuppressive regimens mirror those used in younger, more immunoreactive patients. The OPTIMIZE trial (Sanders JF, et al, JASN, 2025) tested whether older recipients might benefit from reduced calcineurin inhibitors exposure combined with everolimus - preserving efficacy while limiting toxicity. Prior data from TRANSFORM (Berger SP, et al, Am J Transplant, 2019) and SENATOR (Brakemeier S, et al PLoS One, 2019) hinted at such potential, but were underpowered for elderly-specific outcomes. OPTIMIZE aimed to fill the gap by directly examining age-appropriate immunosuppression.
How was it done and What did it show?
OPTIMIZE enrolled 379 de novo kidney transplant recipients aged ≥ 65 years from seven centers. Participants were stratified into two groups: those receiving kidneys from older deceased donors (≥ 65 years, stratum A) or from younger/ living donors (stratum B). All received basiliximab induction and corticosteroids, then randomized to:
TEP ( low-dose tacrolimus+everolimus+prednisolone) or
TMP (standard tacrolimus+mycophenolate mofetil + prednisolone).
The primary endpoint (2-year patient and graft survival with eGFR ≥ 30 mL/min/1.73m2) was achieved in 50% TEP vs 57% TMP, risk difference -7% (95% CI -17 to 3, P=0.18)
Figure 3. Absolute risk difference and 95% CI of the primary and secondary study end points between the TEP and TMP treatment groups, from Sanders JF, et al, JASN, 2025
Rates of treated biopsy-proven acute rejection (tBPAR) were 11% vs 8%, and eGFR trajectories overlapped across groups. Infection incidence (81% vs 83%) and viral events (55% vs 54%) were similar, contrasting prior mTOR studies. Edema occurred more often with TEP (59% vs 43%), whereas nonskin malignancies trended lower (4% vs 8%).
Table 3. Adverse events of special interest, from Sanders JF, et al, JASN, 2025
What does it all mean?
The open-label design, modest sample size, and broad confidence intervals limit inference. The study was powered to detect only large differences (≥ 14% success rate) and was designed as a superiority and not a non-inferiority trial. Conduct during the COVID-19 pandemic may have masked infection-related distinctions. A two-year follow-up was likely too short to capture chronic CNI nephrotoxicity or malignancy divergence.
OPTIMIZE demonstrates that in older recipients, less immunosuppression is not necessarily more effective. Yet, its safety findings suggest that reduced CNI plus everolimus can maintain outcomes without excess rejection - though be careful drawing non-inferiority inferences from a negative superiority trial. The logical next step is a longer, event-driven study integrating frailty and pharmacokinetic profiling to define “immunologic age”. For now, figures 4-5 reinforce the pragmatic message: age alone should not dictate under-immunosuppression, but it should demand precision-lighter where biology allows, vigilant where risk accumulates.
Figure 4. Kaplan–Meier plot for prespecified secondary outcomes patient survival (A), graft survival (B), and death-censored graft survival (C) in the TEP and TMP treatment groups, from Sanders JF, et al, JASN, 2025
Figure 5. eGFR (mean and 95% CI) over time in the study population stratified by treatment group and stratum, a secondary outcome, from Sanders JF, et al, JASN, 2025
When filtration markers disagree: which GFR should we believe?
JAMA. 2025 Nov 7. doi: 10.1001/jama.2025.17578. Online ahead of print.
Discordance in Creatinine- and Cystatin C-Based eGFR and Clinical Outcomes: A Meta-Analysis
Michelle M Estrella, Shoshana H Ballew, Yingying Sang, Morgan E Grams, Josef Coresh, Aditya Surapaneni, Natalia Alencar de Pinho, Johan Ärnlöv, Hermann Brenner, Juan-Jesus Carrero, Teresa K Chen, Debbie L Cohen, Mary Cushman, Ron T Gansevoort, Shih-Jen Hwang, Lesley A Inker, Joachim H Ix, Keiko Kabasawa, Tsuneo Konta, Jennifer S Lees, Kevan R Polkinghorne, Michael G Shlipak, Robin W M Vernooij, David C Wheeler, Ashok Kumar Yadav, Andrew S Levey, Kai-Uwe Eckardt; Chronic Kidney Disease Prognosis Consortium Investigators and Collaborators
PMID: 41202182
Why was the study done?
Discrepancies between creatinine- and cystatin C-based eGFR estimates have long puzzled clinicians, with an epidemic of papers indecent years as cyststain C use becomes more common. We know that creatinine is confounded by muscle mass and diet, whereas cystatin C reflects inflammation and adiposity. Whether these differences carry prognostic meaning has remained uncertain (see McCoy et al AJKD 2025 | NephJC blog). The Chronic Kidney Disease Prognosis Consortium undertook the largest analysis to date to test where discordance between eGFRcys and eGFRcr predicts clinical outcomes beyond numerical noise. (Estrella MM, et al, JAMA, 2025)
How was it done and What did it show?
The meta-analysis included 821,327 outpatients and 36,639 inpatients from 25 cohorts across 12 countries. The mean outpatient age was 59 years, 48% were female, 13.5% had diabetes, and 40% hypertension; inpatients were older (67 years) with more comorbidities. On average, eGFRcys was 6 mL/min/1.73m2 lower than eGFRcr. A large negative discordance- eGFRcys ≥ 30% lower than eGFRcr- occurred in 11% of outpatients and 35% of inpatients.
Figure 1. Weighted distribution of the eGFRdiff across each cohort, from Estrella MM, et al, JAMA, 2025
Over a mean 11-year follow-up, outpatients with large negative discordance (aka eGFRcys lower than eGFR Cr) had higher rates of all-cause mortality (28.4 vs 16.8 per 1000 person-year, HR 1.69 [95% CI 1.57–1.82]), cardiovascular mortality (6.1. vs 3.8, HR 1.61 [1.48–1.76]), ASCVD (13.3 vs 9.8, HR 1.35 [1.27–1.44]), heart failure (13.2 vs 8.6, HR 1.54 [1.40–1.68]), and kidney failure (2.7 vs 2.1, HR 1.29 [1.13–1.47])
Figure 2. Association of eGFRdiff with risk of long-term adverse outcomes in the outpatient cohorts, from Estrella MM, et al, JAMA, 2025
Associations were consistent across general, CKD, and clinical cohorts. Large positive discordance- eGFR higher than eGFRcr- was uncommon (2.8%) and not clearly linked to outcomes.
What does it all mean?
The analysis, though massive, remains observational. Cohort heterogeneity, residual confounding from inflammation or malnutrition, and limited inpatient outcomes data restrict causal inference. Is it something about the kidney function estimation that drives the result (eg creatinine overestimation of GFR in frail sarcopenia) - or is cystatin C picking up a nonrenal signal (eg inflammation) that is driving this (see Lees et al Nat Med 2019 | NephJC Summary) ? Cystatin C assays, despite International Federation of Clinical Chemistry and Laboratory Medicine calibration, still vary slightly across laboratories. Reclassification analyses (eTable 2) also show that discordance frequency rises with lower eGFRcr, complicating interpretation near the CKD threshold.
This study firmly establishes discordant eGFR as a prognostic biomarker rather than a laboratory artifact. Integrating cystatin into CKD staging could define risk prediction, particularly in frail or sarcopenic populations where creatinine-based estimates overstate kidney function. Implementation studies are needed to test whether acting on discordance- such as earlier nephrology referral or medication adjustment- improves outcomes. For now, figures 1 and 2 remind clinicians that when creatinine and cystatin disagree, the lower estimate likely whispers the truth.
Kidney karma, or just osmotic physics? The flozins meta-analyses redraw the map
JAMA. 2025 Nov 7. doi: 10.1001/jama.2025.20834
SGLT2 Inhibitors and Kidney Outcomes by Glomerular Filtration Rat and Albuminuria- a Meta-Analysis
Brendon L. Neuen, Robert A. Fletcher, Stefan D. Anker, Deepak L. Bhatt, Javed Butler, David Z. I. Cherney, Kieran F. Docherty, Silvio E. Inzucchi, Meg J. Jardine, Kenneth W. Mahaffey, Finnian R. McCausland, Darren K. McGuire, John J. V. McMurray, Bruce Neal, Milton Packer, Siddharth M. Patel, Vlado Perkovic, Marc S. Sabatine, Rebecca J. Sardell, Scott D. Solomon, Muthiah Vaduganathan, Christoph Wanner, David C. Wheeler, Faiez Zannad, Richard Haynes, Natalie Staplin, William G. Herrington, Hiddo J. L. Heerspink, for the SGLT2 Inhibitor Meta-Analysis Cardio-Renal Trialists’ Consortium (SMART-C)
and
JAMA. 2025 Nov 7. doi: 10.1001/jama.2025.20835. Online ahead of print.
Effects of Sodium Glucose Cotransporter 2 Inhibitors by Diabetes Status and Level of Albuminuria: A Meta-Analysis
Natalie Staplin, Alistair J Roddick, Brendon L Neuen, Stefan D Anker, Deepak L Bhatt, Javed Butler, David Z Cherney, Kieran F Docherty, Robert A Fletcher, Silvio E Inzucchi, Meg Jardine, Kenneth W Mahaffey, Darren K McGuire, John J V McMurray, Bruce Neal, Milton Packer, Siddharth M Patel, Vlado Perkovic, Marc S Sabatine, Scott Solomon, Muthiah Vaduganathan, Christoph Wanner, David C Wheeler, Faiez Zannad, Richard Haynes, Hiddo J L Heerspink, William G Herrington; SGLT2 Inhibitor Meta-Analysis Cardio-Renal Trialists’ Consortium (SMART-C)
PMID: 41202026
The empire of filtration levels strikes back
For years, nephrologists drew a boundary line: SGLT2 inhibitors for diabetic, albuminuric, CKD; everything else, uncharted. The JAMA twin meta-analyses by Neuen BL et al and Staplin N et al, under the SMART-C consortium, quietly erased the map. Together, they analyzed over 120,000 participants across 15 trials, spanning diabetic, non-diabetic, albuminuric, and non-albuminuric kidney disease, with eGFR dipping below 20 mL/min/1.73,2. The question was deceptively simple: does the benefit fade as the filtration falters or albuminuria wanes? The TL;DR answer: no. These drugs go beyond kidney protection in familiar high-risk settings; they smooth the entire risk gradient.
Neuen BL et al: the slope that refused to fall
Pooling 10 randomized trials (N= 70,361), this meta found SGLT2 inhibitors reduced kidney disease progression by 38% (HR 0.62, 95% CI 0.57-0.68)- a uniform effect from eGFR ≥ 60 down to < 30 mL/min/1.73 m2 (p-trend=.49) (fig 1) and across all albuminuria strata (fig 2).
Flozins slowed annual eGFR decline by 1.3 mL/min/1.73m2, a modest-sounding change that accumulates into years of dialysis delay. Importantly, the benefit extended even to patients without overt proteinuria or with preserved filtration, where event rates were low but directionally identical. The class effect held steady whether background therapy was RAAS blockade or not.
Staplin N et al,: diabetes is optional, benefit mandatory
The companion meta-analysis merged eight SGLT2 inhibitors trials(N= 58,816) and reframed the debate: do glycemia or albuminuria meaningfully modify the benefit? Across all strata, the answer was again steady. For diabetic CKD, progression fell from 48 to 33 events per 1000 person-years (HR 0.65, 95% CI 0.60-0.70); for non-diabetic CKD, 46 to 31 (0.74, 95% CI 0.63-0.85). Heart failure hospitalization and all-cause death also declined, with proportional benefits mirroring kidney outcomes. Albuminuria amplifies absolute gains, but doesn’t dictate the direction. The gradient is economic (not biological!): same relative effect, larger absolute return in higher baseline risk (Fig 1-4)
One drug class to rule them all? Not quite…
These analyses dispel two persistent myths: first, that SGLT2i “ need proteinuria to work”, and second, that they are “diabetic drugs with renal side effects”. In reality, the osmotic and hemodynamic mechanisms (tubulo-glomerular feedback, afferent vasoconstriction, and improved oxygen handling) seem universal. As Gregg et al, note in their editorial, the constancy of relative effect across risk groups contrasts sharply with diverging absolute benefits: a profound NNT of approx 16 for albuminuric diabetes versus approx 120 for non-albuminuric, nondiabetic CKD (Gregg LP et al, JAMA 2025). That is of course why one should use relative risks and not absolute risks or NNTs at all! See previous discussion (Manasi Bapat, NephJC stats post).
Global, regional, and national burden of chronic kidney disease in adults, 1990–2023, and its attributable risk factors: a systematic analysis for the Global Burden of Disease Study 2023
GBD 2023 Chronic Kidney Disease Collaborators†
Lancet 2025 Published November 7, 2025 DOI: 10.1016/S0140-6736(25)01853-7
The Global Burden of Disease (GBD) study has been methodically and regularly churning out widely cited and deeply data-filled studies covering almost all health aspects one can think of. In the latest edition, we get to see this applied to CKD. The GBD studies are a sophisticated mix of systematic reviews, registry data, all pooled together to synthesize information in an easy to understand, though data-dense fashion. In the present analysis, CKD was defined as either GFR < 60 or albuminuria (KDIGO A2 or higher) if GFR > 60. Some of the salient data points follow (but it’s easy to get lost in the rich tables and figures provided to get a sense of CKD across the world).
Figure 1 . Global age-standardised total chronic kidney disease prevalence (A) and death rates per 100 000 (B) in 2023 in people aged 20 years and older
CKD Prevalence
Globally in 2023, an estimated 788 million people aged 20 years and older were living with CKD. If we take age > 20 as 70% of the world’s 8 billion people, that represents about 14% or 1 in 7 people living with CKD.
The proportion of CKD patients in the population is highest in the Middle East and North Africa (see heat map below) but the absolute numbers are highest in China (152 million; and India (138 million).
Most of these are CKD 1-3, and only 0.6% (of the total 14.2% age-standardized prevalence) is stages 4 and 5.
Higher income countries have a lower prevalence of CKD than lower income countries - but a higher proportion of them are on dialysis and transplant, reflecting the underlying health equity and access issues with CKD care.
CKD Mortality
In 2023, globally, there were 1.48 million deaths for which CKD was the underlying cause in adults aged 20 years and older.
The global age-standardised death rate due to CKD was 26.5 deaths per 100 000 which makes it the ninth highest cause of death overall.
Of dubious honor, CKD (along with Diabetes and Alzheimer) is the only cause of death that has inched up the rankings since 1990.
The ten countries with the highest age-standardised mortality rates due to CKD were Nauru, Marshall Islands, Egypt, Tuvalu, El Salvador, Mauritius, Saudi Arabia, Northern Mariana Islands, American Samoa, and Nicaragua.
Disability (DALY)
In 2023, globally, CKD resulted in 769 age-standardised DALYs (higher is worse) per 100 000 among adults aged 20 years and older.
The regions with low dialysis and transplant access had the highest DALYs (e.g. CKD accounting for 1408 DALYs per 100 000 in sub-Saharan Africa, and 1321 DALYs per 100 000 in Latin America and the Caribbean)
Risk Factors
The five leading risk factors contributing to age-standardised DALYs due to CKD were high fasting plasma glucose (32%), high systolic blood pressure (25%), high BMI (24%), diet (18%), and non-optimal temperature (4.4%).
What happens when community health workers join the dialysis team for Hispanic and Latino patients- can navigation close the gap?
JAMA November 7, 2025. doi:10.1001/jamainternmed.2025.5305
Community Health Worker Support for Hispanic and Latino Individuals Receiving Hemodialysis: The Navigate-Kidney Randomized Clinical Trial
Lilia Cervantes; Elizabeth Juarez-Colunga; Neil R. Powe, Jennifer E. Flythe; John F. Steiner; Daniel Cukor; Romana Hasnain-Wynia; Seth Furgeson; Ladan Golestaneh; Claudia Camacho; Lauren McBeth; Brenda L. Beaty; Jiayuan Shi; Emily Bacon; Michel Chonchol,
Why was this Study Done?
Dialysis patients mostly die of CV causes, though it may be more heart failure and sudden death than plaque rupture, and interdialytic weight gain (IDWG) and missed dialysis sessions do contribute to this. Minimizing IDWG is not easy (it’s about salt and not fluid intake for starters!) and requires understanding of diet and nutritional information. Dialysis patients also do not miss dialysis for the fun of playing truant - it is a complex interplay of social, economic, educational factors. These factors are highly prevalent in the Hispanic/Latino population - as the authors put it ‘ (they)face a disproportionate burden of multilevel social and structural challenges that contribute to kidney health disparities—including at the individual (low patient activation, poverty, poor access to healthy foods), interpersonal (lack of culturally responsive care), community (distrust, food deserts), and societal (immigration policies, exclusion from health insurance) levels’. These are not easy problems to address. The authors hypothesized that community health workers (CHWs) could provide culturally tailored education and help overcome modifiable social barriers, such as access to healthy foods, which would then lead to lower IDWG and missed sessions.
How was it done and What did it Show?
This was a parallel-group RCT to determine whether the Navigate-Kidney intervention reduces IDWG compared to standard care. Participants were enrolled by an experienced CHW at 5 urban dialysis centers in Denver, Colorado, between October 2020 and April 2022, then followed up through August 2022. Eligibility was chronic HD patients who self-identified as Latino or Hispanic. Individuals were excluded if they had cognitive impairment, active suicidal intent, psychosis, bipolar disorder, pregnancy, or were expected to undergo kidney transplantation within 3 months. Participants assigned to intervention were seen by a CHW within 2 weeks of enrollment and then every 2 weeks for a total of 6 or more visits within the first 3 to 4 months of enrollment. CHWs were trained and supervised to emphasize 4 core intervention functions (termed functions by the form-function framework) - see figure below.
The study was scheduled to launch in March 2020 with a primary outcome of depressive affect. March 2020 was the start of COVID in North America (when all of us became quite depressed) and this was changed to IDWG. For the outcome analysis, IDWG was measured on a session-to-session basis, in a 90-day period prior to randomization and then during a 180-day period following the end of intervention. IDWG was calculated as the difference in standing weight at the beginning of each hemodialysis session (preweight) minus the standing weight after (postweight) the previous hemodialysis session, divided by the prescribed estimated dry weight - thus IDWG = [preweight − postweight]/estimated dry weight, expressed as a percentage. With an assumption of a conservative target effect size of 0.50 for this measure, they determined that a sample size of 128 participants would provide 80% power to detect a difference of 0.5 SD units. This was inflated to 141 assuming some dropouts and loss to follow up.
Of the 174 assessed for eligibility, a whopping 139 were enrolled, roughly 70 to each group. Even though the intervention was designed to deliver at least 6 visits to each participant, participants had a range of visits from 4 through 19, with a median of 6.5 visits. The mean age was 57 years, half females. Notably, 95 (68%) reported limited English proficiency, 103 (74%) had less than high school education, 96 (72%) had household income below $25 000, 39 (28%) experienced housing insecurity, and 35 (25%) and 42 (30%) reported food or health care access issues in the past year - highlighting the population was quite appropriate for this trial.
There was a statistically significant difference in postintervention IDWG between the intervention and standard care, corresponding to a difference of −0.46 percentage points in IDWG at the end of the follow-up between Navigate-Kidney vs standard care (mean IDWG, 3.26% [95% CI, 2.83%-3.68%] vs 3.72% [95% CI, 3.30%-4.14%]) as seen in table 2 below.
Though the rate of missed sessions was no different (median 0 sessions missed in both groups - ?Hawthorne effect) the rate per 30 days of shortened dialysis sessions was lower in the Navigate-Kidney group vs standard care between preintervention and postintervention periods (median [IQR], 0.1 [−1.2 to 1.1] vs 0.6 [−0.5 to 1.8]; P = .02). More results - including a higher ‘Patient Activation Measure’ was also noted. There is also an accompanying study (Rizzolo et al, JAMA Network Open 2025) which explores the patient preferences in a subset of 24 participants who underwent detailed interviews.
Figure. Schema of the key relationships between themes (functions) and subthemes, from Rizzolo et al, JAMA Network Open 2025
What does it all mean?
The authors conclude that a community-partnered and designed, CHW-led intervention modestly lowered IDWG, reduced shortened hemodialysis sessions, and improved patient activation for Hispanic and Latino participants with hemodialysis-dependent kidney failure. This seems like a low-risk and safe intervention that seems well-liked by patients (e.g. high enrollment, use of CHW for a longer median number than planned). Generalizability would be quite limited - given these were in a single city. There is a possibility of contamination, as the intervention and control groups were in the same 5 units, which might imply the true effect might even be larger. The intervention is itself perhaps complex, and in an accompanying paper (Cervantes et al, KI Reports 2025), the authors describe the framework for designing the intervention properly using implementation science.
Figure ADAPT steps. CHW, community health worker; CSC, community steering committee, from Cervantes et al, KI Reports 2025
This might seem like an intensive and cumbersome intervention - but reducing IDWG is not an easy thing to achieve (especially to show this with a small N in 270 days). Some of the things the authors describe might be useful to understand the effect. E.g., the CHW acted as a cultural broker, sharing with the dietitian important, otherwise unappreciated, details such as types of food in participant homes and information about who shopped and cooked for the participant. They explained the connection between the salt content of food, thirst, water consumption, and subsequent IDWG and used motivational interviewing to teach participants how to reduce IDWG through careful grocery shopping and cooking. In one case, a CHW helped a participant understand that adding chicken bouillon to cooking increased the salt content, driving thirst, and potentially increasing IDWG. We could also discuss these things directly - rather than spend time discussing phosphate levels and fretting over Kt/V?
Summary by Cristina Popa, and Swapnil Hiremath