Research report
A research report is the sourced material a domain
dossier is synthesized from — generated on a plan and a cadence, one topic per file.
A report carries no confidence tags. Its bracketed markers say who might have
an incentive to shade a line; none of them says anyone checked it. To reach a score, a line has
to be drafted onto a candidate as unverified and pass the Verifier or the
Corroborator, like everything else.
Section epidemiology · Version 2026-09-01 · Cadence annual · Evidence class 2 published · Sources 6 · Supersedes none
Domain: potassium-monitoring-ckd · Scope: public primary sources reachable by this repo's connectors, WebSearch and WebFetch. Excludes CPT descriptors (AMA-licensed), fda.gov guidance documents, veterinary datasets, and non-US findings, which are named as such rather than folded into a US trend claim.
Sourcing: six peer-reviewed NHANES- or claims-based US population studies, every one single-source, including two on the same diabetic-CKD subpopulation that report different trend directions over overlapping but non-identical windows
The seed dossier's epidemiology section anchored CKD and hyperkalemia prevalence to single-site Mayo Clinic cohorts and left national trend direction, incidence, and demographic concentration open. This pass fills part of that gap with US population-representative sources: CKD point prevalence (stage 3-4) in NHANES rose from 3.9% in 1988-1994 to 5.2% by 2003-2004 and then held roughly flat through 2015-2016, with disparities by race/ethnicity and socioeconomic status persisting across all 28 years [1]. All-cause mortality among adults with CKD stayed roughly twice that of adults without CKD across the same NHANES-tracked period, but the sex-specific trend diverged sharply — declining among men, rising among women [2]. A national commercial claims analysis put 2014 US hyperkalemia prevalence at 1.57% overall and 6.35% among patients with CKD and/or heart failure, and stated the rate had risen since 2010 [4]. A genuine incidence figure — new cases per population-year, as opposed to a point-in-time prevalence snapshot — was searched for and not found for either CKD or heart failure in this pass; every quantity below is prevalence or mortality, not incidence, and that distinction is carried through deliberately.
_Baseline (v1). No prior version; this establishes the starting point for future diffs._
The largest published NHANES trend analysis for CKD stage 3-4 prevalence used repeated cross-sections of 54,554 US adults aged 20 and older, spanning 1988-1994 and biennial cycles from 1999 to 2016 [1] [single-source]. Its central trend finding:
"The age-, sex- and race/ethnicity-adjusted overall prevalence of stage 3 and 4 CKD increased from 3.9% in 1988-1994 to 5.2% in 2003-2004 (difference, 1.3%; 95% CI, 0.9%-1.7%; P < .001 for change) and remained relatively stable thereafter at 5.1% in 2015-2016." — PMID 32672828
The trend was not uniform across subgroups. Prevalence among Mexican American adults "remained stable between 1988-1994 and 2003-2004 but nearly doubled (difference, 2.1%; 95% CI, 0.9%-3.3%; P = .001 for change) between 2003-2004 and 2015-2016" [1] [single-source], while prevalence gaps by education and income (5.8% vs. 4.3% low- vs. high-education, 4.3% vs. 3.1% low- vs. high-income, both in 1988-1994) "remained largely consistent throughout the entire period" [1] [single-source].
Two NHANES-based papers examine CKD prevalence trend specifically among US adults with diabetes, over overlapping but non-identical windows, and do not read the same way. The first, covering 1999-2018 (n=7,320 adults with diabetes), reports a steady decline:
"The prevalence of chronic kidney disease among individuals with diabetes decreased from 40.5% to 35.4% over 20 years (p=0.02)." — PMID 41659235 [single-source]
The second, covering 2001-2020 (n=4,200 adults with diabetes, representing an estimated 29.0 million persons), reports the decline reversing in the most recent years:
"The prevalence of any CKD decreased from 34.1% to 25.3% by 2009-2012, then increased to 30.6% by 2017-2020." — PMID 41240072 [single-source]
The two studies use different CKD definitions (eGFR<60 or ACR≥30 mg/g in one; a staged G3a/G4-5 breakdown in the other), different adjustment models, and different end years (2018 vs. 2020) — so this is not read here as one paper confirming the other, nor as one refuting the other. It is read as two single-source estimates of the same underlying trend that disagree on direction in the years each study's window does not share, which is itself the honest state of the evidence.
A national NHANES cohort (1999-2016 survey participants, mortality follow-up through December 2019, CKD defined as eGFR<60 or urinary albumin-creatinine ratio ≥30 mg/g) found overall age-standardized mortality among adults with CKD essentially unchanged across the study period, but with opposite trends by sex [2] [single-source]:
"Age-standardized mortality rate was 23.0 deaths/1,000 person-years (95% CI, 19.8-26.2) among adults with CKD surveyed 1999-2004 and 21.8 deaths/1,000 person-years (95% CI, 18.8-24.9) for those surveyed 2011-2016 (APC, -0.4% [95% CI, -1.4 to 0.5]). In contrast to men whose mortality declined (APC, -3.0 [95% CI, -3.5 to -2.6]), mortality rates increased among women with CKD (APC, 2.3 [95% CI, 1.7-3.0])." — PMID 41881385
The same study put the adjusted relative risk of CKD-associated mortality at "approximately 2-fold higher compared with adults without CKD across all time periods" [2] [single-source].
A separate NHANES cohort of 32,848 US adults surveyed 1999-2020 examined age- and sex-related disparities in cardiovascular-kidney-metabolic (CKM) syndrome staging — the CKD/cardiovascular/ metabolic overlap this domain's population sits inside [3] [single-source]. Its finding on convergence between older men and women:
"Elderly women were increasingly classified in stages with cardiovascular organ damage over the last 20 years, reaching the same prevalence as men in the most recent period (25.3% [95% CI, 20.0%-30.6%] of women vs 30.5% [95% CI, 25.7-35.3%] of men aged > 65 years)." — PMID 41659825
Read narrowly: this is CKM-stage classification, a broader construct than a CKD or heart failure diagnosis code, and the paper itself discloses a limitation of its own outcome measure: NHANES data "allow for assessing CKM stages with cardiovascular organ damage mainly based on self-reporting during interviews" [3] [single-source] — a limitation the study discloses about its own outcome measure.
The seed dossier's only hyperkalemia-prevalence figures come from a single Mayo Clinic health system (2.6%-4.8% among stage ≥3 CKD patients across three sites). A separate, broader estimate exists from a national US commercial claims database of 2,270,635 patients with serum potassium data from 2010-2014 [4] [single-source]:
"The annual prevalence of hyperkalemia in the overall population was 1.57% in 2014, with higher rates observed in patients with chronic kidney disease (CKD), heart failure, diabetes and hypertension. Among patients with CKD and/or heart failure, the 2014 annual prevalence was 6.35%. Among patients with hyperkalemia, 48.43% had CKD and/or heart failure in 2014 ... An estimated 3.7 million US adults had hyperkalemia in 2014, and this prevalence rate has increased since 2010." — PMID 29368958
This is a different operational definition again (≥2 potassium readings >5.0 mEq/L, or one ICD-9-CM 276.7 diagnosis code, or one sodium polystyrene sulfonate fill) from the Mayo cohort's ECG-paired ≥5.5 mEq/L threshold, so the two figures are not directly comparable rungs of the same ladder — they are two distinct, single-source estimates using two distinct case definitions on two distinct populations, both real and both worth carrying forward separately rather than averaged.
[1] National Trends in the Prevalence of Chronic Kidney Disease Among Racial/Ethnic and Socioeconomic Status Groups, 1988-2016 — JAMA Network Open (published 2020-07-01; accessed 2026-09-01). PMID 32672828 — https://pubmed.ncbi.nlm.nih.gov/32672828/ [peer-reviewed] [2] National All-Cause Mortality Trends Among Adults With and Without CKD in the United States, 1999-2019 — American Journal of Kidney Diseases (published 2026, month not given in the PubMed record; accessed 2026-09-01). PMID 41881385 — https://pubmed.ncbi.nlm.nih.gov/41881385/ [peer-reviewed] [3] Gender-related and Age-related Disparities in Prevalence of the Cardiovascular-Kidney-Metabolic Syndrome Among US Adults From 1999-2020: An Analysis of the NHANES Survey — Kidney Medicine (published 2026, month not given in the PubMed record; accessed 2026-09-01). PMID 41659825 — https://pubmed.ncbi.nlm.nih.gov/41659825/ [peer-reviewed] [4] The prevalence of hyperkalemia in the United States — Current Medical Research and Opinion (published 2018-06; accessed 2026-09-01). PMID 29368958 — https://pubmed.ncbi.nlm.nih.gov/29368958/ [peer-reviewed] [5] Trends in prevalence of chronic kidney disease and risk management among US adults with diabetes, 2001 to 2020 — Acta Diabetologica (published 2026, month not given in the PubMed record; accessed 2026-09-01). PMID 41240072 — https://pubmed.ncbi.nlm.nih.gov/41240072/ [peer-reviewed] [6] Decreasing Prevalence of Chronic Kidney Disease Among U.S. Adults With Diabetes, 1999-2018 — AJPM Focus (published 2026, month not given in the PubMed record; accessed 2026-09-01). PMID 41659235 — https://pubmed.ncbi.nlm.nih.gov/41659235/ [peer-reviewed]
Well established: nothing in this report rests on two independent, non-affiliated primary sources agreeing on the same quantity — see Thin, below.
Thin: every finding in section 3 rests on exactly one source and is marked [single-source]. Two of the six ([5] and [6]) examine the same diabetic-CKD subpopulation over overlapping NHANES windows and report the prevalence trend moving in different directions in the years their windows do not share — this is disclosed in section 3 as a discrepancy, not smoothed into a single number. [3]'s outcome measure (CKM syndrome staging with cardiovascular organ damage) is partly self-reported by the study's own admission. [4]'s and the dossier's Mayo cohort's hyperkalemia figures use different case definitions on different populations and are not comparable rungs of one ladder.
Rescoped from class 3: none. Every question this report answers is a registry or published quantity, not a question about what a person would do.
Out of scope: non-US findings, per the scope line above — a global CKD burden-and-forecast study (PMID 42662501, GBD 2023 data, 1990-2040 projection) was located during this pass and is named here rather than folded into a US-specific trend claim, since it reports worldwide rather than US figures. CPT descriptors and fda.gov guidance documents did not arise as relevant sources for this topic and are noted only for completeness with the scope boundary.
Not searched vs. not found:
[inference] The demographic subgroups shown here to carry disproportionate or rising CKD/CKM burden — Mexican American adults, lower-income and lower-education adults, and, in the CKM staging construct, elderly women — are not the same population this dossier's HF/CKD comorbidity and RAASi-utilization findings were drawn from (a hospitalized GWTG-HF cohort and a multinational CKD registry, respectively). No source retrieved in this pass or the seed dossier directly links these demographic concentrations to who is actually titrated on RAASi therapy or monitored for hyperkalemia in US outpatient practice; the overlap asserted here is the writer's own reading of two separate literatures, not a finding either source states.
| Proposition | Evidence class | Resolvable identifier | Dossier section |
|---|---|---|---|
| Age-, sex- and race-adjusted CKD stage 3-4 prevalence rose from 3.9% (1988-1994) to 5.2% (2003-2004) and held at 5.1% (2015-2016) in a 54,554-adult NHANES sample | 2 published | PMID 32672828 | 1 |
| CKD prevalence among Mexican American adults nearly doubled (2.1 percentage-point increase) between 2003-2004 and 2015-2016 in the same NHANES sample | 2 published | PMID 32672828 | 1 |
| Age-standardized all-cause mortality among US adults with CKD was approximately 2-fold higher than adults without CKD across NHANES survey periods 1999-2016, with mortality declining among men (APC -3.0%) and rising among women (APC +2.3%) | 2 published | PMID 41881385 | 1 |
| An estimated 3.7 million US adults had hyperkalemia in 2014, per a 2,270,635-patient commercial claims cohort, with 2014 annual prevalence at 6.35% among patients with CKD and/or heart failure | 2 published | PMID 29368958 | 1 |
| CKD prevalence among US adults with diabetes declined from 34.1% to 25.3% between 2001 and 2009-2012, then rose to 30.6% by 2017-2020, in a NHANES sample of 4,200 adults with diabetes | 2 published | PMID 41240072 | 1 |
| CKD prevalence among US adults with diabetes declined from 40.5% to 35.4% over 1999-2018 in a NHANES sample of 7,320 adults with diabetes | 2 published | PMID 41659235 | 1 |