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 reference-products · Version 2026-08-31 · Cadence quarterly · Evidence class mixed · Sources 9 · 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, and any non-US or non-EU jurisdiction.
Sourcing: three peer-reviewed studies and six FDA registry records; the abandonment evidence rests on one large paired study plus one open device recall, which agree
Potassium and creatinine monitoring during RAASi titration is served almost entirely by venous draw onto a central-lab integrated chemistry analyzer, with a small set of cleared point-of-care analyzers at the edges. Capillary sampling is the recurring shortcut, and it is the recurring failure: the largest published paired evaluation found capillary and venous results clinically comparable for 16 of 18 routine analytes but singled out potassium as one of the two that were not [1], and FDA's device recall database records an open, classified recall of a capillary tube kit for a systematic positive bias across the reportable potassium range [4]. Both point at the same mechanism — in-vitro hemolysis raising measured potassium — and that mechanism is also what governs whether a specimen is reportable at all: rejection rates swing from under 1% to 76% depending purely on where a laboratory sets its hemolysis-index cutoff [2]. The market is not short of analyzers. It is short of a sampling route that survives the pre-analytical step.
_Baseline (v1). No prior version; this establishes the starting point for future diffs._
The reference product in this domain is not a device anyone markets to a nephrology clinic. It is the venous draw plus a high-throughput integrated chemistry platform, and its potassium channel is cleared under product code CEM. Recent clearances in that code are analyzer and electrode submissions rather than new sampling formats — Roche's cobas pro indirect-ISE potassium (K253491, decided 2026-02-12), Radiometer's ABL90 FLEX PLUS (K241037, decided 2025-01-14) [8], and Medica's EasyStat 300 (K220396, decided 2023-09-26). The competitive motion in this code is throughput and electrode chemistry, not the specimen.
The creatinine side is a separate generic type (product code CGL) with a much thinner clearance history: over the last ten years openFDA records exactly two, Abbott's i-STAT Crea cartridge with the i-STAT 1 System (K261039, decided 2026-06-26) [7] and the Nova StatSensor Creatinine Hospital Meter System (K171059, decided 2017-05-26) [6] [single-source]. Both are handheld or cartridge formats already used at the point of care. Neither carries a potassium channel in the same clearance, which is the structural point: a two-analyte fingerstick device is arguing two predicates in two generic types, not extending one.
The largest published evaluation of venous–capillary interchangeability is a single-centre paired study of 480 adults yielding 960 capillary specimens across fingertip lancet and upper-arm microneedle collection and three matrices, run against matched venous samples on a Cobas Pro platform under CLSI EP09 and EP35 [1] [single-source]. Its finding is specific and it is not favourable to this domain:
"Capillary and venous results were clinically comparable for 16/18 analytes when specified HIL cutoffs were applied, although potassium and chloride showed limited interchangeability." — PMID 42446434 [1]
The same study reports that fingertip collections yielded larger volumes and fewer collection failures than the microneedle device but showed higher hemolysis [1] — which locates the failure in the pre-analytical step rather than in the measurement.
Whether a potassium result is reportable at all is a threshold a laboratory sets, and the threshold moves rejection rates by more than an order of magnitude. In a single-institution analysis of hemolytic indices for 80,795 specimens tested on Cobas chemistry analyzers in one month, applying the manufacturer's revised reporting guidance would have produced specimen rejection rates of 76% in the NICU, 41% in the emergency department, 16% for inpatients and 9% for outpatients; the institution's own criteria held those to 3%, 2%, 2% and under 1% respectively [2] [single-source]. The authors' conclusion is directly relevant to any device that proposes to suppress a result rather than report it:
"The new recommendations would lead to unacceptably high specimen rejection rates." — PMID 35038719 [2]
That trade-off is not only statistical. A published case report describes a hemolysed paediatric specimen whose suppressed potassium of 3.9 mmol/L was followed by 2.1 mmol/L on redraw, and argues that withholding the result delayed treatment for hypokalaemia [3] [single-source]. Suppression has a cost on both sides, and the cost is asymmetric by care setting.
Two independent records show capillary potassium failing in the field rather than in a study.
FDA's device recall database carries an open, classified recall of Medica Corporation's capillary tube kit for use with the EasyBloodGas and EasyStat systems, initiated 2025-12-31, whose stated reason is:
"A systematic positive bias of up to 20% was identified across the reportable potassium range, resulting in false elevation results of patient samples when using Medica's capillary tubes." — Recall Z-1370-2026 [4]
Separately, Inpeco's FlexLab automation carries an open, classified recall initiated 2025-09-03 because the Sample Integrity Module's hemolysis–icterus–lipemia function "has not been evaluated for safety and effectiveness" for that use [5] [single-source]. The two records point at opposite ends of the same problem: one is a capillary specimen biasing potassium high, the other is the automated check meant to catch exactly that being un-evaluated.
On the clinical side, the incumbent monitoring regime's own failure is measurable. A retrospective cohort of 905 outpatients on RAASi therapy, mean follow-up 29.8 months, found hyperkalaemia (K+ ≥5.1 mmol/L) in 32.8% (95% CI 30–36%), with RAASi discontinued in 6.2% of patients and down-titrated in a further 4.5% because of it [9] [single-source]. That is the revealed-behaviour twin of "would clinicians value easier monitoring": a measured rate at which the drug is actually stopped.
[1] Venous-capillary blood agreement for routine chemistry analytes: effects of collection site and sample matrix — Clinical Chemistry and Laboratory Medicine (published 2026, month not given in the PubMed record; accessed 2026-08-31). PMID 42446434 — https://pubmed.ncbi.nlm.nih.gov/42446434/ [peer-reviewed] [2] Hemolysis Index and Potassium Reporting — American Journal of Clinical Pathology (published 2022, month not given in the PubMed record; accessed 2026-08-31). PMID 35038719 — https://pubmed.ncbi.nlm.nih.gov/35038719/ [peer-reviewed] [3] Rejection of hemolyzed samples can jeopardize patient safety — EJIFCC (published 2020, month not given in the PubMed record; accessed 2026-08-31). PMID 32256285 — https://pubmed.ncbi.nlm.nih.gov/32256285/ [peer-reviewed] [4] Device recall: Medica Capillary Tubes; EasyBloodGas/EasyStat Capillary Tube Kit — FDA (recall initiated 2025-12-31; accessed 2026-08-31). Z-1370-2026 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRES/res.cfm [federal-registry] [5] Device recall: Inpeco FlexLab (FLX) System, Potassium Test System — FDA (recall initiated 2025-09-03; accessed 2026-08-31). Z-0106-2026 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRES/res.cfm [federal-registry] [6] 510(k) Nova StatSensor Creatinine Hospital Meter System — FDA (decision 2017-05-26; accessed 2026-08-31). K171059 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K171059 [federal-registry] [7] 510(k) i-STAT Crea cartridge with the i-STAT 1 System — FDA (decision 2026-06-26; accessed 2026-08-31). K261039 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K261039 [federal-registry] [8] 510(k) ABL90 FLEX PLUS System — FDA (decision 2025-01-14; accessed 2026-08-31). K241037 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K241037 [federal-registry] [9] Real-world insights into hyperkalemia burden and RAASi discontinuation: a cohort study — Current Medical Research and Opinion (published 2025, month not given in the PubMed record; accessed 2026-08-31). PMID 41126484 — https://pubmed.ncbi.nlm.nih.gov/41126484/ [peer-reviewed]
Well established: that capillary sampling for potassium fails on specimen integrity rather than on measurement rests on two independent primary sources that agree — a 480-adult paired study [1] and an open FDA recall of a capillary tube kit for positive potassium bias [4]. Neither is affiliated with the other.
Thin: every quantity in this report rests on one source and is marked [single-source]. The rejection-rate figures [2] are one institution's specimens on one manufacturer's platform; the discontinuation rates [9] are one 905-patient cohort at a single centre in Jeddah, which is not a US population and should not be read as one. The clearance counts for product codes CEM and CGL are openFDA query results over a ten-year window and are floors, not totals — the CEM recall query alone reported 37 total matches against a limit of 8.
Rescoped from class 3: "would clinics adopt a fingerstick potassium draw" was not researched, because nothing published answers it. Its revealed-behaviour twin — the rate at which the incumbent regime causes RAASi itself to be stopped or down-titrated — was researched instead and appears above [9]. [inference] The substitution is the writer's own and the two questions are not equivalent: a measured discontinuation rate says the current burden has a cost, not that a new device would be taken up.
Out of scope: CPT descriptors, which are AMA-licensed and are not reproduced anywhere in this repo; FDA guidance documents, which gate automated requests intermittently so that a 401 page parsed as content would read as "nothing found"; and every jurisdiction beyond the US.
Not searched vs. not found:
[inference] The report treats "the reference product" as the venous draw plus a central-lab analyzer rather than as any marketed device. No source states that framing; it follows from the CEM clearance history being analyzer submissions rather than sampling formats.
Propositions a checker could attempt. This report does not append them — /research drafts them onto a candidate with engine/scripts/append_research_claims.py, which constructs the unverified tag itself.
| Proposition | Evidence class | Resolvable identifier | Dossier section |
|---|---|---|---|
| Potassium and chloride showed limited interchangeability between capillary and venous specimens in a 480-adult paired study of 960 capillary specimens run under CLSI EP09/EP35 | 2 published | PMID 42446434 | 5 |
| Applying the manufacturer's revised hemolysis-index reporting guidance to 80,795 specimens would have rejected 76% of NICU and 41% of emergency-department potassium specimens at one institution | 2 published | PMID 35038719 | 5 |
| FDA recorded an open Class-classified recall of Medica capillary tubes for a systematic positive bias of up to 20% across the reportable potassium range | 1 registry | Z-1370-2026 | 5 |
| The Nova StatSensor Creatinine Hospital Meter System was cleared under product code CGL with a decision date of 2017-05-26 | 1 registry | K171059 | 5 |
| The i-STAT Crea cartridge with the i-STAT 1 System was cleared under product code CGL with a decision date of 2026-06-26 | 1 registry | K261039 | 5 |
| RAASi therapy was discontinued in 6.2% and down-titrated in 4.5% of a 905-patient outpatient cohort because of hyperkalaemia over a mean 29.8 months | 2 published | PMID 41126484 | 5 |