For diagnostics makers and EQA schemes
A neutral count of what a conversion does to patients.
When a laboratory moves to a new analyser or reagent, its director asks one question before signing off: what happens to our patients at our limits? We answer it the same way whoever asks, so a conversion can be planned rather than argued about.
01Analyser and reagent makers
For manufacturers
OfferOffered to makers. No manufacturer pilot has run yet, and no maker has agreed to one.
A conversion pilot puts the patient-impact count into a small number of the maker’s own conversions, chosen by the maker, over a few months. We produce the count the same way whoever pays, and the laboratory keeps its data. The count describes one laboratory’s own change; we never rank one maker’s system against another’s.
What a pilot delivers
For each conversion, a report to the laboratory: the bias at each of its own decision limits with its interval, how many patient results cross each limit against what repeat testing alone would move, a re-test band for the weeks both analysers run, and a note the laboratory can send to requesting clinicians.Who runs it
The laboratory. The program runs on the laboratory’s own computer, on the paired results it already produced for its method comparison. A laboratory that prefers can instead share coded paired results with us under its own written agreement with us.What the laboratory sees
Its full report, delivered to it directly, before or with any copy to anyone else. The laboratory owns its data and its report, and the report says on its face who paid for it and that the payer had no role in its method or findings.What the maker sees
Aggregates only: a copy of a laboratory’s report where that laboratory agrees in writing, and a summary pooled across the pilot’s laboratories that never names one. No patient row and no per-sample file reaches the maker.Neutrality
The method, the decision limits and the wording are the same whoever pays, and the maker cannot review, edit, delay or withhold a report. The same report is available to any laboratory, maker or channel, and no maker gets exclusivity. A change that moves nothing is reported as such: see a public example.
02EQA and proficiency testing
For EQA providers
OfferProposed to schemes; not yet run with any scheme.
We propose research cooperations the scheme runs on its own systems: how far results move when a participant declares a new analyser, reagent or software version, with the code handed over and only tables the scheme approves leaving it.
OfferTested so far on public instrument files, not on any scheme’s data.
We train a detector on your scheme’s own history to flag measuring-system changes participants did not declare. The per-laboratory flags stay with you. We propose it as a research cooperation, with the method published. Its first test on raw instrument files failed and its sealed re-test passed; both are on the changelog.
How a cooperation works
The scheme chooses one of two shapes. In the first, the scheme runs the study on its own systems with code we hand over, and only aggregate tables it approves leave. In the second, used only where the scheme’s own terms and its participants allow, the scheme shares an anonymised extract under a written research agreement. In both, participants’ results stay theirs and the scheme’s, and no laboratory is named.- Designed
The Registry route
A scheme may choose to pool the tables it releases in the Analyser-Change Registry: aggregates by analyte, measuring system and period, never anything that identifies a participant, and no scheme named without its consent. In return it sees what the same systems and changes did across every contributor. The Registry is designed, has not started, and no laboratory or scheme has joined yet.
03Lots
Reagent and calibrator lot changes
OfferDescribed, not a track record.
The same method applies to a new reagent or calibrator lot: shorter windows, the same question. How many results moved across a limit, against what repeat testing alone would move.
Guide: Reagent lot changes
04Evidence
What a managed switch looks like
- Sample
Sample analyser-change report, CD4
One platform replaced by another, on public data, sized and then managed with a re-test band. It shows three of the seven report parts in full and three in part; the other one needs more than one site. The sample page lists each part and what is not shown, with the reason. Reported without adjustment, 535 of 1,885 results change CD4 category, against about 238 expected from repeat testing alone. Nobody commissioned it; it is there so you can see what you would receive.
535results changed category
- Public data
Analyser change explorer
34 method comparisons at 14 stand-in sites, from 12 open-access studies. Of the 34 method comparisons, 21 get a full re-test band and 5 an indicative one, because they have fewer pairs than the minimum we set; the other 8 are shown as a comparison only or are too small for a band. In 6 of the 34 comparisons, the change moved fewer results across a limit than repeat testing alone would. All five checks we fixed before the run passed. The checks are ours, not an outside review.
10,579paired results
05Evidence
When there is nothing to manage
The same count also says when a change does not matter. In this public comparison one analyser switched creatinine method; the change moved fewer results across the decision limits than repeat testing alone would. The report says so, and the laboratory has nothing extra to manage.
Creatinine at 1.2, 1.5 mg/dL
Crossed after the change: 23
Expected from repeat testing alone: about 41
Fig. 1. Central lab analyser, enzymatic creatinine to same analyser, kinetic Jaffe creatinine, 529 paired results. 23 results changed category at the decision limits, against about 41 expected from repeat testing alone.
Source Schmidt RL, Straseski JA, Raphael KL, Adams AH, Lehman CM (2015). A Risk Assessment of the Jaffe vs Enzymatic Method for Creatinine Measurement in an Outpatient Population. PLoS One. doi:10.1371/journal.pone.0143205 (opens in a new tab); PMC4657986 (opens in a new tab). Licence: CC BY 4.0 (opens in a new tab). Changes: Paired results re-analysed by NakedSignal's program; instruments described generically; only derived results are shown. Repeat-test imprecision is assumed, not measured. Open this comparison in the explorer.
Show the numbers
| Measure | Value |
|---|---|
| Paired results | 529 |
| Decision limits (mg/dL) | 1.2, 1.5 |
| Results that changed category after the change | 23 |
| Expected to change from repeat testing alone | 40.83 |
| Mean difference, new minus old, as % of old | 0.34 |
Planning conversions, or holding scheme data?
Tell us what is changing and where. We will say what the count would show and what it needs.