Pillar 02
Diagnstics.
Diagnostics fail quietly. An instrument drifts, a buffer ages — and the number on the screen still looks perfectly plausible.
The discipline
Both halves of the measurement
A diagnostic result is the product of two things: the instrument that reads the signal, and the chemistry that produces it. Own only one and you inherit the other's assumptions without being able to test them.
Buffers, in our view, are not a consumable afterthought. They set the pH, the ionic strength and the stability window inside which an assay is either reliable or merely repeatable — and the difference between those two words is the whole of diagnostics. A buffer formulated for the assay it serves behaves differently from one adapted to it, particularly in month eighteen.
Designing the instrument and the chemistry alongside each other is how the two stay in agreement across a product's life. It also means that when a result drifts, there is somewhere definite to look.
Capability
What we do
We take on instrumentation work, chemistry work, or the matching of the two. Where a system already exists and is misbehaving, the first deliverable is usually a clear account of why.
- Diagnostic instrumentation — specification, development and evaluation of measurement hardware.
- Buffer and reagent systems — formulated for the assay they serve, not adapted to it.
- Assay support chemistry — stabilisers, diluents and controls developed as a matched set.
- Method validation — precision, linearity, robustness and the limits stated honestly.
- Instrument–reagent matching — the two designed together, so that calibration holds.
- Lifecycle and stability — how the system behaves in month eighteen, not just on day one.
- Failure investigation — drift, bias and irreproducibility traced to a cause rather than a symptom.
Why it matters
A plausible wrong number is the expensive one
Detectable failure
Systems designed so that when something goes wrong it announces itself — through controls that fail visibly rather than degrade quietly.
Stated limits
Every method carries an honest account of where it stops working. A validation that reports only successes has not been a validation.
Agreement over time
Instrument and reagent developed as one system, so calibration holds across batches, sites and shelf-life rather than only at release.
A diagnostic that will not hold calibration?
Describe the system and what it is doing. We are glad to put confidentiality in place before the detail begins.