The Race Nobody Can Win

In routine clinical chemistry, the speed war is over — and it ended in a draw. The throughput specifications that ten years ago separated a premium analyzer from a budget one have converged so closely that the difference no longer changes a clinical decision. But if you listen to the vendors, you will still hear them talk about tests per hour as if that were what is at stake. It is not. The real question has moved to a different floor of the laboratory.

I have spent enough years walking laboratories to recognise the moment a category becomes an engineering commodity. Routine clinical chemistry reached it. The precision envelopes, the linear ranges, the throughput curves that once drew hard lines between platforms have narrowed until, for nine out of ten routine panels, the results are clinically indistinguishable. A glucose, a liver enzyme set, a renal profile — run them on the incumbent flagship and on the new entrant, and the numbers land in the same clinical neighbourhood. The engineering is mature, the reagents are well-characterised, and the assays have been refined for so long that doing them badly now takes genuine effort.

Routine chemistry decided who ran fastest. Specialised immunoassay decides who runs in the right direction.

The floor where the race ended

This convergence is not an accident, and it is not a story about any one platform declining. It is what happens when a technology matures. In the earlier essay on the two design philosophies, I argued that the old East–West trade-off of quality for price has been replaced by a different bargain: evidence and standardisation for speed and flexibility. Routine chemistry is precisely where that new bargain has levelled the ground. The Asian design philosophy — optimise price, throughput, menu breadth, and flexibility — has done its work most completely here, because in routine chemistry those are the variables that decide the deal. A Western incumbent can still charge a premium, but it is increasingly a premium for assurance rather than for a measurably different result.

This is worth saying plainly, because it cuts against both the marketing and the pride of the industry: for the majority of routine chemistry, the analyzer is now good enough across the field. The question “which platform is more accurate for my basic metabolic panel” has largely been answered, and the answer is “all of them, within the noise that matters.”

The floor where the race still matters

Move one floor up — into the specialised immunoassay menu — and the picture inverts. Here the assays are younger, the standardisation thinner, the reference materials incomplete, and the clinical consequences of being wrong larger. A difference in precision, in lot-to-lot harmonisation, or in the depth and recency of the evidence base is not a footnote on a specification sheet. It is the difference between a result a clinician can act on and one that quietly misleads.

Consider thyroid function. A patient on levothyroxine is monitored for years against a fixed reference interval. When the TSH assay drifts — slowly, silently, lot over lot — the dose gets titrated against a moving baseline. The clinician does not see the drift; the patient feels it. That is not a throughput problem. It is a standardisation problem, and no speed specification on earth solves it.

Where the real evaluation begins

This is why the fourth reading in the vendor evaluation framework — evidence depth, for the assays that matter — is the one that separates a purchase you can defend from one you cannot. The principle was simple: demand deep evidence where a wrong number hurts a patient. A thick correlation dossier is unnecessary for a robust, well-characterised routine analyte; it is indispensable for the assay a clinician uses to manage a patient over years against a fixed reference interval.

The buyer who scrutinises the thyroid menu and accepts the routine chemistry on trust is making the rational bet. The buyer who does the reverse — auditing the glucose method and taking the endocrinology on faith — is buying the wrong floor.

A platform earns or loses its reputation in these niches. A laboratory earns or loses its clinical credibility there too. And yet most evaluations still spend the majority of their attention on the routine floor, where the answers have already converged, and treat the specialised menu as an appendix to be signed off in passing.

The niches worth walking

I have watched a laboratory discover, months into a contract, that its dengue panel did not cover the serotypes actually circulating in its region. The panel existed. The assay ran. But the epidemiology had moved and the menu had not moved with it. That is the nature of specialised immunoassay: the platform is not judged by what it can run today, but by whether what it runs still matches the clinical reality outside the window.

Or take fertility hormones — AMH, estradiol, progesterone — measured in concentrations where a shift between lots reshapes a treatment cycle. The decision those numbers inform is not reversible. A pregnancy deferred or an embryo transfer mis-timed is not a data point you correct on the next run. The imprecision that would be noise in a routine assay is, in that room, a clinical event.

Infectious disease panels age the same way. The epidemiology moves, the resistance patterns shift, and a menu that was defensible at installation can be quietly obsolete by year three of a commodato. The question is not whether the platform could run the assay, but whether the assay still answers the question the clinician is actually asking.

These are the floors where a platform proves itself. They are also where a careful buyer should begin — not end — an evaluation. Walk them before you sign, and you will know more about a platform than any throughput specification can tell you. The race on the ground floor is finished. The one that matters is still being run, one assay at a time, in the rooms where a result meets a patient.