The Biological Passport Catches Doping Without Finding a Drug

Conventional testing had a losing structure

The traditional anti-doping test asks: is a prohibited substance present in this sample?

That question has a structural weakness the sport spent decades losing to. Every substance has a detection window, and the window is usually short. Time your use correctly and the test finds nothing — not because the test failed, but because there was nothing there to find.

Worse, the test can only look for substances on a list. A compound not yet known cannot be searched for. See how doping bans work for what the list-based system does once something is found.

The passport asks a different question entirely: has this athlete’s body changed in a way that doping would explain?

Your own values are the reference range

The critical move is that the comparison is to yourself.

Population normal ranges are wide, because people differ enormously. A haemoglobin value that would be extraordinary in one athlete is unremarkable in another, so a population threshold catches almost nobody without also catching innocent outliers.

An individual range is far tighter. Your own haemoglobin, measured repeatedly over years, varies within a narrow band — and a value outside your band is meaningful even if it sits comfortably inside the population’s.

That is why it is called a passport. It is a document about one person, accumulated over time, and it becomes more powerful the longer it runs.

The two working modules

The haematological module, introduced in 2009, detects blood manipulation — both transfusions and erythropoietic stimulating agents.

It tracks haemoglobin, haematocrit, red cell count, reticulocyte number and percentage, mean corpuscular values, and the OFF-score, a composite designed to expose the specific signature of blood doping.

The OFF-score deserves a note. Boost red cell production artificially and the body responds by suppressing its own — so reticulocytes, the immature red cells, fall while haemoglobin stays high. That combination is physiologically odd and is what the score is built to detect. It is a fingerprint of the response, not of the drug.

The steroidal module, added in 2014, tracks urinary testosterone, epitestosterone, androsterone, etiocholanolone and related metabolites, together with the urine’s specific gravity.

Administered testosterone is chemically identical to the body’s own, so a straight test for its presence proves nothing. What changes is the ratio between related compounds — and ratios, tracked over time against your own baseline, are detectable where the substance is not.

An endocrinological module has been in development.

The adaptive model does the judging

Values are assessed by a Bayesian adaptive model.

Bayesian is the right word rather than jargon. The model starts with a population prior — what an athlete’s values are likely to be, knowing nothing about them — and updates it with each sample, so the expected range narrows toward that individual.

It also accounts for confounders with proven or potential influence: altitude exposure, among others. An athlete returning from a training camp at 2,500m has genuinely elevated blood values, and a model that did not know this would flag every distance runner in the sport.

So the output is not “this value is high”. It is: given everything known about this athlete and their circumstances, how improbable is this reading? A profile is referred to expert review only when the improbability crosses a defined threshold.

Why it changed the enforcement problem

Three consequences follow, and they are large.

Timing stops protecting you. A doper who avoids every test still leaves the physiological trace in whatever samples are collected. The window they were exploiting was the substance’s, and the passport is not looking for the substance.

Unknown drugs are covered. A compound nobody has identified still produces effects, and effects distort the profile. The passport searches for the consequence rather than the cause.

The evidence is a pattern, not an event. That makes a passport case closer to a world record ratification than to a police finding — a conclusion assembled from measurements and procedure rather than from a single object.

That is also its weakness in a hearing. A passport case argues from statistical improbability rather than from a substance in a bottle, which is harder to present and easier to contest. Such cases are fought on expert evidence about confounders — illness, altitude, hydration, pregnancy — and they take a long time.

And it is why whereabouts exists

The passport’s requirement is unforgiving: it needs a long series of samples, collected unannounced, over years.

A profile built from samples the athlete could anticipate would be a profile of their prepared state, which is worthless.

So the most scientifically elegant tool in anti-doping depends entirely on the most intrusive administrative one — the whereabouts rule, an hour a day, every day, for a career.

The two are frequently discussed as separate things. They are the same programme: the passport is what the samples are for, and whereabouts is how the samples are obtained.