What the Vienna and Maribor Consensus Actually Ask of Your Lab

Most IVF laboratories can name the Vienna consensus. Far fewer can say where their own laboratory sits against it this month.

That gap is not negligence. The consensus documents were written for an expert audience, they run to dozens of pages of definitions, and they were never designed as an operational dashboard. The result is a framework almost everyone cites and comparatively few use as intended.

This is what the indicators actually ask of a laboratory, and how to read your own position against them.

Two consensuses, two scopes

Clinics frequently say “Vienna” and mean both documents. They are separate, and the distinction matters because monitoring one leaves half the process unmeasured.

Vienna 2017 was an expert meeting supported by ESHRE and Alpha Scientists in Reproductive Medicine, covering performance indicators for the ART laboratory (Human Reproduction Open). It defined 19 indicators in total — 12 key performance indicators, five performance indicators and two reference indicators — informed by surveys of laboratory directors and clinical embryologists across 18 countries.

Maribor followed. The expert meeting was held in November 2019 and published in 2021, and it covers performance indicators for clinical practice in ART (Human Reproduction Open). It was written explicitly as a complement to the 2017 laboratory indicators, with statements accepted where at least 70% of responding ESHRE members agreed.

So the pairing is deliberate: one document measures what happens in the laboratory, the other what happens clinically. A laboratory performing at benchmark inside a clinic with weak clinical indicators still produces poor outcomes, and the reverse is equally true. That is why they are cited together.

Competency is not benchmark, and the difference is the whole point

This is the distinction most often collapsed, usually into a single vague idea of “the standard”.

Competency is the minimum standard required to achieve proficiency and maintain ongoing performance. It is a floor. Any laboratory operating properly should clear it.

Benchmark is an externally referenced result expected under optimal conditions. It is aspirational, and it describes what a laboratory can reach when everything is favourable — including its case mix.

Three consequences follow, and they are worth stating plainly because getting them wrong distorts how a laboratory reads its own numbers:

• Sitting between competency and benchmark is normal. It is not underperformance. A laboratory permanently at benchmark across every indicator should ask whether its patient population is unusually favourable.
• Sitting below competency is a signal, not a verdict. One result below the floor is a prompt to look. A sustained trend below it is a finding.
• Each indicator carries its own pair of values. There is no single pass mark across the framework. Published examples — a competency of 60% against a benchmark of 75%, for instance — belong to specific indicators, and applying one indicator’s thresholds to another is a straightforward way to reach a wrong conclusion about your own laboratory.

That last point is worth guarding against, because summarised versions of the consensus circulate widely and often flatten the values into a single number.

Not all 19 indicators do the same job

The framework separates its indicators deliberately, and the categories are not decoration.

Key performance indicators are the ones deemed essential — for evaluating the introduction of a technique or process, establishing minimum proficiency, and monitoring ongoing performance within a quality management system.

Performance indicators sit alongside them, useful for monitoring but not carrying the same weight in judging whether a process is under control.

Reference indicators describe context rather than performance. They help interpret the others.

The practical implication is that not every indicator deserves equal attention on a Monday morning. A laboratory trying to watch all 19 with the same intensity will watch none of them well. The 12 KPIs are where a review meeting should start, with the reference indicators used to explain what the KPIs are showing rather than treated as targets in their own right.

What below-competency actually obliges you to do

Here is where the framework connects to something larger, and where most clinics under-read it.

Both documents frame performance indicators as an element of the quality management system — not as a scorecard, and not as a marketing number. Maribor is explicit that the competence levels are intended for use within each ART centre’s QMS.

That framing carries an obligation. Inside a quality system, a result below competency initiates a defined sequence: establish what occurred, analyse the possible causes, identify whether a system deficiency is responsible, act, and then verify the action worked. It is not satisfied by an explanation at a meeting.

Which is why the honest question for a laboratory is not do we know our KPIs but:

1. Would we notice a KPI crossing below competency, and how quickly?
2. Is there a defined path from noticing to investigating, or does it depend on who happens to be in the room?
3. Can we show, afterwards, what we did about it?

A laboratory that can answer all three is using the framework as designed. One that computes the numbers annually for an accreditation file is reporting, not monitoring.

Why quarterly measurement defeats the purpose

The indicators were built for monitoring, and monitoring has a tempo.

An indicator calculated at the end of a quarter describes cycles that concluded weeks ago. If fertilisation rate drifted below competency in the first month, the laboratory learns about it after two further months of cycles have run on the same conditions. The number is accurate and useless — a post-mortem rather than a control.

There is a second, quieter problem. A figure assembled retrospectively is assembled by someone, from records that were not designed to be aggregated, usually under time pressure before a meeting. That process introduces its own errors, and it makes the indicator expensive enough that nobody wants to compute it more often — which entrenches the tempo that caused the problem.

The fix is not a faster spreadsheet. It is that the underlying clinical and laboratory events have to be captured as structured data when they happen, so an indicator is read rather than reconstructed. That is a data-model question, and it is the same one behind what your EMR should and shouldn’t be tracking in the IVF lab: information recorded as narrative cannot be aggregated, no matter how diligent the person doing the aggregating.

Where the events are already structured, the computation stops being a project. AI & Analytics is where that sits in MedART.

Where to start

You do not need a quality programme to find out where you stand. Take last quarter’s cycles and attempt three things:

Pick three KPIs and locate your own value. Not the national average, not last year’s audit figure — your laboratory, last quarter. If producing it takes more than an afternoon, that difficulty is itself the finding.

Check them against the consensus values for those specific indicators. Each one individually, rather than against a remembered single threshold.

For anything below competency, ask what happened next. If the answer is that nobody noticed until you looked just now, the gap is in monitoring rather than in laboratory performance.

Most laboratories doing this exercise discover their performance is broadly fine and their visibility is not. That is a considerably easier problem to fix, and it is the one the consensus was written to address.

For the wider reporting picture, the IVF success-rate analytics dashboard covers outcome reporting, and IVF lab management software covers the evaluation criteria around laboratory systems generally.