PCR is automated. What surrounds it, far less so.

In a PCR department, the instrument is rarely the weak point. Extraction systems, pipetting robots and thermocyclers are precise and well controlled. What happens around them is far less so: tube reception, the plate layout, the transcription of controls, the validation of the run.

Studies in laboratory medicine have been saying the same thing for years. The latest figures are clear.

Errors happen before and after analysis

A team at a US core laboratory recorded every error in its activity between January 2022 and May 2023: around 11 million samples and 37.7 million results. It counted 87,317 errors. Most were due to haemolysis, meaning red blood cells damaged during sample collection. Setting haemolysis aside, 94.6% of the remaining errors occurred before analysis, 1.7% during and 3.7% after (Lin et al., Clinical Chemistry and Laboratory Medicine, 2025).

The work of Mario Plebani at Padua University Hospital describes the same distribution: the analytical phase has become the safest, and the phases around it remain the most exposed (Plebani, EJIFCC, 2015).

These figures come from routine laboratory testing. They do not carry over directly to PCR. But molecular biology reaches the same conclusion.

In molecular biology too

A review published in Clinical Biochemistry on the molecular diagnosis of infections says so in its abstract: “The majority of laboratory errors occur in the preanalytical phase.” The specimen type, the transport medium, the duration and temperature of storage, a contamination: each of these can produce a false negative or a false positive. The authors draw a practical conclusion: each laboratory must validate its own procedures, because the manufacturer’s recommendations and the adaptations made on site do not always match (Conrad et al., Clinical Biochemistry, 2023).

Four US molecular biology laboratories described two real cases (Qin et al., Current Issues in Molecular Biology, 2022). In the first, two discordant results raised fears of a sample mix-up. It turned out to be a misread abbreviation. In the second, an extraction reagent contained DNA from a strongly positive sample. It produced false-positive results in patients in remission. The authors sum it up: a sample identification error is a serious medical error, and “it is difficult to realize such errors during laboratory bench work”.

What is hard to see at the bench must be traceable afterwards. That means knowing, for every tube, where it has been.

Why these steps resist

A specialty laboratory department is a unit that carries out complex analyses, often in batches. Each assay needs several processing steps, sophisticated instruments and its own controls. Its results are read in the context of the whole run, not as isolated measurements.

The instruments in these specialty laboratories keep getting more powerful. The processes around them remain manual and fragmented. Sample preparation, batch management, quality control, interpretation, traceability and documentation rely on systems with no link between them: each instrument’s software, a spreadsheet, a form, a printed sheet. Every switch from one medium to another is a chance to lose information.

The profession is aware of this. In its strategy published at the end of 2025, the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) makes structured data one of its four priorities: “Promote the standardization and structured formatting of laboratory data to enhance interoperability.” (Plebani, Zima, Gruson, Clinical Chemistry and Laboratory Medicine)

Qualified staff are increasingly scarce

These steps rely on highly qualified people. In Germany, medical laboratories have been sounding the alarm for several years. In 2022, the association of accredited laboratories ALM summed up its position in one sentence: “Ohne Fachkräfte in den Laboren keine angemessene medizinische Versorgung.” Without qualified staff, no proper medical care (Deutsches Ärzteblatt, July 2022). A few months later, its chair saw “ein massives Risiko für die Versorgung mit Labordiagnostik insgesamt” if the training of laboratory technicians did not keep up (Deutsches Ärzteblatt, December 2022).

Every hour spent copying a result, looking for a plate layout or piecing together the history of a control is an hour less for the tasks where these people’s experience is really needed.

What to take away

The gain in quality now lies in the organisation around the instrument: identifying every tube at every step, keeping the plate layout, the controls and the raw data in one place, validating a run in its context. That is where digitalising a specialty laboratory has the most to offer.

The Mobiolink App follows every tube, from its arrival in the department to the export of validated results to the LIS. See the seven steps.