Lucy Letby: We spent years covering the case – here’s why experts are still arguing about it

The medical condition of the babies also fitted with the lab results. In both cases, the babies’ blood sugar levels had plummeted, which is what you would expect to see with insulin poisoning. And while no-one saw Letby poisoning either of the two babies, she was there when they started experiencing symptoms.

Of all the allegations in the case, this one looked like the most solid. In court, Letby herself accepted the scientific evidence that the babies had been given dangerous quantities of insulin. She just denied being responsible. Her lawyers were more cautious. They did not accept the insulin evidence, but they did not say it was incorrect either.

For the prosecution, the insulin evidence was fundamental to the entire case. It seemed to prove that someone on the neonatal unit at the Countess of Chester Hospital was a poisoner.

If jurors could be persuaded on this point, it wouldn’t be so difficult for them to conclude that Letby was the culprit.

And so they did. Of all the allegations in the case, the jury returned unanimous verdicts of guilt on just three – and two of these were the insulin cases.

However, since then, sceptics have questioned whether the lab test used to measure insulin and C-peptide in the Letby case was as robust as the prosecution had claimed. It is called an immunoassay test, and works by using antibodies to detect and measure substances.

Critics argue there are circumstances in which the test can mistake another substance for insulin. It is called interference and it could result in a false positive. The critics say the only way to be sure that the substance being measured is indeed insulin is to use a more precise method of analysis – such as mass spectrometry.

We spent months examining this argument. Our conclusion, having spoken to leading experts on all sides of this debate, is that, while the immunoassay method is not perfect, it is usually accurate and the circumstances in which interference might occur are extremely unlikely in the context of the babies in the Letby case.

It is even more unlikely that two lab tests conducted within months of each other would both be wrong.

In this week’s Panorama, we reveal new evidence on the insulin allegations and the question of whether Lucy Letby really poisoned babies.