Here.
“A staffing rota also showed she had been on duty for every suspicious death or collapse between June 2015 and June 2016. The rota was a key part of the case – a striking visual symbol of the case against her. But a number of statisticians have publicly questioned its usefulness.
One is Peter Green, a professor of statistics and a former President of the Royal Statistical Society. "The chart appears to be very convincing, but there are a number of issues with it," he said. "A big thing is that it only describes 25 of the bad events which happened in this period. It doesn’t include any of the events that happened when Lucy was not on duty."
There were at least six other deaths and numerous collapses.
Prof Green said the chart also does not reflect the fact that Letby was working extra shifts.
"It’s a natural human thing. We all see patterns that are not there," he said.
"The danger is that this evidence can be very compelling to the non-professional, and over interpreted."
Another crucial part of the prosecution's case were blood samples from babies who had collapsed with low blood sugar.
They showed exceptionally high levels of insulin and low levels of a substance called C-peptide.
That combination is only generally seen when the body takes in synthetic insulin, leading to the charge that Letby had deliberately poisoned the babies by adding it to their nursery feed bags.
An expert in paediatric diabetes told the nurse's trial fluctuations in the readings were unusual.
Prof Alan Wayne Jones, an expert in forensic toxicology, is one of those who has challenged the results.
He pointed out that the test used measures the body’s reaction to insulin rather than the substance itself.
"The problem is that the method of analysis used [in these two cases] was probably perfectly good from a clinical point of view, but not a forensic toxicology point of view," he said.
"That test cannot differentiate between synthetic insulin and insulin produced by the pancreas."