The Insulin Nurse Six

In the UK, over the last half century, six nurses have been convicted of intentionally poisoning those in their care with insulin. It has become increasingly clear in recent years that the prosecution proof of the presence of exogenous insulin in each case is not nearly as strong as the respective juries were told it was. Other evidence is also weak in each case. None of the convictions of the six nurses have been overturned. Five of the nurses remain in prison.

by Peter Elston

Introduction

Insulin is a hormone produced naturally in the body to lower blood sugar.

Synthetic insulin is a lifesaver for those whose bodies do not make enough natural insulin or have a resistance to it.

Insulin can kill because too much of it reduces "the amount of glucose in the blood to a level where the brain is unable to function properly".

Insulin overdose can be undiagnosed if its effects (hypoglycaemia, coma, death) are attributed to natural medical issues (sepsis, organ failure, tumours, etc).

Hypoglycaemia can also result if glucose is not being delivered because a line is tissued (a line is in the tissue not a vein so glucose is not reaching the blood stream).

Thus, insulin overdoses can be undetected but, also, a fall in blood sugar can be wrongly attributed to an insulin overdose.

If a fall in blood sugar is accompanied by an insulin test result that is believed, wrongly, to be reliable, a miscarriage of justice can result. It is likely that this is what happened to all six nurses.

Detecting insulin

Nowadays, accurate detection of insulin requires an expensive test known as LC-MS/MS (liquid chromatography followed by tandem mass spectrometry).

A precipitous fall in blood sugar alone is very weak evidence of exogenous insulin as there are natural processes that could cause it.

A positive LC-MS/MS test result is very strong (assuming sample is reliable, storage/transport protocols have been followed, etc).

In between are tests like immunoassays and the polyethylene glycol (PEG) test. Positive tests from either might in a clinical setting raise a suspicion of exogenous insulin administration but are far from being proof of it beyond reasonable doubt in the courts.

Immunoassays, invented in 1959, work like a molecular lock-and-key system, using specialised proteins called antibodies to target, bind with, and measure specific substances in a sample.

However, they are not infallible, because they rely on antibodies that might accidentally lock onto the wrong, similar-looking molecules or get blocked by unexpected substances in a patient's sample.

As for the PEG test, it can yield unreliable or false-positive results due to the co-precipitation of certain proteins.

Think of both as inexpensive, accessible screening tools.

As an analogy, imagine an x-ray that shows a hazy area the size of a tumour. Would you immediately operate to remove the tumour? No. You would want to confirm that the hazy area was in fact a tumour, whether it was benign or malignant, etc. In other words, you would order further and more reliable tests.

Interpretation of immunoassay results is not clear cut, and laboratory guidelines will state the various interpretations for certain outcomes. For example:

"Interpretation of C-peptide with concurrent plasma insulin:

Low C-peptide, raised insulin: Insulin administration, insulin receptor antibodies (IR-A)."

Warnings in other immunoassay laboratory guidelines state explicitly the need for further testing (the one below was in red letters):

"Please note that the insulin assay performed at RLUH is not suitable for the investigation of factitious hypoglycaemia. If exogenous insulin administration is suspected as the cause of hypoglycaemia, please inform the laboratory so that the sample can be referred externally for analysis."

As mentioned, to prove the presence of exogenous insulin in a sample beyond a reasonable doubt, LC-MS/MS is required. And sample protocols must be adhered to.

To determine whether an athlete has been cheating, the World Anti Doping Agency insists on use of LC-MS/MS. This is because it knows immunoassays are not sufficiently reliable. And yet, as we shall see, British courts it appears have relied on them to lock people up, sometimes forever.

Types of caregivers

Nurses are professional caregivers, while there are also what you might call non-professional or domestic caregivers e.g. parents, offspring and other relatives, nannies, babysitters, etc.

There are many cases in the UK of actual and suspected wrongful conviction of non-professional caregivers. These are just as appalling as nurse cases but the focus of this article is the latter.

And there are other types of professional caregiver e.g. doctors. Dr Errol Cornish's trial collapsed while the convictions of Dr Hadiza Bawa-Gaba and Dr David Sellu were quashed. Again, these cases were appalling but are not considered in this article.

My plan is to write a follow up article to this one at some point about caregiver cases in general. And perhaps include the many international cases.

Actual and suspected cases of nurse wrongful convictions and accusations

By my count, there are only two nurses in the UK whose convictions relating to allegations of inflicted harm have been overturned, Jessie McTavish and Amanda Jenkinson. Then there is Rebecca Leighton, who was was wrongly accused (charges were brought then dropped, during which time she spent six weeks in prison on remand).

Listed below are the 14 nurse convictions that I have come across in my research plus the instance of charges against Rebecca Leighton that were dropped. If you know of any others in the UK, please let me know.

Name	Capacity	Case involved allegation(s) of insulin poisoning?	Strength of conviction/accusation
Jessie McTavish	Nurse	Yes	Overturned
Amanda Jenkinson	Nurse	No	Overturned
Rebecca Leighton	Nurse	Yes	Charges dropped
Kenneth Barlow	Husband	Yes	No opinion
Deborah Winzar	Wife	Yes	Weak
Julia Knight	Daughter	Yes	No opinion
Beverly Allitt	Nurse	Yes	Weak
Barbara Salisbury	Nurse	No	Weak
Ben Geen	Nurse	Yes	Weak
Colin Norris	Nurse	Yes	Weak
Victorino Chua	Nurse	Yes	Weak
Lucy Letby	Nurse	Yes	Weak
Catherine Hudson	Nurse	No	Strong
Charlotte Wilmot	Nurse	No	Strong
Olive Tchapong	Nurse	No	Weak

Five of the 15 cases did not involve allegations of insulin poisoning. In terms of both method of alleged harm and strength of evidence, they are all very different and so it is hard to identify commonalities. There is little to be learned by, say, comparing Amanda Jenkinson's overturned conviction with the convictions of Catherine Hudson and Charlotte Wilmot (where the prosecution evidence was strong and which have not been overturned).

I also wish to exclude from further analysis in this article four of the remaining 10 cases that did involve allegations of insulin poisoning:

• Kenneth Barlow's case was too long ago for there to be sufficient trial material to appraise.

• The charges against Rebecca Leighton were dropped so her innocence has been proven.

• Julia Knight did not deny administering insulin to her mother (though her mother had leukaemia and there was a clear case for Knight having wished to end her suffering).

• Jessie McTavish's conviction was overturned so her innocence has been proven.

I wish to focus on the other six cases, those of Deborah Winzar, Beverly Allitt, Ben Geen, Colin Norris, Victorino Chua, and Lucy Letby. Call them "The Insulin Nurse Six". Their cases all involved allegations of insulin poisoning. All six nurses denied the allegations, whether those involving alleged insulin poisoning or otherwise. They all protested their innocence. None of their convictions have been overturned.

The charges against the nurses did not just involve allegations of insulin poisoning. Other alleged methods of harm included injection of air, overfeeding, smothering, poisoning with some other drug, etc.

This article, however, focuses on the insulin-related charges because that is where the biggest problem lies. Juries it seems have been particularly persuaded by laboratory insulin test print outs, but they have not been told that the tests are unreliable.

Evidence of insulin poisoning in the six nurse cases

The table below summarises the sample test evidence with respect to the 39 counts (37 guilty, 2 not guilty) that involved alleged insulin poisoning in the cases of Deborah Winzar, Beverly Allitt, Ben Geen, Colin Norris, Victorino Chua, and Lucy Letby. There is also a column for weakness of non-test evidence e.g. no puncture mark, no insulin missing from the store, convoluted bag spiking, alleged poisoning not witnessed, etc.

Nurse and charge count number/alleged victim or method/jury verdict	Hypoglycaemia (very low blood sugar)	Positive immuno-assay	Doubts about reliability of sample?	Positive LC-MS/MS	Doubts about reliability of sample?	Non-test evidence weak?	Proof beyond reasonable doubt?
Winzar #1/Dominic McCarthy/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Allitt #3/Becky Phillips/guilty	No	Yes	Yes	No	n/a	Yes	No
Allitt #7/Paul Crampton/guilty	Yes	Yes	Yes	No	n/a/	Yes	No
Allitt #25/Dorothy Lowe/not guilty	Yes	No	n/a	No	n/a	Yes	No
Geen #1/David Onley/guilty	No	No	n/a	No	n/a	Yes	No
Geen #2/Anthony Bateman/guilty	No	No	n/a	No	n/a	Yes	No
Norris #1/Vera Wilby/guilty	Yes	No	n/a	No	n/a	Yes	No
Norris #2/Doris Ludlam/guilty	Yes	No	n/a	No	n/a	Yes	No
Norris #3/Bridget Bourke/guilty	Yes	No	n/a	No	n/a	Yes	No
Norris #4/Irene Crookes/guilty	Yes	No	n/a	No	n/a	Yes	No
Norris #5/Ethel Hall/guilty	Yes	Yes	Yes	Yes	Yes	Yes	No
Chua #1/Josephine Walsh/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #2/John ‘Jack’ Beeley/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #3/Linda McDonagh/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #4/Joseph MacDonald/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #5/Anthony Smith/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #6/Joyce Atherton/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #7/Beryl Hope/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #8/Doreen Brace/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #9/Tracey Arden/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #10/Kathleen Murray/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #11/Lilian Baker/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #12/Beatrice Humphreys/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #13/Mary Cartwright/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #14/Eileen Armstrong/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #15/Philip Jones/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #16/Derek Weaver/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #18/Arnold Lancaster/not guilty and Chua #19/Arnold Lancaster/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #20/Grant Misell/guilty	Yes	Yes	Yes	Yes	Yes	Yes	No
Chua #21/William Dickson/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #22/Contaminated saline ampoules/guilty	n/a	Yes	Yes	No	n/a	Yes	No
Chua #23/Daphne Harlow/guilty	Yes	No	n/a	No	n/a	Yes	No
Chua #24/Zubia Aslam/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Chua #25/Glucose bag/guilty	n/a	Yes	Yes	No	n/a	Yes	No
Chua #26/Saline bag/guilty	n/a	Yes	Yes	Yes	Yes	Yes	No
Chua #27/Metronidazole bottle/guilty	n/a	Yes	Yes	No	n/a	Yes	No
Letby #6/Baby F/guilty	Yes	Yes	Yes	No	n/a	Yes	No
Letby #15/Baby L/guilty	Yes	Yes	Yes	No	n/a	Yes	No

General discussion

As citizens, we should be appalled about the weakness of the evidence used by our country's justice system to lock up the six nurses.

In all six cases, the justice system appears to have been blind to the fact that those who need care, whether in a hospital or domestic setting, are far more likely to suffer collapse and death. It has ignored the obvious explanations for spikes in mortality in favour of spurious and lurid ones.

The nurse cases were medical cases, and so involved a tool used by doctors known as the differential diagnosis (essentially a list of possible medical diagnoses in order of likelihood).

While DNA, fingerprints and other forensic sciences are precise, medical diagnosis is an imprecise activity, half art, half science. It often isn't even forensic in the sense that doctors engaged as medical expert witnesses are often not working from physical samples but from paperwork: lab print outs, medical records, original postmortem reports, etc.

Non-medical evidence in these cases was also naturally weak. In none of them was there strong evidence of guilt. The nurses were never witnessed harming a patient. There was no CCTV. There were no confessions.

Therefore, to catch these murderous nurses, we must lower the evidence standard threshold.

And we have given permission to our justice system to do just that.

Police are allowed to instruct biased and unqualified medical experts. The CPS is allowed to lower the bar with respect to charging decision standards. Judges are allowed to rule evidence inadmissible and make other decisions that hamper the defence.

If you are an innocent nurse accused of harming those in your care, the deck is stacked against you.

Although the Court of Appeal has a better record with respect to the quashing of wrongful domestic caregiver convictions, there is still a failure to learn from the mistakes that led to them. This may be because it is argued that it is hard to learn from individual cases of wrongful conviction as the mistakes may be seen as idiosyncratic not systematic. However, when the cases are put alongside each other, the commonalities become apparent. This goes for both domestic and professional caregiver cases.

The six nurse cases, and indeed the many domestic caregiver cases, all exhibit certain commonalities, e.g.:

• A spike in deaths/collapses (in the case of Deborah Winzar and domestic caregiver cases in general, the number of deaths is small) that are deemed unnatural

• An accusation of harm that is not properly assessed by independent experts at an early stage

• A biased police investigation

• Weak prosecution evidence (medical and other) and expert testimony/no direct evidence

• Prosecution doctors who do not provide a differential diagnosis but a singular cause of collapse: inflicted harm

• A biased trial judge

• A large body of circumstantial evidence that is considered strong but which, when dependencies and alternative explanations are accounted for, is weak

• Lack of motive

The sanctity of the jury decision is also cited. The argument is made that because one is not allowed to know how juries reached their decision, it is impossible to learn lessons from their incorrect verdicts. This, however, is not a sound argument.

It is clear that jurors have been convinced by prosecution professional witnesses and medical experts who tell them that a positive immunoassay result is indisputable proof of insulin poisoning. They cannot be blamed for not knowing that that is not true.

If certain safeguards were put in place, wrongful conviction of nurses whose cases involved alleged insulin poisoning would very likely be a thing of the past.

The obvious safeguard is that immunoassays should not be used by courts as proof of insulin poisoning. Also, if there is a LC-MS/MS result that indicates the presence of exogenous insulin, there must be clear evidence that all test protocols were strictly followed. It is not contentious to suggest that the burden of proof in criminal cases should be higher than for the World Anti Doping Agency.

Alongside the obvious safeguard in relation to insulin tests used by courts, there are others that should be introduced to address the systematic factors in the list above. Some are already in place but not used. Others may be new.

Closing remark

Deborah Winzar has been released from prison but in many other respects is still serving her sentence. Beverly Allitt, Ben Geen, Colin Norris, Victorino Chua and Lucy Letby remain in prison.

Their cases are, of course, all different. However, as I hope this article has shown, they share common ingredients. Putting the cases side by side rather than appraising them individually should make it apparent that we have a systematic problem when it comes to locking up nurses in this country.

Peter Elston is campaigner for victims of injustice, with a focus on wrongly convicted caregivers. He is a fellow of The Geological Society and of The Royal Statistical Society.