Nappy Pads on Ceiling Sewage Leaks – Did Infection Kill the Letby Babies?

When we review the sorry saga of the Countess of Chester Neonatal Unit, there’s one niggling question outstanding…

Shoo Lee’s panel of experts have found likelier causes than harm by Letby for every infant death or collapse. Care was substandard, with the unit working far beyond its competence. These conclusions accord with a 2016 report by the Royal College of Paediatrics and Child Health. This noted that consultants only undertook twice-weekly ward rounds and that the unit was under-skilled for a Level 2 site. Prof Lee was blunter, observing that, “If this was a hospital in Canada, it would be shut down.”

The smoking gun of the insulin cases – upon which the other convictions are predicated – is spiked by the facts that the insulin:C-peptide concentration ratios in infants do not conform to those of adults; that the test used was not licensed for infants and that the Liverpool Laboratory had recently failed an external quality assurance test. Having received these anomalous insulin results, Chester then failed to have them rechecked by a reference laboratory.

The end of the string of deaths and collapses in mid-2016 is easily explicable by the Chester unit being downgraded from Level 2 to Level 1. It no longer took such sick babies. Letby’s ‘confession notes’ are no more than jumbled, contradictory jottings, written during therapy. Her habit of keeping piles of handover notes at home contravenes data protection and patient confidentiality regulations, but hardly proves any major crime. To call them ‘trophies’ was absurd but cruelly effective.

In short, Letby’s guilt was not proven ‘beyond reasonable doubt’ and the strong likelihood is that she is the victim of a terrible miscarriage of justice. So, what’s the niggling question? Simply this: “Why did the string of deaths and collapses begin in 2015?” The unit had been a Level 2 for several years; Letby had joined in 2012; Drs Brearey, Jayaram and Gibbs remained the senior medics; and nothing that emerged at Letby’s trials or the Thirlwall Inquiry indicated why care should have suddenly deteriorated in 2015. Yet, whereas there were one to three deaths in each year from 2010 to 2014, there were eight in 2015 and five in the first half of 2016.  

One possibility, impossible to refute, is simple bad luck. Statisticians point out that the death rate in the Chester Unit from mid-2015 to mid-2016 were no worse than the typical worst neonatal unit of the year. And, in any year, one unit must be the worst, with its failings not ordinarily attributed to a homicidal member of staff. A second explanation – for which there is clear and obvious evidence – is that, as the deaths mounted, trust among the staff collapsed, impacting teamwork and worsening outcomes. A third possibility – though it does not fit that timing well – is that University of Liverpool trainee paediatricians, many doubtless competent, stopped circulating through the Chester Unit in 2013/14. 

The fourth explanation is infection – which may be a frequent factor in the statisticians’ ‘worst unit of the year’. Tiny infants lack a fully-developed immune response. They succumb easily and swiftly, sometimes without the classical signs and symptoms of infection that you’d see in an adult. Those born very prematurely, before 32 weeks, do not fully benefit from maternal immunity – meaning the transfer of protective antibiotics across the placenta from the mother. What is more, the unit was cramped, with the infants’ cots close together. This facilitates transmission of pathogens from baby to baby. Assertions that some were ‘well’ before they encountered Letby – and therefore able to fight off pathogens – are refuted by Prof Lee’s analysis. And, if they were well, why were they in a Level 2 unit, not in cots beside their mothers’ beds?

Given the unit’s deficiencies, one may wonder how reliably or assiduously infection was sought. What samples went to the lab, and when? We don’t know for sure, and obtaining a good sample of blood or sputum from a tiny infant can be difficult. What we do know, without details of the pathogens, is that infection was implicated in babies C, D, G and I – all of whom Letby was convicted of murdering or attempting to murder. In addition, necrotising enterocolitis (meaning invasion of the body tissues by gut bacteria) was originally suspected for baby E, who had no post mortem. In other words, there is evidence that at least one third of the ‘Letby’ babies had infections – and we know nothing about the concurrent deaths for which she was not charged. Were some of them dismissed from the charge sheet because infection was clear and obvious?

What is sure is that the unit was bacterially hazardous. The sole defence witness at Letby’s first trial was the hospital plumber, who told how he was regularly called to deal with ‘foul water’ leaks from the ward drains above. Subsequently it was reported that a tap inside the unit was colonised with Pseudomonas aeruginosa, a water-loving bacterial species notorious both for forming biofilms in pipework and for causing difficult-to-treat infections in the vulnerable. The biofilms, extending back into the plumbing, are near-impossible to eradicate and infections, when they arise, are resistant to many standard antibiotics. In 2012 a tap-water-associated P. aeruginosa outbreak in Belfast killed three babies, whilst an outbreak in Bristol infected 12 and killed one.

The Telegraph now reports, following a leak of a different sort, that a second tap was also colonised with P. aeruginosa. Filters, fitted to prevent the bacteria entering the output water, were apt to fall off according to a former member of the Estate Management Team. Why? Were they the wrong filters for the type of tap? Or badly fitted? Or were they used for too long, so that they clogged, then were dislodged by back pressure? A British Standard was published in 2022, long after the events in Chester. Among other aspects it stresses the need for competent fitting, cleaning and changing. Was this done?  

But this isn’t the worst of what has now seeped to the newspaper. The worst is that the unit’s immediate method for responding to the sewage leaks from above was to stick an absorbent nappy pad on the ceiling. No, I couldn’t make it up, either. Leaked e-mails show Eirian Powell, the unit manager, writing to senior managers that the department was “taking too many risks” and “compromising patient safety”. It is hard to disagree. 

This was an outbreak waiting to happen, whether due to tap-water Pseudomonas, or Escherichia coli from sewage, or both. And the fact that excess deaths and collapses were bunched together in 2015-16 suggests to me that an outbreak did occur. Deaths due solely to substandard care would be more widely scattered.

There are just two caveats. First, the occurrence of an outbreak does not mean that it caused all deaths – not, for example, Baby O, who was stabbed in the liver during a consultant’s botched attempt to release gas. Second, as in the case of ‘died with COVID-19’, infection is often the final straw that pushes the vulnerable patient over the mortal edge. Suggesting it as a factor here does not dispute other underlying causes or deficiencies of care, and I defer to Shoo Lee’s panel on these.

David Livermore is a retired Professor of Medical Microbiology, University of East Anglia.

Stop Press: A powerful new TV documentary reveals that Lucy Letby wept repeatedly in the arms of a close colleague as she tried to cope with allegations that she was harming babies in her care, according to the Mail.