Anton Joseph

Weekend death rates: there’s an association, but cause and effect is far from settled

It is not immediately clear why the Editor-in-Chief of the BMJ delayed her public response to Mr Hunt’s perceived misrepresentation of an academic article on the issue of death rates in hospital.

The article that has raised all this controversy is entitled: ‘Increased mortality associated with weekend hospital admission: a case for expanded seven day services?’. It appeared on 5 September 2015. From the title I had assumed that there was some justification for the case being made.

I hold absolutely no brief for Jeremy Hunt, the Health Secretary, but if the BMJ carries an article suggesting a case might have been made for expanded seven day service,  it is hardly surprising that Mr Hunt, after all a politician, made the most of it. The article claims to have shown a clear association between weekend admission and worse patient outcomes.

There are however several caveats. ‘It is not possible to ascertain the extent to which these excess deaths may be preventable; to assume that they are avoidable would be rash and misleading’ and ‘Our analysis shows that we need to determine exactly which services need to be improved at the weekend to tackle the increased risk of mortality’.

Sir Bruce Keogh in July 2015, when questioned by the Parliamentary Health Select Committee, well before the publication of the article, pointed out that excess mortality, “is a statistical construct which is a statistical excess number. It does not mean avoidable. We need to be absolutely clear about that before it gets misused.”

Admittedly there is a warning on the package, but should the toxicity of the contents not have been more explicitly stated in the information sheet. The referees should have required clarification of the alternatives for the association between weekend admissions and worse patient outcomes without leaving the issue wide open.

As it reads, the average reader would be excused for associating excess deaths with reduced services at the week end. This was even more pertinent since leaked figures in the article were being widely used in the propaganda by the government and others as evidence for a weekend effect in hospitals and a case for expanded seven day service. Should the propaganda that preceded publication with the leaking of the figures have raised any concerns with the referees?

Despite the declaration of no competent interest, it is worth noting that the authors of the controversial article state that it ‘arose from a request by BK to update our earlier analyses with more recent data’. Sir Bruce Keogh medical director NHS England has campaigned for NHS services seven days a week since 2013 and clearly wished to obtain further evidence for the ‘week end effect’ in hospitals and ‘a case for expanded seven day services’.

We are assured that the allowing for case mix and other confounding factors that there is a statistical significance for these figures irrespective of the underlying cause. Working at the fringes of statistical reliability, is the statistical evidence robust enough to substantiate the claims given the confounding factors. The association has been established, but the cause and effect is far from settled and may I say tenuous.

It is not too uncommon to hear that someone was knocked down and killed by a car while crossing a very busy road. Do we conclude that crossing the road was the cause of death? Or, for that matter, the majority of people die in bed – hopefully that association has no relationship to causality. The authors should have been required to clarify the variables.

Many would condemn the opportunistic behaviour of Mr Hunt, but perhaps before condemning him for the misinterpretation of the figures should we not question whether there should have been greater demand placed on the authors to discuss the possible range of causes for the excess deaths. The profession I think would have expected more from a reputed journal.

Before entertaining ‘expanded seven day services’ as a possible solution should there not be a clearer understanding of the concept of expanded services before it could be claimed to reduce excess deaths. Before such major changes are contemplated with all its implications it would be important to establish whether all or specific components would have the desired impact.

This government has already accepted that the NHS reform was their worst mistake in power. The NHS cannot withstand another disastrous reform.

There are many loose ends regarding which the journal should have demanded clarification. The editor’s admonition of Mr Hunt though late is most welcome. It is hoped that the journal will continue to demand total justification for the proposed reforms: the very least that the profession and the health service would expect.

Bookmark and Share

Post a Comment

Enter this security code

Submit Comment for Moderation