Although it is a place of healing, there could not be a bloodier political battleground than the NHS. Jeremy Hunt is, by now, a veteran of skirmishes in this war – his policies have been met by fierce and stalwart opposition from members of the public and healthcare professionals.
Recently, wounds were re-opened when Stephen Hawking re-entered the fray.
Prof Hawking’s involvement in NHS-related debates is not nascent. In 2016, he joined other academics in co-signing a letter emphasising the need for more conclusive evidence regarding the ‘weekend effect’: the claim that mortality rates are higher for patients admitted at the weekends.
In his most recent letter, Hawking was more broadly critical of the overall direction the NHS is heading in: it’s a tug of war between public and private interests, and multinational health corporations look like they are winning. But in this article, Hawking again brought up the weekend effect, accusing Jeremy Hunt of ‘cherry-picking’ data to support his drive for a 7-day NHS.
First, a bit of background. Currently, a subset of services such as ambulances, A+E and some GP practices work over a 7-day period. However, in-hospital diagnostics, outpatient care and routine operations work on a 5-day schedule. Jeremy Hunt has previously cited evidence claiming that 6,000 more patients lose their lives on weekends, and has linked this to a dearth of service provision on Saturday and Sunday.
But the weekend effect lacks a precise definition. You could think of a ‘strong’ weekend effect: a higher risk of death on a Saturday or Sunday compared to a weekday. Another, less obvious and unclear (‘weak’) effect would be a higher risk of death for patients who are admitted on Saturdays or Sundays vs. weekdays. This measure – typically calculated over 30 days after admission – is a little muddier. A lot of things can happen in 30 days, so causality is more difficult to establish.
To assess the Health Secretary’s claim, I think three questions need to be answered:
- Is this a strong or weak weekend effect?
- Where does the 6,000 figure come from?
- Can the figure be clearly linked to a lack of service provision at weekends?
Strong or weak weekend effect?
One of the key studies regarding the weekend effect was published in 2012 (Freemantle et al., 2012), based on assessment of admissions during 2009/10. The retrospective analysis looked at all admissions (14.2 million) and followed them up for 30 days. The authors counted just over 187,000 deaths during the 30 day period.
Hunt and the Department of Health emphasised one finding from the study: admission on a weekend was associated with an increased risk of subsequent death within 30 days. Specifically, at any particular time after admission, patients admitted on Saturdays had an 11% increase in risk and on Sundays a 16% increase in risk of death compared to patients on weekdays. It’s worth noting that the increase in risk wasn’t restricted to weekends: it was also higher for Mondays and Fridays.
But hospital stays on weekends were associated with a lower risk of death than on midweek days. The values were 8% lower for Sunday vs. Wednesday, and 5% lower for Saturday vs. Wednesday.
So, the data in this study suggest that there might be a weak effect, but this isn’t restricted to weekends. Furthermore, the risk of death on Saturdays and Sundays is actually lower than on weekdays (the opposite of a ‘strong’ weekend effect). It is irresponsible that Hunt focused on just one aspect of this study, neglecting to mention the other conclusion.
The story doesn’t get any better for the Health Secretary. Even evidence for a weak weekend effect is inconsistent: a large prospective study published in The Lancet looking at stroke patients found variation in 30 day survival across the entire week, with no significant difference between weekends and weekdays.
Stroke patients are an important group – they account for a large number of admissions across a spectrum of clinical severity. Importantly, effective acute management of stroke greatly impacts survival. If weekend and weekday outcomes do not differ, it suggests that weekday performance is at least maintained at weekends. This begs the question: will increasing staff levels on Saturday and Sunday improve patient outcomes?
Where is the 6,000 figure from?
“6,000 more people lose their lives because we don’t have a proper 7-day service.” (Jeremy Hunt, 2015)
Scary? Figures like this one are daunting, and add objective – as well as emotive – weight to arguments. But figures need to be critically appraised.
The (now infamous) 6,000 figure has been questioned by numerous sources, and its precise origin has proven to be elusive. Sir Bruce Keogh – director of NHS England (and one of the co-authors of the 2012 study) – clarified that the figure was derived from existing data. In other words, it is not an empirically measured fact.
The analysis in the 2012 paper has since been updated to reflect the 2013-14 admission period and the results were published in 2015 (Freemantle et al., 2015). This paper did not mention 6,000 excess deaths, although it did find that 11,000 more people died each year “within 30 days of admission to hospital on Friday, Saturday, Sunday or Monday compared with other days of the week.”
So, there is no direct evidence to suggest that the 6,000 figure reflects current mortality data for weekends compared to weekdays. Furthermore, despite the Health Secretary using this dubious figure to support a 7-day proposal, the authors of the 2015 study have emphasised that it is not possible to determine if these deaths are preventable. To do so would be making an unjustifiable assumption.
Any policy based on flimsy interpretation of these data as a foundation is surely a leap of faith.
Can the figure be linked to a lack of service provision on weekends?
It is not unreasonable to suggest that care standards may be poorer at the weekend: staffing ratios are lower and diagnostic availability is reduced. One study found an increased risk of hospital acquired conditions, cost and length of stay associated with weekend admissions.
So yes, there could be a link.
But it is far from clear. The authors of the 2015 paper suggest that linking the increased number of deaths to staffing shortages “would be rash and misleading.”
There may instead be a different cause. A recent study from the University of Manchester found that smaller numbers of sicker patients are admitted at weekends. This is because hospitals apply higher severity thresholds when assessing patients on Saturday and Sunday.
This has two pertinent implications. Firstly, the smaller number of admissions means lower numbers of absolute deaths over weekends. Secondly (and particular pertinent to this debate), the higher death rates may reflect sicker patients rather than inadequate service provision. The weekend effect may therefore be a statistical artefact, reflecting the variable characteristics of patient groups.
Is there any clarity?
Uncertainty is the only certainty in this debate. Data regarding the weekend effect are murky: we don’t know exactly how many more patients die due to being admitted on weekends; we don’t know whether this applies to all conditions or just a few; and we can’t directly link increased mortality rates to service provision.
The debate about the weekend effect is intimately linked to Hunt’s proposal to a 7-day NHS. Whilst the principle of extended service provision is a noble one, it won’t come cheap. The Conservative government has promised to inject £8 billion into the NHS over the next five years, but even if government spending is maintained in line with inflation, the NHS faces a funding gap of £30 billion. This is without taking into account the cost of extending 5-day services to cover the whole week. There is too much financial risk to commit to extending NHS services to full 7-day coverage when we are completely uncertain regarding the extent of the benefits (if indeed there will be any benefit at all).
Prof Hawking clearly has a point: Jeremy Hunt has at best been negligent and at worst deceptive regarding the weekend effect. As well as an apparently negligent disregard for counter-evidence, Hunt has manipulated accounts of primary evidence and presented extrapolated data as fact.
What characteristics and skills would you want in your Health Secretary? Honesty? Integrity? The ability to critically assess data and derive a balanced, well-considered conclusion? Mr Hunt has not demonstrated any of these traits. This leads us to only one conclusion: Jeremy Hunt lacks the credentials needed to serve as an effective health secretary. The egregious negligence he has displayed here surely means his position as Health Secretary is in jeopardy.