08 / 09 / 2015
Seven-day services – analysis and commentary
British Medical Journal, September 2015
Nick Freemantle and colleagues (including Sir Bruce Keogh) have published an updated evaluation of weekend admissions, following a previous study of cases for 2009-10. This new piece of research looked at more than 14.8 million admissions in 2013-14, with the article highlighting that admissions on a Saturday and a Sunday only amounted to eight and six per cent respectively of all cases reviewed.
It is noted that not only were 50 per cent of Saturday admissions and 65 per cent of Sunday admissions found to be emergency cases, but also the “cohort admitted at weekend included a greater prevalence of patients with higher predicted mortality risk than those admitted during the week.” Thirty-day mortality across all admissions for 2013-14 was found to be 1.8 per cent, with the relative risk of death within that timescale judged to increase by 10 per cent (i.e. 1.98 per cent mortality) for a Saturday admission and 15 per cent (i.e. 2.07 per cent mortality) for a Sunday admission. Freemantle et al confirm that their 2013-14 findings were “qualitatively similar to the corresponding results from our previous analyses”, which found relative risk increased by 11 and 16 per cent respectively.
The article also considers the impact of weekend admission for those with cardiovascular disease and oncological conditions (“similar results” overall, although it is noted that for the latter group, there was also an increased risk on Mondays and Fridays). Freemantle et al suggest there is now “a more generalised ‘weekend effect’” covering Fridays and Mondays, with around 11,000 extra deaths annually found for cases admitted over that four-day period in 2013-14. Other conclusions from the study include:
- Patients admitted over the weekend judged to be at the highest risk of predicted mortality have a longer length of stay if they survive the initial episode of care than those with similar conditions admitted during the week.
- The researchers do confirm “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.” Nevertheless they also stress that the ‘weekend effect’ is “not otherwise ignorable.”
- Freemantle et al are unequivocal that patients “should never have to accept an increased risk because of the way healthcare services are designed and delivered.”
The BMJ has simultaneously published an editorial, feature and observation piece about the seven-day working issues.
In his editorial, Professor Paul Aylin argues that research contradicting studies akin to the above “tend to be smaller, carried out in single hospitals, and lack statistical power.” He emphasises the distinction between ‘relative risk’ and ‘attributable risk’, while also noting the evidence pointing to a “weekend increase in avertible errors in care leading to other complications.” The article cites research that Aylin himself was involved in, which found an increasing mortality risk from planned procedures as the week progressed and that the care received by emergency stroke patients at weekends was often lacking crucial elements, e.g. thrombolysis and prompt computed tomography. He concludes by advocating a “comprehensive systematic review” of studies into the ‘weekend effect’ and more research into the “complex relation” between staffing levels and services and safety.
Helen Crump from the Nuffield Trust has examined the evidence around the Secretary of State’s proposals. Crump cites the research from both Freemantle et al and Aylin et al discussed above, while also noting the University of Manchester estimate of an absolute increase of 0.3 per cent for weekend mortality. Furthermore, she stresses that there are “some major questions about the cost”, highlighting that the HFMA estimate of 1.5-2 per cent of trust income “could be enough to push any remaining hospitals just about managing to maintain a surplus into deficit.” The article points to suggestions that the costs of seven-day services would amount to more than £20,000 per quality adjusted life year (QALY) and also contends that there could be “potentially serious consequences across other services” if overall staffing levels are not increased. This prompts Crump to conclude that “pumping more money into the acute sector…goes against the grain of other policies” and that “ramping up weekend elective activity makes no sense if the funding is not available to increase total activity.”
Professor Martin McKee expresses concern about the validity of the Government’s claims that seven-day working could save 6,000 lives annually. He highlights that the previous analysis by Freemantle et al, published in 2012, found that those patients already in hospital “had a significantly reduced risk of dying” on both days of the weekend. Moreover McKee warns of the possibility of “confounding by indication” in relation to the risks faced by those admitted for elective procedures on Sundays and emphasises the “many other possibilities” beyond reduced consultant cover, for example access to community nursing or a delayed decision to admit a frail elderly person. He also cites the figures regarding QALYs for seven-day services, before concluding “it would be useful to have a better understanding of the reasons for any increase in deaths at weekends, details of what is being proposed, and evidence to justify any changes.”