Study finds no evidence linking weekend staffing levels in hospitals to mortality
The ‘weekend effect’ – that patients admitted to hospital over the weekend are at an increased risk of death – overshadows a much more complex pattern of weekly changes in quality of care, which are unlikely to be addressed by simply increasing the availability of hospital doctors on Saturdays and Sundays.
The paper, published in The Lancet, found no association between weekend senior doctor staffing levels and mortality, questioning whether increasing senior doctor weekend working would impact mortality rates.
Studies on the weekend effect have had a major impact on health policy. However, evidence about why the quality of care may be worse at the weekend, or whether there is a direct link between mortality and the availability of consultants remains largely speculative.
Researchers from the High-intensity Specialist Led Acute Care (HiSLAC) project led by the University of Birmingham, collected data on senior doctor input into emergency admissions at 115 NHS trusts at two time points – Sunday 15June and Wednesday 18 June 2014 , and compared this with weekend admission mortality rates average over one year.
Hospital consultants completed a survey recording the number of hours they had spent between 8am and 8pm caring for patients who had been admitted as an emergency. A similar survey was completed by clinical service directors and the findings were compared to national hospital mortality data.
There were substantially fewer senior doctors present and providing emergency care on Sunday (1667) compared to Wednesday (6105). This was partly offset by the fact that doctors spent on average 40% more time caring for emergency admissions on Sunday (5.74 hours) compared to Wednesday (3.97 hours).
Once the number of admissions per hospital was taken into account, the findings show that emergency admissions on a Sunday collectively receive on average less than half the input of senior doctors compared to patients admitted on Wednesday (21.90 total specialist hours per ten emergency admissions on Sunday; compared to 42.73 hours on Wednesday).
The researchers then looked at mortality data for all 115 trusts and compared this to staffing levels at each trust. Across all trusts, there was a slight increase in the risk of death associated with weekend admission (10% relative risk increase). However, there was no evidence to suggest that increased mortality was linked to senior doctor staffing levels.
Lead author Professor Julian Bion, from the University of Birmingham, explained, “Patients admitted to hospital over the weekend are likely to receive less time with consultants, and do indeed have a slightly higher risk of death. Both problems need to be addressed to provide consistent standards of high quality care. But to say that lower staffing is the cause for increased mortality is far too simplistic and not supported by the evidence. Policy makers should be extremely cautious when attributing the weekend effect directly to the lack of consultants at the weekend.”
The authors warn that the study represents preliminary evidence from a longer term project and that finding a lack of association now cannot discount it in the future. However, they urge caution in linking mortality directly to senior doctor staffing levels, and say that more research is needed to understand the key factors affecting mortality in such a complex system.
Notes to editors
For interview requests, a copy of the full paper or for more information, please contact Luke Harrison, Media Relations Manager, University of Birmingham on +44 (0)121 414 5134.
For out of hours media enquiries, please call: +44 (0) 7789 921 165
 Sunday and Wednesday were selected as they are associated with the highest and lowest mortality rates; June was chosen as it has no public holidays, and is unaffected by winter pressures.
HiSLAC is funded by National Institute for Health Research Health Services and Delivery Research Programme.
The paper will be available here: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30442-1/abstract