In a recent letter to junior doctors, secretary of state Jeremy Hunt again argued that “tackling higher mortality rates at weekends” was of utmost priority and that he would be pushing pay reform through to tackle the “link between higher weekend death rates and reduced weekend services”.
Six studies have found a link between higher weekend death rates and a reduced weekend service, the most recent of which found that 11,000 more people die each year if admitted over the extended weekend, from Friday to Monday, compared with those admitted on midweek days.
Hunt’s solution is to increase weekend cover, with more senior consultants on hand to back up junior doctors, together with a full range of diagnostic and support services. His plan will work. By increasing weekend staffing levels, in time there will be a smaller difference between midweek and weekend mortality rates. But the gap may not close for the reason Hunt intends.
Closing the gap
The figure of 11,000 excess deaths was calculated by comparing the rate of deaths to admissions over the weekend with the midweek rate. There are two ways by which a rate can be reduced. One is to reduce the number of deaths of those admitted over the weekend. This is what Hunt wants. It may be that people are dying unnecessarily because there aren’t enough staff to care for them. If so, making more staff available at the weekend should help save lives.
But the weekend mortality rate can also be reduced by increasing the number of weekend admissions. If more people are admitted, then the weekend mortality rate would fall, even if the same number of people were still to die.
At the moment, simply because fewer staff are available, the chance of being admitted over the weekend is lower than during midweek. And because the chance of admission is lower, those people that are admitted at the weekend tend to be sicker than those admitted during the week. And, of course, sicker people are more likely to die.
To some extent, the higher likelihood of dying is taken into account in the calculation of mortality rates. But the calculation is imperfect because there is limited information about how sick patients are. Taking account of this missing information has been shown to be critical in the analysis of hospital readmission rates, and efforts are being made to do something similar in analysing admission from accident and emergency departments.
Probably both the number of deaths and the number of admissions will change as a result of more staff being available at the weekend. So Hunt is likely to achieve his goal. By making more staff available, the weekend mortality rate is likely to fall. But the reduction is likely to be driven not so much by a fall in the number of deaths but mainly by an increase in weekend admissions.
This wouldn’t be an altogether desirable achievement. For one thing, reductions in the weekend mortality rate will come at the expense of efforts to reduce emergency admissions. For another, it may detract from more cost-effective ways to save lives.
It turns out that hospitals have been getting better at keeping people alive over the past decade. This can be seen from data about hospital death rates published by the Health and Social Care Information Centre, the latest release covering 2003-13.
There have been significant improvements in survival in the 30 days following coronary artery bypass graft, elective surgery and treatment for heart attack and leg fracture. But the most dramatic improvement has been in stroke survival, especially following the launch of the national stroke strategy in December 2007.
The trends in survival are cause for celebration. When it comes to keeping people alive, the NHS has been making big improvements for quite some time. These improvements have not happened by chance, but reflect the success of efforts such as the national stroke strategy to improve prevention, diagnosis and treatment. A similar strategy is needed to address higher weekend mortality, underpinned by a better understanding of its causes and a review of how these might be addressed. Without this, the weekend mortality rate might well fall, but we’ll be no better off as a consequence.