March 22, 2017

expert reaction to study reporting no weekend effect on mortality in emergency surgery

A paper published in the British Journal of Surgery suggests that there is no difference in short or long-term mortality following emergency general surgery at the weekend compared to mid-week.

Prof. Julian Bion, Professor of Intensive Care Medicine, & Chief Investigator of The High-intensity Specialist Led Acute Care (HiSLAC) Project, University of Birmingham, said:

“This is a robust piece of research and the data is likely to be reliable. The methods are based on previous work in this area but the authors have taken care to remove possible sources of error.  They have also compared the effect of day of admission and day of surgery separately.

“As with all research in this area, the authors can only report an association (or lack of an association), they cannot determine causation, or show why a relationship between day of surgery and outcomes does not exist.

“There are several possible reasons why the authors have not found a relationship between surgery at a weekend and mortality.  First, their method of adjusting for risk may be more precise than that of other studies, all of which depend on showing that given comparable patients, the mortality rate is higher for weekend surgery.  So if previous studies have not been able to ensure that patients were indeed comparable (diagnosis, age, severity of illness and so on) earlier reports may have identified a surplus mortality attributed to weekend surgery when it should in fact have been attributed to patients being more severely ill.

“Second, hospital care might be better in Scotland than in England or other countries where this effect has been identified.  There is no way to determine this from the study.  However, others have shown that higher levels of staffing are associated with better outcomes [Ozdemir et al, British Journal of Anaesthesia 2016; 116(1): 54-62], and Scotland has more doctors per 100K population than the rest of the UK [https://www.nuffieldtrust.org.uk/chart/number-of-hospital-doctors-per-1-000-population], so perhaps patients get better treatment in Scotland.  The authors do report shorter times to surgery for weekend admissions, but this is likely to reflect the fact that at weekends there is less elective operating, so the system is more ‘receptive’ to emergency admissions.

“Finally, as the cause of the weekend effect may lie in the pre-hospital phase – in the community – it is possible that Scotland has better community care than England and Wales, as health spending is higher in Scotland than other regions of the UK, benefiting in part from the Barnett formula.”

Prof. Paul Aylin, Professor of Epidemiology & Public Health, Imperial College London, said:

“I would suggest that the study is small, and therefore lacks statistical power. This study looked at 50,844 admissions, and Fremantle’s last study looked at nearly 15 million admissions. As a rough figure, you need a study sample of around 100,000 emergency admissions to be able to detect a 10% increase in mortality.  So, actually the study does find an increase in risk by day of admission (OR 1.14; 95% CI 0.98-1.30), which is entirely consistent with previous studies showing an increased risk of death in patients admitted at weekends, but it is not statistically significant (because of relatively small sample size).

“Given that the day of surgery is dependent on resources (staffing/theatre availability etc.), clinical decision making and workload, I would question whether the day of surgery is the most appropriate point of analysis. The day of admission is perhaps more relevant. Also, by restricting analysis to patients who had an operation, the authors of the study have excluded patients who were admitted as an emergency, but who died before they could get to the operating table.

“Many patients who die in hospital or within 30 days are elderly, so looking at long term survival as an indicator has less relevance, and the longer you follow patients up for, the more the background mortality will swamp any hospital effect.

“I would question the study’s rational of only focussing on first admissions in 3 years- this automatically excludes a number of deaths, because by definition, if you are readmitted, you cannot have died in the index admission.”

* ‘The effect of day of the week on short- and long-term mortality for emergency general surgery: a population-based cohort study’ by Gillies et al. will be published in British Journal of Surgery on Wednesday 22 March. 

Declared interests

Prof Julian Bion: “The High-intensity Specialist Led Acute Care (HiSLAC) project is evaluating differences in weekend and weekday care of emergency admissions to hospital in England.  We are specifically examining quality of hospital care to see if there is a difference between weekends and weekdays, as well as examining severity of illness at the point of admission to hospital to see if pre-hospital care is a contributor to the weekend effect.”

Prof. Paul Aylin: ‘I am co-director of the Dr Foster Unit. The Unit is an academic unit within the Department of Primary Care and Public Health, within the School of Public Health, Imperial College London. The unit receives research grant income from a range of funders, including the National Institute of Health Research and Dr Foster Intelligence, an independent health service research organisation (a wholly owned subsidiary of Telstra). The unit is affiliated with the Patient Safety Translation Research Centre at Imperial which is funded by the National Institute of Health Research. We are also grateful for support from the NIHR Biomedical Research Centre funding scheme.’