Higher senior staffing levels at weekends and reduced mortalityBMJ 2012; 344 doi: https://doi.org/10.1136/bmj.e67 (Published 10 January 2012) Cite this as: BMJ 2012;344:e67
- 1Medical Workforce Unit, Royal College of Physicians, London NW1 4LE, UK
- 2Faculty of Medical Leadership and Management, London, UK
Dr Foster Intelligence recently published a report that shows a clear association between reduced numbers of senior doctors in hospitals in the United Kingdom and increased mortality at the weekend.1 How should patients, doctors, and commissioners interpret this finding and what can be done to improve patient outcomes at the weekend?
It has been recognised for many years throughout the Western world that mortality is 10% higher in patients admitted to acute hospitals at the weekend than during the week.2 3 Medical conditions, especially cancer and cardiovascular disease,2 account for most of this excess mortality, but increased weekend mortality has also been shown for surgical diagnoses such as ruptured aortic aneurysm.
The Dr Foster report compared hospital standardised mortality ratios (HSMRs) for patients admitted to English hospital trusts on two weekends in April 2011 with those admitted in the week. They then used self reported data on staffing levels from hospitals to assess the effect of numbers of doctors—both resident and on-call—on weekend mortality. Data were collected on all grades of doctor (and nurses), but interestingly only the number of senior doctors (registrars and consultants) was associated with a difference in weekend mortality.
The hypothesis that early assessment and intervention by experienced clinicians results in improved weekday mortality seems to be a “no brainer,” and patient and professional organisations have called for seven day working to be put in place in all hospitals.4 Lower levels of medical staffing in UK hospitals at nights and weekends are well documented,5 and a large geographical variation exists. Indeed, although not commented on in the Dr Foster report, the preponderance of hospitals with increased weekend mortality in the north of England is striking. However, low staffing levels are only one of the factors that potentially explain increased mortality in patients admitted at the weekend.
Evidence indicates that patients admitted at the weekend are sicker than those admitted during the week, and hospital coding (on which HSMRs rely) is not sensitive enough to allow correction for this. However, studies in which severity scores have been measured show that differences in mortality between the weekend and week disappear for many conditions when the severity of those conditions is adjusted for.6 7 HSMRs have been shown to be highly sensitive to the variability of the coding process, and many clinicians do not trust them. Future measures of mortality need to be accurate and consider severity as well as comorbidity. Equally, clinicians need to recognise the importance of accurate coding.
Although patients with certain conditions—such as trauma, alcohol associated conditions, and self harm—are more often admitted at the weekend, the main reason that sicker patients are admitted at the weekend is variation in referral practice. Out of hours primary healthcare services have changed dramatically over the past decade in the UK, with increasing reliance on “emergency medical services” rather than patients’ own general practitioner. A better understanding of referral practice and medical staffing in the community would be useful when considering variation in mortality between hospitals. This offers a key improvement opportunity for the developing clinical commissioning groups in England.
Provision of hospital support services is reduced at the weekend, so fewer interventional procedures, such as percutaneous coronary intervention for acute coronary conditions and endoscopy for upper gastrointestinal bleeding, are performed.8 9 Increased mortality has clearly been associated with reduced provision of percutaneous coronary intervention at the weekend in parts of the United States.8
The provision of full support services is challenging for many hospitals, both in terms of the workforce and finances (especially in the current economic climate). This has led to the call for the creation of networks through rationalisation of services in parts of the UK. The Dr Foster report shows a reduction in HSMR for stroke in London since the formation of such a network, as well as lower weekend mortality.1 London is unique in the number of large and small hospitals that are within close proximity to one another, but in many other areas the network model for trauma, vascular surgery, and stroke is well advanced. How successful such networks will be in terms of patient outcome, safety, and cost effectiveness remains to be seen.
Even when these other contributory factors are considered, the observation of increased mortality and low staffing levels cannot be discounted and poses a serious problem for the NHS. The process of increasing the number of doctors within the UK, especially senior doctors, is slow and expensive. Short term increases using doctors from outside the UK will probably have an effect on healthcare workforce planning in other European Union countries.10 It is unclear which specialties and grades of doctor need to be increased in number for weekend mortality to be reduced. Future studies need to investigate this question because workforce planners will need to know—for example, whether more accident and emergency doctors or more vascular surgeons are needed to improve weekend outcomes.
Continuity of care must be maintained when remodelling weekend staffing. Emerging data show that working patterns for consultants influence mortality. Hospitals in which the admitting consultants work blocks of more than one day have lower excess weekend mortality than those with a “physician of the day” model (D Bell, personal communication, 2012).
Interestingly, Dr Foster defined first year registrars as senior doctors in their report,1 and although this may be debated by some, it shows the importance of registrars in the provision of out of hours services. Indeed, the number of medical registrars (who run most hospitals at night) could easily be the defining predictor of hospital mortality. The current plan for the UK is to reduce the number of registrar posts in both surgery and medicine.11 This may need to be re-thought but, given the potential profound impact on clinical outcomes, decisions must be based on sound evidence.
The Dr Foster report raises more questions than it answers. Hospitals and commissioners, with their clinicians, need a better understanding of the potential factors that cause higher mortality at the weekend. Among these—and with profound implications for planning—are community out of hours services, hospital staffing, and workforce configuration. All need to be reviewed against the knowledge of which conditions are associated with increased mortality at the weekend. This is an opportunity that, if tackled intelligently, will improve the care of some of our sickest patients for many years to come.
Cite this as: BMJ 2012;344:e67
Competing interests: Both authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Commissioned; not externally peer reviewed.