Weekend emergency admissions and mortality in England and Wales.

Increased mortality for hospital admissions at weekends has been reported for emergency admissions overall and for specific disorders, although the size of this effect varies across reports. No evidence exists that compares a wide range of emergency disorders or, for confirmatory purposes, is based on two independent information sources. Further evidence is needed to defi ne which disorders are susceptible to the weekend effect. A Wellcome Trust project of mortality after emergency admissions across England and Wales was used to investigate disorder susceptibility to the weekend effect across two different health-care systems and two independent information sources. Systematic record linkage of national administrative inpatient and mortality data was used for emergency admissions to all public hospitals across England and Wales. Mortality at 30 days was established for admissions on weekends, and compared with admissions on weekdays, for emergency disorders overall and for 15 major circulatory, gastrointestinal, respiratory, and trauma disorders from Jan 1, 2004, to Dec 31, 2012. Logistic regression modelling was used to adjust mortality for patient age, sex, and comorbidities (table, appendix). Overall mortality at 30 days after emergency admissions in England was slightly lower than in Wales (5·59% vs 5·64%). The increased mortality for weekend admissions compared with weekdays was similar in England and Wales. Mortality was higher in England than in Wales in 2004–06, similar in 2007–08, and lower in England than in Wales in 2009–10 and 2011–12. The sizes of the weekend eff ects on mortality in England and Wales were consistent for all 15 disorders and the Pearson’s correlation for each disorder across the two countries was 0·57. The weekend eff ect was strongest for abdominal aortic aneurysm followed by other disorders with very high mortality during the acute phase; pulmonary embolism, stroke, and subarachnoid haemorrhage. Little or no weekend eff ect was observed for acute myocardial infarction and less acute disorders; chronic obstructive pulmonary disease, pneumonia, hip fracture, acute pancreatitis, and inflammatory bowel disease. No signifi cant variation was observed in the weekend effect over time or across patient age groups. These data provide new evidence as to the emergency disorders that are most strongly affected by the weekend effect and show that findings are quite consistent across two health-care systems. The weekend eff ect is most apparent for disorders with very high mortality that often require access to specialist investigation and care during critical acute phases. We declare no competing interests. We acknowledge support from the Wellcome Trust (093564/Z/10/Z). We thank Judy Williams for clerical assistance, Alan Watkins for statistical advice, and the Health Information Research Unit (Swansea, UK) for access to the Secure Anonymised Information Linkage databank.