The "Friday effect": Can epidemiology tell us when to operate?

Abstract

Concerns have long been voiced over the higher rates of adverse outcomes in patients emergently admitted to hospital over weekends. Indeed, more recent work addressing the effect of the day of the week as a factor in surgical outcomes has suggested that the odds of death are higher following elective surgery on a Friday compared with other days of the week. This ‘‘Friday effect’’ and the accompanying concerns over reduced staffing levels on weekends have been suggested as the main explanation for the differences in outcomes and have led to calls for increased weekend staffing. For example, in the United Kingdom (UK), the clinical director of the National Health Service has called this need for more staffing his numberone priority. Though most studies addressing the impact of the day of the week have focused on mortality as their outcome, the risk of death is very low for many patient subgroups and other outcomes may be of equal importance. In this issue of the Journal, McIsaac et al. merged several administrative databases in Ontario in an analysis of Emergency Department (ED) visits and hospital readmissions following ambulatory surgery. Results of their analysis showed that one in ten patients had an ED visit or hospital readmission within 30 days of discharge after their surgery. In their composite outcome of ED visit or readmission, they found a small but statistically significant elevated risk for operations on Fridays and Tuesdays (7% and 5% higher hazard ratio, respectively) when compared with Mondays. Their Friday effect was driven specifically by ED visits. Furthermore, subanalysis by procedure revealed much larger effects for both shoulder surgery and transurethral resection of the prostate (TURP). In studies focusing on mortality, the hypothesis has been that poorer weekend care led to some additional adverse outcomes. In contrast, McIsaac et al.’s hypothesis was not that patients having surgery late in the week have higher rates of complications – though the authors did not specifically report rates of complications – but that patients may find it harder to access care in a surgeon’s or family physician’s office over the weekend, thus necessitating an ED visit. A prior report from their group had shown such difficulties in access following ambulatory surgery. In effect, the ED became the substitute ‘‘office’’. This theory is supported by their finding of an increased risk of an ED visit during the week following shoulder surgery. Shoulder surgery is a particularly painful procedure which the authors propose leads patients to think that they need to seek immediate help, perhaps in the ED rather than in the surgeon’s office. A limitation of the Ontario database is the lack of information on the specific reason for the ED visit; however, arguably an even more important missing piece of data is the reason for postoperative family physician or surgeon visits during the week. Researchers have long been frustrated by the lack of linkage between primary and secondary care records beyond local pilot studies; however, such linkage is increasingly available in many other A. Bottle, PhD (&) Dr Foster Unit, Department of Primary Care and Public Health, School of Public Health, Imperial College London, 3 Dorset Rise, London EC4Y 8EN, UK e-mail: robert.bottle@imperial.ac.uk