The weekend effect in neurosurgery: the Leeds general infirmary experience

Abstract

Objective

In this article, we have studied what the impact of the 2016 contract has been on the weekend mortality rate in a single UK Neurosurgery centre for emergency admissions.

Methods

All adult neurosurgery admissions and mortality data from Leeds General Infirmary in 2016 and 2018 was included. Weekday was defined as between 00:01 am Monday and 23:59 Friday. Weekend was defined as anything outside this timeframe. In the first part of the analysis, we excluded all public holiday admissions and compared mortality risks between weekday and weekend admissions. A Cox proportional hazard model was used to examine the time to in-hospital death or censorship. From the model, we compared the hazard ratio of weekend-vs.-weekday admissions for 7-day, 30-day and overall mortalities as well as compared the hazard of mortality on each day of the week to Wednesday admission. In the second part of the analysis, we compared mortality risks of weekday admissions versus public holiday admissions. Finally, to further evaluate whether there was any change in service standard from 2016 to 2018, we assessed the odds ratio of mortality between admission in 2018 and 2016 on weekends and weekdays excluding public holidays.

Results

At 95% confidence interval, no significant difference in hazard ratio was found between admissions on different days in the week when compared to Wednesday in 2016 and 2018. There is a higher weekday admission 7-day mortality hazard ratio in 2018 compared to 2016 but overall there is no statistically significant difference in mortality hazard ratio between the two years. There is, however, a statistically significant difference in hazard ratio when comparing public holiday mortality in 2018 to weekday mortality.

Conclusions

There was no weekend effect in our unit in 2016 or in 2018, however there is a public holiday effect in 2018.

Keywords:

• Weekend effect
• neurosurgery
• junior doctor contract
• mortality rate

Disclosure statement

No potential conflict of interest was reported by the authors.