Is there a weekend effect in obstetrics?

Recent research published in The BMJ and elsewhere brings renewed attention to the “weekend effect,” suggesting higher rates of adverse outcomes associated with hospital admissions and procedures performed at weekends than on weekdays. Findings are not uniform among studies and fields of medicine, and persistent questions remain about whether significant findings reflect differences in case mix severity during the weekend or staffing and volume factors that are likely to influence outcomes among the patients at highest risk. The weekend effect is particularly under-studied in obstetrics, with decidedly mixed results from the small number of studies. A study by Palmer and colleagues (doi:10.1136/bmj.h5774) helps to fill this evidence gap, presenting a thoughtful analysis of adverse birth outcomes in a retrospective cohort from the United Kingdom. This study found that some adverse outcomes were slightly but significantly more common among weekend deliveries, most notably perinatal mortality. Although the magnitude was small (an unadjusted absolute increase of 0.9 deaths per 1000 deliveries (0.73% v 0.64%); adjusted odds ratio 1.07, 95% confidence interval 1.02 to 1.13), the gravity of this outcome demands our attention. Despite some notable examples of null findings, enough evidence now exists for us to reasonably suspect that out of hours deliveries are at higher risk for adverse outcomes. The evidence for higher risks among night time deliveries is even stronger.However, additional well designed studies are needed to determine whether these findings are robust within and across populations. In particular, although Palmer and colleagues controlled for patients’ characteristics in regression models, more work should examine potential differences in case mix between weekend and weekday deliveries. 14 Observational data are often the best available option to study this topic, but every effort should be made to rule out uncontrolled confounding. These authors are to be commended for analyzing multiple quality metrics (for example, puerperal and neonatal infections, birth trauma), and the concordance of several of these findings lends credibility to results. Given ongoing debate about the most appropriate quality metrics, 16 future studies should analyze currently accepted metrics, even as we continue to refine definitions. In particular, calls have been made for researchers to stop using severe perineal lacerations as a quality metric in obstetrics. Even in the most rigorous studies, the most likely mechanism underlying the weekend effect is systems factors (for example, staffing, resource availability, hospital policies). Such factors are specific to a healthcare system, and even to a hospital. Therefore, finding evidence of an effect in one population does not guarantee that the association will persist in others or even within a single population over time. This heterogeneity of effect complicates research but also offers important opportunities to identify levers with potential to improve outcomes for women and babies. The weekend effect in obstetrics fits within the broad concept of “capacity strain” in healthcare systems—the process bywhich performance of a clinical unit can deteriorate above a certain threshold of patient volume, complexity (acuity), or both. 18 Given the decreased levels of staffing and availability of resources that characterize most hospitals at the weekend, a lower threshold above which capacity strain threatens patients’ outcomes is likely. Evidence is emerging that other factors related to capacity strain such as busy days, holidays, and doctors’ absence at conferences affect patients’ outcomes, in addition to weekend effects. 19 20 More research is needed in obstetrics to explore capacity strain: how to define and measure it, which obstetric outcomes are sensitive to it, and when. Most importantly, when a capacity strain effect is found, we must identify effective strategies to safeguard maternal and infant outcomes during such vulnerable times. Factors that may help to mitigate the weekend effect and other forms of capacity strain in obstetric units include specific staffing models, such as the obstetric hospitalist model and other flexible models of care, and hospital policies including condition specific protocols. Unfortunately, several recent studies, including this one, have found no association between outcomes and staffing or use of protocols. 23 Still, we must continue to explore the factors that differentiate obstetric units from one another, analyze how and when adverse outcomes “out of hours” are associated with these factors, and apply the findings to clinical practice and hospital policy.