Biases in detection of apparent "weekend effect" on outcome with administrative coding data: population based study of stroke.

Li L, Rothwell PM; Oxford Vascular Study. BMJ. 2016;353:i2648.

The "weekend effect" refers to the fact that mortality for several common conditions is higher in patients admitted on weekends compared to weekdays. While the mechanism for this effect is unclear, it likely varies for different disease processes. For example, prior studies have postulated that a weekend effect exists for patients with acute stroke. However, this study analyzed a large British database and found that many patients with a history of stroke who were later hospitalized for other reasons had their admission diagnosis inaccurately documented as acute stroke. This inaccuracy occurred more frequently in patients admitted on weekdays. Because the weekday admissions included many patients who were hospitalized for less morbid conditions, mortality appeared lower for patients admitted on weekdays than on weekends. When data was reanalyzed to include only those patients with a true acute stroke, no weekend effect was found. This study demonstrates the limitations of administrative data in analyzing patient safety issues.

PubMed citation Free full text