Medical researchers have long been interested in the so-called “weekend effect” — the fact that patients admitted on weekends seem to fare more poorly than those admitted on weekdays. Now, in a paper in the British Medical Journal, a team led by Dr. William L. Palmer of Imperial College London has published what it says is the “most comprehensive assessment of its type of the ‘weekend effect’ in obstetric care.”
The researchers took British health data on 1,349,599 births that occurred from April 2010 through March 2012, and tested to see whether the date of admission or birth made a difference with regards to all sorts of different outcomes — not just infant mortality, but injuries and illnesses sustained by the mother as well. On one of the most important outcomes, perinatal mortality (that is, the chance of a baby being stillborn or dying within a week of birth), the researchers found that babies born on weekends were 7 percent more likely to die than those born on weekdays, which translates to an increased mortality rate of about one death per 1,000 deliveries.
This is an interesting public-health mystery — the researchers aren’t sure what’s going on, and couldn’t find any clear relationship between perinatal mortality and staffing levels at the hospital in question, which is the most obvious culprit. It’s also worth pointing out that it was actually Thursday births that had the highest perinatal mortality risk (about 20 percent higher than the lowest day, which was Tuesday), and a couple other weekdays were in the same ballpark. To the extent there’s a trajectory here, Monday and Tuesday saw the lowest risk, and then the rest of the days of the week really were all in the same general ballpark — it’s just that when you pool all the weekdays together, the weekend effect does indeed emerge.
Given all the other more statistically important and controllable stuff that goes into healthy births, and the relatively small differences in risk levels, this days-of-the-week thing isn’t really worth parents-to-be worrying about on a practical level. It’s just an interesting, potentially important quirk of the data that seems to keep popping up in different areas of health research.