Seasonal Variation in Emergency General Surgery: Why Is February Underrepresented
We read with interest the recently published article “Seasonal Variation in Emergency General Surgery.”1 We consider the approach to distinguish seasonal patterns in admission of patients for emergency surgery very interesting. A methodological remark may however be embraced with regard to the marked decrease in emergency admissions in February. Since February has had only 28 days in each year except for 2004 and 2008 during the investigated time span, we observe an average of 2.5 less days available for the recruitment of such patients in comparison to all the other months in the respective years, which accounts for a more than 8% difference in the raw numbers of all counted cases. The authors uniformly point at a difference ranging from 17.2% up to 22.1%, the case load in February falls short of the respective maximum month figure, depending on the processing of the data—that is, whether the number of cases, the adjustment for estimated monthly US population, or the adjustment for 100,000 inpatients is analyzed. However, we wonder why no adjustment per day for every month was performed, which would have allowed for a much more specific and valid analysis. If the monthly US population and the average daily admission rate remain approximately the same in February in comparison with the rest of the year, we would explain roundabout half of the difference discerning February from the respective summer month (July or August) with the maximum emergency case load.
With respect to the current literature, February has already been identified to be the month with the least emergency admissions during the year. A report from Taiwan exclusively covers appendicitis2 but fails to adjust the data for a correction according to the number of days in the month. Concerning comprehensible reasons for seasonal fluctuations of emergency surgery, we suggested a barometric influence upon the incidence of aortic aneurysm rupture.3 The results in international publications are contradictory; an early experience from Ireland indicated a contribution of weather conditions,4 whereas other reports could not confirm such influences.5 Eventually, it may be interesting to provide valid data on this topic, but we will only be able to expect reasonable results if the representation and the reporting of the data are both standardized and free of bias.