Statistics from Altmetric.com
I read the paper by Brinjikji et al1 with great interest. The authors concluded that the annual number of aneurysm (ruptured or unruptured) treatments (clipping or coiling) performed in the USA has leveled off, and the proportion of cases treated at ‘high volume’ centers (defined as 50+ treatments during the sampled year) has decreased. I found the conclusion not fully valid for a multitude of reasons.
Firstly, to confirm a real change (as opposed to random variability) in recent time, one is obliged to test the findings objectively with statistical trend. However, the authors did not conduct any statistical trend analysis.
Secondly, the per cent proportion of cases undergoing any treatment at a ‘high volume’ center (for any given volume cut-off) is a function of the total cases treated and the proportions treated at centers of different volumes. Thus the per cent treated at a ‘high volume’ center will decrease if the total number of cases decreases, even if the distribution of the proportion of patients at various centers remains constant, a result of centers with annual volumes close to the cut-off not meeting the cut-off due to an overall decrease in numbers. Of note, the authors found a decrease in overall volume of coiling during 2009–2010 and, therefore, a decrease in the proportion of …