In the article by Harris et al,1 I was struck by the data for Swan-Ganz (S-G) catheterization. In their Table 3, this item appears with a significance level of P=.03. This is not significant because of the number of comparisons made, as properly pointed out by the authors. However, considering the Mid America Heart Institute–Cardiac Care Unit (MAHI-CCU) score weight of 3 for S-G catheterization and the sheer differential number (n=49, 9.4% of the usual care group), it struck me that this may be a possible source of the reported group difference.
Price JM. Does Prayer Really Set One Apart?. Arch Intern Med. 2000;160(12):1873. doi: