Dr Barrie1 states that the study contains a type II statistical error that makes the findings "meaningless." A type II error occurs if it is claimed that no difference exists when, in fact, there are true differences. The most common reason for a type II error is insufficient numbers within the study. We generated an estimate of trial numbers to have a reasonable expectation of avoiding a type II error. We based this estimate on an expected rate of wound infection of 10%. In this study, the total wound infection rate proved to be 4.3%; therefore, we recognized that, although no significance in frequency of major infections was seen between the 3 test groups, we would be unable to state that no true difference existed unless we studied many more patients than was feasible. Thus, we were careful to avoid a type II error.
However, significant differences
Paul E. O'Brien, Simon Woods, Gillian Anderson. Rational and Appropriate Use of Prophylactic Antibiotics-Reply. Arch Surg. 1997;132(3):321. doi:10.1001/archsurg.1997.01430270106026