In Reply:—
We take this opportunity to thank Dr. Beach for his comments and showing keen interest in our recent article.1Fisher exact test indeed returns exact one- and two-tailed P values for a given frequency table, whereas the chi-square test of independence, which is also used in such situations, provides an approximation. However, it is an accepted practice to use the chi-square test to analyze the data in two-by-two contingency tables to test the null hypothesis when the sample size is sufficiently large and there is no imbalance in the size of the groups. Otherwise, the exact probability test devised by Fisher, Irwin, and Yates is used.2Because the groups were similar in size and none of the cells had an expected frequency of less than 5, the Fisher exact test was not used to analyze the difference in outcomes between the two groups and was accordingly not provided in the article. The Fisher exact test was used to calculate the difference between the groups in other categorical data. We retrieved the files of the four patients who were excluded because surgery was not extensive to find out that one patient had a wound infection and another patient had a urinary tract infection, both of which were from the liberal group. As correctly suggested by Dr. Beach, although the number of patients with complications (which was defined as the primary end point of the study) was significantly lower in the restrictive group compared with the liberal group, no significant difference was observed in the overall rate of complications (P = 0.21, 95% confidence interval, 0.05–0.37). Dr. Beach also commented that advanced statistical methods may have been used to model the data and explain the impact of relevant covariates. We had, in fact, considered using logistic regression. The decision to avoid complex modeling stemmed from the number of patients included in the study, which would have limited the number of covariates to be included in the model. We thought that such analyses would be more appropriate in studies including a greater number of patients.
*Hadassah Hebrew University Medical Center, Jerusalem, Israel. idit_matot@yahoo.com
