Figure 2

A simplified, hypothetical example of a forest (meta-analysis) plot. This figure represents a simplified, hypothetical example of a forest (meta-analysis) plot demonstrating eight RCTs comparing laparoscopic versus open appendectomy with respect to postoperative wound infections. Each RCT is represented by a square (the odds ratio found for this trial) and a horizontal line, which represents the 95% confidence interval. If the square is to the left of the vertical line of no effect (odds ratio = 1, e.g. studies 1, 2, 3, 5, 6, and 8), the study favors laparoscopic appendectomy; if the square is to the right of the line (e. g. studies 4 and 7), then open appendectomy is favored. If the 95% confidence interval crosses the line of no difference (odds ratio = 1), then the trial is not statistically significant (e.g., studies 1, 2, 3, and 6). Conversely, if the 95% confidence interval does not cross the line of no effect (odds ratio = 1), then this trial yields a statistically significant difference. Studies 4 and 7 found a significant advantage in favor of open appendectomy, whereas studies 5 and 8 found significantly less wound infection in the group randomized to laparoscopic appendectomy, indicating considerable heterogeneity. The size of the squares varies with respect to the sample size of each individual trial: the larger the sample size, the larger the square will be. An overall (pooled) effect is represented by the diamond. In this case, the overall results demonstrate a statistically significant reduction of postoperative wound infections in the group randomized to laparoscopic appendectomy (Forest plot created with StatsDirect v. 2.7.2; StatsDirect Ltd., Cheshire, UK).