We were interested to read the article by Linden et al (2017) that claims to provide evidence to "support the safety of omitting a pelvic examination in women with a confirmed intrauterine pregnancy," (p. 833).1 In light of previous studies questioning the utility of the pelvic exam, this trial addresses a critical question for patients with threatened abortion. We commend the authors for conducting the largest randomized trial on pelvic exams to date, but several methodological issues raise questions about whether clinicians can make reliable inferences from this study.First, the implemented experiment does not seem to have adequate statistical power to detect equivalence. The results suggest the intervention (no pelvic exam) reduced "composite morbidity" by -0.024, with a 90% confidence interval of (-0.118, 0.071). This is a noisy estimate, and the 90% interval is not contained inside the chosen equivalence range (-0.08, 0.08). This does not provide evidence of equivalence.Although the study's power calculations suggest 0.80 power with 720 participants, only 221 of the 1,280 patients deemed eligible for participation were randomly assigned. Our simulations (Figure 1), suggest the idealized study would detect equivalence about 84% of the time, whereas the implemented study never would. Why does the failure to detect equivalence support the safety of omitting the pelvic exam?Although the authors suggest this result supports omitting a pelvic exam, the power of the reported experiment was approximately zero. Although reported power calculations based on 720 participants suggest 0.80 power, only 221 of the 1,280 patients deemed eligible for participation were randomized. As illustrated in Figure 1, the implemented experiment (N = 221) would fail to detect equivalence in all of the 10,000 simulations we conducted.As shown in Figure 1, the equivalence range determines the decision rule for the test -we conclude equivalence if and only if the 90% interval falls inside this range. The wider the equivalence range, the more likely we are to declare equivalence. The choice of this margin is widely recognized as one of the most important decisions in the design of an equivalence study [2][3][4] . Why was an 8% change from a 15% baseline chosen as the appropriate equivalence range?