LAUSR

1 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Dear Editors: We congratulate van Engelenburg-van Lonkhuyzen et al on their study on the effectiveness of pelvic physiotherapy in children with functional constipation. This randomized, controlled trial showed that pelvic physiotherapy was more effective than standard medical care in the treatment of children with constipation after 6 months of follow-up. This is the first study exploring a promising nonpharmacologic treatment for functional constipation in children with a longer follow-up. The results are in line with an earlier study by Silva et al that showed a significantly positive effect of a 6-week physiotherapy program on defecation frequency. Unfortunately, van Engelenburg-van Lonkhuyzen et al could include only 53 of the targeted 367 children, which means that their findings need to be interpreted with caution. Owing to the small sample size, there is the danger of a type I error, namely, that a statistically significant difference between the 2 treatment groups would not have been found if the sample size had been as calculated before the start of the study (false-positive result). The authors report the chance of a type I error is small because all primary and secondary outcomes, with the exception of the SDQ, showed significant results favoring pelvic physiotherapy, the dropout rate in the trial was lower than expected, and the absolute risk reduction in the trial was 30% instead of 15% used in the power calculation. However, small and underpowered studies can only detect significant effects that happen to be large. Therefore, even when the significant effect found in this study is not a false-positive result, it is likely the estimate of the magnitude of the effect is exaggerated (winner’s curse). The authors showed that 92.3% of the children undergoing pelvic physiotherapy and 63% of the children receiving standard medical care had been treated successfully after 6 months. This means an absolute risk reduction of 29.3% with a 95% confidence interval (CI) of 0.07%– 48.8%. The wide CI shows a lack of precision in estimating the effect of the intervention. This means that the possibility that the effect will be less than the 15% risk reduction the