Authors’ Reply

  1. Top of page
  2. Authors’ Reply
  3. References


We appreciate Dr Ismail’s letter with its constructive comments and suggestions.1

We agree that the early period after a traumatic event such as childbirth is an interesting period to assess the acute effects of delivery on the pelvic floor. Ultrasound assessment allows us to evaluate the pelvic tissue immediately after acute trauma, especially before the onset of tissue transformation. Hence, the possibility of encountering artefacts related to remodelling mechanisms is low. Tunn et al.2 demonstrated a change in the magnetic resonance imaging signal intensity of levator ani muscle (LAM), a decrease of 27% of urogenital hiatus area and biochemical changes of pelvic floor muscle from the postpartum period to 6 months. In the minimal criteria for sonographic diagnosis of LAM avulsion, Dietz et al.3 defined it as ‘abnormal insertion’, rather than ‘absence of muscle’, taking into account the possibility of seeing scar tissue connecting the retracted muscle to the sidewall. Moreover, two cases of ‘healing’ of levator trauma after 2–3 years have been described.4

Despite this, we are also aware of the possible increase of bias in the evaluation of LAM at this time. As stated by Dr Ismail,1 LAM denervation and reduction in pelvic floor muscle strength postpartum (as the result of pain caused by episiotomy, perineal lacerations or caesarean wound) affect the extent of muscle shortening during contraction. Consequently, we evaluated the LAM at rest, lacking the dynamic volumes, during which the avulsion seems to be more defined.5 Additional potential explanations for the higher rate of LAM defects and the unexpected LAM abnormalities after emergency cesarean section have been discussed in the original article.

We concur with Dr Ismail’s statement regarding the role of labour in the aetiology of LAM damage: the publication of our counterintuitive data aimed to stimulate potential new hypotheses regarding the occurrence of LAM trauma postpartum, which need to be confirmed or contradicted, as well as the involvement of the second stage of labour.

A total of 292 women had their first delivery in the study interval: 157 (53%) were included in this trial. This number is reasonably consistent with the high number of foreigners who live in Germany and who have insufficient knowledge of German or English. During this time, we performed 111 (38%) caesarean sections, 75 (67%) elective and 36 (32%) emergencies, which was consistent with the sample rate included in our trial.

To determine the sample size necessary to test the study hypothesis, we considered the minimal difference between the occurrence rate of LAM trauma after vaginal delivery and caesarean section reported in the literature available at the time of our study. We assumed a 36% occurrence rate of LAM trauma after vaginal delivery.6 Avulsion of LAM after caesarean section had not been previously described.

Hence, the sample size calculation was based on detecting a difference in LAM avulsion occurrence rate from 36% to 0. Assuming a two-tailed α of 0.05, 37 women per group would provide 95% power to detect this difference. Although 157 patients might be regarded as a relatively small number, it was double the required sample size.

The subgroup comparisons (see Supporting information, tables S1 and S2, in original paper7) were carried out to give descriptive additional information regarding this population.

Median and interquartile range for continuous data that did not fit a normal distribution were provided in tables 1 and 2, and S1 and S2, in our original paper.7


  1. Top of page
  2. Authors’ Reply
  3. References