The relationship between postpartum levator ani muscle avulsion and signs and symptoms of pelvic floor dysfunction

Authors


Abstract

Objective

To establish the relationship between postpartum levator ani muscle (LAM) avulsion and signs and/or symptoms of pelvic floor dysfunction (PFD).

Design

Observational longitudinal cohort study.

Setting

District General University Hospital, UK.

Population or sample

Primigravida at 36 weeks' gestation and 3 months postpartum.

Methods

Pelvic floor muscle strength (PFMS) and pelvic organ prolapse were assessed clinically using validated methods. Transperineal ultrasound was performed to identify LAM avulsion and measure hiatus dimensions. Validated questionnaires evaluated sexual function, urinary and faecal incontinence.

Main outcome measures

PFD signs and symptoms related to LAM avulsion.

Results

Two hundred and sixty nine primigravida without LAM avulsion participated and 71% (n = 191) returned postpartum. LAM avulsion was found in 21% of vaginal deliveries (n = 30, 95%CI 15.1–28.4%). Women with minor and major avulsion had worse PFMS (P < 0.038) and more anterior compartment prolapse (maximum stage 2; P < 0.024). Antenatal hiatus antero-posterior diameter on ultrasound was significantly smaller in women sustaining avulsion (= 0.011). Postnatal measurements were significantly increased following avulsion. Women with major avulsion were less sexually active at both antenatal and postnatal periods (P < 0.030). These women had more postnatal urinary incontinence and symptoms such as reduced vaginal sensation and ‘too loose vagina’. No postnatal differences were found for faecal incontinence, prolapse symptoms or quality of life. The correlation of differences in variables was only slight–fair with avulsion severity.

Conclusions

Twenty one percent of women sustain LAM avulsion during their first vaginal delivery with significant impact on signs and symptoms of PFD. As avulsion has been described as the missing link in the development of prolapse; longer term follow-up is vital.

Introduction

Pelvic organ prolapse (POP) is a common condition and the life-time risk of a woman undergoing surgery for POP is 11–20%.[1-3] Anatomical POP recurrence occurs in 40% of women in the operated compartment, with a re-operation rate of 9.7% due to symptomatic recurrence.[4] The main contributor for POP is vaginal delivery with damage to the levator ani muscle (LAM).[5, 6] A recent review has shown that this damage, diagnosed on transperineal ultrasound (TPUS) a few months following childbirth, occurs in 13–36% of women.[7] A 2.4–2.9 fold increase of anatomical cystocele recurrence has been shown in women with LAM avulsion, although not all women were symptomatic.[8, 9]

In addition to POP,[5, 6, 10] women with LAM avulsion are at risk of a reduction in pelvic floor muscle strength (PFMS)[11-13] and an increased vaginal hiatus.[14-16] Although the literature is ambiguous, these anatomical changes may lead to symptoms of pelvic floor dysfunction (PFD). Although the relationship between LAM avulsion and faecal[17-20] and urinary incontinence[21-25] has been evaluated, sexual function has not been studied in relation to LAM avulsion. We are not aware of a study in the literature that has utilised validated methods for objective and subjective assessment of PFD in relation to LAM avulsion before and after childbirth.

The aim of this study was to establish the relationship between postpartum LAM avulsion and signs and/or symptoms of PFD.

Methods

Between January 2011 and May 2012, primigravid women were invited to participate in an observational, longitudinal, cohort study to establish the prevalence of LAM defects during childbirth and to correlate these with pelvic floor symptoms and muscle strength. We invited consecutive primiparous women, to create a sample representative for the normal population. The inclusion criteria were a singleton pregnancy, maternal age >18 years, no previous history of pregnancy of more than 20 weeks' gestation, and being able to read and understand English. The recruitment process has been described previously,[26] (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]), and our study sample was representative for the local population.[26] This study was approved by the National Research Ethics Service South West London committee (REC 10/H0806/87). Clinical examination was carried out in the supine position with knees semi-flexed. Women were asked to empty their bladder prior to the examination. PFMS was assessed by digital palpation, inserting the index finger approximately 4 cm into the vagina. The strength was graded using the Modified Oxford Score (MOS), on a six-point scale (0–5),[27] for which substantial inter-rater agreement has been found.[28] POP assessment was performed using the validated International Continence Society POP-Q staging method.[29, 30] Rigorous investigator training and observation by the principal investigator (RT) minimised measurement and technique variability.

3D/4D TPUS was performed using the GE Voluson 730 system with a 4–8 MHz transabdominal curved array volume transducer, with an acquisition angle of 85°. Imaging was performed at rest, at maximum pelvic floor muscle contraction and at maximum Valsalva manoeuvre. Blind offline analysis was performed using 4D view version 10.2. The cineloops were reviewed in the midsagittal view to identify the plane of the minimal hiatal dimensions.[31] This is the minimal distance from the posterior margin of the symphysis pubis to the anterior margin of the levator plate.[28, 31] Tomographic ultrasound imaging (TUI) on maximum pelvic floor muscle contraction was used to assess the entire LAM and its attachment to the inferior pubic ramus as previously described (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted].[32] Two independent investigators (KvD and KK), blinded for findings on assessment and each other's results, analysed the TUI in postnatal scans for LAM avulsion. Consensus was reached by a third investigator (RT). Using direct visualisation, the central three slices were scored as positive or negative for LAM avulsion, scoring left and right side separately.[32, 33] The unilateral score ranged from 0 (no avulsion) to 3 (complete LAM avulsion).[33] Reliability analyses have shown excellent agreement between two raters (Cohen's kappa 0.83, 95% confidence interval [CI] 0.59–1.0) when diagnosing LAM avulsion using TUI on TPUS at maximum pelvic floor muscle contraction.[34] A summed total score for either side (0–6) was assigned and categorised as no LAM avulsion (score 0), minor LAM avulsion (score 1–3) or major LAM avulsion (score 4–6, or a unilateral score 3; Figure S1) (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted].[10, 35]

The rendered image (Figures 1 and 2) was used to perform hiatus measurements in the axial plane of the minimal hiatal dimensions. Sectional planes could be used instead, as the rendered volume is not available on all ultrasounds. Hiatus area and hiatus antero-posterior diameter (AP) were measured at rest, at maximum pelvic floor muscle contraction and at maximum Valsalva manoeuvre.[31] All measurements in the rendered image were performed by one investigator (KvD) and a test–retest series was done by a second investigator (KK).

Figure 1.

Normal antenatal levator hiatus in rendered volume, at rest. PB, pubic bone; U, urethra; V, vagina; A, anus; L, levator ani muscle.

Figure 2.

Abnormal postnatal levator hiatus in rendered volume, at rest. PB, pubic bone; U, urethra; V, vagina; A, anus; L, levator ani muscle. Arrows indicate bilateral LAM avulsion.

Previously validated questionnaires were administered to assess bowel, urinary and sexual function in the third trimester of pregnancy and 3 months postpartum. The St Mark's incontinence (SMIS) scoring system was used for faecal incontinence: a total score was calculated adding up the separate scores of frequency of faecal urgency, faecal incontinence, flatus incontinence and impact on lifestyle (range 0–24).[36, 37] The International Consultation Incontinence Questionnaire Short Form (ICIQ-SF) was used for urinary incontinence: a total score was calculated including frequency, amount that leaks and interference with everyday life (range 0–21) and urinary incontinence was defined as ICIQ-SF >0, and sub analysis was performed to assess when urine leaks.[38] ICIQ-VS was used for vaginal symptoms and sexual matters and all subscales of this questionnaire were analysed separately.[39]

Statistical analysis

Based on previous studies on LAM avulsion following childbirth, we enrolled 269 women to detect 14% incidence of LAM avulsion (including allowance for a 30% drop-out rate) with a precision of 2.5% (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]. The same cohort of primiparous women was used to establish the relationship between LAM avulsion and PFD in the present study.

All analyses were performed by definition of the three groups: no LAM avulsion (score 0) versus minor LAM avulsion (score 1–3) versus major LAM avulsion (score 4–6, or a unilateral score 3). Outcomes of digital assessment, ultrasound assessment and the validated questionnaires were analysed before and after childbirth. To assess antenatal and postnatal differences between the three independent groups, we used analysis of variance (anova), applying a post-hoc least significant difference procedure for inter-group comparison, Kruskal–Wallis test, chi-square test and Fisher's exact test where appropriate. To assess differences between the antenatal and postnatal visit for each variable, comparisons within groups were performed using paired Student t test and Wilcoxon Signed rank test where appropriate. We hypothesised that the differences between antenatal and postnatal PFD assessment would depend on the severity of LAM avulsion. To assess whether there was an association between the severity of pelvic floor dysfunction and the severity of LAM avulsion, Pearson's rho was used for normally distributed continuous data, Spearman's rank for continuous data that were not normally distributed and Kendall's tau b for categorical data. An increasing rank correlation implies increasing agreement between two variables. Correlation ranges from −1 (perfect disagreement) to +1 (perfect agreement), where 0 refers to completely independent rankings. spss version 20.0 was used (SPSS Inc., Chicago, IL, USA) and two-sided P < 0.05 were considered statistically significant.

Results

In all, 269 primigravid women participated at a median of 36 weeks of gestation (range 34–41 weeks) and 71% (n = 191) returned for follow-up at a median of 13 weeks (range 10–26 weeks). All except three women underwent examination and filled in the validated questionnaires. Two women did not undergo the antenatal examination and dropped out, and one woman declined vaginal examination at follow-up. Women who attended the 3 months follow-up had a mean age of 30.7 years (SD 5.5), the mean BMI was 25.3 (SD 5.5), the majority was from a white ethnic background (55%, n = 106) (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]. In all, 75% (n = 143) delivered vaginally and 25% (n = 48) had a caesarean section. A more detailed description of demographics and obstetric details to assess differences in the follow-up and lost to follow-up group has been described previously (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted].

Test–retest analyses for the performed measurements on TPUS in the rendered image revealed moderate–good correlation for hiatus area (ICC 0.45–0.61) and good correlation for hiatus antero-posterior diameter (ICC 0.64–0.80).

None of the antenatal women had LAM avulsion and LAM avulsion was not found after caesarean section (n = 48). The overall incidence of LAM avulsion following first vaginal delivery was 21.0% (n = 30, 95%CI 15.1–28.4%); 4.9% (n = 7, 95%CI 2.2–9.9%) for minor LAM avulsion and 16.1% (n = 23, 95%CI 10.9–23.0%) for major LAM avulsion (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]. Further analysis of the present study is based on three groups of primiparous women divided into: no avulsion (n = 113), minor LAM avulsion (n = 7), and major LAM avulsion (n = 23).

Clinical assessment

Antenatal and postnatal PFMS were lower in women with minor and major LAM avulsion (mean 3.0 and 2.4, respectively), compared with women without LAM avulsion (mean 3.6 and 3.1, respectively; P < 0.038) (Table S1). PFMS of women with no or major LAM avulsion decreased significantly following childbirth. However, the differences between antenatal and postnatal assessment did not correlate with LAM avulsion severity. No significant differences between the three groups were found on POP-Q examination performed during pregnancy. Postpartum, significantly more prolapse was found in women with no and major LAM avulsion on POP-Q assessment for points Ba (anterior compartment), C (central compartment) and Bp (posterior compartment). Women with major LAM avulsion had more anterior compartment prolapse following childbirth (P < 0.024). However, LAM avulsion severity correlated only slightly with differences between antenatal and postnatal anterior compartment prolapse. Maximum POP-Q stage was 2. Genital hiatus increased in all groups following childbirth and perineal body length decreased in all groups, without differences between groups; no correlation with LAM avulsion was found.

Ultrasound assessment

On antenatal TPUS, no difference in hiatus area measurements was found between the three groups. However, a significantly smaller hiatus antero-posterior (AP) diameter at rest was found in women who later sustained LAM avulsion (= 0.011; Table 1).

Table 1. Hiatus measurements on ultrasound assessment in women with no, minor or major LAM avulsion
 No avulsion (n = 113)Minor avulsion (n = 7)Major avulsion (n = 23)P-value
  1. Continuous variables are given as means with standard deviations (SD); categorical variables are given as numbers with percentages (%). For antenatal and postnatal variables: anova applying least significant difference if P < 0.05. Statistically significant differences between no and major avulsion (A), minor and major avulsion and no and minor avulsion, respectively. All P-values are two-sided. Pearson's ρ was used to assess the association between the difference of antenatal and postnatal variables in relation to LAM avulsion.

  2. AP, hiatus antero-posterior diameter.

  3. a

    One scan of the group could not be analysed.

  4. b

    Significant difference between antenatal and postnatal variables per group.

  5. c

    Two scans of the group could not be analysed.

  6. d

    Four scans of the group could not be analysed.

Area – rest (cm 2 )
Antenatal16.1 (SD 3.1)14.1 (SD 2.8)15.0 (SD 4.4)a= 0.15
Postnatal15.3 (SD 3.3)15.2 (SD 2.6)15.6 (SD 3.8)a= 0.91
Difference–0.7 (SD 2.7)b1.0 (SD 2.9)0.8 (SD 3.5)ρ = 0.21
AP – rest (cm)
Antenatal5.7 (SD 0.7)5.2 (SD 0.8)5.3 (SD 0.9)a= 0.011A
Postnatal5.5 (SD 0.8)5.3 (SD 0.6)5.3 (SD 0.7)a= 0.61
Difference–0.2 (SD 0.5)b0.2 (SD 0.7)Δ 0.1 (SD 0.7)ρ = 0.22
Area – contraction (cm 2 )
Antenatal12.6 (SD 2.4)a12.4 (SD 2.8)11.9 (SD 3.5)a= 0.58
Postnatal12.5 (SD 2.7)a13.9 (SD 2.7)14.2 (SD 3.2)= 0.020A
Difference–0.04 (SD 2.2)1.4 (SD 2.6)2.3 (SD 4.5)bρ = 0.32
AP at contraction (cm)
Antenatal4.7 (SD 0.7)a4.4 (SD 0.6)4.4 (SD 0.9)a= 0.25
Postnatal4.6 (SD 0.7)a4.7 (SD 0.7)4.7 (SD 0.6)= 0.81
Difference –0.1 (SD 0.5)0.3 (SD 0.5)0.3 (SD 0.8)ρ = 0.22
Area at Valsalva (cm 2 )
Antenatal20.5 (SD 6.0)d17.3 (SD 3.7)20.0 (SD 7.2)a= 0.38
Postnatal20.5 (SD 6.4)a21.0 (SD 5.6)23.7 (SD 6.5)c= 0.049A
Difference1.6 (SD 46.2)3.7 (SD 3.6)b4.1 (SD 7.2)bρ = 0.29
AP at Valsalva (mm)
Antenatal6.3 (SD 1.0)d5.6 (SD 0.9)6.0 (SD 1.1)a= 0.20
Postnatal6.1 (SD 0.9)a6.0 (SD 0.9)6.2 (SD 1.0)c= 0.76
Difference–0.2 (SD 0.8)b0.3 (SD 0.7)0.2 (SD 0.9)ρ = 0.19

Following childbirth, hiatus area decreased in women without LAM avulsion and increased in women with LAM avulsion. Postnatal hiatus area at rest was not significantly different between the three groups. Significantly larger areas were found for women with LAM avulsion on images acquired at maximum pelvic floor muscle contraction and maximum Valsalva manoeuvre. Fair correlation was found between the increase in hiatus area and LAM avulsion severity. Following childbirth, hiatus AP diameter decreased in women without LAM avulsion and increased in women with LAM avulsion. Fair correlation was found between the increase of hiatus AP diameter and LAM avulsion severity. Overall, the changes in AP diameter were less distinct than the changes in hiatus area measurements.

Subjective assessment

A significant increase in faecal incontinence was seen in women with and women without LAM avulsion, mainly due to an increase in flatus incontinence (Table S2). As such, no trend was found between faecal incontinence and LAM avulsion severity. Impact on quality of life was not different between the three groups.

Urinary incontinence score was higher in women who later sustained LAM avulsion. Women without LAM avulsion showed a significant improvement in urinary symptoms following childbirth, which was not found in women with LAM avulsion. The latter had significantly more urinary incontinence 3 months postnatal, with a higher total score. No trend was found between urinary incontinence and LAM avulsion severity. Before and after childbirth, the majority of women had urinary incontinence related to stress urinary incontinence (n = 54, 74%; and n = 32, 64%, respectively).

No antenatal differences were found for vaginal symptoms. Following childbirth, women with major LAM avulsion had a significant increase in the troublesome symptom of a reduced vaginal sensation, and the difference between antenatal and postnatal assessment correlated slightly with LAM avulsion severity. Furthermore, women with LAM avulsion had significantly more symptoms of a ‘too loose vagina’ following childbirth and the differences between antenatal and postnatal assessment correlated slightly with LAM avulsion severity. In all three groups, prolapse symptoms did not differ significantly following childbirth. Significantly fewer women who later sustained a major LAM avulsion during delivery were sexually active in the third trimester of pregnancy (39% major LAM avulsion versus 86% minor LAM avulsion versus 65% no LAM avulsion, = 0.029). Fewer women with major LAM avulsion had resumed sexual intercourse within 3 months following delivery (43% major LAM avulsion versus 100% minor LAM avulsion versus 71% no LAM avulsion, = 0.004). However, none of the vaginal symptoms reported interfered with their sex life. Overall, antenatal interference of vaginal symptoms with everyday life was significantly higher in women who later sustained a major LAM avulsion. This was not significant any longer following childbirth, and the differences between antenatal and postnatal assessment did not correlate with LAM avulsion severity.

Discussion

Main findings

This study shows the relationship between postpartum LAM avulsion and PFD using validated techniques for objective and subjective assessment. A smaller antenatal hiatus antero-posterior diameter was associated with a significantly increased risk of postnatal LAM avulsion. Less PFMS, more anterior compartment prolapse and a larger hiatus were found in women with minor and major LAM avulsion following childbirth. Women with major LAM avulsion had more urinary incontinence but no differences in faecal incontinence. Furthermore, women with major LAM avulsion had more troublesome vaginal symptoms, and had less sexual intercourse before and after childbirth.

Strengths and limitations

The strengths of this study are the prospective design and the use of validated methods to assess PFD related to LAM avulsion. We evaluated PFD objectively and subjectively, as signs and symptoms do not always correlate.[40, 41] When analysing LAM avulsion and hiatus measurements on ultrasound, the investigators were blinded to delivery details, clinical examination and each other's results. Furthermore, the statistically significant differences presented during postnatal assessment often were not related to changes between antenatal and postnatal assessment. Therefore, differences between antenatal and postnatal assessment did not correlate well with LAM avulsion severity. This highlights the merits of performing prospective studies, as we have done.

We acknowledge the limitations of this study. We could not perform a power calculation based on the validated assessment techniques, as they have not been previously used in relation to LAM avulsion and childbirth. The current sample size was based on a power calculation to detect the incidence of LAM avulsion following first delivery, which will be published elsewhere (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]. Secondly, we acknowledge that the size of the group of minor LAM avulsion was small.

Interpretation

Antenatal and postnatal PFMS were significantly less in women with LAM avulsion. The postnatal difference did not correlate with LAM avulsion severity, suggesting that there may be another mechanism responsible for the worsening PFMS. Although our psychometric properties of PFMS were good, we acknowledge that there is debate in the literature and therefore this should be taken into consideration during interpretation of this score. An increase of PFMS during pregnancy, followed by a reduction in strength postpartum, has been described previously.[42] Although worsening of POP following vaginal delivery has been demonstrated previously in prospective studies,[18, 41, 43] this was not related to LAM avulsion. Our study revealed more anterior compartment prolapse for women with LAM avulsion. Although women were asymptomatic and maximum POP-Q stage was 2, this study provides baseline data to evaluate future POP development. Previous studies in prolapse patients have revealed a reduced PFMS in women with LAM avulsion.[11, 12] Moreover, supervised pelvic floor muscle training to increase PFMS in women with POP can improve severity of prolapse and reduce prolapse symptoms.[44]

Postnatal hiatus area was significantly larger in women with major LAM avulsion (except at rest), compared with women without LAM avulsion. The explanation is two-fold. First, diagnosis of LAM avulsion appears to be most reliable on maximum contraction,[34] as LAM avulsion becomes more obvious on contraction. Secondly, as the LAM has been stretched to a greater extent when avulsion has occurred,[45] we speculate that this can lead to muscle that stretches more easily even 3 months later. However, the correlation between increase in hiatus area and LAM avulsion severity was fair. Another new finding is that women with a smaller antenatal antero-posterior diameter at rest are at greater risk of sustaining LAM avulsion during delivery. Although this is only a geometric measure, it does support the hypothesis that as LAM has to stretch more to allow passage of the fetus, this would increase the risk of avulsion from the inferior pubic ramus.[45]

Our findings on objective assessment are in accordance with the literature, where POP, enlarged hiatus and LAM avulsion are inter-related.[5, 6, 10, 14, 16, 46]

In contrast to another study performed postpartum,[18] we found that women both with and without major LAM avulsion had significantly more faecal incontinence, without differences between groups. As shown previously we expected women with major avulsions to report more faecal incontinence, as the incidence of OASIS was higher (n = 8/30) (K. W. M. Van Delft, R. Thakar, A. H. Sultan, N. Schwertner-Tiepelmann, K. B. Kluivers [accepted]. Although two studies among older women showed an association between faecal incontinence and LAM avulsion,[17, 19] another study in a tertiary referral centre showed the opposite.[20]

The incidence of urinary incontinence prior to childbirth can be explained by the physiological changes that occur during pregnancy and the load of the gravid uterus predisposing women to urinary incontinence. However, we could not explain why women that were later to sustain LAM avulsion had more antenatal urinary incontinence. Women with LAM avulsion persisted in having more stress urinary incontinence than women without LAM avulsion. Other postpartum studies found that women with LAM avulsion had more urinary incontinence, although no antenatal values were available.[21, 22] Contradictory findings have been published regarding urinary incontinence and LAM avulsion in prolapse patients.[23-25] We speculate that women with LAM avulsion have an earlier onset of stress urinary incontinence that might balance out with age.

Women with LAM avulsion were less sexually active prior to childbirth and their vaginal symptoms interfered with their everyday life. We cannot offer an explanation for this and attribute it to a coincidental finding. A recent review reports a 41–83% prevalence of postpartum sexual dysfunction, especially following instrumental vaginal delivery and perineal trauma.[47] However, vaginal symptoms and sexual function have not been evaluated previously in women with LAM avulsion postpartum. Lack of vaginal sensation has been shown to occur after vaginal delivery,[43] and is worse in women with LAM avulsion. Furthermore, these women feel that their vagina is too loose, which concurs with the enlarged hiatus found on TPUS. Women with these vaginal symptoms and associated major LAM avulsion were less likely to resume sexual intercourse postpartum. This might not necessarily be an independent factor, as women with LAM avulsion sustained more severe perineal trauma, which could explain delayed resumption of sexual intercourse. Future follow-up may show whether overall trouble with vaginal symptoms including sexual intercourse in women with LAM avulsion persists.

Previous studies have shown that minor LAM avulsion behaved similar to no LAM avulsion.[33] However, differences between antenatal and postnatal objective and subjective assessment were not obviously associated with LAM avulsion severity. We therefore can not draw conclusions regarding the impact of minor LAM avulsion. Furthermore, the minor LAM avulsion group was small, as stated in the limitations, which hampers conclusions in this respect.

A smaller antenatal hiatus antero-posterior diameter could be a predictor of LAM avulsion, although this requires further evaluation in future studies. The presented changes in objective and subjective assessment can be incorporated in the management of women with LAM avulsion. We can now target women with LAM avulsion and provide intensive lifestyle modification, education and pelvic floor muscle training.[44] Together with weaker pelvic floor muscles, we found more anterior compartment prolapse and larger hiatus in women with major LAM avulsion. We also found more vaginal symptoms and urinary incontinence in women with LAM avulsion. These symptoms can be addressed during postnatal counselling.[47] Longer term follow-up is vital to establish the pattern of resolution, on-going and de novo symptoms of PFD. Furthermore, as LAM avulsion has been described as the missing link in the development of POP, we need to establish whether women with LAM avulsion progress to overt prolapse, and therefore longer term follow-up is planned.

Disclosure of interest

All authors declare no conflicts of interest.

Contribution to authorship

KWMvD: Project development, recruitment, data management, data analysis, manuscript writing, approved the final manuscript. AHS: Protocol development, Project development, manuscript editing, approved the final manuscript. RT: Protocol development, project development, data analysis, manuscript editing, approved the final manuscript. NS-T: Protocol development, recruitment, approved the final manuscript. KBK: Data analysis, manuscript editing, approved the final manuscript.

Details of ethics approval

This study was approved by the National Research Ethics Service South West London committee (REC 10/H0806/87) on 17 November 2010.

Funding

KvD was funded by the Mayday Childbirth Charity Fund.

Acknowledgement

We would like to thank J. IntHout, Statistician from the Radboud University Medical Centre, for her help with the statistical analyses.

Mini commentary on ‘The relationship between postpartum levator ani muscle avulsion and signs and symptoms of pelvic floor dysfunction’

  • HP Dietz

  • Sydney Medical School Nepean, Nepean Hospital, Penrith, Australia

In this issue of BJOG, Van Delft et al. from Croydon University Hospital provide the best data yet on prevalence and short-term clinical relevance of pelvic floor muscle trauma in the UK. They use methodology that has become the de facto standard in this field. Their definition of avulsion using a magnetic resonance (MR) method is less well validated than the alternative tomographic ultrasound method, which has been validated against symptoms, signs and MR diagnosis (Dietz et al. Int Urogynecol J 2011; 22: 699-704; Zhuang et al. Am J Obstet Gynecol 2011; 205: 232.e231–8). Fortunately, the authors corrected for the main weakness of the MR scoring system by rating a complete unilateral avulsion as ‘major avulsion’. This allows comparisons with other publications, as ‘major LAM [levator ani muscle] avulsion’ is almost fully equivalent to ‘full’ or ‘complete avulsion’ in the tomographic scoring system.

Van Delft et al. found no avulsions in women who had not delivered vaginally, consistent with the literature, as is the prevalence of 16.1% for complete avulsion. The paper is focused on symptoms and signs of pelvic floor dysfunction and presents interesting data, although it may be prudent not to attach too much importance to results due to the relatively small data set and the large number of statistical tests performed.

As to be expected, and consistent with the literature, there was increased pelvic organ descent and increased hiatal dimensions in women with avulsion. As regards symptoms, Van Delft et al. found an increased prevalence of stress urianry incontinence (SUI) in women with avulsion. While this is consistent with other postnatal data, this relationship seems absent in middle-aged and elderly women (Dietz, et al. Int Urogynecol J 2009; 20: 967–72). Findings regarding vaginal sensation are also in agreement with previously published data (Thibault- Gagnon et al. Int Urogynecol J 2012; 23: S183–5).

The last 8 years have seen the publication of about a dozen perinatal imaging studies from around the world, demonstrating that major levator ani muscle trauma is common after childbirth. We have also learnt that such trauma is associated with female pelvic organ prolapse and recurrence after prolapse surgery. In addition, and not the least due to the efforts of the Croydon University Hospital group, we know that trauma to the anal sphincter is much more common than previously thought, and that it often has long-term consequences for the parturient's future, just like levator trauma.

It is time to consider maternal birth trauma a key performance indicator of obstetric services. The long mean latency between trauma and clinical manifestations such as prolapse and faecal incontinence (Thomas et al. Ultrasound Obstet Gynecol 2013; 42: 39) is not an excuse for ignoring it. Unfortunately, virtually all levator trauma and probably about 80% of all anal sphincter trauma escapes detection at birth. Attempts at improving the performance of obstetric services in this regard will require routine postnatal follow-up by imaging. This would enable early intervention, which is seen as essential in the treatment of musculoskeletal injuries. Surely our mothers deserve no less.

Disclosure of interests

H.P.D. has accepted unrestricted educational grants from GE Medical.

Ancillary