Randomised controlled trial comparing early home biofeedback physiotherapy with pelvic floor exercises for the treatment of third-degree tears (EBAPT Trial)




To compare early home biofeedback physiotherapy with pelvic floor exercises (PFEs) for the initial management of women sustaining a primary third-degree tear.


Single centre, randomised trial.


National Maternity Hospital, Dublin, Ireland.


A total of 120 women sustaining a primary third-degree tear.


Women were randomised in a one to three ratio: 30 to early postpartum home biofeedback physiotherapy and 90 to PFEs.

Main outcome measures

Differences in anorectal manometry results, Cleveland Clinic continence scores and Rockwood faecal incontinence quality of life scale scores after 3 months of postpartum treatment.


The mean anal resting pressure was 39 ± 13 mmHg in the early biofeedback physiotherapy group and 43 ± 17 mmHg in the PFE group. The mean anal squeeze pressure was 64 ± 17 mmHg in the biofeedback group and 62 ± 23 mmHg in the PFE group. There was no significant difference in anal resting and squeeze pressure values between the groups (= 0.123 and = 0.68, respectively). There were no differences in symptom score and quality of life measurements between the groups.


This study demonstrates no added value in using early home biofeedback physiotherapy in the management of women sustaining third-degree tears. Poor compliance may have contributed because women found it difficult to designate time to using biofeedback.


Women who sustain a third-degree anal sphincter tear are at increased risk of subsequent faecal incontinence.[1] Previous work has highlighted specific risk factors that predispose to third-degree tears[2, 3] and the optimum method for their repair.[4] There is less evidence concerning optimum treatment following repair of the tear and long-term follow up.

Biofeedback physiotherapy is a technique that uses external equipment to demonstrate and alter physiological events, with either visual or auditory biofeedback. It results in good initial symptomatic improvement[5, 6] that is maintained for at least 2 years following treatment of both anal[7, 8] and urinary[9] incontinence. Previous work from this group has demonstrated that intra-anal electromyographic biofeedback augmented with electrical stimulation is superior to transvaginal sensory biofeedback alone in women with altered faecal continence following obstetric anal sphincter injury.[10] Following on from that study, the group demonstrated that intra-anal biofeedback physiotherapy alone was associated with improved anal squeeze pressure, continence and quality of life in women with altered postpartum faecal continence, with no further benefit gained with the addition of electrical stimulation (augmented biofeedback).[11]

At the National Maternity Hospital, Dublin, women who sustain a third-degree tear are managed with immediate repair, oral antibiotics, oral analgesics and stool softeners. Women are educated in pelvic floor exercises (PFEs) and followed up in a dedicated Perineal Clinic at 3 months postpartum. Symptomatic women and those with abnormal anal manometry scores are referred for home biofeedback physiotherapy.

This study tested the hypothesis that a group of women who had sustained a third-degree tear and were treated with home intra-anal biofeedback physiotherapy would have superior anal manometry results at 3 months postpartum than a group treated with PFEs alone. As a secondary outcome, an improvement in continence scores in the home biofeedback physiotherapy group was anticipated.


The trial was registered as the Early biofeedBAck PhysioTherapy versus pelvic floor exercises in women who sustain third-degree tears (EBAPT) Trial, ISRCTN70419152. Primiparous women, fluent in English, who sustained a primary third-degree tear were included whereas women with an infant in the special care unit, a history of alcohol or illicit drug abuse, a positive viral status (hepatitis virus, HIV) and not fluent in English were excluded. The diagnosis of a third-degree tear was made by a Registrar (post-MRCOG) or Consultant and when suspected, senior assistance was always called upon. Endo-anal ultrasound was not routinely used as it adds little in the immediate postpartum period.

Women who met the inclusion criteria were identified on the labour ward on the day of delivery and approached for recruitment. Information regarding the trial was given and women were re-approached 24 hours later when informed consent to register for the trial was obtained. Women who declined participation were given a standard follow-up appointment for 3 months postpartum in the Perineal Clinic as per hospital policy. Recruited women were randomised using sealed, opaque envelopes generated from a computer randomisation program with 30 women randomised to early biofeedback physiotherapy and 90 women randomised to pelvic floor exercises (1:3 randomisation).

A total of 145 women who had sustained a primary third-degree tear were approached (February 2007 to December 2009) for recruitment to the trial, of whom 120 were randomised to either early postpartum home biofeedback physiotherapy (n = 30) or conventional PFEs (n = 90) (Figure 1). All underwent immediate repair of the identified tear and were prescribed a 5-day course of oral antibiotics, oral analgesics and stool softeners. The repair was performed by Registrars or Consultants. All Registrars were trained specifically in sphincter repair at the commencement of their tenure by Consultants with a specific interest in sphincter injury and pelvic floor repair.

Figure 1.

CONSORT diagram for the EBAPT trial.

Women randomised to early biofeedback physiotherapy were educated on the use of the portable biofeedback machine (CombiStim XP, Neurotech®, Galway, Ireland) in the Perineal Clinic before discharge. Women randomised to PFEs were educated on the ward before discharge. The protocol for each intervention was to perform two work periods each day. Biofeedback training was undertaken in the Perineal Clinic and was delivered by either a Specialist Obstetrician (CM, MF) or a Specialist Nurse (MC), often a combination of both. All questions were answered and a successful work period was achieved. Written information was not provided. PFE education was given by senior postpartum midwives or physiotherapists on the ward and written information was provided.

Follow up was carried out 3 months postpartum in the dedicated Perineal Clinic where anorectal manometry, endoanal ultrasound, Cleveland Clinic continence score[12] and Rockwood faecal incontinence quality of life scale[13] were all completed. All postpartum investigations were carried out by the same staff using the same equipment. Subgroup analysis of third-degree tear type (a, b or c) was considered but the numbers in the trial were insufficient to allow for this. The authors believe that anal continence at 3 months postpartum is reflective of long-term continence.

The biofeedback machines were programmed to the electromyography (EMG) setting alone. The work period was set to ten contractions, each of 5 seconds duration, with a ten-second rest between each contraction. The EMG target setting (threshold) was set to automatic and visual biofeedback was achieved when this was reached. Women were educated on how to turn the machine on and off, how to place the intra-anal probe (in the left lateral position) and how to clean and maintain the probe. The PFE protocol was to perform standard Kegel exercises for a 5-minute period. Women in both groups were requested to complete two work periods each day for the 3-month period before their scheduled Perineal Clinic appointments and to keep a diary regarding adherence to the protocol. Poor adherence was defined as performing less than 70% of the intended work periods. If a woman did not attend their scheduled Perineal Clinic appointment they were contacted by telephone and were re-scheduled for the following week's Clinic.

The power calculations were based upon the mean manometry values from the Perineal Clinic using a standard deviation of 16 mmHg derived from a previous publication.[11] The hypothesis was that women receiving home biofeedback physiotherapy would have a mean anal squeeze pressure value 13 mmHg greater than those performing PFEs alone, giving the study a power of 87%. A one to three randomisation was used as there were only five biofeedback machines available to randomised women and so a one to one or one to two randomisation would have led to the trial being of longer duration. The data were analysed using the Mann–Whitney U test and an unpaired Student's t test. Statistical analysis was performed using SPSS software (SPSS® version 12, SPSS Inc., Chicago, IL, USA).


Demographics concerning use of epidural anaesthesia (= 0.123, Fisher's exact test), use of episiotomy (= 0.145, Fisher's exact test) and length of second stage of labour (= 0.537, Student's t test) were similar between both groups, although those in the biofeedback group delivered heavier babies (= 0.039, Student's t test) (Table 1). All women were educated in their relevant treatment before postpartum discharge, with all women in the biofeedback group completing a successful work period.

Table 1. Group demographics
Epidural (%) 6271
Episiotomy (%) 5750
Total duration (minutes)364434
Second stage duration (minutes)6259
Mode (%)   
Birthweight (kg) (range) 4.033.82

A complete follow up was achieved and all recruited women attended the Perineal Clinic for their 3-month postpartum appointment. No women in the biofeedback group reported any adverse events regarding use of the machine. The mean anal resting pressure was 39 ± 13 mmHg in the early biofeedback therapy group and 43 ± 17 mmHg in the PFE group. The mean anal squeeze pressure was 64 ± 17 mmHg in the biofeedback group and 62 ± 23 mmHg in the PFE group. There was no significant difference in anal resting and squeeze pressure values between the groups (p = 0.123 and p = 0.68, respectively, unpaired Student's t-test) (Figure 2). There were no incomplete repairs identified on endoanal ultrasound—all women had evidence of external anal sphincter scarring, but none had major defects requiring consideration of a delayed repair or referral to a colorectal surgeon. No undiagnosed fourth-degree tears were present at 3-month follow up.

Figure 2.

Scatter plot of anal resting and squeeze pressures of all women (blue dots, women performing PFE; green dots, women given biofeedback physiotherapy).

There was no difference in the Cleveland Clinic continence score (0 = full continence, 20 = complete faecal incontinence) between the two groups (= 0.88, Mann–Whitney U test, Figure 3). Similarly, no difference was demonstrable in Rockwood faecal incontinence quality of life scale across all domains: lifestyle (= 0.29), coping (= 0.27), depression (= 0.89) and embarrassment (= 0.51) (Mann–Whitney U test, Figure 3).

Figure 3.

Mean Wexner symptom scores and Rockwood faecal incontinence quality of life scale scores at 3 months postpartum (Mann–Whitney U test).

In the biofeedback physiotherapy group, seven of 30 women reported poor adherence. Further analysis of manometry values (between women with good adherence to the protocol and those with poor adherence) demonstrated no significant difference in either anal resting or anal squeeze pressures (= 0.22 and = 0.24, respectively, paired Student's t-test).


Main findings

This study investigated whether early postpartum incontinence might be ameliorated by an intervention. It demonstrated no significant difference in either anal squeeze or resting pressures at 3 months postpartum in women who sustained a third-degree tear and were managed with home biofeedback physiotherapy compared with those managed with PFEs alone. No difference was seen in the symptom scores and quality of life scores between the two groups. The work presented here focused on an area not previously studied: the use of immediate postpartum biofeedback physiotherapy, i.e. akin to using biofeedback as a prophylactic measure, whereas biofeedback has traditionally been used only as a therapeutic measure.

Strengths and weaknesses

There was a complete follow up (i.e. no dropouts) and the interventions employed are both well proven. Clear and readily reproducible protocols were followed for each intervention. Adherence was a major issue as seven of 30 women did not use the machine as per the trial protocol. A lack of time to use the machine and an unwillingness to use it in the absence of symptoms of faecal incontinence were the reasons offered by women with poor adherence to the protocol. Two work periods per day were chosen as it was felt to be pragmatic and was in keeping with the daily time allocated to biofeedback in previous referenced trials from the group. It is possible that participants may have benefited more had they undertaken more biofeedback training, but the protocol instituted was felt to be reasonable and would result in good compliance in women with recently repaired sphincter injuries and newborn infants at home. It may be that a longer or more intense period of immediate biofeedback might have resulted in significant manometric or quality of life differences; however, the present study suggests a clinically significant difference in outcome would be highly unlikely. It has also been shown that home biofeedback is as efficacious as hospital based treatment[9] so the current protocol was valid. All women who sustain a third-degree tear at the National Maternity Hospital are currently followed up at 3 months postpartum with further routine appointments precluded due to current workload and resources. Only those with ongoing or worsening symptoms at initial follow up are routinely reviewed.

There were no antepartum data collected on the women subsequently randomised so it is not possible to comment on improvement or deterioration in continence or manometric results over the trial period. However, as previous studies have shown a benefit in the use of postpartum biofeedback physiotherapy in symptomatic women,[10, 11] it appeared justified to specifically investigate whether immediate biofeedback would result in a significant increase in anal manometric pressures postpartum. Almost a quarter of the biofeedback group reported poor adherence, although there was no demonstrable manometric effect at per protocol analysis. It must be assumed that there was also a variation in the use of the machine among women with good adherence. The cost of each machine is approximately € 700 and this must also be factored in if a potential change to the treatment algorithm were to be considered.


Obstetric injury to the pudendal nerve is known to be one of the principal causes of faecal incontinence in women.[2] Previous work in animal models has demonstrated a reduction in somatosensory evoked cortical potentials following injury to the nerve supply of the anal sphincter complex[14] and atrophy of external anal, internal anal and external urethral sphincters is evident following simulated childbirth injury.[15] There is a wealth of evidence that early neuromuscular (re-)training leads to improved functional outcomes.[16] Hence, early biofeedback physiotherapy may improve both motor and sensory function and lead to an increase in cortical awareness of the sphincter complex.

It has previously been shown that the risk of obstetric anal sphincter injury is greatest at the time of first vaginal delivery.[17] Women with documented transient faecal incontinence or a persistent alteration in faecal continence following their first delivery are more likely to develop worsening symptoms or frank faecal incontinence following a subsequent, i.e. second, vaginal delivery. Therefore, it is essential to maximise postpartum faecal continence in all women sustaining anal sphincter damage, with immediate biofeedback physiotherapy a potential option, although no demonstrable benefit has been shown in the current study.


There is no evidence from this study that routine use of early biofeedback physiotherapy should be introduced immediately following diagnosis and repair of third-degree tears so current management protocols should not be altered. The study was not designed to selectively treat those women who were symptomatic and the possibility remains that early intervention in this group might be therapeutic; hence, a trial focusing specifically on symptomatic women in the early postpartum period would be beneficial. Biofeedback physiotherapy remains a treatment option for symptomatic women at 3 months postpartum.

Disclosure of interests

The authors have no conflicts of interests to disclose.

Contribution to authorship

Acquisition, analysis and interpretation of data and drafting of article was done by CP. CM and MF helped with conception and design, acquisition of data and revision of the article. MC helped with acquisition and analysis of data and drafting of article while LD helped with analysis and interpretation of data and revision of the article. Both PRO and CO contributed to conception and design, interpretation of data and revision of article.

Details of ethics approval

The study was approved by the Ethics Committee of the National Maternity Hospital, Dublin, Ireland on 21 January 2007.


The work was funded by a grant awarded from the Health Research Board (HRB), Ireland.


The authors wish to acknowledge the staff of the Physiotherapy Department and the Perineal Clinic.

Commentary on ‘Randomised controlled trial comparing early home biofeedback physiotherapy with pelvic floor exercises for the treatment of third-degree tears (EBAPT Trial)’

Continued attention on helping women who experience a perineal tear is welcome. Although guidelines have considerably improved immediate management, the long-term results and natural history of symptoms after a tear are still unclear. We need to know how to prevent symptom development. However, this paper does not inform prevention strategies. Taking women immediately after sphincter repair, with unknown symptom status prepregnancy, during pregnancy or postnatally and intervening for only 3 months with no longer term follow-up predictably led to no significant results.

It is difficult to imagine women being keen to use an intra-anal device immediately after a repaired tear. The distraction of a new baby and likely soreness from a perineal wound could be seen as major disincentives to using an anal probe, or even to exercising: addition of a compliance monitor to the biofeedback machine would have been useful. The study compares the instruction to exercise with the instruction to use a biofeedback machine; self-reported compliance was 70%+ in most of the biofeedback group and unreported in the exercise group.

The outcome measures used are anal pressures as a proxy for symptoms, but with no baseline for comparison and no normal range quoted to enable judgement of improvement or whether the intervention brought women into likely continent ranges. Using the Wexner faecal incontinence score for an essentially asymptomatic cohort (where a score of 1 equates to accidental loss of flatus less than once a month) within normal population ranges, it is unsurprising that no difference was found between the groups. Likewise, for faecal incontinence quality of life: if women have no symptoms then faecal incontinence quality of life is not relevant. Asymptomatic women had limited capacity to benefit on these outcome measures from either intervention. Participants were not followed long enough to see if they developed symptoms. The time for comparing these women will be in future decades when they encounter menopause, obesity, frailty or other comorbidities that typically trigger symptoms.

Happily the faecal incontinence rate in this study was low, although results presented as mean scores may include a few with significant symptoms. However, some women have very low resting and/or squeeze pressures after intervention (see figure 2). We do not know if these women had symptoms nor how to prevent them from developing later symptoms.

The authors conclude that there is no added benefit from biofeedback, but the interventions chosen for comparison were very different: different therapist, different instructions and different exercise times (2 minutes 30 seconds versus 5 minutes) between groups. If a difference had been found, it would have been impossible to tell if the addition of biofeedback had made the difference. We need well-controlled trials of pelvic rehabilitation exercises, varying one element at a time. Comparing different complex interventions tells us little, whether or not interventions are effective. With no true control group who did nothing, the natural history is unknown: neither intervention may be needed. The authors rightly state that we do not know how best to manage women following repair. That remains the case following this study.

Disclosure of interest

I declare no conflict of interest in publishing this commentary.

  • C Norton

  • Imperial College Healthcare NHS Trust & King's College London, London, UK

Journal club


On the postnatal ward, a patient who had a repaired third-degree tear asked if she should purchase a portable biofeedback machine to help her recovery. As her attending physician, what would be your recommendation?

Structured question

ParticipantsPrimiparous women who sustained primary third-degree tears during childbirth
InterventionEarly biofeedback physiotherapy (using a portable biofeedback machine)
ComparisonPelvic floor exercise education

Primary: differences in anorectal manometry at 3 months postpartum

Secondary: Cleveland clinic continence scores and Rockwood faecal incontinence quality-of-life scale scor

Study designRandomised controlled trial

Discussion points

  1. Objective and background: What are the long-term outcomes of women who sustained obstetrics anal sphincter injuries (see suggested reading)?
  2. Methods: What are the differences between the internal validity and the external validity of a clinical trial? How does the lack of blinding in this trial affect its validity? Are there alternative methods to address this issue?
  3. Statistics: Do you agree with the method of power calculation used in this study (see suggested reading)?
  4. Do you agree with the authors' belief that ‘anal continence at 3 months postpartum is reflective of long-term continence’?
  5. Results: What are the implications of poor adherence observed in this study?
  6. Take-home message: In a single sentence, how would you describe the available evidence on early biofeedback physiotherapy in the prevention of complications after obstetric anal sphincter injuries? (Data S1)

Suggested reading

Bland JM. The tyranny of power: is there a better way to calculate sample size? BMJ 2009;339:b3985.

Fornell EU, Matthiesen L, Sjödahl R, Berg G. Obstetric anal sphincter injury ten years after: subjective and objective long term effects. BJOG 2005;112:312–16.

  • EYL Leunga & D Siassakosb

  • aWomen's Health Research Unit, Centre of Public Health and Primary Care, Queen Mary, University of London, London, UK, Email: elaine.leung@qmul.ac.uk

  • bSchool of Clinical Sciences, Southmead Hospital, Bristol, UK

A complementary slide set for this journal club accompanies the online version of the paper.

Join us at #BlueJC: You can collaborate with us to host a #BlueJC session. Find out more on our Journal Club page at www.BJOG.org. Follow @BJOGTweets to stay updated on scheduled #BlueJC sessions. For background of #BlueJC, please refer to BJOG 2013;120:657–660. http://bit.ly/10VaiRZ