Pelvic floor muscle exercises are suggested as a method to improve pelvic muscle control in the antenatal and postnatal period (Dinc et al, 2009; Bo and Haakstad, 2011; Langeland-Wesnes and Lose, 2013), and are defined as the repetitive contraction of the pelvic muscles performed with an intent to strengthen, increase endurance and coordinate muscle activity, in order to prevent urinary incontinence (Hay-Smith et al, 2008). During pregnancy, pelvic floor exercises may help to counteract the pressure caused by the fetus, and the increased laxity of ligaments in the pelvic area (Hay-Smith et al, 2008). It is recognised that this information is provided by midwives at booking as part of their role in public health. The National Institute for Health and Care Excellence (NICE) (2008) recommends that pregnant women are advised about pelvic floor exercises at their first booking appointment; however, there is no data collected that would indicate if this advice takes place in a standardised way locally, nationally or internationally.
There is consensus that approaches to public health combine science and art, and definitions include phrases linked to protecting health and wellbeing, preventing ill-health, and prolonging life via the efforts of a supportive community (Faculty of Public Health, 2010; Lueddeke, 2016). Theories associated with how to promote health originate from paternalistic or facilitative perspectives (Lueddeke, 2016). The facilitator role of the midwife has also been recognised by policy makers and midwives themselves; for example in a study by Hunter et al (2015: 44), one midwife comments that, ‘delivering information is to deliver it in a non-biased way, giving them (women) informed choice’. There is no doubt that midwives have an important role in public health promotion; however, public health messages can be complicated by issues such as limited time, training and resources, or midwives feeling that they ‘bombard’ women with a ‘wall of information’ (Sanders et al 2016: 257). Further limitations influence public health messages when strategic prioritisation means that midwives focus on the essential aim to reduce maternal and neonatal death—such that prioritising information about sepsis, flu or streptococcus may overshadow information related to pelvic floor exercises (Hunter, et al 2015; Knight et al, 2016) when time is limited and there are reduced staffing levels. In this context, morbidity associated with reduced pelvic floor functioning becomes hidden, which may be the reason why in Hunter et al (2015), little emphasis is placed on pelvic floor exercises by midwives and the topic is mentioned only briefly. However, Gerrard and Hove (2013), in their joint statement for the Royal College of Midwives (RCM) and Chartered Society of Physiotherapy, believe that pelvic floor exercises need to be addressed by midwives as a key issue.
To complicate issues related to the midwife's public health role, there is no consensus to suggest when antenatal pelvic floor muscle exercises should begin. The Bladder and Bowel Foundation (2017) advise pelvic floor exercises to be performed at least 3 times per day, varying in intensity of contraction to strengthen each of the pelvic muscles. A scoping search was performed to try and understand more about pelvic floor exercises and included Cochrane, CINAHL and Discover More, software that enables access to 55 000 professional journals. Search terms included ‘pelvic floor exercises AND urinary incontinence AND antenat* AND postnat* AND health promotion AND after birth AND education’, which highlighted 211 relevant studies. Delimiting the search to include peer reviewed titles provided 28 papers to inform this literature review.
Why are pelvic floor exercises important?
Pregnancy and childbirth factors have been considered to be key for the development of urinary stress incontinence (Whitford et al, 2007a; 2007b; Dinc et al, 2009; Ko et al, 2010). Stress incontinence relates to the leakage of urine when the bladder is under pressure from activities such as coughing or sneezing (Hay-Smith et al, 2008). From a comprehensive literature review, it was suggested that pelvic floor exercises are used commonly as a prevention method for urinary incontinence (Dinc et al, 2009). A systematic review by Boyle et al (2012) included 22 randomised controlled trials related to pelvic floor exercises, where the intention was to prevent or treat urinary incontinence in antenatal and postnatal women (n=8485). Systematic reviews use a pre-defined selection of best available evidence to ensure the highest quality of research is used for practice (Rees, 2011). Boyle et al (2012) included a large sample of women from a variety of backgrounds and ethnicities, which supports generalisation of the findings. The systematic review found that pregnant women without previous urinary incontinence were 30% less likely (95% CI) to report urinary incontinence up to 6 months after delivery if they engaged with intensive pelvic floor exercises. This finding increased to 40% (95% CI) for women who experienced persistent postnatal urinary incontinence and went on to report the change for up to 12 months.
Studying the effect of pelvic floor exercises on urinary incontinence
Another review by Dumoulin et al (2014) compared pelvic floor exercises to no treatment, a placebo, sham treatments and inactive control treatments. The review included 21 randomised controlled trials and involved 1281 women. Although the sample was comparably smaller to the previous systematic review (Boyle et al, 2012), the study can still be considered large (Rees, 2011). In this meta-analysis, women who performed pelvic floor exercises were eight times more likely to report improvement or cure (Dumoulin et al, 2014), although the review suggested that findings were undermined due to the small sample sizes in the individual trials. A further limitation of the review was that Dumoulin et al (2014) did not identify whether any participants were pregnant or post-delivered. Therefore, it was not possible to generalise these findings to a particular population (Aveyard et al, 2015).
A randomised controlled trial by Ko et al (2010) found that, from a population of 300 women, those who performed pelvic floor exercises (n=150) had significantly lower self-reported urinary incontinence in late pregnancy (36 weeks gestation) and 6 months postpartum compared to the control group (P=<0.01).
Another small investigation (Panhale and Mundra, 2012) carried out a cross-sectional survey involving 50 women from an Indian population and found a significant association between frequency of stress incontinence in the antenatal and postnatal period and frequency of pelvic floor exercises. Generalisability of the findings was limited, as quantitative research requires large samples to ensure that research is able to identify a level of statistical significance (Aveyard et al, 2015). The size of Panhale and Mundra's (2012) study therefore suggests that the findings could have occurred by chance, as the study was underpowered.
From women's perspectives, Dinc et al (2009) used an experimental design to collect data, using interviews together with urinary diaries and digital palpation to measure pelvic floor strength. Experimental designs allow researchers to observe and compare two groups of participants; enabling them to see whether a particular variable has had an effect on the study group (Rees, 2011). Burns et al (2013) identify that three elements confirm an experimental design, namely randomisation, researcher manipulation of a variable, and creation of a control group, all of which were present in the research by Dinc et al (2009). Although the sample of 80 participants was reduced to 54, due to withdrawal and low adherence rates, the intervention group measured a significant decrease in urinary incontinence during pregnancy (P=0.008), and increased pelvic floor strength postnatally (P=0.014).
‘Midwives have an important role in public health promotion; however, public health messages can be complicated by issues such as limited time, training and resources, or midwives feeling that they ‘bombard’ women with a ‘wall of information’’
Dinc et al's (2009) results should be treated with caution, due to attrition and the small sample size, which suggests that the finding could be due to coincidence (Rees, 2011). In addition, it was not clear where and how women were recruited, meaning that the ethical process is unclear (Rees, 2011; Parahoo, 2014). However, the combination of results from multiple study designs, various sample sizes and various populations suggests a strong association between pelvic floor exercises, stronger pelvic floor muscles and reduced urinary incontinence.
Variation in education
Education provider
Education plays an important role in reducing urinary incontinence (Dinc et al, 2009; Ko et al, 2010; Mason et al, 2010; Bo and Haakstad, 2011; Panhale and Mundra, 2012). Guerrero et al (2007) used a survey to examine how women wished to be taught pelvic floor exercises in the antenatal period. The survey method used questionnaires, which are suggested to be a quick and cheap form of research that enables wide geographical distribution (Rees, 2011). A total of 54 women and 75 health professionals (21 obstetricians, 29 midwives and 25 GPs) were included. Guerrero et al (2007) found that 36% of GPs failed to discuss pelvic floor exercises with antenatal women, although the sample size was small. GPs reported that they did not see women often, which was suggested to inhibit the provision of pelvic floor exercise information. GPs also felt that pelvic floor exercise education was the midwife's role. The majority (n=30; 57%) of women wanted a discussion in the antenatal period and 76% (n=41) of women wanted midwives to teach pelvic floor exercises. However, 48% (n=14) of midwives felt that continence advisors or physiotherapists could perform the task better, due to their increased knowledge and training. Lack of confidence in providing pelvic floor exercise education creates barriers for women. This finding was reflected in an Australian study (Butterfield et al, 2007), which gathered responses from 225 midwives. Although set in Australia, the wide grading of midwives in the sample was reflective of midwifery grades in the UK. Butterfield et al (2007) found there were anomalies within midwifery knowledge related to frequency of assessment for incontinence, and that midwives omitted to link how antenatal incontinence was a factor for postnatal incontinence. Lack of knowledge was therefore found to create missed opportunities for health promotion around pelvic floor exercises.
Optimum start date
In Ko et al (2010), Mason et al (2010) and Bo and Haakstad's (2011) research, physiotherapists educated women about pelvic floor exercises. However, Dinc et al (2009) failed to identify what type of professional provided education related to pelvic floor exercises in their study, and simply identified the educator as ‘the researcher’. There is consequently a variation in pelvic floor exercise education, which is provided by different health professionals using a variety of training methods. There is additional variation regarding when to begin pelvic floor exercises and, in many studies, there was no specific gestation date identified (Dinc et al, 2009; Ko et al, 2010; Mason et al, 2010; Bo and Haakstad, 2011; Boyle et al, 2012; Panhale and Mundra, 2012). However, Bo and Haakstad (2011) suggested that pelvic floor exercises should be performed two to three times per week for 12 weeks, which was supported by Ko et al (2010). Level of significance differed between the two studies (Ko et al 2010; Bo and Haakstad, 2011) and one explanation for the discrepancy may be attributed by differing sample sizes. Variations such as these therefore reduce the ability to make comparisons between studies effectively.
Information received
A further aspect of variation may be in the education provided, whereby some women are given more information than others. Mason et al (2010) recruited 286 pregnant women from two hospitals in north west England, who were assigned to an intervention group (n=141), or a control group (n=145). Participants received varying information, from one-to-one exercises with an instructor, to leaflets, brief reminders, or nothing at all. The participants assigned to the intervention group performed pelvic floor exercises more times than the control group, although results were not significant. The authors concluded that women who had more information about pelvic floor exercises were more likely to perform exercises. This finding was similar to a study by Whitford et al (2007a) where younger, first time mothers from deprived backgrounds were less likely to be informed of, or to practise, pelvic floor exercises. The varied information provided by Mason et al (2014) to the intervention and control groups would lead to ethical questions about equity and withholding information.
Despite limitations, the intervention group was found to have fewer episodes of urinary incontinence, although findings were not significant.
Barriers
Lack of knowledge
Women expressed different perceptions about the importance of pelvic floor exercises and identified barriers that influenced their exercise performance (Melville et al, 2008). Melville et al (2008) used a cross-sectional survey to consider the etiology of urinary incontinence (n=1458). All participants were asked a series of questions on their knowledge of pelvic floor exercises, if they recalled being informed about pelvic floor exercises, and how they were taught. Responses were open to socially desirable answers and it was possible that the participants may not have remembered everything due to recall bias (Bowling, 2014), potentially undermining the validity of the findings (Rees, 2011). However, the use of open-ended questions may have encouraged the participants to complete the questionnaire using the terms they believed described their own experiences (Rees, 2011). A key theme identified that almost one-third (n=370) of the 1458 participants attributed their urinary incontinence to the weakening or loss of control of their pelvic floor muscles (Melville et al, 2008). Participants reported this as a suspected physiological change which occurred from childbirth, reporting, ‘I just assumed that delivering a baby a few years ago made me undergo some anatomical changes’ (Melville et al, 2008:1095). Different perceptions about the anatomy of the pelvic floor and women's lack of knowledge about pelvic floor exercises suggested that more education was required for women.
Reasons for non-compliance
Fine et al (2007) aimed to identify reasons why pelvic floor exercises were not performed (n=759). Reasons included forgetting, being too tired, being too busy, and perceiving exercises as unimportant or unnecessary. Secondary analysis from a prospective multicentre cohort study was used (Fine et al, 2007). When using secondary analysis there may be issues with availability and quality of data, which may be less rigorous and objective (Kumar, 2011); however, Borello-France et al (2013) corroborated that forgetting to perform pelvic floor exercises was the most frequently reported barrier for women. Other barriers included sickness, travelling/vacations, being tired, work hours, personal conflict and boredom with the exercises (Borello-France et al, 2013). These factors influenced the frequency of pelvic floor exercises and contributed to participants not performing pelvic floor exercises at all. Borello-France et al (2013) included 296 participants, which would make a level of significance clearer, although there may have been limitations to the study as a self-administered questionnaire was used for data collection. In studies using self-administered questionnaires, participants must be able to read and write and those who return them may have a particular attitude and opinion on the subject (Kumar, 2011), meaning that there is potential for a low response rate that may then reduce representativeness of the findings. In Borrello-France et al's (2013) study, the response rate was 88% (n=132) in the combined group and 85% (n=134) in the behavioural intervention group. Parahoo (2014) identifies a response rate of more than 70% must be achieved to enhance reliability of findings.
Pelvic floor exercises in subsequent pregnancies
Whitford and Jones (2011) found that those women who performed pelvic floor exercises before and after a pregnancy were more likely to perform pelvic floor exercises in subsequent pregnancies. Whitford and Jones (2011) performed a longitudinal cohort study, in which 289 women were recruited. Longitudinal cohort studies allow the researcher to follow the same participants over a period of time, which helps to establish possible influences to specific conditions (Rees, 2011). However, attrition is a problem with longitudinal design. In Whitford and Jones' (2011) study, the sample originally consisted of 438 participants, decreasing by 44% to 247 due to factors such as changes in appointments, women having no time or being in early labour. The follow up questionnaire was received from 163 women (37% of the original sample size). Despite attrition, Whitford and Jones (2011) concluded that midwives needed to stress the importance of pelvic floor exercises to primiparous women, as this may influence whether or not they perform pelvic floor exercises in the future.
Conclusion
Differences in provision and education
Urinary incontinence is a frequent complaint for women before and after childbirth, but in both instances, women who practise pelvic floor exercises reduce the incidence of morbidity attached to continence. Findings indicated that antenatal education about pelvic floor exercises benefitted women and reduced urinary incontinence, although the provision of education was varied. Differences exist as to which professional delivers education to women, with midwives appearing to lack confidence and knowledge regarding health promotion, and it was difficult to identify how pelvic floor exercises education was provided and how training techniques differ. There was also variation around when and how often women should perform pelvic floor exercises.
Variation around pelvic floor exercises affects the level of significance and reliability of findings, and suggests that women receive information in a non-standardised format, so that some women may benefit from pelvic floor exercises education and others may not. The type and style of information is therefore essential and influence whether or not women perform pelvic floor exercises.
Barriers to performing pelvic floor exercises
Women did not know the reason for performing pelvic floor exercises. Those who were young, primiparous and from deprived areas were less likely to perform pelvic floor exercises, as they had no access to information or believed the exercises were unnecessary. There were differences in women's perceptions of the origin of their urinary incontinence, with many believing that it was normal after childbirth. Women showed minimal understanding of the anatomy of the pelvic floor, highlighting an unmet need for education.
Women appear to be bored with pelvic floor exercises and there is a need to identify more innovative ways to engage women. Pelvic floor exercises could be incorporated into popular exercise classes such as Zumba or yoga, for example (Domene et al, 2016), which may interest a wider group of pregnant or postnatal women. Alternatively, innovative teaching may be developed via telephone applications and/or alarms to enable women of different ages to have easy access to pelvic floor exercises education in their home environment. New innovations need to be explored, tested or investigated with research before they are advocated.
More definitive research is required to clarify approaches to pelvic floor exercises. Research should identify an appropriately qualified health professional to deliver information, standarise the timing of pelvic floor exercise education, and identify a specific gestation date, with an appropriate rationale for this decision provided. Larger samples are required in research to increase reliability, representativeness and generalisability.