CHAPTER 52 Urethral sphincter incompetence
stress incontinence
Terminology and Definitions
The symptom of incontinence is the complaint of involuntary loss of urine (Abrams et al 2002). However, there are several types of incontinence symptoms. These have been defined by the International Continence Society’s Standardization of Terminology Committee and are shown in Box 52.1. This chapter will focus on ‘stress’ urinary incontinence.
Box 52.1 International Continence Society’s (ICS) standardization of terminology
Other types of incontinence (not specifically defined by the ICS)
Epidemiology
In the epidemiological assessment of the prevalence of incontinence, it is important to define the population because incontinence has been shown to vary with age and race. The prevalence of incontinence in an institutional elderly care setting is much higher than in a community-dwelling young population. Prevalence estimates for the most inclusive definitions of incontinence (e.g. ‘have you ever experienced urinary incontinence’) vary greatly from 5% (van Oyen and van Oyen 2002) to 69% (Swithinbank et al 1999); however, most studies fall within the range of 25–45% (Thomas et al 1980, Yarnell et al 1981, Holst and Wilson 1988, Rekers et al 1992, Lara and Nacey 1994, Hannestad et al 2000).
Another factor which greatly impacts on the prevalence of urinary incontinence is bothersomeness. Swithinbank et al (1999) demonstrated that although urinary symptoms are common among the female population, they are not always perceived as a problem. Stress incontinence was experienced, at least occasionally, by 60% of the women, but only half of them felt that it was a problem.
There is less variation when studies asked about daily incontinence. Three studies conducted in women under 60 years of age reported a prevalence of 4–7% (Burgio et al 1991, Miller et al 2000, Samuelson et al 2000), whilst the prevalence in women over 65 years of age was 4–14% (median 9%) (Wetle et al 1995, Nakanishi et al 1997, Brown et al 1999).
Most epidemiological studies assess all types of urinary incontinence, and few have attempted to assess the prevalence of stress incontinence. Overall, approximately half of all incontinent women have stress incontinence. Hannestad et al (2000) demonstrated that there is a regular rise in the proportion of cases of urge incontinence compared with stress incontinence from the age of 40 years. Hunskaar et al (2004), in a large epidemiological study of 29,500 women in four European countries, demonstrated a similar change in the prevalence of different types of incontinence with age. The relative prevalence of mixed urinary incontinence increased with age and that of stress incontinence decreased.
Pathophysiology
Possible variations in pathophysiology of incontinence can be made considering each of these components of continence (see Chapter 49).
Aetiology
The possible risk factors for stress incontinence are shown in Box 52.2.
Some studies have demonstrated age as a significant risk factor for stress incontinence (Goldberg et al 2003), whilst others have only shown age to be significant in urge incontinence (Nygaard and Lemke 1996, Hannestad et al 2000, Samuelson et al 2000).
There are factors associated with age which may be responsible for incontinence rather than age alone. Resnick (1996) describes these using the mnemonic ‘DIAPERS’: Delerium, Infection, Atrophic changes, Pharmacological, psychological, Excess urine output, Restricted mobility and Stool impaction. Certainly, pharmacological agents such as α-blockers (e.g. doxazosin, an antihypertensive) can cause stress incontinence, and angiotensin-converting enzyme inhibitors can cause chronic cough which may aggravate stress incontinence.
There has been controversy about the role of oestrogens in the treatment of stress urinary incontinence, and a meta-analysis concluded that oestrogen did not improve the symptoms of stress incontinence (Ahmed Al-Badr Ross et al 2003).
Pregnancy, regardless of the mode of delivery, is associated with an increase in the prevalence of incontinence, although the majority of cases improve in the puerperium. Burgio et al (2003) found that 60% of women experienced urinary leakage during pregnancy, but this decreased to 11% by 6 weeks post partum. However, women who experience incontinence during pregnancy, even if it resolves in the puerperium, are more likely to experience incontinence in later life than those who have not had incontinence in pregnancy (Viktrup and Lose 2001).
In a large study involving over 15,000 women, Rortveit et al (2003) demonstrated that women who had a vaginal delivery were at greater risk of developing stress incontinence in later life compared with women who delivered by caesarean section (odds ratio 2.4). The evidence regarding the impact of other obstetric factors is less clear; however, there is some evidence that use of forceps increases the risk of urinary incontinence (Farrell et al 2001, Nelson et al 2001).
There is good evidence that obesity has a causal role in the development of stress incontinence. Several studies have demonstrated an association between obesity and stress incontinence (Brown et al 1999, Hannestad et al 2000, Viktrup and Lose 2001, Goldberg et al 2003), which has been confirmed by intervention studies of bariatric surgery (Deitel et al 1988, Bump et al 1992) and weight loss programmes (Auwad et al 2008). In the treatment group, there was a reduction of stress incontinence from 61% to 12% (Deitel et al 1988).
Epidemiological studies have demonstrated that the prevalence of stress incontinence is lower in Black African women (27%) compared with White women (61%) (Bump 1993). Thom et al (2006) found similar differences when comparing Black American women with White American women, and they adjusted for age, parity, hysterectomy, oestrogen use, body mass, menopausal status and diabetes. They suggested the presence of a protective factor in Black women.
Inherent differences in connective tissue may predispose to stress incontinence. Young nulliparous premenopausal women with stress incontinence had a decreased collagen I : III ratio compared with controls (Keane et al 1997).
Assessment
The assessment involves history, examination and investigations.
A detailed history is required using language that the patient understands. The International Continence Society has produced a document which standardizes the terminology of lower urinary tract symptoms (Abrams et al 2002), although these are not terms which patients will necessarily understand. The confusion caused by the term ‘stress incontinence’ has been highlighted. It is preferable to use a standardized validated symptom questionnaire to elicit the patient’s symptoms. Terms which are not well understood are usually removed in the validation process. Such tools can be completed by the physician or the patient. They ensure that symptoms are not missed and may help patients to describe symptoms which they find embarrassing.
It is important to ask about the impact of stress incontinence on sexual function. In one study, 68% of women stated that their sex life was affected by their urinary symptoms (Ward and Hilton 2002).
Several validated symptom questionnaires exist, and the main differences are content. Some only address symptoms of stress incontinence; for example, the Severity of Symptoms Index developed by Black et al (1996). Others, such as the Kings Health Questionnaire (Kelleher et al 1997) and the Bristol Female Lower Urinary Tract Symptom (BFLUTS) Questionnaire (Jackson et al 1996), assess all lower urinary tract symptoms, and some assess all pelvic floor symptoms (Barber et al 2001, Radley et al 2006).
Recently published national guidelines in the UK (National Institute for Health and Clinical Excellence 2006) and the USA (Agency for Health Care Policy and Research 1992, 1996) even suggested that patients who only had the symptoms of pure stress incontinence could be treated surgically without the need for twin channel cystometry.
Harvey and Versi (2001) evaluated the symptoms and signs of stress incontinence in predicting the presence of urodynamic stress incontinence, and found that the isolated symptom of stress incontinence had a positive predictive value of only 56% for the diagnosis of pure urodynamic stress incontinence. This implies that history alone is not reliable in establishing a diagnosis. In practice, most urogynaecologists and urologists still perform urodynamics prior to surgery. If a detailed symptom questionnaire is used to elicit the history, few patients have ‘pure’ stress incontinence alone.
Some symptom questionnaires, such as BLFUTS and ePAQ (Electronic Personal Assessment Questionnaire), separately assess the degree of bother that each symptom causes and the severity of the symptom (Jackson et al 1996, Radley et al 2006). Epidemiological studies have shown that although symptoms of leakage are very common, occurring in up to 60% of the population, less than half of women are bothered by them (Swithinbank et al 1999). Hence, an assessment of bother is important. The degree to which a symptom causes bother may be specific to the patient’s lifestyle and personality. Each individual patient will have their own expectations about treatment and goals which they hope to achieve. In clinical practice, it is important to establish whether or not the patient’s goals are attainable.
In research or audit, a formal assessment of the impact of incontinence on a patient’s quality of life can be made using a quality-of-life questionnaire. There are a large number of validated quality instruments available. The International Consultation on Incontinence (ICI) has appraised and rated the existing instruments (Donovan and Bosch 2005).
Examination
Physical examination of all patients with stress incontinence is important. Body mass index (BMI; kg/m2) should be recorded because there is good evidence that links incontinence with high BMI (Hannestad et al 2000). General abdominal examination may reveal striae suggestive of underlying collagen disorders. There may also be evidence of benign joint hypermobility with hyperextension of joints. A general assessment of the patient’s mobility and dexterity should be made with a view to considering if they would be capable of performing self-catheterization post operatively if necessary. A general neurological assessment should be performed because the presence of neurological disease may indicate an early need for urodynamics instead of conservative treatment.
Pelvic examination is performed to assess:
The levator ani can be assessed digitally just inside the posterior forchette on the left and right of the midline. The Modified Oxford Scale (Laycock and Jerwood 2001) (Table 52.1) provides a useful grading scale of the muscle, and has been shown to correlate well with surface electromyography and manometry (Haslam 1999).
Table 52.1 The Modified Oxford Scale for pelvic floor muscles (PFM)
| Muscle grade | External observation | Internal examination |
|---|---|---|
| 0 | No indrawing movement of perineal body | No activity detected |
| 1 | A mere flicker of activity | |
| 2 | A weak contraction but no PFM lift | |
| 3 | An indrawing movement of perineal body | A moderate PFM lift but without resistance |
| 4 | A good PFM lift against some resistance | |
| 5 | An ability to lift PFM against more resistance with strong grip of examing finger |
The presence of prolapse should be documented using a standardized assessment such as the pelvic organ quantification system (POP-Q) (Bump et al 1996). The advantage of POP-Q is that it has been shown to be reproducible.
Investigations
Pad tests
The main use of pad tests is to assess the outcome of treatment or to demonstrate urine loss. There are two types of pad test: short (≤2 h in duration) and long (>24 h). Although the only standardized test is the International Continence Society’s 1-h pad test, its use as an outcome measure is limited due to its poor reliability. Simons et al (2001) demonstrated that test–retest reliability of the 1-h pad test over a period of 3–10 days, even when similar bladder volumes were used, was clinically inadequate as the first and second pad test could differ by −44 to +66 g.
The report from the second ICI (Artibani and Cerruto 2005) concluded that 24-, 48- and 72-h pad tests had better reproducibility than the 1-h test. Compliance decreases with increasing duration of the test, and for this reason, extending the test beyond 24 h was not recommended. A pad weight gain of 1.3 g or more was considered a positive test. Attempts have been made to objectively quantify the severity of incontinence using pad tests, but there was poor correlation between the short duration tests and subjective assessments of severity (Reid et al 2007). Correlation with the 24-h pad test was only moderate (Sandvik et al 2006). Pad tests cannot be used to diagnose the type of incontinence.
Pyridium pad test
Pyridium (phenozopyridine hydrochloride) is a compound which, when ingested, causes urine to be dyed a bright yellow/orange colour. This test can be useful to determine if the leakage experienced is urine. The test needs to be performed very carefully due to the high risk of false-positive results from perineal staining after voiding. Nygaard and Zmolek (1995) reported nearly 100% Pyridium staining in asymptomatic women during exercise; however, the mean stained area was only 2.6 mm. Some women who have excessive vaginal discharge perceive this to be urine. A Pyridium test can be useful to distinguish the two conditions.
Urodynamics
Urodynamics are discussed in detail in Chapter 51. It has not yet been proven that preoperative evaluation of women with stress urinary incontinence improves the outcome of surgery. Despite this, urodynamic studies are regularly used to assess women with stress urinary incontinence in an attempt to identify preoperative risk factors for failure or postoperative voiding dysfunction. The presence of detrusor overactivity, ISD and voiding dysfunction are theoretically associated with lower cure rates. However, the ability to measure these parameters reliably using current urodynamic techniques remains debatable, and there is little robust evidence that diagnosis of these factors should alter surgical management. There is no agreed definition of ISD. Two commonly used definitions are a maximum urethral closure pressure (MUCP) of less than 20 cmH2O and Valsalva leak point pressures (VLPP) of less than 60 cmH2O; however, both have limitations. MUCP is a measurement made at rest, not during the dynamic stress phase. Although VLPP is a measure of pressure in the dynamic phase during ‘stress’, there is variation in the pressures recorded depending on catheter size, patient position and bladder volume. Another significant problem with VLPP is that up to 66% of women with stress incontinence experience leakage on coughing but do not leak with Valsalva (Sinha et al 2006).
Treatment
Conservative treatment
Patients with stress incontinence may benefit from lifestyle advice. They should be encouraged to maintain a normal BMI. There is evidence that obesity is an independent risk factor for incontinence, and that weight loss in the morbidly (Deitel et al 1988, Bump et al 1992) and moderately (Subak et al 2002, Auwad et al 2008) obese will improve incontinence.
There is some evidence of an association between chronic straining due to constipation and pelvic floor damage which increases the risk of incontinence (Diokno et al 1990, Spence-Jones et al 1994). However, there have been no intervention trials of the treatment of straining and urinary incontinence.
