Fecal incontinence can be due to uncontrollable defecation as a result of inflammatory or functional disease of the rectum and more proximal bowel. Diarrheal disorders associated with urgency include any form of proctitis, but in clinical practice ulcerative colitis or Crohn’s disease are the most important. Infective proctocolitides, dysentery, is usually self-limiting or responds to treatment, but it is necessary to consider this in the differential diagnosis of incontinence, particularly because the incidence of infective proctocolitis is increasing in Western society, especially in institutions (see Chapter 33). Organisms responsible include among others Shigella, Clostridium difficile, and Campylobacter. C. difficile and ulcerative colitis can present in an identical manner. There has been an “alarming” increase in C. difficile- associated disease (CDAD) in the United States. This has been dramatic in patients with inflammatory bowel disease (IBD). Rodemann and colleagues reported that between 1998 and 2004 in the United States, CDAD in ulcerative colitis patients rose from 18.4 to 57.6 (more than threefold) per 1,000 admissions.³⁷⁹ In endemic areas, Entamoeba histolytica should be borne in mind.

Urgency may be the predominant feature in some patients with irritable bowel syndrome. It is important to recognize this because surgery has no role to play in the treatment, although neuromodulation may have a place.

Urgency incontinence may also occur as a result of inadequate capacitance of the rectum or “neorectum” in patients who have undergone restorative surgery or radiotherapy. In colorectal practice where many patients will have had a low anterior resection, for example, this form of incontinence is fairly common. Functional symptoms after anterior resection are referred to as the anterior resection syndrome and include the frequent passage of small-volume stool with passive and sometimes urgency incontinence. Those with this disorder may also have a weak sphincter (see next section).

The anal sphincter mechanism may be unable to retain stool and flatus if it suffers damage to its nerve supply or as a consequence of direct trauma. Myopathic degeneration can also occur. There is a gradual physiologic weakening with age that can be identified by physiologic testing.²⁰⁵

Idiopathic incontinence is the term used to describe denervation injury to the sphincter muscle that occurs for no apparent cause. Habitual straining during defecation, nerve entrapment, the preprolapse state, and the descending perineum syndrome have all been implicated. It may be seen after traumatic obstetrical delivery.

The condition was first formally defined by Parks and colleagues³⁵² in a group of patients with incontinence or rectal prolapse associated with diffuse weakness of the anal sphincter muscles. Biopsies taken from the external sphincter showed histologic evidence of atrophy of type I fibers. These patients also had neurophysiologic evidence of denervation as determined by single-fiber electromyographic (EMG) studies. This technique enables an estimate of the ratio of nerve to muscle fibers in the motor neuron unit. When denervation occurs, there is reinnervation by branching of intact nerve axons in an attempt to reach denervated muscle fibers. Consequently, the ratio of nerve to muscle fibers increases, and this can be detected neurophysiologically by direct recording of a needle electrode introduced through the skin into a nerve ending.³²⁰ In addition, measurement of the speed of conduction in the pudendal nerve showed the latency to be increased.²²⁷ Further studies in patients with idiopathic incontinence identified that most also have degenerative changes in the internal sphincter.^245,266 It was postulated that the neuropathic changes might be due to damage to the pudendal nerve through stretching, thereby causing a degree of apraxia as might occur with excessive descent of the pelvic floor, such as in the descending perineum syndrome or during childbirth. It is unlikely that stretching occurs at the level of the ischial spine because at this level the pudendal nerve has not yet branched to supply the sphincter, perineum, and clitoris. Computer modeling has hypothesized that potential stretch during delivery is more likely to damage the inferior rectal nerve.²⁶³

Objective evidence of a sensory deficit can be found in patients with idiopathic neuropathic incontinence, as would be expected in the presence of nerve damage. Others have supported the hypothesis that sensory function may be an independent factor contributing to continence.^22,31

The normal anal canal is extremely sensitive to temperature. Moreover, there appears to be a temperature gradient between the rectum and the anal canal.^308,309 Ambulatory pressure monitoring has shown that the sphincter relaxes several times an hour, allowing equalization of rectal and anal pressures, which permits entry of small amounts of rectal contents into the anal canal.³⁰⁷ This so-called sampling reflex may be lost with neuropathic change. Patients with idiopathic incontinence have reduced thermal sensation at all levels within the anal canal when compared with normal individuals.³⁰⁶ Another report from the same unit involving patients with hemorrhoids and others with incontinence showed reduced sensation in incontinence but also in those with hemorrhoids, albeit to a lesser degree.³⁰⁸ Others have challenged the importance of thermal sensitivity on the basis that the maximum temperature difference was merely 0.13° and was in the region of the anorectal junction.³⁸¹ Such a
difference was considered too small in their view to have a meaningful role in sampling.

Internal sphincter degeneration was described by Vaizey and colleagues in 1997.⁴⁶⁶ They reported a series of 45 patients (35 females) who presented with passive incontinence associated with reduced resting anal and normal squeeze pressures. The pudendal nerve terminal motor latencies (PNTMLs) were normal. The internal sphincter on ultrasound was thinned and showed hypoechoic changes (Figure 16-4). The increased thickening seen with normal aging is not seen. Such patients may be suitable for neuromodulation (see later).

Scleroderma is a chronic multisystem disorder characterized by excess deposition of connective tissue in skin and internal organs, associated with microvascular changes and immunologic abnormalities leading to nerve and muscle degeneration. This results in motility disturbances of the gastrointestinal tract. Thus, gastric emptying is delayed in up to 75% of patients, with small bowel involvement occurring in 17% to 57% of patients. The migrating motor complexes are reduced or absent, predisposing to bacterial overgrowth. Barium enema examination demonstrates pancolonic involvement in 10% to 50% of patients associated with wide-mouthed diverticula. Pseudoobstruction may occur. Complications include pneumatosis cystoides intestinalis, stercoral ulceration, and perforation.

Systemic sclerosis may produce incontinence on either a neurologic or a myopathic basis.^66,257 The condition affects the internal sphincter, which shows morphologic changes compatible with both neuropathy and myopathy, with atrophy and fibrosis. Resting tone is reduced, and the RAIR may be absent.¹¹⁴

Anal manometry may demonstrate a reduction or loss of the RAIR. Treatment includes biofeedback, sacral nerve stimulation, and surgery.^114,232

Soiling may occur in anorectal disease from such conditions as hemorrhoids, fissure, and anal fistula, even without surgical intervention. Prolapse of rectal mucosa or hemorrhoids and true procidentia may interfere with closure of the anal canal. With time, the protruding mass stretches the sphincter and may lead to further weakening. In addition, the rectal prolapse is likely to be associated with internal sphincter atrophy and pudendal neuropathy, both of which are factors responsible for its occurrence in the first place. The incontinence is often ameliorated with definitive treatment of the prolapse (see Chapter 21).³⁹⁵

Failure of embryologic development of the hindgut, pelvic floor, and related structures results in varying degrees of anatomical and functional deformity. The condition occurs in approximately 1 in 5,000 births. There are several classifications that mostly divide patients into those with a low, intermediate, or high anomaly. A high anomaly is often associated with a degree of sacral agenesis and in addition to anatomical loss of the lower gut tube; there is also sensory and functional loss of that part of the rectum that is present. The anal sphincter mechanism is not developed and may be too weak to maintain continence even after surgical procedures, such as pull-through operations and posterior sagittal anorectoplasty (PSARP). A megarectum may be present.

The adult colorectal surgeon may encounter patients with a congenital anomaly after leaving pediatric care. Many will have undergone surgery in infancy, either in the form of a pull-through procedure or a PSARP.^62,355,356

Those with a low anomaly may have been satisfactorily managed by surgical recanalization when the lumen is absent by dilatation of the stenosed anal segment. The functional results may be disappointing. The condition is described in greater detail in Chapter 3.

In this condition, there is an abnormal communication between the intestine above the level of the anal sphincter and the outside, usually the perianal skin. Etiologies and associated conditions include neoplasms, inflammatory conditions (Crohn’s disease, diverticular disease), and anastomotic leakage. A fistula from an intra-abdominal organ to the perineum or vagina will result in fecal incontinence. A pouch-vaginal fistula is one of the most common conditions encountered in colorectal surgical practice.^185,412 Management is directed to the cause of the fistulization. This can be located in the intestine, the female genital tract, or to the urologic system. Figure 16-5 demonstrates a fistula to
the perineum from a segment of diverticular disease of the sigmoid colon. The diagnosis and management of extrarectal fistula will be dealt with in Chapter 14.

The anal sphincter mechanism can be damaged by direct injury in the form of surgery, itself, or by civilian or military trauma. It can also be damaged during childbirth. In the practice of colon and rectal surgery, childbirth injury is now the most common cause of traumatic incontinence.

Encopresis, or psychogenic soiling, is defined as the passage of formed or semiformed stool in a child’s underclothes (or other inappropriate places) that occurs regularly after the age of 4 years. It is essentially an involuntary evacuation of the bowel not caused by organic factors. Encopresis is at least four times more common in boys than in girls and is analogous to enuresis as it pertains to urinary incontinence.

The condition was first described by Weissenberg who recognized that this form of fecal incontinence was associated with emotional disturbance. Behavioral factors that may contribute to the problem include the following:

In time, the increasing retention of feces leads to attenuation of the rectal wall, lax sphincter contractility, progressive constipation, obstipation, and fecal impaction. Anal fissure and hemorrhoidal difficulties may develop. Loening-Baucke observed a common clinical history with children exhibiting chronic constipation and soiling, often many years of infrequent and abnormal stools, a dilated rectal ampulla, and the presence of an abdominal fecal mass.

Treatment is usually directed toward bowel management, stress reduction, and child and family psychological counseling. Laxatives, enemas, and dietary regimens are recommended, as well as encouraging the child to sit on the toilet for 10 minutes twice daily at the same time each day. The goal is to establish a practical time for defecation and, ultimately, a spontaneous bowel evacuation habit. Uridine-5-triphosphate has been suggested to have some limited success. Although the mechanism of action of this drug has not been ascertained, it is believed to stimulate the cortical substance of the brain to make the child more aware of the need to defecate.

Loening-Baucke used anorectal manometry to evaluate 20 healthy children, 12 with constipation, and 20 with chronic constipation and encopresis. Mean values for anal resting tone and anal pull-through pressure were lower in the constipated and the encopretic children than in the controls. The study was repeated up to 4 years after treatment for the condition; abnormal anorectal function was still apparent even years after cessation of treatment and apparent recovery. In a later study of 97 children, also by Loening-Baucke, the author reported that 57% had not recovered. Using a host of training techniques, Loening-Baucke noted that there was no difference in recovery rates for boys and girls, and that the likelihood of success or failure could not be predicted a priori.

Trauma to the perineum can injure the sphincter mechanism (Figure 16-12). Impalement on a spike or pole as well as social injuries (e.g., fist fornication) may result in rupture of the sphincter and contamination of the extrarectal spaces.^{53,89,265,343,494} Sepsis can supervene and lead to excessive scar formation with a resultant patulous anal canal and an incompetent sphincter (see also Chapters 17 and 18).

Patients with fecal impaction often are incontinent owing to the reflex development of internal sphincter relaxation. This results in passive leakage of stool associated with a physiologically patulous sphincter.

Fecal soiling around the anus may be noted. On spreading the patient’s buttocks, it will be possible to determine whether the anus is patulous, implying a weak internal sphincter from neuromuscular impairment. Inspection will determine whether the anal sphincter ring is intact or whether it is ruptured. This is one of the most important signs that will determine whether a sphincter repair is likely to help the patient. The presence of a scar alone does not indicate disruption if, for example, it represents the site of a previous repair or an episiotomy and when the sphincter ring is intact. A scar with a disrupted ring, however, may be a sign that a repair may be necessary or appropriate. The most obvious example of this abnormality is the presence of a cloacal deformity. This is usually caused by complete disruption of the anterior sphincter into the posterior vaginal wall (see Figure 16-6). A deliberate attempt to identify a mucosal ectropion (Figure 16-14) or a rectal prolapse (Figure 16-15) must be made because this will involve a different management plan from other causes of incontinence. If a prolapse is suspected from the history but is not initially evident during the examination, the patient should be asked to strain while sitting on the toilet (see Chapter 21). The presence of perineal descent during staining should be noted. Spontaneous opening of the anal canal or sphincteric relaxation is said to be indicative of anoreceptive sexual intercourse or it may suggest neurologic impairment.

Palpation allows the resting tone of the anal canal to be assessed. Most of the effects are due to contraction of the internal sphincter. The patient is then asked to “tighten up,” and the degree of contraction or the “squeeze pressure” is evaluated. The examiner will be able to judge whether the contraction is normal or reduced. In extreme cases, it may be absent. The key question is whether any weakness felt is diffuse or localized. Diffuse weakness is a feature of pudendal neuropathy as occurs in idiopathic incontinence (see previous discussion). A localized weakness may be the site of a sphincter injury. In patients with disruption of the anal ring, an assessment of the quality or contractility of the divided external sphincter can only be obtained by palpation. With the finger in contact with the muscle, it is possible to gauge the power of its contraction. This is not possible by the physiologic tests currently available because pressure recording in a disrupted anal canal will not reflect the activity of either sphincter.

In neurologic conditions, including lesions of the spinal cord and cauda equina, the tone may initially seem to be normal. However, when gentle traction is applied to the anorectal ring, the anal orifice is seen to gape (see Figure 16-3).¹⁷⁷ Thus, Hill and colleagues performed a prospective study on 237 patients with idiopathic fecal incontinence and were able to demonstrate that an informed history and digital examination can predict the manometric findings and specialized anorectal physiologic studies with a high degree of accuracy.¹⁸⁸ Others have reported the same conclusions.^173,221 Sensory testing of the perineal skin is essential in patients with neurologic disease. Spinal and cauda equina lesions result in reduced sensation of sacral segments, S4 and S5, which will be apparent by soliciting numbness of the perineum and buttocks.

At the end of the examination, it should be possible to state whether the anal ring is intact, whether the anal muscle is diffusely weak, and whether the muscle is normal but interrupted by a defect. If a defect is present, the quality of the divided external muscle will have been assessed.

Anal manometry can be undertaken using various techniques, such as open-tipped or closed-tipped catheters, perfused catheters, macroballoons, and microballoons (see Chapter 7). Most units today use an open-tipped perfused catheter (Figure 16-16). Variations in recording instrument diameter consistently affect measurements of resting anal canal pressure and the maximum squeeze pressure (MSP).²⁹⁶ Interpretation of physiologic studies is made difficult by incomplete knowledge of the physiology of defecation.⁸⁰ Ambulant manometry is technically difficult to accomplish and remains a research tool.¹⁵

Pedersen and Christiansen studied 78 healthy volunteers to determine the range of normal physiologic variations with anal manometry.³⁵⁴ They found that the maximum intraindividual variations in the length of the anal high-pressure zone, the resting pressure, and the squeeze pressure were 10, 26, and 68 mm Hg, respectively. The median length of the pressure zone was 4 mm (14 mm Hg resting and 48 mm Hg squeeze). No gender difference was found in the length of the high-pressure zone, whereas resting pressures and squeeze pressures were higher in men than in women. One study has shown that male patients with so-called idiopathic fecal seepage may have a long anal sphincter with an abnormally high resting tone.³⁴⁷ Standard tests of anorectal sensorimotor function have been reproducible by different investigators, suggesting that comparison of data from different institutions is probably valid, provided the methods employed are clearly defined.³⁸²

The working party agreed that maximum resting anal pressure (MRP) should be used to denote the highest recorded pressure at any site in the anal canal at rest, and the MSP should be defined as the highest recorded pressure at any site in the anal canal during maximum voluntary (squeeze) contraction.²²³ Despite their recommendation to report pressure in kilopascals (100 cm of water = 9.8 kPa), most continue to use centimeters of water or millimeters of mercury.

Anorectal manometry is a useful objective measure to assess the power of the internal sphincter in patients with an intact sphincter ring and of the external sphincter during voluntary contraction. In patients with a disrupted ring, the pressure in the anal canal at rest or during maximal squeeze will not be a reflection of the intrinsic function of either sphincter.

Generally, patients suffering from problems with bowel control have lower anal canal pressures at rest and during maximum voluntary contraction than normal. There is also overlap of the value of squeeze pressures between asymptomatic
patients and those with impairment for bowel control. Many factors contribute to the mechanism for continence, and impairment of one may be compensated by the combined function of others.²³⁸

This technique was developed and introduced into clinical practice by Bartram who modified the probe used for endorectal ultrasound, replacing the balloon with a sheath of adequate diameter to permit acoustic contact with the anal canal lining (Figure 16-18).²⁴⁹ The space between the sheath and the sensor on the probe, itself, was filled with degassed water. This technique has revolutionized the diagnosis of many anal and pelvic floor disorders, and in the case of incontinence, it allows identification of both internal and external sphincters. It can demonstrate the thickness of the former and the integrity of the sphincter ring. A normal anal sonograph is shown in Figure 16-1. Defects appear as amorphous areas of varying echogenicity that interrupt the normal striated pattern (see Figure 16-8).²⁴⁸

The technique was compared with electromyography in 15 women who had sustained an obstetrical injury, and the results demonstrated for the first time that an objective picture of the sphincter could be obtained, including the presence of a defect.²⁴⁷ The examination was well tolerated, much more so than electromyography, and a subsequent study revealed significant abnormalities in most of 44 patients with incontinence.²⁴⁸ Its ability to detect defects was confirmed in a second publication.⁵² Voyvodic and coworkers demonstrated a correlation between the size of the tear and anal pressures.⁴⁸³ Endoanal ultrasound has been used extensively for the identification and detection of defects in the anal sphincter as described previously.^{90,99,116,120,128,133,376,436} Sultan and colleagues have also used vaginal endosonography to image the anal sphincters.⁴⁴⁰ Anterior internal and external sphincteric defects can be clearly identified with both techniques, but obviously vaginal endosonography is limited to the anterior sphincter.

BK Medical Systems (Brüel & Kjær, Wilmington, MA) introduced the 2050 Transducer with built-in, three-dimensional imaging capability (Figure 16-19). The scanning head is moved along a 60-mm distance inside a fully encapsulated probe by using two control buttons on the handle of the transducer.¹⁶⁹ Thus, the technique differs from conventional two-dimensional ultrasound in that multiple images are acquired during automatic withdrawal of the probe in short steps. Interpretation is not made in real time as it is with twodimensional ultrasound. The images of three-dimensional ultrasound can be obtained by a technician and interpreted by the radiologist or clinician at a convenient time later.

Gold and coworkers at St. Mark’s Hospital used this multiplanar imaging technique to reveal the length and radial extent of a sphincter tear (Figure 16-20).¹⁵⁹ Twenty controls and 24 patients with fecal incontinence were studied. They were also able to clearly demonstrate the sex differences in sphincter configuration. Bollard and colleagues attempted to quantify the nature, characteristics, and frequency of variations in female anal sphincter anatomy.³⁸ They observed that nulliparous women have a variable natural “defect” occurring along the anterior length of the sphincter, a factor that may contribute to overinterpretation of the existence of defects. Three-dimensional ultrasound imaging may ultimately prove to be the most useful diagnostic tool in the assessment of an individual with anal incontinence.

The application of high-resolution imaging of the anal sphincter mechanism has been achieved by means of an endoanal coil.¹⁰⁰ The St. Mark’s Hospital group affirmed that an external sphincter injury can be readily assessed by means of endosonography.⁴⁹² Williams who also compared this technique with that of three-dimensional endosonography analyzed the anal anatomy at similar levels by a graphic. Williams and colleagues used an overlay technique, whereby the images of three-dimensional ultrasound and magnetic resonance
imaging (MRI) were combined in the same patient.⁴⁹¹ There was an excellent correlation with the images of the external and internal sphincters but not with the longitudinal muscle.

Magnetic imaging is expensive and given the accuracy and easy availability of ultrasound, it has remained primarily a research tool. It may, however, be a method of quantifying atrophy of the external sphincter. Briel and colleagues were the first to use MRI to assess this.⁴⁴ They studied 20 female patients aged 50 years (range, 28 to 75) and found atrophy in 8 (40%). Its presence was related to the clinical outcome. They subsequently studied 25 women with fecal incontinence who had a sphincter repair following obstetrical injury. Biopsies of the external sphincter were taken at operation. The MRI showed atrophy in 9 (36%), which was confirmed histologically in 8.⁴⁵ Williams and colleagues studied 25 female patients, none of whom had a normal resting pressure and 16 had a low squeeze pressure. The respective cross-sectional area of the external sphincter was 240, and the mean fat content was 23%.⁴⁹²

The measures to be taken will depend to a considerable extent on the type of incontinence and the age and social circumstances of the patient. General principles include education of the patient as to the anatomy and function of the pelvic floor as well as the mechanism of defecation. Weight reduction in the obese and daily exercise for everyone should be advised.^23,402 Smoking may reduce intestinal transit time, a habit that may aggravate a tendency to urgency.⁴⁰⁶ If individuals are living in community-based accommodations, particularly when they require help from a caregiver, it is essential that a routine be established. This includes regular visits to the bathroom and the maintenance of a clean and well-ordered environment. Unfortunately, this is not always the case when it comes to many institutions. Therefore, education of the management personnel and caregivers is an important part of the program for improving the patient’s life. Such a regimen should emphasize the importance of avoiding constipation; clearly, if impaction develops, incontinence will ensue. It is also important to avoid urgency incontinence by anticipating the individual’s need for nearby bathroom access. In selected individuals, rectal irrigation may have a place (see later).⁶⁹

Drugs that the patient is taking should be reviewed. This will include any nonstandard medications such as herbal treatments. Laxatives and products which thin the consistency of the stool should be identified. It should be recognized that patients with incontinence often manipulate their diets.¹⁷⁶ This may involve avoiding meals or reducing the volume of food ingested. It is important to bear the possibility of nutritional deficiencies in mind and to check the blood count for anemia. In general, any dietary advice should be personalized with inquiry made as to foods that in the patient’s experience may tend to induce loose stools. There is some evidence that increasing dietary fiber may be beneficial,³⁵ but in another study it was not found to be helpful.²⁴⁶

Drug treatment for incontinence includes antidiarrheal agents as well as those used to treat constipation. Loperamide is generally considered the drug of choice for diarrhea. It has no irreversible side effects, although it can be associated with bloating and abdominal cramps in some patients. In those with passive incontinence, phenylephrine theoretically might be beneficial through its direct effect on the smooth muscle of the internal sphincter. There is precious little information on this, however.

In 1950, Kegel suggested an exercise regimen that appeared to be beneficial in both fecal and urinary incontinence.²²² Since that time, numerous articles have been published attesting to the validity of this approach. Although it probably is not possible to increase internal anal sphincter tone by perineal strengthening exercises, muscle bulk and voluntary contractility of the external anal sphincter, puborectalis sling, and levatores may be improved by such a regimen.

This is a difficult area because of the numerous subjective influences to which behavioral treatments in general are liable
and, in addition, to the particular effect the therapist is likely to have on the patient. Biofeedback involves the combination of monitoring of anal pressure, which the patient can relate to his or her own attempts at contraction of the pelvic floor and the attention of a therapist (Figure 16-21). An interesting observation on the value of exercise was made by Engel and colleagues in a report of six patients who had severe fecal incontinence from diverse causes.¹²⁵ They inserted a 50-mL balloon into the rectum and recorded the response to sphincter contraction on a polygraph. By verbal reinforcement, each patient was able to sense the rectal distension and associated this stimulus with attempts to tighten the sphincter. This technique became known as biofeedback and has become one of the mainstays of conservative treatment. It is emphasized, however, that it is one part of a routine consisting of many features as outlined previously.

There are useful reviews on biofeedback, one of the first being that of Heymen and colleagues.¹⁸⁶ Norton and associates reviewed the published trials of biofeedback and found 70 noncontrolled and 11 controlled studies.^334,336 All reported that biofeedback resulted in improvement, but the dropout rate was approximately 20%. The treatment groups often contained patients with various conditions, and few papers revealed long-term results. Most of the studies contained small numbers, and all were from single centers. In the 11 controlled studies, there were 592 patients; improvement occurred in 70%.

In a recent randomized, controlled trial aimed at determining whether the availability of biofeedback added value to pelvic floor exercises, there were 108 patients (83 females) with incontinence of at least 1 teaspoon of feces per week.¹⁸⁷ They were randomized to pelvic exercises alone and pelvic exercises plus biofeedback. The patients entered a 4-week period of education and medical treatment. Those who experienced success were then excluded from the trial. The remainder went on to exercises (63) and exercises plus biofeedback (45). They each had six 1-hour training visits once in 2 weeks for 3 months. The patients were then evaluated (53 vs. 40) at 3 months. Those who had a successful result were then assessed at 1 year. The biofeedback included display of rectal and anal canal pressures to the patient, with the end points of perception of a 10-mL rectal balloon and a squeeze of 125 mm Hg maintained for 10 seconds. At 3-month assessment on an intention to treat basis, 20 (44%) of 48 biofeedback and 13 (23%) of 63 exercises only patients were free of incontinence. The frequency of days per week with fecal incontinence was 0.83 ± 1.5 and 1.6 ± 2.0 days per week (P = .083). At 12 months, 24 of the 45 biofeedback patients had “adequate relief” compared with 22 of the 63 exercises only patients.¹⁸⁷

Electrical stimulation in incontinence has also been the subject of a Cochrane review.¹⁹⁹ These authors found four eligible trials containing a total of 260 patients. The results were contradictory, although symptoms improved in the stimulation group in each report. The authors concluded, however, that it was not possible to determine that stimulation was the factor responsible.

Electrical stimulation with biofeedback was compared with biofeedback alone in a multicenter (six center) randomized trial. One hundred fifty-eight patients with incontinence for any reason (sphincter damage, rectocele, fistula surgery, previous hysterectomy, etc.) were randomized to receive amplitude-modulated, medium-frequency stimulation with EMG-biofeedback (“triple treatment”) or EMG-biofeedback alone twice daily for 9 months. At randomization, there were 79 patients in each group, but this had fallen to 52 and 62 at 6 months and to 19 and 43 at 9 months. Patients were lost to follow-up largely because they were satisfied with the results, because they had become discouraged, or did not have the time. The Cleveland Clinic scores at baseline were 10.9 ± 4.2 and 11 ± 4.8 in the EMG-biofeedback alone and the triple treatment groups, and these were 7.8 ± 5.1 versus 6.0 ± 5.3 at 6 months and 7.3 ± 5.2 versus 4.8 ± 5.7 at 9 months.⁴⁰³ In an invited editorial, Norton comments on the high attrition rate of the study.³³³ In addition, the patients had various causes for their incontinence, and the variance of the continence scores was quite considerable. Thus, the conclusions from the study that stimulation adds to biofeedback were not secure.

In 1984, Prager described a device for control of feces from an end colostomy.³⁶⁸ Although this fell from use owing to the occurrence of local pressure necrosis in some cases, the balloon used was adapted for anal use. Mortensen and Humphreys investigated three different designs of an anal continence plug made of polyurethane sponge coated with a water-soluble surface in 10 patients (Figure 16-22).³¹⁰ Patients were able to tolerate them for a median of 12 hours, indicating their potential usefulness in selected patients with anal incontinence.

Deutekom and Dobben reviewed the literature and found four trials of the anal plug.¹⁰¹ Two compared plug to no plug,⁴⁷² one compared two plugs of the same brand,³³⁵ and one compared two plugs of different brands. There was no report of the outcome related to the severity of symptoms. Moreover, there were methodologic defects, including incompleteness of follow-up, failure to blind (three studies), and failure to assess on an intention to treat basis (three studies). Of 20 individuals tested, 14 were not able to tolerate the plug because of discomfort.³³⁵ There is a suggestion that plugs made of polyurethane function are better than those of polyvinyl alcohol.¹⁰¹

The procedure can be carried out in the office or in a day center. The bowel is prepared with a Fleet enema (CB Fleet Co., Inc., Lynchburg, VA). The procedure is covered by a single dose of antibiotic such as a second-generation cephalosporin with metronidazole. With the patient in the left lateral position under mild sedation and after infiltration of a local anesthetic, a finger is inserted in the anus. An 18-gauge, 2.5-in. needle attached to a ratchet gun or other convenient syringe, depending on the viscosity of the material, is placed through the skin just to the side of the anus. It is advanced under digital control into the submucosa or intersphincteric space and also into any obvious defect (Figure 16-27).

Durasphere is delivered using a prepacked syringe with an 18-gauge, 4-cm long needle. Injections are made at four evenly spaced positions using about 2 mL at each site. The technique of injection through a proctoscope was modified by Tjandra and colleagues by the use of ultrasound guidance.⁴⁵⁷ This improves the accuracy of placement of injectable silicone PTQ implants (Bioplastique), with textured polydimethylsiloxane elastomer particles suspended in a bioexcretable hydrogel carrier of polyvinylpyrrolidone (PVP, povidone).

Maeda and colleagues have carried out a Cochrane review²⁷³ and indentified four published randomized, controlled trials.^{278,416,455,457} In the trial by Tjandra and colleagues, 82 patients were randomized to intersphincteric silicone injections with (42) or without (40) endosound guidance.⁴⁵⁷ Of the 82, 71 had an intact internal sphincter, and 60% of the patients had an increase in the PNTML. At a mean of 6 months (range, 1 to 12), the respective proportions of patients who had more than 50% improvement in the Wexner score were 69% and 40%, a statistically significant difference. A high proportion of patients in each group (93% and 92%) had an increase in the quality-of-life score of more than 50%. The authors concluded that ultrasound guidance improved the results.⁴⁵⁷