14 Fecal Incontinence
Abstract
Fecal incontinence (FI) is not only prevalent but socially devastating. In any of its manifestations, from mild soilage and inadvertent flatus, to frank stool loss, it can be intolerable. The competent surgeon must therefore be familiar with all treatment options. This chapter focuses on the etiology, diagnosis, management, and outcomes of fecal incontinence.
14.1 Background
Fecal incontinence (FI) is a socially crippling disorder. Soiling, the escape of flatus, and the inadvertent passage of stool are embarrassing situations few people can tolerate. It therefore behooves surgeons who care for these individuals to be familiar with any treatment options that might be available. Anal continence is dependent on a complex series of learned and reflex responses to colonic and rectal stimuli, and the considerable individual variation in bowel habits makes clear distinction of derangement of continence difficult. Normal continence depends on a number of factors: mental function, stool volume and consistency, colonic transit, rectal distensibility, anal sphincter function, anorectal sensation, and anorectal reflexes. 1 The patient who has lost complete control of solid feces has complete incontinence. The patient who complains of inadvertent soiling or escape of liquid or flatus has partial incontinence. Less fastidious individuals may not complain of partial incontinence; therefore, careful questioning of the patient may be necessary. In an effort to classify the severity of symptoms, Browning and Parks 2 proposed the following criteria: category A, patients continent of solid and liquid stool and flatus (i.e., normal continence); B, patients continent of solid and usually liquid stool but not flatus; C, acceptable continence of solid stool but no control over liquid stool or flatus; and D, continued fecal leakage. Numerous other grading scales exist. All these severity scores are simple to use. However, they mainly reflect sphincter function. The worse the function is, the higher the score. Thus, incontinence to solid stool is always considered worse than incontinence for liquid stool. Unfortunately, this assumption does not necessarily reflect the subjective experience of the patient. Furthermore, the reliability and validity of these grading scales are questionable. Because of these drawbacks and the lack of precision of the grading scales, they are no longer recommended as the sole method of categorizing patients and monitoring outcome of treatment. 3 Some of the deficiencies of grading scales can be addressed by summary scales. These scales produce multilevel summative scores. The values for each type of incontinence are assigned according to the frequency of incontinent episodes, as frequency is one of the factors contributing to the severity of incontinence. Several scales also include items such as cleaning difficulties, the use of pads, and lifestyle alterations. Numerous summary scales have been designed, such as those according to Rockwood, Wexner/Cleveland Clinic Florida Fecal Incontinence Score (CCF-FIS), Pescatori, Vaizey/St. Marks, and many others. Some scales also attempt to assess parameters unrelated to the sphincter, such as urgency and use of antidiarrheal medication. The assignment of values to types and frequencies of incontinence varies between scales. The most frequently cited CCF-FIS is outlined in ▶ Table 14.1. 4 This scale has been globally validated in numerous languages. Moreover, a statistically significant correlation exists between scores > 10 and decreased quality of life (QOL). 5
In some summary systems, equal values are assigned to the same frequencies of the different types of incontinence, whereas in other scales variable weights are given. However, the lack of patient perspective in this assignment of values compromises the comparability and validity of these summary scales. To address this problem, Rockwood et al developed the Fecal Incontinence Severity Index (FISI). This index assigns values to various frequencies and types of incontinence on the basis of subjective ratings of severity. 6 The matrix includes four types of leakage commonly found in the fecal incontinent population—gas, mucus, and liquid and solid stool—and six frequencies—never, one to three times per month, once per week, twice per week, once per day, and twice per day.
Given the subjective nature of incontinence, the incorporation of patient values into severity measurement has been a major step forward. Although it is important to know the severity of FI, it is also important to measure the impact of incontinence and its treatment on QOL. To assess QOL for patients with FI, generic QOL scales such as the SF-36 and condition-specific scales such as the Fecal Incontinence Quality of Life Scale (FIQLS). 7 The FIQLS, developed by the American Society of Colon and Rectal Surgeons (ASCRS), is an instrument that has been studied well and seems to be very useful.
The FIQLS is composed of a total of 29 items; these items form four scales: lifestyle (10 items), coping/behavior (9 items), depression/self-perception (7 items), and embarrassment (3 items). Detailed questions are listed in ▶ Table 14.2.
Each of the four scales of the FIQLS is capable of discriminating between patients with FI and patients with other gastrointestinal problems. The scales in the FIQLS demonstrated significant correlations with the subscales in the SF-36. The psychometric evaluation of the FIQLS showed that this FI-specific QOL measure produces both reliable and valid measurement. However, unfortunately, there was discord between items deemed important by clinicians versus variables felt to be important by patients. Due to this disconnect between clinicians and patients, the well-intended FIQOL is not as widely used as had initially been envisioned. Major incontinence has considerable social consequences and demands an effort at some form of definitive therapy.
The exact incidence of FI is unknown. However, recent reviews estimate that this disorder is much more prevalent than previously believed. Some literature suggests that this may affect up to 18% of the adult population, and is a major factor in nursing home admissions. 8 , 9 Brown et al, in conjunction with the Nielsen group, queried over 5,800 U.S. women 45 years or older and found a prevalence of nearly 20% when asked about at least one episode of FI within the previous year. A further questionnaire surveyed what specific issues regarding the FI were most prevalent as well as what was most bothersome. Unsurprisingly, 97% of the women surveyed were bothered by the frequency of the leakage, but the most troublesome symptom was not the leakage itself, but rather the associated urgency. 10 Groups of individuals at high risk for incontinence include the elderly, the mentally ill, institutionalized patients, those with neurologic disorders, and parous women. In order to characterize factors associated with the negative impact of accidental bowel leakage (ABL) on QOL, Brown et al conducted an internet survey aimed at identifying the most important factors involved. Issues related to frustration, emotional health, and ability to interact socially showed the greatest negative impact, with over 39% identifying this as “severe.” 11 When characterizing factors associated with seeking care in women with ABL, less than 30% of women with ABL sought care, and if they did, the majority spoke to their primary care physicians. Such discussions were more likely to occur if the CCF-FIS was > 10. 12
In hospitalized geriatric and psychiatric patients, an incidence of 26 and 31%, respectively, has been reported. In 30 residential homes for the elderly, FI occurred at least once weekly in 10.3% of the residents, of whom 94% had evidence of organic brain damage. 13 Thirty-nine percent of Wisconsin nursing home residents have FI, 14 while a 46% incidence was reported from a Canadian long-term hospital. 15 Incontinence of stool is the second most common cause for institutionalizing an elderly person. 16 , 17
Macmillan et al 18 conducted a systematic review to investigate FI in the community. A total of 16 studies met the inclusion criteria. These could be grouped into the definitions of incontinence that included or excluded incontinence of flatus. The estimated prevalence of FI (including flatus incontinence) varied from 2 to 24%, and the estimated prevalence of FI (excluding flatus incontinence) varied from 0.4 to 18%. The prevalence estimate of FI from these studies was 11 to 15%.
Most discussions of etiology of FI have been based on the assumption that women, particularly women younger than 65 years of age, are more at risk for FI than men. Obstetric injury to the pudendal nerve or sphincter muscle is described as the primary risk factor, irritable bowel syndrome as second (a disease thought to be more prevalent in women), and other etiologies such as diabetes were listed as a distant third. 14 Yet, each population-based survey of FI prevalence, including that by Nelson et al, has shown a high prevalence in men. Clearly, etiologies other than childbirth must be sought.
In an excellent review of the subject, Sangwan and Coller 16 detail the tremendous socioeconomic and psychological burden of FI on society. Their comprehensive but often not discussed list of concerns includes the annual cost to patient, the prevention and management of skin breakdown, the increased incidence of female genital infection, social alienation, a personal sense of inadequacy with depression, pessimism, and low self-esteem, embarrassment about odor, and fear of coital incontinence with decreased libido and sexual dysfunction. The combination of these factors produces a social impact that is impossible to quantify. In a long-term care hospital, it was estimated that the annual cost of incontinence per patient was $9,771. 15 It was reported that, even in 1996, over $400,000,000 was spent each year for adult diapers. 19 In the United States, in 1994, the economic impact was estimated at $16 to $26 billion annually. More recent calculation placed the annual average total cost for FI at $4,110 per person. 20 Any discussion of the impact of this disabling disorder must address not only the prevalence and impact of the problem, but also the fact that it is disabling and prevalent, as well as that we are probably only seeing the tip of the iceberg; more than two-thirds of patients with FI do not even seek care, in some cases as few as 15%, 12 , 21 , 22 and of those that do, more than one-half are speaking only with their primary care physicians—clinicians who may not have a complete understanding of the full capabilities for treatment. 11
Therapeutic recommendations for incontinence can be made best when the anatomy and physiology of the anorectal region are understood, as detailed in Chapters 1 and 2.
14.2 Etiology
The exact percentage of incontinence attributable to each of the various causes is unknown. In one series, the most common causes of FI were injury sustained to muscles and nerves during operation (48%) and peripheral nerve injuries associated with systemic disease such as diabetes. Spinal cord injuries or defects involving spinal cord injuries accounted for 22% of cases. 23 In most series, obstetric and operative injuries account for most cases of incontinence. 24 , 25 The variation often depends on the type of referral practice and the special interests of the authors.
14.2.1 Previous Operative Procedures Previous Anal Operations
Lindsey et al 26 characterized the patterns of anal sphincter injury in 93 patients with FI after manual dilatation, internal sphincterotomy, fistulotomy, and hemorrhoidectomy. The internal sphincter was almost universally injured, in a pattern specific to the underlying procedure. One-third of patients had a related surgical external sphincter injury. Two-thirds of women had an unrelated obstetric external sphincter injury. The distal resting pressure was typically reduced with reversal of the normal resting pressure gradient of the anal canal in 89% of patients. Maximum squeeze pressure was normal in 52%. They concluded incontinences after anal operations are characterized by the virtually universal presence of an internal sphincter injury, which is distal to the high-pressure zone, resulting in reversal of the normal resting pressure gradient in the anal canal.
Internal Sphincterotomy
Lateral internal sphincterotomy is highly effective in the treatment of chronic anal fissure. However, this procedure results in a permanent defect in the internal anal sphincter, which may lead to impairment of fecal continence. The exact incidence of this complication is not known. During the first two decades, after the introduction of lateral internal sphincterotomy, several studies were conducted, aimed at evaluating the sequelae of this procedure. Impaired continence was observed in only a minority of the patients, most of them having temporary incontinence to flatus. In these retrospective studies, the patients were followed by chart review or telephone interview and not by mailed questionnaire. The duration of the follow-up was short and neither grading scales nor QOL scales were used. Some reviews, emphasizing the importance of long-term follow-up, have shown higher incontinence rates. In the series of Khubchandani and Reed, lack of control of flatus was the most common complaint (35%), followed by soiling of underclothing (22%) and accidental bowel movements (5%). 27 A significantly higher proportion of patients who had accidental bowel movements were aged over 40 years. Similar figures have been reported by Garcia-Aguilar et al. 28 A significantly lower incidence of continence disturbances has been reported by others. Pernikoff et al observed an overall incidence of 8%. 29 In their series of 265 patients, Hananel and Gordon encountered impairment of continence in 1.2% of the patients, most of them having only temporary problems. 30 In a prospective study among 35 patients, Hyman assessed continence prior to and 6 weeks after lateral internal sphincterotomy using the FISI. 31 The FIQLS was administered to patients with a FISI score > 0. Three patients had worsening of their FISI score after surgery. Only one of them reported an evident deterioration in FIQLS. Based on these data, the author concluded that lateral internal sphincterotomy is a safe procedure. Anecdotal reports illustrate that incontinence for solid stool, although very rare, can occur after lateral internal sphincterotomy. This complication is often attributed to division of an excessive amount of internal anal sphincter or inadvertent injury to the external anal sphincter. Most recently, Liang and Church reported on a prospective series of 57 patients undergoing lateral internal sphincterotomy for chronic fissure. Only 2 (4%) reported any changes incontinence and overall satisfaction in this cohort was 9.7 ± 0.9 out of 10 (p < 0.001). 32
Coexisting occult defects of the external anal sphincter in multiparous women seem to be another risk factor. 33 When comparing office records and response to a postal survey, Casillas et al found that significantly more patients had incontinence to gas after lateral internal sphincterotomy than that reported in their medical records. This problem was encountered by 29% of the multiparous female patients who underwent this procedure. Incontinence for solid stool was not observed. Among their patients, the overall QOL scores were in the normal range. 34 Sultan et al performed anal endosonography before and 2 months after lateral internal sphincterotomy. They found that this procedure in most females tends to be more extensive than intended in contrast to division of the internal anal sphincter in males. According to the authors, this problem is probably related to the shorter anal canal in females. They also found that lateral internal sphincterotomy may further compromise continence, especially in females with occult sphincter defects. 35
Fistula Surgery
Fistula surgery is the anorectal procedure most commonly followed by postoperative incontinence. Significant FI may be avoided if the anorectal ring is preserved. However, minor defects in continence may follow if even a small amount of sphincter muscle is severed. This complication can be reduced by avoiding wide separation of the severed ends of the sphincter mechanism. This goal is accomplished either by placing a seton or by “coring out” the fistulous tract, with subsequent sparing of the sphincter mechanism or, in the case of tracts crossing the sphincter mechanism at a high level, by the adoption of the advancement flap technique (see Chapter 10).
Although the transanal advancement flap repair is designed to minimize damage to the anal sphincters, impairment of continence after this procedure has been documented. The reported incidence of this complication varies between 8 and 35%. 36 , 37 , 38 , 39 It has been suggested that anal stretch caused by the use of a Parks retractor is a major contributing factor. 40 It has been demonstrated that the use of a Parks retractor has indeed a deteriorating effect on fecal continence.
Because this side effect is not observed after the use of a Scott retractor, this type of retractor has been advised in fistula repairs. 41 The last decade has seen a significant shift in sphincter-sparing approaches to fistula surgery. The development of these techniques has been driven specifically by the need to develop safer approaches to fistula surgery. The development of the anal fistula plug, a scaffolding of porcine intestinal submucosa, was greeted with high initial enthusiasm, with a near-impeccable safety profile and initial success rates in the mid 80% range. 42 However, more mature data have shown a much lower durable success, with rates as low as 14%. 43 Similarly, the use of fibrin glue to seal fistulas has essentially a zero chance of worsening continence; however, success rates are poor, as low as 14%. 44 In 2007, Rojanasakul et al 45 published their series on the LIFT procedure (ligation of the internal fistula tract), and in 2009, Rojanasakul published a large series 46 with an impressive success rate of over 90% and no reported incidence of incontinence. Seven years and more than 50 publications later, this technique continues to show reliable success, ranging from 60 to 94%, and almost no reported risk of decreased continence. 43 , 44 , 45 , 46 , 47 , 48 , 49
Hemorrhoidectomy
In modern surgery for hemorrhoids, incontinence is a rare complication. However, if the sphincter mass is inadvertently injured (in a blind-clamping technique in which the internal sphincter is grasped by a clamp), incontinence may result. Minor alterations in continence may be due to the removal of the hemorrhoidal tissue, a tissue that has been described as possibly functioning as a corpus cavernosum of the anus. 50 When incorrectly performed, the Whitehead operation leads to eversion of the rectal mucosa onto the anoderm. This abnormal anatomy results in incontinence through destruction of the normal sensory mechanism and mucosal leak from the exposed mucosal surface onto the perineum. Rarely, a circumferential scar will form after hemorrhoidectomy, which may lead to improper closure of the anal canal, causing some degree of FI.
Manual Dilatation of Anus
Forceful dilatation of the anal canal for the treatment of any anorectal pathology can result in varying degrees of incontinence. The disadvantages and consequences of this form of treatment are discussed fully in Chapter 8.
Sphincter-Saving Procedures
In the usual anterior resection, normal continence for flatus, liquid, or solid feces is generally maintained. However, when a distal anastomosis is performed, impairment of normal continence is not unusual. Incontinence of liquid or flatus often follows, and the patient may be unaware of a sudden bolus of stool. These problems are frequent in the early postoperative period, but they subside within 6 months in the great majority of patients. The lower limit at which an anastomosis can be created without interfering with the mechanism of incontinence is the uppermost level of the anal canal at the top of the anorectal ring, which in most individuals is approximately 4 cm from the anal verge. The circular stapler has made it technically possible to perform extremely low rectal anastomoses. However, if the anorectal ring is disturbed, partial or total incontinence may result. The severity and duration of the dysfunction are not predictable.
Goligher et al 51 reported that of 62 patients who underwent a low anterior resection, all of the 12 patients with the anastomosis less than 7 cm from the anal verge initially had less than perfect continence. With time, however, five developed perfect continence and three nearly perfect continence, but four remained with imperfect continence. Abdominoanal pull-through resection of the rectum, as popularized by Hughes, results in a high incidence of partial incontinence. In a review of his results with this procedure, he found that only 29% of the patients had normal function postoperatively, 23% had severe incontinence, and the remaining 48% had minor incontinence. 52
Parks and Percy 53 described a coloanal sleeve anastomosis for the treatment of rectal lesions. Of 70 patients who underwent this operation, 1 was incontinent, whereas 30 others experienced increased frequency of stool. Enker et al 54 reported that 64% of patients who could be evaluated in their series of 41 patients who underwent coloanal anastomosis had good or excellent function. Vernava et al 52 reported that 87% of their 16 patients were normally continent. Intersphincteric resection (ISR), designed to push the envelope for sphincter presentation by establishing adequate distal margins with partial sacrifice of the internal sphincter followed by handsewn coloanal anastomosis, exemplifies the adage: “Just because we CAN doesn’t mean we always SHOULD.” In a prospective comparison of 77 patients who underwent ultralow low anterior resection (LAR) with coloanal anastomoses, Bretagnol et al compared 37 patients with conventional anastomosis to 40 who underwent ISR. They found that compared to the conventional approach, while there was no difference in stool frequency, fragmentation, or urgency, patients with ISR had significantly worse CCF-FIS scores (10.8 vs. 6.9, p < 0.001). In addition, QOL was significantly decreased when using FIQL scoring. 55
After ileorectal and ileoanal anastomosis, varying degrees of incontinence may develop. In the former case, the cause is usually the loss of reservoir function, but the situation may be compounded by a weakened sphincter. In the latter case, intraoperative manipulation by necessity may stretch the sphincter mechanism.
14.2.2 Childbirth
FI has a female-to-male preponderance of 8:1, consistent with childbirth as the principal causative factor. In 1993, Sultan et al published their well-known article entitled “Anal-sphincter disruption during vaginal delivery.” In their paper, they described the results of an endosonographic study among 79 primiparous women. Endoanal ultrasound (EAUS) was performed 6 weeks before and 6 months after routine vaginal delivery. After childbirth, sphincter defects were detected in 35% of these females. A similar study was performed in 23 primiparous women who underwent a cesarean section. None of these women had a sphincter defect after delivery. 56 In a study by Eason et al 57 of 949 pregnant women 3 months after delivery, 3.1% reported incontinence of stool and 25.5% had involuntary escape of flatus. Incontinence of stool was more frequent among women who had delivered vaginally and who had third- or fourth-degree perineal tears than among those who had delivered vaginally but had no anal sphincter tears (7.8 vs. 2.9%). Forceps delivery (relative risk, 1.45) and sphincter tears (relative risk, 2.09) were independent risk factors for incontinence of flatus or stool or both. Anal sphincter injury was strongly and independently associated with first vaginal births (relative risk, 39.2), median episiotomy (relative risk, 9.6), forceps delivery (relative risk, 2.3), and vacuum-assisted delivery (relative risk, 7.4), but not with birth weight (relative risk for birth weight 4,000 g or more, 1.4) or length of stage of the second labor (relative risk for second stage 1.5 hours or longer compared with less than 0.5 hours, 1.2).
The reported incidence of occult sphincter defects after normal vaginal delivery varies between 7 and 41% (▶ Table 14.3). Oberwalder et al conducted a meta-analysis in order to determine the incidence of anal sphincter defects after vaginal delivery. Their Medline search yielded five studies with more than 100 women who underwent endoanal ultrasonography after childbirth. All these women were also questioned about symptoms of FI, not including urgency. The incidence of sphincter defects in primiparous women was found to be 27%. In multiparous women, the incidence of new sphincter defects was 8.5%. Overall, 30% of the defects were symptomatic. Only 3% of the women experienced impairment of continence without any sphincter defects. Based on the results of this study, it is clear that sphincter damage during vaginal delivery is quite common in primiparous women. In 70% of these women, the sphincter defects are asymptomatic in the postpartum period. 58 The question is whether women with an occult and asymptomatic sphincter defect are at increased risk for FI with aging. According to Rieger and Wattchow, it seems likely that many women remain asymptomatic, because the number of occult sphincter defects is far greater than the documented prevalence of FI in the community. 59 Oberwalder et al examined elderly females with late-onset incontinence, all of whom had vaginal deliveries. The authors observed sphincter defects in more than 70% of their patients. 60 A similar finding has been reported by others. 21 Despite these findings, it is still not possible to determine the exact risk for asymptomatic women with a sphincter defect to develop FI later in life. More studies, including control groups of equal parity and age, are mandatory. During the last decade, attention has also been focused on the risk factors for obstetric sphincter defects. Donnelly et al conducted a prospective study among primiparous women. After caesarian section, even when performed late in labor, none of the women experienced impairment of continence. Neither induction of labor nor its augmentation with oxytocin influenced the risk of sphincter injury or postpartum impairment of continence. Instrumental delivery was associated with a more than eightfold increased risk of anal sphincter damage and a more than sevenfold increased risk of symptoms when compared with unassisted delivery. 61 The increased risk of sphincter defects after instrumental vaginal delivery, especially after the use of a forceps, has also been reported by others (▶ Table 14.4). Not all studies confirm the previous observations that anal sphincter injury is common after forceps delivery. de Parades et al observed sphincter defects in only 13% of 93 females after their first forceps delivery. According to the authors, this observation gives support to the conclusion that forceps delivery is still a safe technique. 62 However, recruitment bias might be a possible explanation for their contradictory finding, because 60% of their patients did not return for postpartum assessment. Except for this single study, all other reports provide substantial evidence for the detrimental effect of forceps delivery on anal sphincter integrity.
FI among primiparous women increases over time and is affected by further childbirth. 79 FI at 9 months postpartum is an important predictor of persistent symptoms. In the study by Pollack et al among women with sphincter tears, 44% reported FI at 9 months and 53% at 5 years. Twenty-five percent of women without a sphincter tear reported FI at 9 months, and 32% had symptoms at 5 years. Risk factors for FI at 5 years were age (odds ratio [OR], 1.1), sphincter tear (OR, 2.3), and subsequent childbirth (OR, 2.4). As a predictor of FI at 5 years after the first delivery, FI at both 5 months (OR, 3.8) and 9 months (OR, 4.3) was identified. Among women with symptoms, the majority had infrequent incontinence to flatus, whereas FI was rare.
Besides instrumental delivery, other obstetric events are also associated with an increased risk of anal sphincter injury. Prolongation of the second stage of labor due to epidural analgesia, midline episiotomy, and perineal tears are well known independent risk factors. After primary repair of third- or fourth-degree perineal tears, persistent sphincter defects have been reported in up to 85% of the cases. 80 , 81 Nine months after primary repair of perineal tears, Pollack et al observed impairment of continence in 44% of the women. Five years later, 53% of the women suffered from continence disturbances. 79 These findings indicate that the damage sustained during third- and fourth-degree tears is much greater than is generally appreciated. Furthermore, it is clear that primary repair does not provide lasting integrity of the anal sphincters. Fernando et al conducted a systemic review and a national practice survey regarding the management of obstetric anal sphincter injury. They identified 11 studies with long-term follow-up (mean duration: 41 months) after primary repair of third-degree tears. In these studies, symptoms of FI were reported by 20 to 59% of the women. 82
Sze 83 , 84 found the proportion of women who had severe incontinence was significantly higher among women who had undergone at least two additional deliveries after sustaining a fourth-degree sphincter tear as a nullipara. Sze 83 also found the rate of FI and severe incontinence similar among women who had zero, one, and two or more additional deliveries after sustaining a third-degree perineal laceration and between women who had one sphincter tear and no additional delivery versus those with two tears and more than two subsequent deliveries.
Increasing awareness among women and health professionals about the sequelae of obstetric sphincter injury has given rise to a debate regarding the protective role of cesarean delivery. It has been postulated that elective cesarean section at term before the onset of labor protects the anal sphincters and prevents FI. 85 , 86 Although cesarean section performed during labor also protects the anal sphincters, it does not prevent FI. This finding indicates neurologic injury to the sphincters during labor. A 2010 Cochrane review of almost 32,000 women, over 6,000 of whom underwent delivery by cesarean section, was performed. This review comprised 21 studies, and in only 1 was there any difference in preservation of anal continence. They concluded that preservation of anal continence should not be used as a criterion for electing cesarean section for delivery. 87
It is certainly noteworthy that women with transient FI or occult sphincter injury after their first delivery are at higher risk of FI after a second delivery, but from a practical point of view, no change of obstetrical recommendation will be made as women will not be advised to avoid having children based on this information nor would it seem reasonable to recommend a cesarean section on the basis of fear of further alteration in continence. Despite the potential for cumulative sphincter injury or pudendal neuropathy, help is available for those individuals who suffer sphincter damage, and it must be remembered that cesarean section has its own potential immediate complications for mother and baby as well as possible late complications of a laparotomy such as adhesive small-bowel obstruction. Modifying this course of action may be the recognition that injury during the second delivery is primarily neurological with prolonged pudendal nerve terminal motor latency (PNTML). Intra-anal ultrasound has recently been touted as the most accurate method of determining occult injury to the sphincter mechanism, but in the absence of symptoms, a battery of investigative modalities that include intrarectal ultrasound, anorectal manometry, and PNTML would not likely be enthusiastically endorsed by most postpartum women. Nevertheless, understanding the potential for injury is useful knowledge for the clinician.
Previous Hysterectomy
Patients undergoing abdominal hysterectomy may have an increased risk for developing mild-to-moderate postoperative FI; this risk is increased by simultaneous bilateral salpingo-oophorectomy. In a study by Altman et al, 88 an increased risk of FI symptoms could not be identified in patients undergoing vaginal hysterectomy.
14.2.3 Aging
A very common form of FI is that associated with old age and general debilitation. Elderly patients with a longstanding history of straining at defecation may cause a stretch injury to the pudendal nerve as well. This often is described as incontinence of neurogenic origin.
14.2.4 Procidentia
Procidentia, or complete rectal prolapse, may chronically impair the internal and external sphincter mechanism. Procidentia is associated with incontinence in more than 50% of cases, 89 usually attributed in part to nerve injury. Repair of the procidentia results in improvement of the incontinence in approximately 50% of patients. Various treatments have been applied in the past, including waiting and hoping that sphincter tone would return, electrical stimulation of the sphincter mechanisms, and various plicating operations.
None of these methods, including the well-known Parks postanal repair, has met with uniform success. Biofeedback might be worthwhile in the treatment of persistent incontinence after repair of rectal prolapse. If biofeedback fails, sacral neuromodulation (SNM) is an alternative. 90 If incontinence remains a problem, a colostomy is the final option.
14.2.5 Trauma
In the case of impalement injuries, the sphincter mechanism is often disrupted. Depending on the extent of the injury, primary repair may be achieved without performing a protective colostomy. However, if the tissues are badly destroyed and there has been a delay in recognition, performing a protective colostomy with later definitive repair is preferable. Insertion of foreign bodies or alternative sexual practices may result in sphincter injury.
14.2.6 Primary Disease
Diarrheal states from any cause at times may overwhelm the normal continence mechanisms and result in temporary transient episodes of FI. Chronic inflammatory processes of the anorectal region, such as those that occur in patients with ulcerative colitis, amebic colitis, lymphogranuloma venereum, progressive systemic sclerosis, infections, or laxative abuse, can result in local sensory derangement, interference of the sphincter mechanism, and/or mucosal irritability, resulting in a loss of the rectal reservoir function. A patient with carcinoma of the anal canal may also present with incontinence caused by either infiltration into the sphincter mechanism or failure of the anal canal to close adequately.
14.2.7 Radiation
A main component of treatment of cervical and uterine carcinomas is by extracavitary and intracavitary irradiation. Similarly, external beam and brachyradiotherapy are frequently employed for prostate and rectal neoplasms. 91 Varying degrees of destruction of the muscular components of the rectum and anal canal occur, resulting in radiation proctitis. A radiation-induced lumbosacral plexopathy has been reported. 92 Initial conservative management with two or three daily cleansing enemas is generally recommended along with a high-bulk diet. SNM, to be discussed later in this chapter, has shown profound success for incontinence, even in cases of radiation-induced sphincter and nerve damage. If the condition becomes intolerable, colostomy is the last recourse. If severe bleeding remains a problem, therapeutic options include the topical application of short-chain fatty acids or 4% formalin. Laser therapy may also prove helpful. In recalcitrant cases, proctectomy may be necessary.
14.2.8 Neurogenic Causes
In cases of myelomeningocele, both the sensory and motor nerve supplies are disturbed in a variety of ways, leading to incontinence. Any form of trauma, neoplasm, vascular accident, infection, or demyelinating disease to the central nervous system or spinal cord can interfere with normal sensation or motor function, leading to incontinence.
Diabetic patients with autonomic neuropathy may have impaired reflex relaxation of the internal sphincter. 93 Diabetics with FI have a higher threshold of conscious sensation than do continent diabetic patients. Late onset of rectal sensation is one cause of FI in diabetics. Pinna Pintor et al 94 reported that somatic neuropathy plays an important role in FI in diabetic patients, combined with sensation threshold impairment as a feature of autonomic involvement.
14.2.9 Idiopathic Incontinence
Clinically, it is possible to identify those incontinent patients who have a sphincter defect. This can be accomplished by careful physical exam, EAUS, and MRI. Disruption of the external anal sphincter is the most common surgically correctable cause of FI. The prevalence of sphincter defects in patients with FI has been assessed with the use of EAUS. Deen et al examined 42 women and 4 men with FI. They found sphincter defects in 87% of their patients. 95 Karoui et al observed sphincter defects in 65% of 335 incontinent patients. 96 Comparable figures have been reported by others. 97 , 98 Based on these data, it is obvious that sphincter defects are present in at least two-thirds of incontinent patients. Less than one-third of the patients do not have any evidence of sphincter defects or other anorectal abnormalities. Their incontinence, formerly termed “idiopathic,” is thought to be secondary to pudendal neuropathy, characterized by a slowed conduction in the pudendal nerve. It is most likely that this prolonged latency is due to stretching of the nerve during straining. The question is whether such pudendal neuropathy is the principal cause of “idiopathic” FI.
Súilleabháin and coworkers reported a prolonged latency in only 60% of the incontinent patients without sphincter defects. Furthermore, they were not able to demonstrate a correlation between the PNTML and the maximum squeeze pressure in this group of patients. According to these authors, the etiology of “idiopathic” incontinence is more complex than damage to the pudendal nerve alone. This nerve is probably not the only one to sustain trauma during vaginal delivery. Neuropathic changes in the internal anal sphincter and abnormal sensation in the anal canal as well as in the rectum have been observed in patients with “idiopathic” incontinence. These findings indicate that the neurologic damage associated with vaginal delivery is not limited to the pudendal nerve, but may also involve damage to the autonomic inferior hypogastric nerves. 99 , 100 , 101 , 102
14.2.10 Congenital Abnormalities
The various operative procedures designed for treating an imperforate anus are based on the type of deformity. The ultimate goal is to establish a perineal opening with adequate sensory and motor control. Rarely are sensory mechanisms preserved; therefore, some defect in continence usually results. Gross incontinence usually can be avoided by careful placement of the colon or rectum through residual sphincter mechanisms such as the puborectalis sling.
14.2.11 Miscellaneous
Overflow secondary to fecal impaction is a frequent cause of incontinence. This problem often is missed because the patient complains of profuse diarrhea. Digital examination usually reveals a rectum full of stool. This problem generally occurs in elderly or debilitated patients, or in young children recovering from surgical procedures (usually anorectal). Thus, physicians must be aware of this potential problem and should routinely institute early preventive measures. In general, hospital patients should be administered a bulk-forming stool softener of a psyllium seed derivative. If impaction occurs, gentle enemas with a combination of tap water, phosphate soda, and hydrogen peroxide may be used. If these measures fail, disimpaction (either with or without administering anesthesia) is the treatment of choice.
In patients with diarrhea, from whatever cause, the normal mechanisms of continence may be overwhelmed, and the patient may experience incontinence.
Soiling rather than complete involuntary loss of rectal contents may occur. For example, large prolapsing third- or fourth-degree hemorrhoids can cause partial incontinence by interfering with the normal closure mechanism of the internal sphincter. This situation can result in the escape of either flatus or liquid feces or in mucosal irritation. After operations for fistula-in-ano or fissures, soiling may occur as well.
A variety of pelvic floor disorders, including descending perineum syndrome, solitary rectal ulcer syndrome, and a nonrelaxing puborectalis muscle, may be associated with varying degrees of incontinence. Psychiatric problems may predispose the patient to the clinical problem of FI.
14.3 Diagnosis and History
As in the investigation of any pathologic condition, obtaining a careful history is necessary. Indeed, treatment recommendations are based on the particular cause of the incontinence together with the assessment of the sphincter status. Particular attention must be paid to the characteristics of the incontinence. Complete incontinence is defined as the uncontrolled passage of solid feces, whereas partial incontinence is defined as the uncontrolled passage of liquid or flatus. True incontinence should be distinguished from perianal leakage, which may be associated with a variety of anorectal disorders. Incontinence also must be distinguished from urgency, in which the patient’s diet or individual bowel habits lead to frequent passage of liquid stool accompanied by a great sense of urgency. In such cases, simple dietary change may be all that is necessary. In addition to consistency, knowing the patient’s frequency of bowel movements helps determine whether an antidiarrheal agent is required. Urge incontinence has been reported to be a marker of external anal sphincter dysfunction. 103 However, because urge incontinence may be due to any and all proctidites, its attempted quantification in the Vaizey/St. Marks incontinence scores obfuscates the quantification of FI based on sphincter insufficiency. Again, the inclusion of antidiarrheal agent use in the St. Marks/Vaizey score hurts its utility, as diarrhea is not related to sphincter integrity or function. Female patients should be asked about childbirth and type of delivery. It is very important to know whether the delivery was instrumental assisted or not. It is also necessary to obtain a history with regard to episiotomy, perineal tears, and continence in the postpartum period. The patient also should be asked about associated problems or conditions such as urinary incontinence, prolapsing tissue, diabetes mellitus, medications, or radiation treatment. Patients with congenital abnormalities such as Hirschsprung’s disease generally present with some form of constipation and megacolon. An accurate history is necessary to distinguish the condition from acquired megacolon in the adolescent and adult age groups. In a patient with acquired megacolon, soiling of the perineum from the overflow incontinence often is secondary to fecal impaction. With Hirschsprung’s disease, incontinence of liquid or flatus is rare because of the constantly closed internal sphincter.
Whether the patient has had a previous anorectal operation or low colon anastomosis must be noted, because these procedures can lead to FI. Also, beverages such as coffee or beer can lead to frequent loose bowel movements. Any history of remote or recent trauma to the anorectal area may aid in establishing the cause of incontinence. Associated motor or sensory symptoms may point to a neurologic lesion. 104 A clue to the severity of the problem is to determine the frequency of the incontinence and the necessity to wear a protective pad.
Grading and scoring the severity of the problem is another important aspect of a careful history. It is worthwhile to know the severity of FI as well as the impact of this problem and its treatment on the QOL. Several aspects of grading scales and QOL scales are discussed in more detail elsewhere in this chapter.
14.3.1 Physical Examination
It must be noted whether a patient’s incontinence is a manifestation of a generalized disease or neurologic disorder or whether it is a local phenomenon. Undergarments should be inspected for staining by stool, mucus, or pus. In addition, the perineum must be inspected. In female patients with a history of vaginal delivery, it is helpful to measure the length of the perineum between anus and vagina. A decreased length of the perineum is frequently associated with a defect of the external anal sphincter. By simple retraction of the gluteal muscles, the large patulous anus that occurs with rectal procidentia can be recognized easily. Also, any large prolapsing hemorrhoids or evidence of pruritus may point to the fact that local anatomic factors may be responsible for the minor soiling by liquid or flatus. Scars from previous operations or episiotomies may also be identified. Sensation to pinprick and the anocutaneous reflex should be checked. The anocutaneous reflex can be checked by stroking the perianal skin and observing the sphincter “wink.” On straining, perineal descent or mucosal or full-thickness rectal prolapse may become obvious. Examination while the patient squats may be necessary to demonstrate prolapse.
Digital rectal examination reveals the strength (resting tone and augmentation on squeeze) or discontinuity of the sphincter muscle. Palpation points out any “keyhole” deformity of the anal canal, which might lead to soiling that may be misinterpreted as partial incontinence. The assessment of anal tone is, at best, a very indistinct barometer of sphincter function. The ability to assess the strength of voluntary sphincter contraction is subjective. Contraction of the puborectalis at the tip of the finger versus contraction of the external sphincter over the midportion of the finger may be distinguished. The anorectal angle can be assessed. The patient’s complaints should provide a more reliable index of incontinence. Anoscopic and proctosigmoidoscopic examinations reveal any inflammatory process or neoplasm contributing to the patient’s complaint.
Many tests are available for the assessment of FI. However, need for those tests has recently been controversial. In daily practice, most investigations do not influence the choice of treatment. In many centers, for example, the initial steps in the treatment of FI consist of medical therapy or biofeedback, irrespective of the underlying cause. However, from a surgical point of view, it is helpful to know whether the external anal sphincter is damaged or not. Physical examination is unreliable for the detection of sphincter defects. In the past, needle electromyography (EMG) has been used to identify defects of the external anal sphincter. The potential discomfort and the inability to identify internal anal sphincter defects are drawbacks of this type of investigation. The use of EAUS has largely supplanted this technique.
14.3.2 Special Investigations
Anal Endosonography
As stated earlier, EAUS has supplanted electromyographic mapping as it is easily available and more comfortable for the patient. It has been shown to be superior for the evaluation of sphincter defects with a sensitivity of detecting defects of 100%, compared with 89% for electromyographic mapping, 67% for anorectal manometry, and 56% for physical examination. 105 Based on these and other findings, EAUS is now considered to be the gold standard diagnostic tool for the assessment of FI (▶ Fig. 14.1). However, interpretation of ultrasound images of the external anal sphincter is rather subjective, operator dependent, and confounded by normal anatomical variations. Because the external anal sphincter and the perianal fat are both echogenic, it can be difficult to assess the thickness of the external anal sphincter and to identify atrophy of this muscle. Discrimination of normal variants from sphincter defects is also difficult, especially in the upper part of the anal canal in female patients, due to asymmetry of the external anal sphincter at that level. 106 In 75% of asymptomatic nulliparous women, Bollard et al found a natural gap in the anterior part of the external anal sphincter, just below the level of the puborectalis sling. According to these authors, this gap explains the difficulties in the interpretation of postpartum ultrasounds. 107 Sentovich et al evaluated the accuracy and reliability of EAUS for anterior sphincter defects. 108 In incontinent, parous women, the sphincter defects, detected by ultrasound, were confirmed at operation in 100% of the cases. A similar accuracy has been reported by others. 95 In continent, nulliparous women, the two ultrasonographers identified sphincter defects in 55 and 75%, respectively. This high false-positive rate could be decreased to 40 and 60% by using the video recording of the ultrasounds. The false identification of defects in normal, intact sphincters might be explained by the existence of a natural gap, as described by Bollard et al. It has been suggested that the false-positive rate might be reduced by the measurement of perineal body thickness. Zetterström et al reported that the perineal body thickness was 6 ± 2 mm in patients with an anterior sphincter defect and 1 ± 3 mm in asymptomatic subjects. 109 A similar finding has been reported by others. 110 EAUS is associated with a substantial interobserver variability with regard to the thickness of the sphincters. It has been shown, however, that the interobserver assessment of sphincter defects is very good. 111 Despite several disadvantages, EAUS is to date the most optimum diagnostic tool for the assessment of FI. During the last decade, the use of three-dimensional (3D) techniques has become more commonplace. Studies have shown that compared to 2D-EAUS, 3D-EAUS (▶ Fig. 14.2, ▶ Fig. 14.3) has an improved concordance with operative anatomic findings. 112 The value and clinical relevance of other tests have been questioned. According to some authors, most of these investigations lack clinical usefulness because they add little additional information to a complete clinical patient assessment. Furthermore, it is thought to be unlikely that these tests result in a significant alteration in a patient’s management plan. Frequently, abnormal values do not correlate with the severity of symptoms. Despite these limitations, several tests are still frequently applied. They have been reported to predict the outcome after medical or surgical treatment, thereby permitting the clinician to provide the patient with sound recommendations and allowing the patient to have realistic expectations. 113 In this section, the investigations that are most frequently used are highlighted.
The “Enema Challenge”
The simplest and most unsophisticated test for incontinence is administration of an enema. The ability to retain a disposable enema is a very useful clinical guide in the assessment of incontinence. If the patient is able to retain a 100-mL water enema, any surgical correction or prolonged treatment plan is likely unnecessary. Reassurance that there is not a more serious problem is all that may be indicated for such a patient.
Anorectal Manometry
Information derived from manometry includes assessment of the resting and squeeze pressures and the anorectal inhibitory reflex. The presence of the reflex eliminates suspicion of Hirschsprung’s disease. Basal pressure is reported to represent mainly the activity of the internal sphincter, and the spontaneous activity of the external sphincter affects maximal basal pressure. 114 Squeeze pressure is reported to be the voluntary function of the external sphincter and the pelvic floor muscles. If both basal and squeeze pressures are low, patients are prone to be totally incontinent. If only the voluntary function is low, the patients are probably partially incontinent. 114 External sphincter function is critical for maintaining continence of solid stool. 115
Penninckx et al 116 studied the relationship between symptoms and the results of manometric data in incontinent patients. Discriminatory values of greater than 40 mm Hg for maximum basal pressure and greater than 92 mm Hg for squeeze pressure could identify continent patients with 96% and incontinent patients with 88% accuracy. The uncontrollable evacuation of a balloon, progressively filled with water at 60 mL/min before the maximum tolerable sensation level was reached, was related to the degree of clinical incontinence. Also, the maximum retained volume and the interval between the first sensation volume and the maximum retained volume (“perceived rectal capacity”) were related to the clinical symptoms. The balloon-retaining test proved to be superior to the rectal saline infusion test for the determination of the severity of incontinence.
Unfortunately, there is a 10% overlap between the manometric values obtained from incontinent and normal persons. 115 Following childbirth, pudendal nerve damage increases the risk of FI in women with anal sphincter rupture, but manometric findings indicate damage to the sphincter apparatus in both continent and incontinent patients. 117 However, overlap in anorectal physiologic data between continent and incontinent patients is so great as to make accurate prediction of FI impossible. Furthermore, the values do not correlate with the severity of incontinence, nor do they predict postoperative results. No correlation exists between the outcome of the operation and the preoperative anorectal manometric studies. 118 Normal manometric findings do not exclude incontinence entirely. 114
In recent years, the relationship between anorectal manometry and EAUS has been studied extensively. de Leeuw et al applied both tests in 34 patients at least 10 years after primary repair of a perineal tear and in 12 asymptomatic women with a history of normal, uncomplicated vaginal delivery. Impaired continence was reported by 22 patients (65%). A persistent sphincter defect was found in 86% of these patients. Because maximum anal squeeze pressure and maximum anal resting pressure showed a considerable overlap between the different groups (with and without impaired continence and with and without a sphincter defect), anorectal manometry provided little additional information. 119
Nazir et al conducted an observational cohort study among 132 patients after primary repair of a third- or fourth-degree perineal tear. The mean time interval between delivery and evaluation was 5 months. All women underwent EAUS and vector volume manometry. They found no difference in manometric values between females without a defect and those with a less extensive defect. Only in women with a large, extensive defect were the manometric values significantly lower. Although they observed a correlation between incontinence scores and manometric variables, there was a large overlap between continent and incontinent females regarding manometric values. No cutoff point could be defined to distinguish continent from incontinent females. 120 Liberman et al designed a study to determine whether anorectal physiology testing alters the management of patients with FI. Manometric findings did not change the pretest management plans. No association was found between manometric results and ultrasound findings. EAUS was the test most likely to change the patient’s treatment plan. 113 Voyvodic et al observed a strong correlation between maximum anal squeeze pressure and the presence or absence of an external sphincter defect. The authors classified the defects into partial versus full-length and narrow versus wide-open. This classification appeared to be of little benefit in defining further functional disability because the squeeze pressures in these subgroups were not significantly different. This might imply that the loss of integrity due to disruption of the external sphincter ring is the most important factor in loss of function rather than the degree of separation of the muscle margins. 121 Bordeianou et al conducted a prospective study aimed at determining the relationship between anal resting pressure and FISI scores in the presence of endosonographic sphincter defects and found that FISI scoring was less sensitive in discriminating between patients with and without sphincter defects; however, the patient had a sphincter defect and a significant decrease in resting pressures. 122
Defecography
The anorectal angle is postulated to be more obtuse in patients with incontinence 118 ; however, voiding defecography or balloon proctography can easily demonstrate this increased angle. This examination will probably add little information regarding the cause of incontinence except perhaps the demonstration of an occult rectal internal procidentia, which may or may not be contributing to the FI.
Pudendal Nerve Terminal Motor Latency
Although the severity of denervation does not appear to influence the severity of incontinence, it seems to affect the outcome of sphincter repair. Assessment of the PNTML provides a useful tool in defining pathology of the pudendal nerves. Prolongation of PNTML is indicative for pudendal neuropathy and is considered to be a hallmark of “idiopathic” incontinence. Roig et al 123 found pudendal neuropathy in 70% of their patients with FI (59% in patients with a sphincter defect and 94% in patients without a sphincter defect). Based on this finding, it seems likely that pudendal neuropathy is an etiologic or associated factor in FI. It is not clear whether a prolonged conduction velocity of the pudendal nerve affects its functional integrity. It has been shown that one out of three patients with bilateral prolonged PNTML has squeeze pressures in the normal range and that almost half of those with a normal PNTML have squeeze pressures below the normal range. 124 Although it has been stated that the information obtained by PNTML testing does not contribute to the management of incontinence in individual patients, it might be of prognostic value when surgical treatment is being considered. Laurberg et al were the first to demonstrate that pudendal neuropathy affects surgical treatment. In their series, the outcome of sphincter repair was successful in 80% of the patients without neuropathy and in only 10% of the patients with neuropathy. 125 This finding has been confirmed by others. 126 Sangwan et al reported that the outcome of sphincter repair was good in patients in whom both pudendal nerves were normal, whereas only one out of six patients with a unilateral pudendal neuropathy had such an outcome. According to these authors, both pudendal nerves must be intact to achieve normal continence after sphincter repair. 127 The relationship between pudendal nerve integrity and successful outcome after surgical repair is not universally accepted. Rasmussen et al, Chen et al, and Young et al were unable to identify any relationship between pudendal neuropathy and a poor outcome after sphincteroplasty. 128 , 129 , 130
Osterberg et al 131 questioned the routine use of PNTML in the assessment of patients with FI. They found pudendal neuropathy and increased fiber density are common in patients with FI. Fiber density but not PNTML was correlated with clinical and manometric variables. The severity of nerve injury correlated with anal motor and sensory function in patients with neurogenic or idiopathic incontinence.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
