Clinical Management of Intractable Constipation

DIAGNOSIS

AND

TREATMENT

Clinical Management of Intractable Constipation Michael Camilleri, MD; W. Grant Thompson, MD; James W. Fleshman, MD; and John H. Pemberton, MD

• Purpose: To review current management of intractable constipation. • Data Sources: Original articles and reviews published in the English-language literature between 1965 and 1993 identified by MEDLINE search. Verbal feedback from attendees after presentation of the document as a clinical symposium at the 14th International Symposium on Gastrointestinal Motility in September 1993. • Study Selection: Key words included constipation, epidemiology, colonic inertia, pseudo-obstruction, pelvic floor dysfunction, and results of therapeutic interventions, particularly the effects of biofeedback training and subtotal colectomy. • Results: In most patients, constipation is usually due to lack of dietary fiber and responds to simple measures to correct these factors, often without consulting a physician. In some, probably fewer than 10% of patients who consult their physicians, structural diseases of the colon and rectum, systemic disease, or medications that slow gut transit should be excluded, and regular exercise, dietary fiber, and an osmotic laxative prescribed. In a series of 277 highly selected patients from a tertiary referral center who had intractable constipation, only 29% had a definable abnormality; identification of abnormal transit facilitates selection of patients for further investigations identifying colonic inertia or pelvic floor dysfunction. • Conclusion: An algorithmic approach can carefully select patients with intractable constipation for behavioral modification and biofeedback; the long-term outcome is excellent, with at least 75% success in several series. In a minority of patients with slow transit constipation unresponsive to medical treatment, subtotal colectomy with ileorectostomy is indicated and effective.

Ann Intern Med. 1994;121:520-528. From the Mayo Clinic and Medical School, Rochester, Minnesota; University of Ottawa and Ottawa Civic Hospital, Ottawa, Ontario; and Washington University, St. Louis, Missouri. For current author addresses, see end of text. 520

© 1994 American College of Physicians

Downloaded From: http://annals.org/ by a Mayo Clinic User on 10/05/2016

Wc reviewed current knowledge and the advances in diagnosing and treating patients who are thought to have intractable constipation. We reviewed the epidemiologic characteristics and definitions of constipation, briefly addressed management of constipation in a primary care setting, and subsequently focused on those few patients who are deemed intractable after appropriate exclusion of reversible underlying disorders or standard medical therapy. Methods Using MEDLINE, we identified English-language original journal articles and reviews published from 1965 to December 1993. The key words included constipation, epidemiology, family or general practice, colonic inertia, pseudo-obstruction, pelvic floor dysfunction, obstructed defecation, pelvic floor retraining, biofeedback, and colectomy. We also used combinations of key words, such as colectomy and constipation, or biofeedback and pelvic floor dysfunction. We reviewed the literature and extracted articles according to our expertise: definition, epidemiology, and management in primary care (WGT); physiologic testing and medical management (MC); anorectal and pelvic floor testing (JWF and JHP); surgical treatment (JWF and JHP); and biofeedback (JWF). After agreeing on the format, content, and recommendations, a summary of this text formed the basis of a symposium on managing chronic constipation held during the 14th International Symposium on Gastrointestinal Motility, 29 August to 3 September 1993, in Minett, Ontario, Canada. After presenting the preliminary document at the symposium, we incorporated verbal feedback from attendees to ensure a comprehensive, state-of-the-art representation of current knowledge and practice in most special referral centers worldwide. Definitions and Epidemiologic Features Constipation is variably defined (Table 1) but usually refers to persistent, difficult, infrequent, or seemingly incomplete defecation. Diagnosis is often arbitrary and may depend on the patient's perception of what is abnormal. The prevalence of constipation depends on the definition and the population studied. Fifty years ago, surveys of British postal workers and general practice lists concluded that more than 95% of persons showed a Gaussian distribution in the frequency of defecation; frequency ranged between three movements per day and three per week (1). This was confirmed in other questionnaire-based studies done on 301 apparently healthy persons in Britain (2) and in the United States (3). The latter study surveyed 1128 young adults about their bowel habits. Burkitt and associates (4) drew attention to the importance of stool weight and transit time, showing in 1972 that Ugandans eating a high-fiber diet produced stools that were four times as heavy and that passed through the gut in half the time of those of British sailors. In the 1980s, "straining at stool" was introduced in surveys of constipation among participants in the community. Constipation, defined as straining more than 25% of the time, occurred in about

Table 1. Components in the Definitions of Constipation Symptoms: straining, hard stools, unproductive call to stool, infrequent stools, incomplete evacuation Stool form and consistency (Heaton et al [11]): loose stools suggest rapid transit; hard, small stools suggest slow transit. Definition of defecation: fewer than 3 bowel movements per week; straining, hard stools, or incomplete evacuation on more than 25% of occasions. Constipation defined by 2 or more symptoms in the previous 12 months (Talley et al [5]) Definition of defecation (the "Rome criteria" for functional constipation): 2 or fewer bowel movements per week; stool weight less than 35 g/d, straining on more than 25% of occasions; hard, lumpy stools on more than 25% of occasions; sensation of incomplete evacuation on more than 25% of occasions. Constipation defined by 2 or more symptoms for at least 3 months (Thompson et al [7]). Physiologic measures: colonic transit time Patient's opinion

6% of adults surveyed (2). Clearly, slight differences in the components considered in the definition of constipation (Table 1) or in the demographics of the persons surveyed can produce vast differences in the prevalence figures for constipation. Talley and colleagues (5) mailed a self-administered questionnaire survey; two or more of the following defined constipation during the 12 months before the survey: fewer than 3 bowel movements per week more than 25% of the time; straining more than 25% of the time; sense of incomplete evacuation on more than 25% of occasions; and hard stools on more than 25% of occasions (5). This survey was of the predominantly white, middle socioeconomic class community of Olmsted County, Minnesota, and the investigators recorded a prevalence of constipation close to 20% in the entire group, whose age range was 30 to 64 years. The "Rome criteria" (6) for functional constipation were developed by an ad hoc working team of five experts, and these criteria are generally accepted by other experts in the field (Table 1). Drossman and coworkers (7) sent a self-administered questionnaire survey to 8250 householders across the United States and found that of the 5430 responders, 3% were constipated (male, 2.4%; female, 4.8%) based on the Rome criteria. The surveyed group consisted of 96% white persons, randomly selected from a list of persons who were willing to participate in marketing surveys, and their mean age ±SD was 49±16 years. In this report (7), there was little increase in the prevalence of constipation with age, but others found that among patients 70 years and older, one third had infrequcncy. straining, or used laxatives frequently (8). Based on epidemiologic analyses of almost 5 million persons in the National Health Interview Survey, the National Hospital Discharge Survey, and the National Disease and Therapeutic Index in the United States, and the Morbidity Statistics from General Practice in England and Wales, the average prevalence of constipation was 2% (9, 10), and this increased exponentially after age 65 years (9), reaching 10% in those persons older than 75 years (9). Constipation was also more common in black persons and those of low socioeconomic status (9). Although age greater than 65 years appears to be a significant factor, it is plausible that differences in the prevalence of constipation among these studies reflect differences in the way the question was posed, the period when symptoms were sur-

veyed (previous 3 months or during the previous year), and the participants' likely differences in their interpretation of abnormal straining and hard stools. More importantly, these studies emphasize that the individual patient's symptoms must be analyzed in detail to determine the meaning of "constipation" or "difficulty" with defecation. Heaton and associates (11) have shown that stool form and consistency are well correlated with the time lapse since the preceding defecation. Hard, pellety stools occur with slow transit, whereas loose, watery stools are associated with rapid transit (11). Small, pellety stools are more difficult to expel than are large ones (12). The patient may state that he or she is constipated yet fulfill none of the definitions in Table 1. In a postal questionnaire study about knowledge, beliefs, and experiences of bowel function conducted in general practice in London, England, in the mid-1980s and involving 171 patients 55 years and older, 10% reported no predictable frequency of bowel movements; at the same time, 90% believed that "regularity" was necessary for good health. Regularity was defined by two thirds of respondents as having a bowel movement every day, and by one third as passing the movement at the same time each day. Sixteen percent of respondents regularly self-treated with over-the-counter laxatives and supplemented their diet with bran, vegetables, or fruits (13). The authors determined that 95% of respondents gave reasonable definitions of regular and diarrhea but that 10% were unsure of the definition of constipation. This is not surprising considering the range of definitions used in the literature or, indeed, when we recognize that little concordance exists between symptoms and stool form, which is used as a surrogate measure of gastrointestinal transit (14). Patients may also perceive that they are constipated, but when formally evaluated by daily diary during a 4-week period, only 20 of 44 constipated patients had a reduction in stool frequency to fewer than 3 bowel movements per week (15). The perception of hard stools or excessive straining is more difficult to assess objectively, and the need for enemas or digital disimpaction are clinically useful markers to corroborate the patient's perceptions of difficult defecation. Psychosocial factors may also be important, and thus persons whose parents attached great importance to daily defecation may become concerned when they miss a daily bowel movement, especially if there is accompanying anxiety and depression. Some children withhold stool to gain attention; drafty outhouses or filthy plumbing may commence a lifetime aversion to defecation; and some adults are simply too busy or too embarrassed to interrupt their work when the cramp-like discomfort or call to have a bowel movement is sensed. One study found that after controlling for dietary fiber intake, those persons with confident, outgoing personalities had larger stools than did their introverted peers (16). Thus, constipated patients cannot be treated successfully without addressing their attitudes and psychosocial circumstances. Approach to Constipation in Primary Care A careful history should explore the patient's symptoms and confirm whether he or she is indeed constipated

1 October 1994 • Annals of Internal Medicine • Volume 121 • Number 7 Downloaded From: http://annals.org/ by a Mayo Clinic User on 10/05/2016

521

based on frequency (such as fewer than 3 bowel movements per week), consistency (lumpy or hard), or excessive straining as shown by prolonged defecation time or need to support the perineum or digitate the anorectum. Between 1965 and 1993, there were no reports in the literature that document the prevalence of colorectal cancer, strictures, depression, and metabolic disorders (hypothyroidism, diabetes) in primary care practices. Generally, investigators believe that in most cases (probably more than 90%), no underlying cause exists and the condition is idiopathic (17) and can be treated empirically with ample hydration, exercise, and supplementation of dietary fiber. A good diet history and attention to the psychosocial issues discussed previously are important. Knowledge of use and misuse of laxatives or intake of drugs that cause constipation are important. A careful history with detailed documentation of the patient's perception of constipation or difficulty with defecation, as well as a physical examination and, particularly, a rectal examination should exclude most of the important diseases for which constipation is a symptom. There appears to be consensus on the selection of patients for further investigation (18). Measurement of serum calcium and thyroid-stimulating hormone levels will identify rare patients with metabolic disorders. Rectal bleeding or anemia with constipation mandates sigmoidoscopy and barium enema or colonoscopy, particularly in patients older than 35 years, to exclude structural diseases such as polyps, cancer, or strictures. Melanosis coli, or pigmentation of the colon mucosa, indicates the use of anthraquinone laxatives such as cascara or senna, but this is usually revealed in a careful history. In patients with rectal bleeding and weight loss associated with constipation, a barium enema should be done, especially in the elderly, to exclude structural causes of constipation; an unexpected disorder such as megacolon or cathartic colon may be detected by the colonic radiographs. Barium enema may be preferable to colonoscopy in the constipated patient because it is less costly and identifies colonic dilatation and all substantial mucosal lesions or strictures that are likely to occur with constipation (19). Excluding rectal and anal disorders, the double-contrast barium enema has been shown to have similar sensitivity to colonoscopy in detecting mucosal lesions resulting in subacute or chronic intestinal bleeding (20). However, if the patient with constipation has weight loss, anemia, or rectal bleeding, colonoscopy is necessary and more cost-effective than barium enema because it provides an opportunity to obtain biopsy specimens of mucosal lesions or to perform polypectomy. Most patients with constipation may be managed empirically with at least 20 or 30 g of dietary fiber daily (21); the most common reason for failure of treatment is noncompliance. Therefore, a measurable amount of bran or psyllium (beginning at 3 tablespoonsful/day) is preferable to an undefined high-fiber diet. Ample hydration and exercise should be ensured. Those with more difficult constipation may be helped by a bowel training regimen (6). This includes discontinuing laxatives; evacuating with enema or glycerine suppository as needed; eating a highfiber diet and, if needed, taking an osmotic laxative; and after breakfast and coffee, a distraction-free 15 to 20 minutes on the toilet rather than sitting on the toilet and 522

1 October 1994

straining. Excessive straining may lead to development of hemorrhoids. In addition, weakness of the pelvic floor or injury to the pudendal nerve may result in obstructed defecation several years later. Nightly ingestion of an osmotic laxative may also help, and examples include lactulose, 30 to 60 mL, or milk of magnesia or magnesium citrate, which is less expensive if the intake of magnesium is not contraindicated. Other therapeutic approaches, such as the use of mineral oil, agents that evoke intestinal secretion (such as docusate), or stimulant laxatives (such as sennosides, bisacodyl, or glycerine suppositories) are not recommended because of the chronicity of the problem and the risk for side effects. The efficacy of prokinetic agents (such as cisapride) is still not definitely proved. An oral colon washout solution (such as 250 to 500 mL polyethylene glycol solution daily) has also been suggested as a relatively safe approach (22); in some centers this treatment follows exclusion of significant pelvic floor or colonic disorders that require more specific treatment. A fuller discussion of the pharmacologic therapy to treat constipation is beyond the scope of this article. Those few who fail the simple measures delineated previously or require long-term treatment with stimulant laxatives with the attendant risk of developing laxative abuse syndrome are assumed to have intractable constipation and should have further investigation. Management of Intractable Constipation This section addresses a small minority, probably fewer than 1%, of all patients with constipation. These patients are most likely seen by gastroenterologists or in tertiary referral centers. Constipation should not be considered intractable until these measures have been followed and then fail. Further observation of the patient sometimes may reveal a previously unrecognized cause, such as laxative abuse, malingering, psychiatric disorder, and the Munchausen syndrome. In the remaining patients, recent studies suggest that evaluation of the physiologic function of the colon and pelvic floor help the physician rationally to choose treatment. The literature is replete with articles addressing these functions using different techniques. Our goal was to summarize the current state of the art to recommend tests that are relatively easy to perform and interpret and to facilitate an algorithmic approach to treating patients with intractable constipation. However, even among these highly selected patients with severe constipation, a cause can be identified only in about 30% of patients (see sections to follow). Physiologic Tests of Colonic Motor Function Two broad categories exist of physiologic tests of colonic motor function. These include transit tests, of which the simplest is the radiopaque marker transit test, and intraluminal testing, which involves placing various probes into the colon and measurement of its myoelectrical or contractile activity. The marker test (described in the next section) is simple and helps the physician decide whether to continue simple, empiric treatment if the transit time is normal, or to refer the patient for more specialized testing if it is prolonged. Such testing should be done only by

• Annals of Internal Medicine • Volume 121


• Number 7

centers with the expertise to interpret and act on the results. Radiopaque Marker Transit Tests and Scintigraphy Radiopaque marker transit tests are easy, repeatable and, generally applicable in evaluating constipated patients in clinical practice. Several variations on the same theme have been suggested, and the spectrum of tests was reviewed previously (23). The method is very simple. For example, 20 radiopaque markers are ingested at time zero, and an abdominal flat film taken on day 5 should indicate passage of 80% of the markers out of the colon (24). These are safe, cheap, reliable methods to document delayed colonic transit; however, they do not provide information regarding the transit profile of the stomach and small bowel, and avoidance of laxatives or enemas during the 4- to 7-day study period is essential. Radioscintigraphy using a delayed-release capsule containing radiolabeled particles has been used noninvasively to characterize normal or abnormal colonic function with low radiation exposure (25-28). There is close agreement between radiopaque marker and scintigraphic measurements of colonic transit (26). Radioscintigraphy provides a summary of colonic function for a period of 24 hours that has good predictive value for accelerated or delayed colonic transit (29). The advantages of this approach are the ability to assess simultaneously gastric and small bowel transit and the ability of scans taken in the first 24 hours to identify accelerated or delayed transit, which makes this test particularly useful in a tertiary referral practice. The disadvantages are the greater cost (five times that of an abdominal radiograph) and need for specific materials and preparation in a nuclear medicine laboratory. Transit measurements are particularly useful in clinical practice, provided normal values and daily variations are considered when interpreting the data. However, transit measurements do not provide a detailed examination of the colon's propulsive mechanisms unless there is simultaneous intraluminal recording by manometry or other methods. Intraluminal Measurement of Colonic Myoelectrical and Motor Function During three decades, intraluminal measurement of colonic myoelectric activity has been investigated with ring electrodes mounted on a polyvinyl tube and manometry with point sensors (30). Long spike bursts appear to have important propulsive function (30) and, along with highamplitude propagated contractions measured manometrically, they show a diurnal variation (31). However, the precise propulsive function of all other colonic electrical and lower-amplitude manometric activity is still unclear. Initial reports suggest that it may be possible to identify patterns of colonic motility that may be diagnostic of underlying pathophysiologic processes (32, 33), but this field is still in its infancy and these reports require confirmation. Colonic dilatation influences the measurement of these contractile patterns; when the colon's diameter exceeds 5.6 cm, the fidelity of amplitude measurements by manometric recordings is reduced (34). Barostat measurements, in which an infinitely compliant bag is apposed to the internal wall of the colon, may provide a way to

Figure 1. Dynamics of normal defecation. The process of defecation requires relaxation of the puborectalis and external anal sphincters, straightening of the rectoanal angle, and an increase in intraluminal pressure usually induced by a Valsalva maneuver to increase intra-abdominal pressure. Obstructed defecation may result if either of these functions is impaired. (Adapted and reproduced with permission from Camilleri M. Four patients with intractable constipation. Gastrointestinal Diseases Today. 1993;2:7-15.)

detect colonic tone and phasic events (28, 35). Thus, combined barostat and manometric recordings have shown marked changes in postprandial tone in carcinoid diarrhea (28) and absence of any alteration of postprandial tone in chronic megacolon (36). Combined methods using manometry and scintigraphy are also very useful for topographic and functional assessment of colonic motor activity and may provide a basis for differentiating pathophysiologic processes causing constipation (32, 33). A transit test such as a radiopaque marker evaluation appears to be the most applicable technique to assess overall colonic motor function in constipated patients, and it can be done by any clinician to exclude slow transit. More sophisticated approaches including scintigraphy and intraluminal measurements are applicable for research and may become part of routine clinical evaluations of colonic motor function. Thus, intraluminal measurements should be regarded as research tools for use in tertiary care centers to explore further the motor function of the colon in health and disease. It appears that combined methodologies are preferred to single approaches.

Anorectal and Pelvic Floor Tests, Balloon Expulsion, and Anorectal Manometry Constipation may result from disturbances in the dynamics of defecation (Figure 1) and thus may be associated with other complaints suggesting pelvic floor dysfunction (37). These include inability to evacuate the rectum, feeling of persistent rectal fullness, rectal pain, pelvic floor descent, and straining. Thus, symptoms such as the need to extract stool from the rectum digitally, application of pressure on the posterior wall of the vagina, or support of the perineum during straining imply a disturbance of the process of defecation; these significant symptoms should be contrasted with the sense of incom-


523

plete rectal evacuation, which is so common in the irritable bowel syndrome. Before further testing, an integral part of the assessment of patients with clinically suspected obstruction of defecation is review by a psychologist: Patients with eating disorders may have constipation. Psychologists also provide relaxation training for these patients and help identify depression that may have been missed. Constipation resistant to the measures discussed previously and associated with features of obstructed defecation should be investigated using available techniques such as defecography, anal manometry, and balloon expulsion. Defecography reveals abnormalities in as many as 50% of asymptomatic healthy persons: Rectal emptying was incomplete in 17 of 32, and in 10 of 32 there was internal intussusception, rectocele, and mucosal prolapse (38, 39). Others have shown a broad range of anorectal angle and pelvic floor descent in a study of 47 healthy young volunteers; these data overlap with reported pathologic states (40). In 187 patients with constipation referred for defecography in one center (39), internal intussusception (30% of patients) (41), nonrelaxing puborectalis (30%) (42), and rectocele (20%) (42) were frequent. The importance of these findings is unclear, but most of the symptoms related to these findings can be treated medically by increasing dietary fiber. In a very small proportion of patients, significant anatomic defects associated with intractable constipation respond best to surgical treatment. These defects include severe intussusception with complete outlet obstruction due to funnel-shaped plugging at the anal canal, or an extremely large rectocele that is preferentially filled during attempts at defecation instead of expulsion of the barium through the anus. Defecography requires an interested and experienced radiologist, and abnormalities are not pathognomonic for pelvic floor dysfunction. More commonly, outlet obstruction results from a nonrelaxing puborectalis muscle, which impedes rectal emptying. The literature is replete with tests to evaluate anorectal and pelvic floor function (Table 2). Which of these are most cost-effective in patients with severe chronic constipation? There is no clear answer because no single study has evaluated this broad spectrum of tests (37). Simplicity, availability, low costs, and practical utility of information obtained from the test are factors that positively affect our consensus. In our experience, a careful history suggesting abnormal defecation, such as excessive straining, finger disimpaction, and futility of enemas, and an abnormal balloon expulsion (see sections to follow) are highly (more than 90% predictive) suggestive of an important disturbance of the defecatory process. A simple test in the office to document a nonrelaxing puborectalis muscle is to have the patient strain to expel the index finger during a digital rectal examination. Motion of the puborectalis posteriorly during straining indicates proper coordination of the pelvic floor muscles. Measurement of perineal descent is relatively easy to gauge clinically by placing the patient in the left lateral decubitus position, and watching the perineum in order to assess either lack of descent (a sign of pelvic floor dysfunction) or ballooning during straining (43). The degree of descent can be measured by a perineometer with ref524

1 October 1994

erence to bony landmarks (typically the ischial tuberosities). One of the most useful tests is balloon expulsion (44). The overall defecatory process and relaxation of the puborectalis muscle is assessed by placing a urinary catheter in the rectum, inflating the balloon to 50 or 60 mL, and determining whether it can be expelled. Two approaches are used: In one, the patient is seated on a toilet and asked to pass the 60-mL balloon (44); in the second, the patient is in the left lateral decubitus position, and a bucket is suspended from the end of the catheter over a pulley at the level of the anal verge. The weight needed to facilitate expulsion of the balloon (normal, 0 to 200 g) is determined (45). Failure to expel the balloon is commonly associated with pelvic floor dysfunction, anatomic defects of the rectum, or anismus. If results of the balloon expulsion test are abnormal, patients with intractable constipation should be assessed by a psychologist to exclude denial or a "need to control" (44). The psychologist may also identify an eating disorder or depression, for which constipation is a predominant symptom. Further tests are indicated in some patients (such as defecography or anal endosonography to detect surgically correctable anatomic defects as prolapse or rectal wall defects, or rarely puborectalis electromyogram to assess spasm or denervation). The next section is a short discussion of the major indications and contraindications of these anorectal and pelvic floor tests. Anorectal manometry is not often contributory in patients with severe constipation. This test identifies rare syndromes, such as adult Hirschsprung disease, by the absence of the rectoanal inhibitory reflex or the presence of occult incontinence (37). An excessively high resting and squeeze anal sphincter tone suggests anismus (37, 45, 46). Dynamic imaging studies such as proctography during defecation or scintigraphic expulsion of artificial stool (45) help measure perineal descent and the rectoanal angle during rest, squeezing, and straining, and scintigraphic expulsion quantitates the amount of artificial stool emptied. Failure of the rectoangle to increase significantly (by about 20 degrees) during straining confirms pelvic floor dysfunction. Neurologic testing is more helpful in patients with incontinence than in those with symptoms suggesting obstructed defecation (47-49). External anal sphincter or puborectalis electromyography and pudendal nerve terminal motor latency prove denervation, but few centers do Table 2. Anorectal and Pelvic Floor Function Tests Balloon expulsion test Defecation proctography Anorectal manometry Perineometry Pudendal nerve terminal motor latencies Sphincter, puborectalis electromyogram Measurement of rectoanal angle Ultrasonography Scintigraphic expulsion of artificial stool Rectal sensation: mechanical, electrical Spinal evoked potentials by rectal stimulation Cerebral evoked potentials by rectal stimulation



• Number 7

these tests routinely (49). Electromyography of the puborectalis muscle is painful because of insertion of wires when concentric electromyography is done to document paradoxical contraction of the puborectalis during straining. Surface electromyography is not quantitative. Although surface electromyography using an intra-anal plug correlates well with invasive tests (50), it is artifactual and sometimes can provide a false reading. However, an increase in the activity recorded by the anal plug during straining is usually consistent with the presence of nonrelaxing puborectalis muscle (51). Among patients with chronic constipation, the absence of neurologic signs in the lower extremities suggests that any documented denervation of the puborectalis results from pelvic (such as obstetric) injury or from stretching of the pudendal nerve by chronic, long-standing straining (52). In patients with severe constipation and a megarectum, sensory perception of the rectal vault to distension may be impaired simply from chronic stretch. This may be documented with measurement of the earliest sensation using an air-inflated balloon or the maximal tolerable rectal volume before the urge to defecate is overwhelming (37, 45). This is most commonly found in idiopathic constipation and may be confused with Hirschsprung disease. During this inflation procedure, anal sphincter relaxation or the rectoanal inhibitory reflex is also assessed to exclude Hirschsprung disease. Sometimes this may require larger volumes of balloon inflation in patients with megarectum and idiopathic constipation because the rectum may have been chronically dilated and thus fails to respond to smaller volumes of inflation. A novel approach to evaluate anorectal structure is ultrasonography (53), which identifies sphincter defects and may help select patients for surgical correction (54). Spinal evoked responses during electrical rectal stimulation or stimulation of external anal sphincter contraction by magnetic stimulation over the lumbosacral cord identifies patients with limited sacral neuropathies with sufficient nerve conduction to attempt biofeedback training (55). A balloon expulsion test is an important screening test for anorectal dysfunction. If results are positive, an anatomic evaluation of the rectum or anal sphincters and a functional assessment of pelvic floor relaxation are the primary means to evaluate patients in whom obstructed defecation is suspected. Treatment of Patients with Severe Constipation in a Referral Practice Optimal management requires an attempt to classify patients with constipation in a specialized center using tests of colonic and pelvic floor function to delineate those with colonic inertia or slow transit constipation who will benefit from aggressive medical or surgical treatment, and those with anismus or pelvic floor dysfunction who usually respond to biofeedback management (Figure 2). Among 277 patients referred to one center with severe intractable constipation (56), but in whom no obvious cause, such as volvulus or megacolon, was found, the tests discussed previously categorized patients as having slow transit (n = 29), pelvic floor dysfunction (n = 37), slow transit and pelvic floor dysfunction (combined disorder,

Figure 2. Algorithmic approach to constipation in a tertiary referral practice. * If the result of a second colonic transit test is abnormal (left side of algorithm), the patient requires further testing similar to patients whose first result of a colonic transit test is abnormal.

n = 14), and constipation-predominant irritable bowel syndrome or normal transit constipation (n = 197). Thus, even among a very highly selected referral population of constipated patients, only 29% had a definable abnormality. Patients who had slow transit constipation are routinely treated with bulk, osmotic, and stimulant laxatives. The approaches in different centers include use of fiber, psyllium, polyethylene glycol (colonic lavage solution), and bisacodyl. Preliminary data suggest that misoprostol (prostaglandin E! analog) may also be useful in these patients (57), but long-term studies are needed. When a 2- to 3-month trial of medical therapy is not efficacious in patients with documented slow transit constipation unassociated with obstructed defecation, surgery is indicated. In the study noted previously (56), 29 patients had abdominal colectomy and ileorectostomy, which was successful in nearly all of the patients during a median follow-up of 2 years, with a mean number of bowel movements of 4 per day and no patients with incontinence. Patients with a combined disorder had pelvic floor retraining (biofeedback and muscle relaxation) first, followed by colectomy and ileorectostomy if results of colonic transit studies were not normal with biofeedback alone. Patients with pelvic floor dysfunction only had biofeedback training with an 80% success rate measured by achieving comfortable stool habits (58). The success rate for colonic surgery for intractable constipation depends on careful selection of patients: Specifically, pelvic floor dysfunction must be excluded, and those with a generalized gastrointestinal motility disorder have a 50% chance of recurrent symptoms or other abdominal complaints after colectomy (5961). Thus, gastric and small-bowel motility studies (transit, manometry) are advocated before performing colectomy for severe chronic constipation. The decision to operate is made more easily if megacolon and megarectum are present (62). Complications after surgery include smallbowel obstruction (11%) (56) and fecal soiling, particularly at night time during the first year after surgery. A recent report (63) from St. Mark's Hospital, London, En-


525

gland, suggests that in two patients with selective delays of transit through the left colon, segmental resection of this region provides symptom relief, with less postoperative diarrhea, which occurs in as many as 30% of patients during the first year after subtotal colectomy with ileo rectostomy (56). Reports of the experience of other institutions with segmental resections are necessary before this approach can be generally recommended. The operative management of obstructed defecation is controversial. Surgical approaches have included partial division of the puborectalis muscle and anorectal myectomy. Wasserman (64) reported excision of a portion of the puborectalis muscle with success in four patients. Initially, others confirmed these findings in a larger group of 33 patients (65). More recently, others have tried to replicate their success; Barnes and colleagues (66) divided the posterior puborectalis muscle in nine patients, but only two were improved and five became incontinent. These same investigators divided the puborectalis muscle in both the left and right lateral quadrants in nine patients and results were also poor (67). Anorectal myectomy similarly fails over time because the operation removes a portion of the involuntary internal anal sphincter. Martelli and coworkers (68) described success with this procedure (importantly, 22% of their patients had aganglionosis or Hirschsprung disease) without excision of part of the sphincter. Others performed myectomy alone and, despite good early results, they later found that two thirds of the patients were no better (69). A nonoperative approach to treat puborectalis spasm is injection of botulinum A toxin into the muscle (70); selective weakening of the muscle has been reported, but symptoms were improved in only one half of the patients and then only temporarily. As noted previously, the preferred method to treat obstructed defecation is pelvic floor retraining (biofeedback), which aims to reeducate the muscles of the pelvic floor to relax during defecation straining (71, 72). Pelvic floor retraining is a multimodality approach aimed at teaching patients to defecate effectively. Dietary manipulation, stool bulkers, biofeedback techniques, balloon pullthrough, and defecation stimulation are all used. Psychological counseling is helpful, particularly in those patients in whom the disorder is a manifestation of a psychological disorder. Attempts to manage pelvic floor dysfunction with operations (internal anal sphincter or puborectalis muscle division) have had only mediocre success. Because the results of biofeedback training, a completely noninvasive approach, are so good (with a mean of 67.7% patients improved [73]), any other form of management would seem irrational. Conclusions In primary care, the cause of constipation is most often related to diet, fluid intake, and psychological factors. The management approach should include a thorough history to exclude structural diseases of the colon and rectum and systemic diseases or medication that may slow gastrointestinal transit. In most cases, use of simple measures such as regular exercise, dietary fiber, and ample fluid intake are successful. This is also true for most consti526

pated patients seen by specialists. Full compliance with a high-fiber diet using bran or psyllium preparations helps most patients. In a few, long-term use of an osmotic laxative such as lactulose and a polyethylene glycol solution may be necessary. Colonic transit time should be measured in patients who are still constipated; this can be done by any physician with the assistance of a radiologist. If the transit time is normal, the probability of a psychological or behavioral disorder must be considered. If the transit time is prolonged (that is, more than 20% of ingested markers are still in the colon after 5 days) despite the measures noted previously, then referral to a center equipped to do specialized studies is indicated. Tests for colonic inertia and pelvic floor dysfunction require careful interpretation. Only a few patients with intractable constipation and persistently slow transit despite medical therapy should be considered for surgery, and the current consensus is to do a subtotal colectomy with ileorectostomy. When test results show pelvic floor dysfunction, behavioral treatments such as biofeedback are successful in about 70% of patients, and results of surgery are not very good. The long-term outcome is excellent when an algorithmic approach is followed and patients are carefully selected for treatment with subtotal colectomy or biofeedback. Nevertheless, this field still requires well-structured clinical and applied research because several questions pertaining to the common symptom of constipation must still be answered: What is the prevalence of structural diseases among patients coming to their primary care physicians with constipation? Is the mechanism of colonic inertia related to alterations in one or more populations of enteric neurones and can these alterations be corrected with pharmacologic approaches? What is the prevalence of pelvic floor dysfunction among patients with constipation-predominant irritable bowel syndrome? Are there pharmacologic approaches to relieve pelvic floor dysfunction? We hope that our summary of the current state of the art will stimulate further research in this important area of clinical gastroenterology. Requests for Reprints: Dr. Michael Camilleri, Gastroenterology Research Unit, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Current Author Addresses: Dr. Camilleri: Gastroenterology Research Unit, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Dr. Thompson: Ottawa Civic Hospital, 1053 Carling Avenue, Ottawa, Ontario, Canada K1Y 4E9. Dr. Fleshman: Division of Colon and Rectal Surgery, Jewish Hospital, 216 South Kings Highway, St. Louis, MO 63110. Dr. Pemberton: Section of Colon and Rectal Surgery, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.

References 1. Connell AM, Hilton C, Irvine G. Variation of bowel habit in two population samples. BMJ. 1965;2:1095-9. 2. Thompson WG, Heaton KW. Functional bowel disorders in apparently healthy people. Gastroenterology. 1980;79:283-8. 3. Sandler RS, Drossman DA. Bowel habits in young adults not seeking health care. Dig Dis Sci. 1987;32:841-5. 4. Burkitt DP, Walker AR, Painter NS. Effect of dietary fibre on stools and transit-times, and its role in the causation of disease. Lancet. 1972;21408-12. 5. Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ 3d. Functional constipation and outlet delay: a population-based study. Gastroenterology. 1993;105:781-90. 6. Thompson WG, Creed F, Drossman DA, Heaton KW, Mazzacca G. Functional bowel disease and functional abdominal pain. Gastroenterology International. 1992;5:75-91. 7. Drossman DA, Li Z, Andruzzi E, Temple RD, Talley NJ, Thompson

1 October 1994 • Annals of Internal Medicine • Volume 121 • Number 7


WG, et al. U.S. householder survey of functional gastrointestinal disorders. Prevalence, sociodemography, and health impact. Dig Dis Sci. 1993;38:1569-80. 8. Campbell AJ, Busby WJ, Horwath CC. Factors associated with constipation in a community based sample of people aged 70 years and over. J Epidemiol Community Health. 1993;47:23-6. 9. Johanson JF, Sonnenberg A. The prevalence of hemorrhoids and chronic constipation. An epidemiologic study. Gastroenterology. 1990; 98:380-6. 10. Johanson JF, Sonnenberg A, Koch TR. Clinical epidemiology of chronic constipation. J Clin Gastroenterol. 1989;11:525-36. 11. Heaton KW, Radvan J, Cripps H, Mountford RA, Braddon FE, Hughes AO. Defecation frequency and timing, and stool form in the general population: a prospective study. Gut. 1992;33:818-24. 12. Bannister J J, Davison P, Timms JM, Gibbons C, Read NW. Effect of stool size and consistency on defecation. Gut. 1987;28:1246-50. 13. MacDonald L, Freeling P. Bowels: beliefs and behaviour. Fam Pract. 1986;3:80-4. 14. Probert CS, Emmett PM, Cripps HA, Heaton KW. Evidence for the ambiguity of the term constipation—the role of irritable bowel syndrome. Gut; [In press]. 15. Ashraf W, Srb F, Lof J, Quigley EM. Idiopathic constipation: subjective complaints vs. objective assessment [Abstract]. Gastroenterology. 1994;106:A461. 16. Tucker DM, Sandstead HH, Logan GM Jr, Klevay LM, Mahalko J, Johnson LK, et al. Dietary fiber and personality factors as determinants of stool output. Gastroenterology. 1981;81:879-83. 17. Murtagh J. Constipation. Aust Fam Physician. 1990;19:1693-7. 18. Blackstone MO, Kirsner JB. Clinical application of diagnostic tests in selected colonic disorders. In: The Large Intestine: Physiology, Pathophysiology and Disease. Phillips SF, Pemberton JH, Snorter RG, eds. New York; Raven Press: 1991; 253-301. 19. Kronborg O, Ostergaard A. Evaluation of the barium-enema examination and colonoscopy in diagnosis of colonic cancer. Dis Colon Rectum. 1975;18:674-7. 20. Thoeni RF, Venbrux AC. The value of colonoscopy and double-contrast barium-enema examinations in the evaluation of patients with subacute and chronic lower intestinal bleeding. Radiology. 1983; 146: 603-7. 21. Cummings JH. Constipation, dietary fibre and the control of large bowel function. Postgrad Med J. 1984;60:811-9. 22. Andorsky RI, Goldner F. Colonic lavage solution (polyethylene glycol electrolyte lavage solution) as a treatment for chronic constipation: a double-blind, placebo-controlled study. Am J Gastroenterol. 1990;85: 261-5. 23. von der Ohe MR, Camilleri M. Measurement of small bowel and colonic transit: indications and methods. Mayo Clin Proc. 1992;67: 1169-79. 24. Hinton JM, Lennard-Jones JE, Young AC. A new method for studying gut transit times using radioopaque markers. Gut. 1969;10:842-7. 25. Proano M, Camilleri M, Phillips SF, Brown ML, Thomforde GM. Transit of solids through the human colon: regional quantification in the unprepared bowel. Am J Physiol. 1990;258:G856-62. 26. Stivland T, Camilleri M, Vassallo M, Proano M, Rath D, Brown M, et al. Scintigraphic measurement of regional gut transit in idiopathic constipation. Gastroenterology. 1991;101:107-15. 27. Vassallo M, Camilleri M, Phillips SF, Brown ML, Chapman NJ, Thomforde GM. Transit through the proximal colon influences stool weight in the irritable bowel syndrome. Gastroenterology. 1992; 102: 102-8. 28. von der Ohe M, Camilleri M, Kvols LK, Thomforde GM. Motor dysfunction of the small bowel and colon in patients with the carcinoid syndrome and diarrhea. N Engl J Med. 1993;329:1073-8. 29. Camilleri M, Zinsmeister AR. Towards a relatively inexpensive, noninvasive, accurate test for colonic motility disorders. Gastroenterology. 1992;103:36-42. 30. Bueno L, Fioramonti J, Ruckebusch Y, Frexinos J, Coulom P. Evaluation of colonic myoelectrical activity in health and functional disorders. Gut. 1980;21:480-5. 31. Narducci F, Bassotti G, Gaburri M, Morelli A. Twenty-four hour manometric recording of colonic activity in healthy man. Gut. 1987; 28:17-25. 32. Bazzocchi G, Ellis J, Villanueva-Meyer J, Jing J, Reddy SN, Mena I, et al. Postprandial colonic transit and motor activity in chronic constipation. Gastroenterology. 1990;98:686-93. 33. Di Lorenzo C, Flores AF, Reddy SN, Snape WJ Jr, Bazzocchi G, Hyman PE. Colonic manometry in children with chronic intestinal pseudo-obstruction. Gut. 1993;34:803-7. 34. von der Ohe M, Hanson RB, Camilleri M. Comparison of simultaneous recordings of human colonic contractions by manometry and a barostat. Neurogastroenterology and Motility. 1994; [In press]. 35. Steadman CJ, Phillips SF, Camilleri M, Haddad A, Hanson RB. Variation of muscle tone in the human colon. Gastroenterology. 1991; 101:373-81. 36. von der Ohe MR, Camilleri M, Carryer PW. A patient with localized

1 October 1994 Downloaded From: http://annals.org/ by a Mayo Clinic User on 10/05/2016

megacolon and intractable constipation: evidence for impairment of colonic muscle tone. Am J Gastroenterol. 1994; [In press]. 37. Pemberton JH. Anorectal and pelvic disorders: putting physiology into practice. J Gastroenterol Hepatol. 1990;5:127-43. 38. Goei R, van Engelshoven J, Schouten H, Baeten C, Stassen C. Anorectal function: defecographic measurement in asymptomatic subjects. Radiology. 1989;173:137-41. 39. Goei R. Anorectal function in patients with defecation disorders and asymptomatic subjects: evaluation with defecography. Radiology. 1990; 174:121-3. 40. Shorvon PJ, McHugh S, Diamant NE, Somers S, Stevenson GW. Defecography in normal volunteers: results and implications. Gut. 1989;30:1737-49. 41. Fleshman JW, Kodner U, Fry RD. Internal intussusception of the rectum: a changing perspective. Netherlands J Surg. 1989;41:145-8. 42. Selvaggi F, Pesce G, Scotto Di Carlo E, Maffetone V, Canonico S. Evaluation of normal subjects by defecographic technique. Dis Colon Rectum. 1990;33:698-702. 43. Pezim M, Pemberton JH, Phillips SF. The immobile perineum: pathophysiologic implications in severe constipation [Abstract]. Dig Dis Sci. 1987;32:A924. 44. Fleshman JW, Dreznik Z, Cohen E, Fry RD, Kodner U. Balloon expulsion test facilitates diagnosis of pelvic floor outlet obstruction due to nonrelaxing puborectalis muscle. Dis Colon Rectum. 1992;35:101925. 45. Pezim ME, Pemberton JH, Levin KE, Litchy WJ, Phillips SF. Parameters of anorectal and colonic motility in health and in severe constipation. Dis Colon Rectum. 1993;36:484-91. 46. Preston DM, Lennard-Jones JE. Anismus in chronic constipation. Dig Dis Sci. 1985;30:413-8. 47. Swash M, Snooks SJ, Henry MM. Unifying concept of pelvic floor disorders and incontinence. J R Soc Med. 1985;78:906-11. 48. Womack NR, Morrison JFB, Williams NS. The role of pelvic floor denervation in the etiology of idiopathic faecal incontinence. Br J Surg. 1986;73:404-7. 49. Lubowski DZ, Jones PN, Swash M, Henry MM. Asymmetrical pudendal nerve damage in pelvic floor disorders. Int J Colorectal Dis. 1988;3:158-60. 50. Parks TG. The usefulness of tests in anorectal disease. World J Surg. 1992;16:804-10. 51. Kuijpers HC, Bleijenberg G. The spastic pelvic floor syndrome. A cause of constipation. Dis Colon Rectum. 1985;28:669-72. 52. Bartolo DC. Pelvic floor disorders. Incontinence, constipation, and obstructed defecation. Perspect Colon and Rectal Surg. 1988;1:1-24. 53. Bartram CI. Anal endosonography. Ann Gastroenterol Hepatol. 1992; 28:185-9. 54. Felt-Bersma RJ, Cuesta MA, Koorevaar M, Strijers RL, Meuwissen SG, Dercksen EJ, et al. Anal endosonography: relationship with anal manometry and neurophysiologic tests. Dis Colon Rectum. 1992;35: 944-9. 55. Herdmann J, Bielefeldt K, Enck P. Quantification of motor pathways to the pelvic floor in humans. Am J Physiol. 1991;260:G720-3. 56. Pemberton JH, Rath DM, Ilstrup DM. Evaluation and surgical treatment of severe chronic constipation. Ann Surg. 1991;214:403-11. 57. Soffer EE, Metcalf A, Launspach J. Misoprostol is effective treatment for patients with severe chronic constipation [Abstract]. Gastroenterology. 1993;104:A585. 58. Juhasz ES, Pemberton JH, Rath DM. Early and late results of pelvic floor retraining for pelvic floor discoordination [Abstract]. Gastroenterology. 1993;104:A530. 59. Loening-Baucke VA, Anuras S, Mitros FA. Changes in colorectal function in patients with chronic colonic pseudoobstruction. Dig Dis Sci. 1987;32:1104-12. 60. Leon SH, Krishnamurthy S, Schuffler MD. Subtotal colectomy for severe idiopathic constipation. A follow-up study of 13 patients. Dig Dis Sci. 1987;32:1249-54. 61. Preston DM, Hawley PR, Lennard-Jones JE, Todd IP. Result of colectomy for severe idiopathic constipation in women (Arbuthnot Lane's disease). Br J Surg. 1984;71:547-52. 62. Stabile G, Kamm MA, Hawley PR, Lennard-Jones JE. Colectomy for idiopathic megarectum and megacolon. Gut. 1991;32:1538-40. 63. Kamm MA, van der Syp JR, Hawley PR, Phillips RK, Lennard-Jones JE. Left hemicolectomy with rectal excision for severe idiopathic constipation. Int J Colorectal Dis. 1991;6:49-51. 64. Wasserman IF. Puborectalis syndrome (rectal stenosis due to anorectal spasm). Dis Colon Rect. 1964;7:87-9. 65. Wallace WC, Madden WM. Experience with partial resection of the puborectalis muscle. Dis Colon Rect. 1969;12:196-200. 66. Barnes PR, Hawley PR, Preston DM, Lennard-Jones JE. Experience of posterior division of the puborectalis muscle in the management of chronic constipation. Br J Surg. 1985;72:475-7. 67. Kamm MA, Hawley PR, Lennard-Jones JE. Lateral division of the puborectalis muscle in the management of severe constipation. Br J Surg. 1988;75:661-3. 68. Martelli H, Devroede G, Ahran P, Duguay C. Mechanisms of idio-


• Number 7

527

pathic constipation: outlet obstruction. Gastroenterology. 1978;75:62331. 69. Yoshioka K, Keighley MR. Anorectal myectomy for outlet obstruction. Br J Surg. 1987;74:373-6. 70. Hallan RI, Williams NS, Melling J, Waldron DJ, Womack NR, Morrison JF. Treatment of anismus in intractable constipation with botulinum A toxin. Lancet. 1988;2:714-7.

71. Kuijpers HC, Bleijenberg G. Assessment and treatment of obstructed defecation. Ann Med. 1990;22:405-11. 72. Wexner SD, Cheape JD, Jorge JM, Heymen S, Jagelman DG. Prospective assessment of biofeedback for the treatment of paradoxical puborectalis contraction. Dis Colon Rectum. 1992;35:145-50. 73. Enck P. Biofeedback training in disordered defecation. A critical review. Dig Dis Sci. 1993;38:1953-60.

Whoever is to acquire a competent knowledge of medicine ought to be possessed of the following advantages: a natural disposition; instruction; a favorable position for the study; early tuition; love for labor; leisure. First of all, a natural talent is required; for, when Nature opposes, everything else is vain. Hippocrates The Law Submitted by: Mumtaz A. Siddiqui, MD The Graduate Hospital Philadelphia, PA 19146

Submissions from readers are welcomed. If the quotation is published, the sender's name will be acknowledged. Please include a complete citation, as done for any reference.—The Editors

528

1 O c t o b e r 1994

• Annals

of Internal

Medicine

• V o l u m e 121


• Number 7