An increasing number of adult patients with severe constipation who have not responded to the usual dietary measures and laxatives are being seen in surgical practice. This is not a trivial symptom and the patient's life both at home and at work is often severely restricted by abdominal pain, distension, a poor sense of well-being, and decreased bowel frequency. The symptom of constipation cannot be precisely defined. Patients have a problem with defecation and may be worried by infrequent and/or hard stools which are difficult to pass. Different types of severe constipation have a different cause and require a different therapeutic approach.

Patients with minor, easily treated symptoms do not require sophisticated investigation. A careful history of the exact nature of the constipation, together with any history of sexual or psychiatric problems, is essential. Abdominal examination will reveal the presence of a palpable faecal mass. Weakness of the pelvic floor is indicated by extreme descent of the perineum on straining. Laxity of the anal sphincter and perianal soiling is usually associated with gross faecal impaction in the rectum.

Patients can generally be divided into those with a colon of normal width and those with an abnormally wide or capacious rectum and/or colon: a barium enema defines into which of these two groups the patient falls. In those patients with rectal impaction, an unprepared radiograph using a water-soluble contrast material will illustrate the rectum and the distal colon with the impacted stools still in place. The rate of transit of contents through the colon can be estimated by a single abdominal radiograph taken 5 days after the patient has swallowed 20 radio-opaque markers. In specialist units, information about evacuation disorders can be obtained by recording the appearance of the rectal and anal canal on video tape or fast film sequences as the patient expels a barium gel or paste. Simple manometric studies can be used to test rectal sensation and the rectoanal inhibitory reflex by distending a balloon in the rectal ampulla.

A small proportion of patients with Hirschprung's disease do not present until adolescence or young adult life. Elliot and Todd reported 39 such patients treated at St Mark's Hospital, London, between 1965 and 1984. There were 26 males and 13 females with a mean age of 23 years. The Duhamel operation produced good results, with complete continence, and daily evacuation, in 36 patients; the remaining three patients had poorer results, with occasional incontinence. Investigation of this disorder should include a barium enema which will reveal the characteristic junction between the dilated proximal colon and the aganglionic segment (Fig. 1) 1110. Anorectal physiology studies will show an absent rectosphincteric reflex, and full thickness rectal biopsy will demonstrate the absence of ganglion cells in the myenteric plexus. A frozen section biopsy of the normal colon during the operation will ensure that normally innervated bowel is brought down behind the aganglionic rectum.

Once Hirschprung's disease and metabolic and other secondary causes of a dilated colon have been excluded there remains a group of patients with a dilated large bowel of unknown cause. This condition affects males and females equally and it may commence in childhood or adult life. Recurrent faecal impaction is the usual presentation, though in childhood it may present with soiling. The nerve plexuses and muscle coats appear grossly normal on histological assessment and the aetiology of this condition remains unknown. The rectal diameter is usually greater than 6.5 cm. In two-thirds of cases the rectum alone is affected, but in one-third of cases the dilation extends into the sigmoid colon.

It is worth trying medical treatment initially, prescribing a combination of enemas, suppositories, and osmotic laxatives which will keep the rectum empty. If laxatives are poorly tolerated or are ineffective and it seems unlikely that the grossly dilated bowel will recover, surgery is indicated. It is not clear which is the best procedure to offer these patients. The alternatives include a preliminary loop ileostomy followed by either a Duhamel procedure or a resection and coloanal anastomosis. In those with a moderately dilated colon a colectomy with ileorectal anastomosis may offer a reasonable compromise. In a recent series of 40 patients undergoing a colectomy for this condition 80 per cent achieved normal bowel frequency and most were relieved of the need for laxatives. A third of the patients, however, continued to experience some abdominal pain. Ileorectal anastomosis produced results superior to those of a caecorectal anastomosis or a sigmoid resection: these last operations had a higher incidence of persistent constipation. In patients with a grossly dilated rectum the most commonly performed procedure was the Duhamel operation. It seems however that results in this situation are less satisfactory than those obtained in patients with Hirschprung's disease. In a series of 20 patients reported by Stabile only half achieved a normal bowel frequency. Seven patients remained constipated and of these five required further surgery. The other alternative is a resection of the grossly dilated rectum and a coloanal anastomosis. This usually has to be performed using a hand suturing technique because of the thickness of the rectal stump. There are no reliable data on the outcome of this procedure.

The patient who remains constipated despite colectomy may be treated by creation of a stoma, bringing out proximal bowel of normal diameter. If there is progressive dilatation of the proximal colon an ileostomy may be necessary and, in very selected cases, a restorative proctocolectomy would be an alternative approach.

Preston and Leonard Jones (1986) reported a series of 64 women complaining of severe constipation. The patients passed about one stool weekly with the aid of laxatives and were greatly troubled by abdominal pain, bloating, malaise, and nausea to the extent that their symptoms were an increasing social disability. A decrease in bowel frequency and other symptoms were often noticed around the age of puberty and slowly became worse until they were at their most severe by the third decade. In a few patients the symptoms appeared suddenly after an abdominal operation or an accident.

A barium enema in these patients shows a normal rectal and colonic diameter; on large bowel transit studies, less than 80 per cent of the 20 shapes or radio-opaque markers ingested on the first day have been passed on the fifth day (Fig. 3) 1112. The rectosphincteric reflex is normal, and a full thickness rectal biopsy is not normally required, though in doubtful cases it is sensible to exclude Hirschprung's disease.

As is the case in patients with megarectum, it is as well to institute medical therapy for as long as possible, aimed at emptying the rectum with osmotic laxatives, suppositories, or enemas. These patients have usually tried the whole range of laxatives and many of them become increasingly constipated despite quite high doses of concurrent laxative medication. Interestingly, high fibre diets often make their symptoms worse.

The surgical options include a segmental colectomy, a colectomy and caecorectal anastomosis, colectomy and ileorectal anastomosis, a proctocolectomy and ileoanal reservoir, or an ileostomy. It is now generally believed that segmental colectomies have nothing to offer: patients tend to become progressively constipated within a year or so of this operation. The most widely used operation is colectomy and ileorectal anastomosis. In a series reported by Kamm et al. (1988) of 44 patients, all female with a median evacuation frequency of once every 4 weeks, 22 patients were reported to be normal, 17 had diarrhoea, and 5 had recurrent constipation after this procedure. A worrying feature was the fact that 71 per cent of the patients had continuing abdominal pain. In the series reported by Roe et al. (1988) 22 patients underwent colectomy and ileorectal anastomosis. There were four failures, two with pain and bloating, one with continuing constipation, and one patient had recurrent adhesions. In the study by Kamm et al. (1988) 10 patients needed psychiatric treatment for severe psychological disorders. While many of these patients are completely normal there is a small subgroup with severe and often occult psychological problems who need to be very carefully investigated and managed prior to surgery. If a colectomy and ileorectal anastomosis fails there is the possibility of creating a stoma. Van der Syp et al. (1990) reported 37 such patients; those who improved most following creation of a stoma were those with idiopathic constipation who had not undergone previous surgery. Of the 10 patients who had had a failed ileorectal anastomosis only 50 per cent were improved by a colostomy. A number of these patients continued to have pain and bloating and a continued use of laxatives. There is the possibility, however, in the colostomy group of colostomy irrigation, and even a colostomy plug.

This is a rare group of patients who have severe constipation and a dilated colon, but also dilatation of the upper intestine. In one-third of patients the disorder shows Mendelian inheritance. Histological studies have demonstrated abnormalities of nerve plexuses and smooth muscle in most patients. Treatment is usually with prokinetic agents to increase intestinal motility. A number of patients also require nutritional support.

Some patients will give a clear history of an evacuatory disorder. Despite what seems to be a normal propulsive effort by the colon with stool in the rectum, the patient is unable to expel it. Evacuatory disorders have been given a number of names including outlet obstruction, anismus, pelvic floor spasm, the solitary ulcer syndrome, rectal intussusception, and the descending perineum syndrome.

The patient gives a characteristic history of evacuatory difficulty. This may involve long periods of time sitting straining at stool, manipulation of the anal margin, pressure on the perineal body, and, in female patients, pressure on the posterior wall of the vagina. In some patients the distress becomes so extreme that the anal canal is digitated and the faecal material manually removed.

The cause of this condition is not known: there may be a number of underlying problems ranging from an inappropriate contraction of the pelvic floor to an occult rectal intussusception. Physiological studies usually show a normal anal sphincter and rectum but the key investigation is the evacuation proctogram. A mixture of some form of starch and barium or a dilute barium paste is inserted into the rectum and the patient is then asked to strain as at stool. The time taken to expel the simulated stool is noted and any abnormalities in the architecture of the rectal wall recorded (Fig. 4) 1113.

The surgical options include rectal prolapse repair, anorectal myectomy, lateral division of the puborectalis, repair of a rectocele, and colostomy. The demonstration of spasm of the pelvic floor led Kamm et al. (1988) to use lateral division of the puborectalis muscle for the treatment of 15 patients with this condition: four of them improved while three had mild incontinence. The operation has now largely been abandoned. Pinho et al. (1989) used a variation of a procedure advocated for patients with Hirschprung's disease. A long myectomy of the internal anal sphincter and the rectal smooth muscle wall was made in the midline, posteriorly. There is, however, no evidence that these patients have an abnormality of the smooth muscle of either the anal sphincter or the rectum. In 63 patients with a 30-month follow-up there was no functional improvement in 70 per cent and 10 per cent of the patients developed mild incontinence. When an occult rectal intussusception is found on evacuation proctogram it is tempting to think that a rectopexy rather as for a rectal prolapse procedure would solve the problem. However, in a group of 17 patients reported by Roe et al. (1986), only four treated by rectopexy required no further treatment, five went on to require a subtotal colectomy, two had an ileostomy, and one a sphincterotomy. Only two patients had a successful result. This is an extremely difficult group of patients to manage satisfactorily. Often a careful explanation of the problem and reassurance that digitation for example does not do any harm will suffice. When it is impossible to distinguish between slow transit constipation and an evacuation disorder a loop ileostomy will allow time for further evaluation. If it is quite clear that there is an evacuatory disorder it is worth creating a colostomy and using colostomy irrigation as a semipermanent solution.

Bartolo DCC, Roe AM, Virjee J, Mortensen NJMcC. Evacuation proctography in obstructed defaecation and rectal intussusception. Br J Surg 1985; 72 (suppl): 111–6.

Elliot MS, Todd IP. Adult Hirschprung's disease: results of the Duhamel procedure. Br J Surg 1985; 72: 884–5.

Hinton JM, Lennard Jones JE, Young AC. A new method for studying gut transit times using radio opaque markers. Gut 1969; 10: 842–7.

Hosie KB, Kmiot WA, Keighley MRB. Constipation: another indication for restorative proctocolectomy. Br J Surg 1990; 77: 801–2.

Kamm MA, Hawley PR, Lennard Jones JE. Outcome of colectomy for severe idiopathic constipation. Gut 1988; 29: 969–75.

Pinho M, Yoshioka K, Keighley MRB. Longterm results of anorectal myectomy for chronic constipation. Br J Surg 1989; 76: 1163–4.

Preston DM, Lennard Jones JE. Severe chronic constipation of young women: ‘idiopathic slow transit constipation’. Gut 1986; 27: 41–8.

Roe AM, Bartolo DCC, Mortensen NJMcC. Diagnosis and surgical management of intractable constipation. Br J Surg 1986; 73: 854–61.

Stabile G, Kamm MA, Hawley PR, Lennard Jones JE. Results of the Duhamel operation in the treatment of idiopathic megarectum and megacolon. Br J Surg 1991; (b) 78: 661–3.

Stabile G, Kamm MA, Hawley PR, Lennard Jones JE. Colectomy for idiopathic megarectum and megacolon. Gut 1991: 32: 1538–40.