2 The procedure

2.1 Indications

2.1.1 The causes of faecal incontinence are diverse. Existing treatment options include medical therapy, biofeedback techniques and surgery in selected patients. Surgical treatments include sphincter repair, sacral nerve stimulation, encirclement procedures and muscle transposition (for example, dynamic graciloplasty). Some patients may require acolostomy if other treatments fail.

2.2 Outline of the procedure

2.2.1 Implantation of an artificial anal sphincter is used to treat severe faecal incontinence. In this procedure, a fluid-filled cuff is implanted around the anal canal. Tubing from the cuff is channelled under the skin of the perineum and connected to a control pump placed subcutaneously in the scrotum or labia. The control pump is connected by tubing to a pressure-regulating balloon implanted in the abdominal wall. The cuff simulates the natural function of the sphincter muscle; when the fluid is displaced from the cuff to the balloon via the patient-controlled pump, defaecation can take place. Once defaecation is complete, the fluid is slowly returned to the cuff and continence is again achieved. For more details, refer to the Sources of evidence section.

2.3 Efficacy

2.3.1 No controlled studies were identified. Some of the studies identified were small and some had high losses to follow-up. Among the studies identified, removal of the artificial sphincter system was required in 19% (10/53) to 41% (7/17) of patients. In patients who had not undergone explantation, all the studies showed improvement in continence. However, different measures of continence were used in the studies. The studies that reported manometric results showed increased mean anal pressures after implantation. For more details, refer to the Sources of evidence section.

2.3.2 The Specialist Advisors considered the main efficacy concern to be the frequent need to remove the implanted artificial sphincter.

2.4 Safety

2.4.1 The largest study identified reported that device-related complications occurred in 86% (99/115) of patients. The most common adverse events reported in this study were: infection 33% (38/115); pain 32% (37/115); erosion 21% (24/115); faecal impaction 18% (21/115); faecal incontinence 18% (21/115); constipation 17% (20/115); surgical injury 13% (15/115); wound problems 10% (11/115); difficult evacuation 9% (10/115); and wound dehiscence 9% (10/115). For more details, refer to the Sources of evidence section.

2.4.2 The Specialist Advisors considered the main safety concerns to be infection, erosion and evacuation difficulties.

2.5 Other comments

2.5.1 The procedure may have a place in the treatment of patients who are unsuitable for sacral nerve stimulation.

2.5.2 There is a significant rate of complications, such as infection, cuff erosion, wound dehiscence and haematoma, and patients may require revisional surgery or removal of the device. Fully informed consent is therefore particularly important.

Andrew Dillon
Chief Executive
June 2004