Sacral neuromodulation in functional urinary retention: an effective way to restore voiding


S.R. Aboseif, Director, Neurourology & Reconstructive Surgery, Department of Urology, Kaiser Permanente Medical Center, 4900 Sunset Boulevard, 2nd Floor, Los Angeles, CA 90027, USA.


Objective  To determine the long-term efficacy and complications of sacral nerve stimulation as an alternative therapy for functional unobstructive urinary retention, often considered to be psychogenic and effectively treated by clean intermittent catheterization, but for which pelvic floor dysfunction has been recognized as a possible cause.

Patients and methods  Twenty patients (17 women and three men, mean age 48 years) with idiopathic, unobstructive functional urinary retention and in whom other forms of therapy had failed, had a pulse generator implanted (Medtronic, Minneapolis, MN, USA) and a sacral nerve implant. Their mean duration of symptoms was 68 months; 13 patients had chronic pelvic and perineal pain associated with their obstructive voiding symptoms. All patients were managed with clean intermittent catheterization and pharmacological therapy (α-blockers) before the procedure. All patients had a percutaneous nerve evaluation before the permanent implant, which showed> 50% improvement in their symptoms. All patients were evaluated at 1, 6 12, 18 and 24 months, then yearly thereafter. The results were assessed both subjectively by patient's symptoms and objectively by checking the postvoid residual volume (PVR) and voided volume.

Results  Eighteen patients were able to void spontaneously with a mean increase in voided volume from 48 to 198 mL, and a significant decrease in PVR from 315 to 60 mL. Eighteen of the patients had a ≥ 50% improvement in their symptoms and said they would recommend the therapy to a friend or relative. Complications occurred in six patients.

Conclusion  Sacral nerve stimulation is an effective and durable new approach to functional urinary retention, with few associated complications. Test stimulation provides a valuable tool for selecting patients.


Functional urinary retention with no evidence of mechanical obstruction has been a challenging problem for many years; the exact cause is poorly understood. Most patients were designated ‘hysterical’ and diagnosed as having ‘psychogenic urinary retention’[1–4]. Patients were initially treated with one or several drugs including α-blockers, muscle relaxants and antidepressants, frequent urethral dilatation and bladder neck incision [5]. Unfortunately, the results in most patients were failure and frustration. The lack of an effective treatment led patients to depend either on indwelling catheters or chronic CISC. Recently, sacral root neuromodulation has been recognized as an effective and safe therapy for restoring bladder function in this difficult group of patients. A large multicentre clinical trial led to FDA approval of this therapy in 1999 for treating chronic urinary retention and frequency/urgency syndrome. The aim of the present study was to determine the long-term efficacy and complications of sacral nerve stimulation in this specific group of patients.

Patients and methods

Twenty patients (17 women and three men, mean age 48 years, range 22–75) with functional unobstructive idiopathic urinary retention had a sacral nerve implant (Interstim, Medtronic Inc., Minneapolis, MN, USA). The mean (range) duration of their symptoms was 68 (18–120) months, and of the follow-up 24 (6–36) months. All patients had been placed on CISC (mean duration 23 months) after all other modes of therapy had failed, including behavioural therapy, pelvic floor exercises with biofeedback, and pharmacological therapy with different agents. Patients underwent a baseline evaluation including a thorough medical history, physical examination and routine laboratory tests. Patients were asked to keep records of their voided volumes and postvoid residual volumes (PVR) for 3 days. In patients with chronic pelvic pain associated with their voiding symptoms, a self-reporting tool for pain assessment was used [6]; patients were asked to grade their pain on a scale of 0–10, with 0 being no pain and 10 being the worst possible pain. Cystoscopy under local anaesthesia and multiple-channel video-urodynamic testing (LaBorie Medical Technologies, Toronto, Canada) were undertaken in all patients to evaluate both bladder and sphincteric function, and to exclude the presence of mechanical obstruction. In all 20 patients the urodynamic findings showed no detrusor activity, with some patients voiding a little urine by the Valsalva manoeuvre.

In all, 32 patients had a temporary percutaneous nerve evaluation (PNE) and 20 were selected to receive a permanent implant, i.e. those who had a> 50% improvement in their symptoms during 3–5 days of temporary stimulation. The implant was placed using previously described bony landmarks [7,8] (Fig. 1) as an outpatient procedure under local anaesthesia. Accurate placement of the lead wire was confirmed both by fluoroscopy and by the clinical response to stimulation. Typical S3 sensory responses include vibration or a ‘pulling’ sensation in the rectum, perineum and labia/vagina in women, and in the scrotum/penile tip in men. Motor responses include tightening of the levators (the bellows response) and plantar flexion of the great toe. Wires were securely taped and attached to an external neurostimulator (Screener 3625, Medtronic). Patients were instructed to maintain the sensory stimulation at a comfortable level for 3 days; they were asked to continue CISC and keep records of their voided volumes and PVR during the trial period.

Figure 1.

Bony landmarks for locating the S3 foramen.

In the first 10 patients the permanent implant was placed under general anaesthesia using short-acting or non-paralytic anaesthetic agents, as previously described [8,9]. In the last 10 patients a modified minimally invasive technique was used. With the patient prone a 22 G 13 cm insulated test needle was inserted in the S3 foramen using the same bony landmarks as previously described. The functional response and fluoroscopy confirmed the correct location of the needle. A small 2.5 cm incision was made along the needle, exposing the underlying lumbodorsal fascia including the site of the needle entrance into the foramen. The top of the needle was cut using a bone cutter and a 14 G angiocath inserted over the needle into the S3 foramen. The permanent quad-lead was inserted through and secured to the lumbodorsal fascia without opening it, using four #0 polypropylene sutures and a ‘twist-in-lock’ device. (Fig. 2). The lead was tunnelled subcutaneously and connected to the implantable pulse generator (IPG, Itral II, model #3623, Medtronic) which was placed in the upper part of the buttocks at a site accessible to the patient. All patients were hospitalized for 1 day; they were placed on intravenous antibiotics for 24 h and oral antibiotics for 1 week. The device was activated after 1 week. The results were evaluated at 1, 3, 6, 12 and 24 months using a standardized and validated questionnaire addressing voiding symptoms, an estimate of the PVR and the pain assessment scale. The data were analysed using the paired Student's t-test.

Figure 2.

Securing the lead to the lumbodorsal fascia using a ‘twist-lock’ device.


Seventeen patients were able to void spontaneously with no need for CISC; one patient required a bilateral implant and can only void with both implants on. Two patients continue to use CISC but less often, There was a mean increase in voided volumes from 48 to 198 mL and a significant decrease in PVR from 315 to 60 mL. Eighteen patients had a ≥ 50% improvement in their quality of life and said they would recommend the therapy to a friend or relative. All patients were able to void with the device switched on.

Although chronic pelvic pain was not the main presenting symptom it was associated with urinary retention in 13 patients; they had a significant improvement in pain level, from 6.0 to 2.5.

The mean hospital stay was 24 h; complications occurred in six patients (30%), the most common being seroma formation in three at the site of the IPG, that resolved spontaneously. There was one superficial wound infection that was managed with antibiotics and local wound care, and two mechanical failures requiring replacement of the device.


Voiding is a physiological phenomenon that requires a complex integration between the sacral micturition centre at S2–4, the higher micturition centre in the pons and the neural pathways in the CNS. The pelvic floor and the external sphincter receive their main innervation via the pudendal nerve; ≈ 70% of the nerves depend on efferent activity in S3, while S2 and S4 contribute the remaining 30%. Various theories suggest that both pelvic floor muscles and the external sphincter are important in controlling bladder function. De Groat [10] showed that sacral pre-ganglionic outflow to the bladder receives inhibitory impulses from the pelvic floor. During bladder filling, increased muscular tone of the pelvic floor and external sphincter suppresses detrusor contractility and inhibits voiding. In addition, the pelvic floor muscles also provide a guarding reflex, which is a protective mechanism against urinary incontinence. This reflex is mediated by a bladder afferent pathway to a urethral efferent; it is normally activated when bladder pressure increases, as during coughing or exercise.

Voiding normally begins by voluntary relaxation of the external sphincter and pelvic floor muscles, initiating the micturition reflex. Pelvic floor dysfunction can result in severe spasticity and hyperactivity of the pelvic floor, and functional detrusor sphincter dyssynergia [11]. This can inhibit detrusor function, resulting in difficult voiding and varying degrees of urinary retention.

Functional urinary retention was initially considered psychogenic but evidence has shown that it is secondary to severe spasticity and hyperactivity of the pelvic floor. Fowler et al.[12] described a syndrome in young women with urinary retention, showing a functionally abnormal EMG of the striated urethral sphincter. The inability of the urethral sphincter to relax results in urinary retention. Although the condition was thought to be limited to young women, in recent years it has been recognized in elderly women and men.

Previously there was no effective treatment for this disorder except CISC; pharmacological agents, including α-blockers and muscle relaxants, gave poor results. Repeat urethral dilatation and bladder neck incisions have been associated with inconsistent results, a high relapse rate and complications.

Sacral neuromodulation has been recognized recently as an effective treatment for restoring bladder function. The early work by Tanagho and Schmidt [13] prompted a large multicentre clinical trial in 1999 which resulted in FDA approval of sacral neuromodulation for treating idiopathic unobstructive chronic urinary retention. In the present study, sacral nerve stimulation restored voiding in most of the patients, supporting results from previous studies [14–16]. The mechanism by which neuromodulation facilitates voiding is not clearly understood. Some consider that sacral neuromodulation allows the patient to re-localize the pelvic floor and regain the ability to relax it [17]. Others suggest that neuromodulation deactivates the urethral guarding reflex. Inhibiting the sphincter indirectly facilitates bladder activity. It is unlikely that sacral neuromodulation restores voiding by inducing detrusor contraction through direct stimulation of the autonomic efferent. This usually requires a higher current than generated for neuromodulation. Urodynamics were not used in the present study and thus we cannot comment on the exact effects of sacral neuromodulation on the detrusor and the urinary sphincter. This would be an interesting subject to address in future studies, to compare the voiding pattern on urodynamics before and after neuromodulation.

The response to neuromodulation in the present patients was impressive; a patient who has not voided for many months or even years is able to void within 24 h of inserting the temporary electrode. Some voided immediately after placement of the temporary lead, before leaving the office.

Some investigators have recommended the routine use of a bilateral neurostimulator [18]; only one of the present patients requires bilateral stimulation. Although this is theoretically appealing, there are currently limited data to suggest that bilateral is better than unilateral stimulation.

Sacral nerve implants have been associated with minimal complications; the most common in previous studies was related to the IPG site [19]. Placing the IPG in the upper buttocks significantly reduced this complication.

In conclusion, sacral nerve stimulation is an effective treatment for patients with refractory idiopathic urinary retention; implantation is a safe and minimally invasive procedure that can be undertaken on outpatients. This technology is a significant addition to the treatment options that urologists need to consider when approaching such a difficult group of patients. The PNE stimulation provides a valuable tool for selecting appropriate patients who would benefit from this therapy.