Incontinence in the Nursing Home

Prevalence and Morbidity

Urinary incontinence affects approximately half of nursing home residents [4, 5]. The prevalence varies among individual facilities depending on the case mix; rates may range from 40% to 70% or even higher in facilities with a functionally impaired resident population. In contrast to urinary incontinence among ambulatory community-dwelling geriatric patients, urinary incontinence among nursing home residents is more severe and more commonly associated with fecal incontinence. Incontinent nursing home residents generally have multiple episodes of urinary incontinence throughout the day and night, and approximately half are also incontinent of stool more than once per week [5, 6].

Urinary incontinence in the nursing home is associated with substantial morbidity and cost. It can predispose patients to skin irritation, make pressure ulcers difficult to heal [7], and result in symptomatic urinary tract infection when urinary retention with overflow urinary incontinence remains undiagnosed or when urinary incontinence is inappropriately managed by long-term use of an indwelling catheter [8, 9]. It may also lead to falls among residents with nocturia and urge urinary incontinence and impaired balance or gait [10]. The adverse psychological effects of urinary incontinence among nursing home residents have been difficult to document systematically [11], but incontinent residents who do not have severe dementia are often embarrassed and frustrated by their urinary incontinence. Nursing home staff generally consider urinary incontinence to be one of the most onerous and difficult conditions for which they care, and they perceive that they spend a disproportionate amount of time on the care of incontinent residents. The economic costs of urinary incontinence in the nursing home have been estimated to be close to $5 billion annually, including the costs of staff time, laundry, and supplies [12].

Types and Causes of Urinary Incontinence

The pathogenesis of urinary incontinence among nursing home residents is often multifactorial, involving urologic and gynecologic conditions, neurologic disorders, behavioral and psychological factors, and functional impairments. Thus, the approach to assessment and treatment must be comprehensive and consider all of these potential factors. The most important factors to consider are those that are reversible. Potentially reversible conditions that can contribute to urinary incontinence in nursing home residents are listed in Table 1. These factors can be recalled by the acronym DRIP (delirium; restricted mobility, retention; infection, inflammation, impaction; polyuria, pharmaceuticals). Although identification and management of these reversible factors may not cure the urinary incontinence, its severity may be reduced and thereby be made more manageable by other interventions. In addition, identification and management of these conditions may have important benefits for the resident's overall functioning and quality of life.

A classification of the basic types of persistent urinary incontinence is shown in Table 2. Three important features of this classification should be noted. First, from a neurologic perspective, this classification is greatly simplified and does not include all of the pathophysiologic types of urinary incontinence. For example, patients with suprasacral spinal cord lesions, such as those that can occur in multiple sclerosis, may have detrusor hyperreflexia with external sphincter dyssynergy that can result in incontinence and incomplete bladder emptying (detrusor-sphincter dyssynergy). Second, many incontinent nursing home residents have mixtures of these types of incontinence. The predominant abnormality of lower urinary tract functioning found among nursing home residents is detrusor hyperactivity (involuntary bladder contractions found on cystometry; also called detrusor instability, unstable bladder, and detrusor hyperreflexia [the latter occurs in the presence of a neurologic disorder]). Although most often associated with urge urinary incontinence, detrusor hyperactivity is commonly seen with sphincter weakness and stress urinary incontinence among women and with obstruction in men with benign or malignant enlargement of the prostate. Interestingly, detrusor hyperactivity is also seen in continent elderly people (men more frequently than women) [13, 14]; thus, its precise role in the pathogenesis of incontinence in the nursing home population is incompletely understood. In nursing home residents, detrusor hyperactivity is commonly associated with impaired bladder contractility that results in incomplete bladder emptying (called detrusor hyperactivity with impaired contractility) [15, 16]. Nursing home residents with this disorder may have symptoms that mimic stress, overflow, or urge incontinence. Third, and most importantly, functional urinary incontinence should be a diagnosis of exclusion because most nursing home residents have impairments of cognitive or physical functioning that may interfere with their ability to use a toilet. These residents may also have other potentially treatable conditions that contribute to their urinary incontinence. Thus, a search for reversible factors and other types of urinary incontinence should be done before a nursing home resident's urinary incontinence is labeled as “functional.”

Assessment

Nursing home residents are heterogeneous, and a realistic and appropriate goal for one type of resident may be unrealistic and inappropriate for another [17, 18]. The approach to urinary incontinence brings this concept into sharp focus. A resident having active rehabilitation after a hip fracture or a stroke may, after a thorough incontinence assessment, benefit from a specific bladder-retraining protocol or pharmacologic therapy for detrusor hyperactivity; incontinence undergarments and indwelling catheters are probably inappropriate for this type of resident. On the other hand, a resident with end-stage dementia and severe agitation may be most appropriately managed by an incontinence undergarment after reversible causes of their urinary incontinence have been excluded. Thus, an important aspect of incontinence care is to determine, through the interdisciplinary care-planning process, if a particular resident has the potential to respond to specific interventions for the urinary incontinence. Because even severely impaired residents may respond well to a prompted voiding program (see below), a bias in favor of thorough assessment and a therapeutic trial is appropriate.

Basic assessment of bladder and bowel function as indicated on the MDS is required for all newly admitted nursing home residents. A bladder and bowel record is helpful in documenting the continence status of new residents and can also be used as part of periodic reassessments. A legible record, such as the one shown in Figure 1, should be used [19]. The specific symbols used are not important, but the record should provide a simple way of documenting wetness, dryness, appropriate toileting, and bowel status and a space for comments. Records such as the one shown in Figure 1 can be reduced so that several records fit on one page. This type of record is also helpful in monitoring responses to therapeutic interventions. Because many newly admitted residents come from acute-care hospitals, they frequently arrive at the nursing home with an indwelling bladder catheter. In this situation, it is essential to determine why the catheter was placed (for example, to monitor urinary output or for urinary retention or management of urinary incontinence) and to consider the resident for a bladder-retraining program. The catheter should be removed unless there is an appropriate indication for retaining it.

Basic Evaluation

After these initial assessments and documentation, incontinent nursing home residents should have a basic evaluation that includes a history, physical examination, urinalysis, and determination of postvoid residual urine volume. Much of this evaluation can be done by nursing home staff and a “physician extender” [nurse practitioner, physician's assistant, or clinical nurse specialist]. This basic evaluation has three objectives: 1) to identify potentially reversible factors (Table 1); 2) to identify potentially serious underlying conditions or conditions that may require further urologic, gynecologic, or urodynamic evaluation (Table 3); and 3) to determine the type of incontinence (urge, stress, overflow, or mixed) and an appropriate management plan.

Both the Resident Assessment Protocol for incontinence and the Agency for Health Care Policy and Research (AHCPR) clinical practice guideline on urinary incontinence [20] contain specific recommendations for the assessment of incontinence in nursing home residents. The information summarized below is consistent with these recommendations, but readers should become familiar with the Resident Assessment Protocol and the AHCPR guidelines as well.

History

Medical records should be reviewed to identify medications, medical conditions, and genitourinary history that may be relevant to the resident's incontinence. A precise history and report of symptoms may be difficult to obtain from many nursing home residents because incontinence is closely associated with dementia in this population. Residents who can reliably report symptoms should be questioned about irritative and other symptoms that suggest urge incontinence (frequency, urgency, nocturia [more than two episodes]); symptoms of stress incontinence (leakage simultaneous with coughing, sneezing, laughing, bending); voiding difficulty (hesitancy, poor or intermittent stream, straining to finish voiding); and other associated symptoms such as dysuria or pain. Pain related to voiding is unusual and should be carefully evaluated.

Nursing home staff can be helpful in identifying many of these symptoms among residents who cannot reliably report them, as well as in assessing the timing and nature of fluid intake and monitoring voiding activity using a record such as the one shown in Figure 1.

Physical Examination

Because impairments of cognitive and physical function are commonly associated with incontinence among nursing home residents, the physical examination should include an assessment of these areas specifically related to the resident's ability to respond to a prompt to urinate, use the toilet, and manage their clothing and hygiene. Physical and occupational therapists can be helpful in completing this assessment.

The general physical examination should exclude volume overload (for example, venous insufficiency or congestive heart failure) and previously undiagnosed neurologic conditions that might contribute to the incontinence (such as signs of parkinsonism or spinal cord compression). Rectal examination is done to assess resting and active sphincter tone, to exclude fecal impaction, and, in men, to assess the prostate. The size of the prostate on physical examination can neither establish nor exclude the presence of urethral obstruction.

The pelvic examination should assess perineal and vulvar skin condition, exclude pelvic masses and marked pelvic prolapse (such as uterine descensus or a cystocele that protrudes through the vaginal introitus), and evaluate the vaginal epithelium for signs of inflammation (erythema, friability, or bleeding) that suggest atrophic vaginitis. A cough test to detect stress incontinence can also be done during the pelvic examination, but it is best done while the patient is standing and has at least 200 mL of urine in the bladder [3].

Urinalysis

The urinalysis is primarily done to exclude significant bacteriuria in residents who have symptoms (other than incontinence) suggesting a urinary tract infection (dysuria, recent onset or worsening of incontinence, or unexplained fever or functional decline) and to exclude sterile hematuria, which may indicate bladder or upper urinary tract disorders.

Clean urine may be difficult to obtain from incontinent, cognitively impaired, immobile nursing home residents. However, it is possible to noninvasively obtain from both men and women a clean-catch urine that accurately matches culture results from a catheterized specimen. For women, the technique involves carefully cleaning the perineum with a sterile preparation kit and then having the patients void into a disinfected collection device (a toileting insert or “hat” or a fracture bedpan) [21]. For men, a clean condom catheter can be applied after the penis is cleaned with a sterile preparation such as Betadine [22].

Although rapid-urine screening methods are generally considered inaccurate in the nursing home population, they can be used to exclude bacteriuria with reasonable accuracy. The combination of negative dipstick test results for both nitrite and leukocyte esterase counts and a negative enzyme-based screen for bacteriuria (Uriscreen, Ventrex Laboratories, Portland, Maine) has a negative predictive value of more than 95% for significant bacteriuria among incontinent nursing home residents [23]. Whether such screening tests are practical depends on the availability of personnel who can reliably collect the urine and do the tests and on the costs relative to sending a specimen to the laboratory for urinalysis.

Estimate of Postvoid Residual Volume

Significant degrees of urinary retention can occur without symptoms or signs of obstruction or bladder contractility problems. In addition, physical examination (suprapubic palpation, bimanual examination) is neither sensitive nor specific in detecting significant residual volume. The postvoid residual is usually determined by catheterization, but a portable ultrasound device is now available (Bladder Volume Instrument; Bard Urological, Covington, Georgia) that can noninvasively measure residual urine volume with reasonably good accuracy in incontinent nursing home residents [24]. It is important to note that postvoid residual volumes can substantially vary within individual patients. The timing of the determination in relation to a void, the degree to which the void was natural for the patient, the nature of instructions given to the patient, the amount of straining, and body position can all influence the postvoid volume obtained. Thus, when moderately elevated urine volumes are detected (for example, 150 to 300 mL), repeated determinations may be appropriate. In these situations, ultrasonography may be especially valuable in avoiding the need for repeated catheterizations. Repeated residual volumes of more than 200 mL should prompt consideration of whether the resident should be referred for further evaluation (Table 3).

Further Evaluation

Selected incontinent nursing home residents may benefit from further urologic, gynecologic, or urodynamic evaluation. Examples of conditions that may require further evaluation are outlined in Table 3. In determining the need for further evaluation, the most important consideration is whether the results will change the management of the resident's incontinence. Thus, for example, a nursing home resident with urinary retention who is not a candidate for surgical intervention is unlikely to benefit from a urodynamic evaluation to diagnose the cause of the retention; such a patient should be managed by intermittent or long-term catheterization. On the other hand, urodynamic evaluation is essential for a resident for whom surgery is being considered (for example, a woman with symptoms and signs of severe stress incontinence or a man with symptoms and signs suggesting obstruction).

Complex urodynamic tests (including cystometry with simultaneous measurement of bladder and abdominal pressure; voiding pressure-flow study; measurement of urethral pressure at rest, during voiding, and with straining; sphincter electromyography; and videourodynamic studies) have been shown to be safe and feasible in the nursing home population [16]. This type of testing is essential for the accurate diagnosis of obstruction, detrusor hyperactivity with impaired contractility, and the type of stress incontinence (that is, primarily caused by urethral hypermobility or by intrinsic sphincter deficiency). Because the testing procedures are relatively expensive, involve some discomfort, and require specialized equipment and trained personnel, they should only be done in incontinent nursing home residents for whom the results are necessary to determine an appropriate treatment plan.

Some simple urodynamic tests, including a pad test for stress incontinence and “bedside” cystometry, do not require specialized equipment [17, 25, 26]. The Resident Assessment Protocol recommends a stress test for nursing home residents whose incontinence persists after the basic evaluation and management of reversible conditions [3]. The stress test is done by having the patient cough forcefully several times, preferably in the standing position, when their bladder is relatively full (but not when the patient has a strong urge to void). The test result is positive if there is immediate leakage similar to the volume and circumstance of the resident's usual incontinence. If the test result is negative and the total bladder volume (voided volume plus residual volume) is less than 200 mL, the test should be repeated. If the postvoid residual determination has been done by catheterization, the bladder can be filled with sterile water (by gravity) to a volume of 200 mL for the stress test. Simple cystometry can also be done immediately after catheterization [25] by filling the bladder in 50-mL aliquots through a 50-mL catheter-tip syringe without the piston, while asking the resident to hold as much as they can. Involuntary bladder contractions are shown by continuous upward movement of the fluid column in the syringe in the absence of abdominal straining (which can sometimes be detected by palpating the abdomen) or by leakage of the water around the catheter. When involuntary contractions do not occur, the volume at which the resident has a strong urge to void (cystometric bladder capacity) can be determined. Although this test can be helpful in identifying involuntary bladder contractions [27], the results may be difficult to interpret in cognitively impaired nursing home residents. Straining may be difficult to detect and may be misinterpreted as an involuntary contraction, and low-pressure involuntary contractions (such as those seen in many patients having detrusor hyperactivity with impaired contractility) are easily missed.

In our studies, we have not found stress tests and simple cystometry helpful in identifying the substantial proportion of incontinent residents (25% to 40%) who respond well to daytime prompted voiding [28]. Other investigators have also shown that urodynamic diagnosis does not predict responsiveness to bladder training in community-dwelling older women [29]. We therefore suggest that if behavioral interventions such as prompted voiding and bladder training are to be used as the initial therapy, urodynamic testing is not necessary before a therapeutic trial. Urodynamic testing to help differentiate stress from urge and mixed (both stress and urge) incontinence should be considered for residents who do not respond well to the behavioral intervention and for whom more specific therapy (such as pharmacologic treatment) is planned.

Approaches to Therapy

Reversible Factors

The first step in treating incontinence is to attempt to reverse the potentially reversible factors (Table 1). Conditions that may interfere with the resident's ability or willingness to get to a toilet or use a toilet substitute should be addressed. New or worsening incontinence, may, for example, be a manifestation of delirium—a condition with myriad potential underlying causes and substantial morbidity and mortality if it remains unidentified and untreated. Immobility caused by musculoskeletal pain (such as hip fracture or exacerbation of arthritis) may precipitate incontinence. In these situations, regular toileting assistance and the optimal use of toilet substitutes (such as fracture bedpans) and other environmental interventions should be used. Physical restraints may also contribute to incontinence and, consistent with OBRA rules [1], should be avoided whenever possible. Unrecognized and untreated depression is another major cause of morbid conditions in the nursing home population and can interfere with a nursing home resident's motivation to be continent. Underlying depression should be considered if a patient repeatedly refuses to cooperate with toileting assistance.

Data do not support treating asymptomatic bacteriuria in incontinent nursing home residents because eradicating bacteriuria does not appear to alter morbidity, mortality, or the severity of chronic incontinence in this population [30-32]. The presence of pyuria in this population (defined as more than 10 leukocytes per high-power field on routine microscopic examination of centrifuged urine) does not necessarily indicate the presence of symptomatic infection that requires treatment [32, 33]. However, manifestations of a symptomatic urinary tract infection in this population may be subtle and include the onset or worsening of incontinence, unexplained low-grade fever, anorexia, and functional decline. Thus, clinicians should not treat asymptomatic bacteriuria on the one hand but have a high index of suspicion for symptomatic infection in the presence of such nonspecific symptoms on the other. Atrophic vaginitis, manifested by inflamed or friable epithelium on pelvic examination, may cause irritative symptoms, including urge incontinence, and should be treated by either a vaginal estrogen cream (1 to 2 g applied 3 to 5 nights per week) or oral conjugated estrogen (0.3 to 0.625 mg/d). No data support any specific treatment protocol [34]. The goal is not necessarily to cure the incontinence but to reduce the frequency of incontinent episodes and possibly make the resident more responsive to other forms of treatment. If a clear response is shown, estrogen therapy can either be continued or discontinued and reinitiated if symptoms and signs recur. Prolonged, unopposed estrogen therapy (that is, lasting longer than 1 to 2 years) in a woman with a uterus should prompt consideration of adding a progestational agent [35].

Fecal impaction is an important reversible cause of urinary (and fecal) incontinence. The mechanisms by which impaction contributes to urinary incontinence has never been fully explained. Impactions can generally be prevented by a bowel regimen similar to the one presented in the section below on fecal incontinence. Factors contributing to polyuria or nocturia should be addressed. Although controlling diabetes in some nursing home residents is difficult, the osmotic diuresis induced by glucosuria can certainly exacerbate urinary incontinence, and better glucose control should be attempted. Edema that is mobilized in the supine position, especially in the evening and night hours, can cause bothersome polyuria and nocturia. This may be hazardous for residents who are prone to falls. It is therefore reasonable in some situations to initially manage the urinary incontinence by adding (or increasing the dose of) a rapid-acting diuretic in the morning to reduce the edema, which will make the resident urinate more often when they have better access to a commode.

Drugs that may be contributing to urinary incontinence Table 1 should be stopped whenever possible, especially if urinary retention is present. If it is not possible to discontinue therapy, reducing the dose or modifying the dosage schedule may help manage the urinary incontinence.

Behavioral Interventions

Most nursing homes have bladder training protocols that consist of simple scheduled toileting every 2 hours [6]. Although this is practical for staff and may help some residents, it is probably not the most efficient procedure and should be combined with other techniques whenever feasible. Behavioral interventions can be divided into two basic types: patient-dependent and caregiver-dependent. Patient-dependent procedures include pelvic muscle (Kegel) exercises, bladder training, biofeedback, and other similar behavioral therapies. All have some demonstrated efficacy among community-dwelling incontinent patients [36-38]. None of these procedures, or other adjunctive therapies such as electrical stimulation [39], has been adequately studied among nursing home residents. The goal of patient-dependent interventions is to restore a normal pattern of voiding and continence, and it is presumed that the resident will be capable of independent toileting once bladder functioning is improved. Because of the high prevalence of dementia among incontinent nursing home residents, the learning and practice required for these techniques to be successful are obstacles to their widespread use in this population. Bladder retraining is a patient-dependent behavioral intervention that is relevant in the nursing home. This intervention is applicable in the common situation in which an incontinent resident has been admitted (or readmitted) to the nursing home from an acute-care hospital with an indwelling bladder catheter that was inserted for either urinary retention or accurate monitoring of urine output. (Under these circumstances, the bladder retraining may serve as a “voiding trial” as suggested by the Resident Assessment Protocol [3].) An example of a bladder-retraining protocol is shown in Table 4. The precise protocol depends on the resident's bladder function. If the catheter was placed to measure urine output or to manage urinary incontinence during a hospitalization, the bladder is probably irritable and has a small capacity; in this instance, progressively increasing the intervals between voiding can be attempted. In highly functional residents, pelvic-muscle exercises and other behavioral techniques may be helpful in this process. If the catheter was inserted for urinary retention, the bladder muscle may be decompensated; in this situation, regular attempts to void should be combined with routine estimation of residual urine volume and postvoid intermittent catheterizations as needed. During this phase, ultrasound determination of postvoid residual volume can be helpful in avoiding repeated catheterizations. It may take weeks for the bladder to begin functioning again. If postvoid residual volumes remain elevated and the resident remains incontinent, a urologic evaluation should be considered. Note that in either situation, clamping the catheter before removal is not necessary; it has never been shown to be beneficial and may be harmful to residents who already have a decompensated bladder after an episode of urinary retention.

Caregiver-dependent techniques, including habit-training and prompted voiding, have been shown to be effective for some incontinent nursing home residents [40-46]. These interventions can be used with demented and physically impaired residents who are not candidates for patient-dependent procedures. Habit-training involves establishing a pattern for assisting residents with voiding and toileting according to their own pattern. This technique may be difficult to implement because it requires that multiple nursing home staff keep track of individual toileting schedules [45], and it is not clear what types of residents respond to this intervention. In prompted voiding, residents are asked (prompted) on a 2-hour schedule if they will attempt to use the toilet. An example of a prompted voiding protocol is shown in Table 5. As opposed to bladder retraining Table 4, the goal is not necessarily to restore a normal pattern of voiding and continence but rather to prevent wetness. Prompted voiding has been shown to be highly effective in reducing incontinence frequency during the day and evening hours when motivated staff implement it properly. Because it appears to be effective in nursing home residents with different types of lower urinary tract dysfunction and has almost no side effects, it is a reasonable initial approach to managing most incontinent nursing home residents after reversible factors have been identified and treated. Data collected in nearly 30 nursing homes suggest that 25% to 40% of incontinent residents respond well, with a reduction in their daytime incontinence frequency from 3 to 4 episodes per day to 1 or fewer [28, 40-42, 46-50]. Nursing home staff can use simple behavioral and other observations from a 3- to 5-day trial of prompted voiding to target those most likely to respond well Table 5[28, 47, 50]. An intense quality-assurance program, based on principles of statistical quality control or other management techniques, is necessary for the ongoing effectiveness of prompted voiding or any other caregiver-dependent intervention in nursing homes [48-51].

The best results from prompted voiding have been documented from 7 a.m. to 7 p.m. Prompted voiding has not been systematically studied between 7 p.m. and 7 a.m., even though evidence suggests that incontinence is at least as severe during this nighttime period [52]. The frequency of incontinence during night hours is similar to that observed during daytime periods, but the volume of the incontinence episodes at night is significantly higher. Although volume overload (congestive heart failure, venous insufficiency with edema) and benign prostatic hyperplasia may contribute to nighttime incontinence, the mechanism or mechanisms underlying increased urine production at night in this population are not well understood. Hormonal abnormalities such as altered secretion of antidiuretic hormone or atrial natriuretic peptide may be responsible, but further research is needed in this area. Nursing rounds during the night to change incontinent residents frequently disrupt sleep [53, 54]. Prompted voiding schedules have not yet been designed to improve nighttime incontinence management without disrupting resident's sleep. It is therefore not clear whether the potential benefits of regular toileting throughout nighttime hours are outweighed by disruption of the resident's sleep. In addition, staffing patterns at night may preclude an effective prompted voiding program, even in targeted residents. Thus, until further data are available, nursing homes should modify the prompted voiding procedures for nighttime hours or should use incontinence undergarments.

Drug Therapy

Because of the close association between functional disability and urinary incontinence among nursing home residents, drug therapy must generally be used as an adjunct to some form of toileting program. Drug therapy in this population is directed at one of two abnormalities of lower urinary tract functioning, or a combination of both. Detrusor hyperactivity (involuntary bladder contractions on urodynamic testing), the most common urodynamic abnormality found in incontinent nursing home residents, is generally associated with urge urinary incontinence [16, 26, 55]. Drugs with anticholinergic effects, including propantheline (15 to 30 mg three to four times per day), oxybutynin (2.5 to 5 mg two to four times per day), and imipramine (10 to 25 mg three times per day) have been most commonly used as bladder relaxants. Several studies suggest that frail and functionally impaired incontinent patients do not respond well to bladder relaxant medications [56-60]. A recent placebo-controlled trial in which oxybutynin was added to prompted voiding suggests that a modest subgroup (approximately 25%) of incontinent nursing home residents with detrusor hyperactivity may benefit from this drug when it is added to a prompted-voiding protocol [61]. If one of these medications is given in a therapeutic trial, careful observation is needed for anticholinergic side effects such as dry mouth, constipation and fecal impaction, blurry vision, urinary retention, and worsening cognitive function. Imipramine can also cause postural hypotension. Women with urge incontinence associated with atrophic vaginitis should be treated with estrogen as described above. New types of bladder-relaxant drugs and long-acting preparations (for example, slow-release capsules and skin patches) are being developed but are not currently available.

For sphincter weakness with stress urinary incontinence in women, drug treatment involves a combination of estrogen and an α-agonist. Estrogen can be given by vaginal cream, 1 g at bedtime, or orally, 0.625 mg conjugated estrogen daily. Few data support the effectiveness of estrogen, especially when given alone, for treating stress incontinence [20, 34]. Either pseudoephedrine, 30 to 60 mg three times per day, or phenylpropanolamine, 75 mg twice per day, can be used as the α-agonist. Phenylpropanolamine has been shown to be as effective as pelvic-muscle exercises among noninstitutionalized older women with stress incontinence [62]. Drug treatment for stress urinary incontinence should be combined with a toileting program (to keep the bladder volume as low as possible) and pelvic-muscle exercises (for residents who can cooperate). Women who have prominent pelvic prolapse or who fail a 3- to 6-month trial of drug or behavioral therapy, or both, should be considered for surgical intervention.

For women with mixed urge and stress urinary incontinence, a combination of the above approaches can be used. Imipramine, in doses of 10 to 25 mg three times per day, may be tried as a combination anticholinergic-α-agonist. Imipramine can worsen cardiac conduction abnormalities and, in addition to its anticholinergic side effects, can cause significant postural hypotension in ambulatory nursing home residents. Thus, this drug should be used cautiously in the nursing home population.

Drug treatment for overflow urinary incontinence has not been adequately studied in incontinent geriatric patients. Cholinergic agonists such as bethanechol have not been consistently effective when given orally on a long-term basis in patients with urinary retention caused by poor bladder contractility [63]. Nursing home residents having bladder retraining after an episode of urinary retention Table 4 may benefit from a therapeutic trial of bethanechol if urinary retention persists. Other drugs, such as metoclopramide, have also been used for this purpose. Both bethanechol and metoclopramide can have significant side effects in frail geriatric patients and must be used cautiously. α-Antagonists (such as prazosin and terazosin) have been somewhat successful in men with obstructive symptoms (hesitancy, straining, frequency, urgency, and nocturia) that are associated with benign prostatic hyperplasia [64-66]. Because these drugs can cause postural hypotension and result in falls and related injuries in frail older men, it is recommended that therapy be initiated with a small bedtime dose and that postural vital signs be monitored as the dose is increased.

Surgery

Surgical intervention for urinary incontinence is a consideration for a small but important subgroup of incontinent nursing home residents. Because urinary incontinence is not a life-threatening problem, elective surgery should be pursued only if the incontinence is bothersome enough to a resident who is a candidate for such treatment.

There are two types of surgery for urinary incontinence. First, women with stress urinary incontinence associated with significant pelvic prolapse and urethral hypermobility may benefit from bladder-neck suspension and repair of the pelvic prolapse. Although studies have not been done in the nursing home population, the short-term (1 to 5 years) success of this type of surgery for properly selected older women is approximately 75% for significant reduction or elimination of wetness [20]. Artificial urinary sphincters have been used for younger patients with stress incontinence associated with intrinsic urethral deficiency. This requires a relatively major surgical procedure and is probably not appropriate for nursing home residents. Periurethral injections of collagen have recently been approved for patients with intrinsic urethral deficiency. Although studies have not yet included very frail patients, periurethral injection has reportedly been effective in younger patients [67].

The other type of surgical intervention for urinary incontinence is the removal of anatomical obstruction, most commonly an enlarged prostate in men or a urethral stricture. Detailed discussion of the evaluation and surgical treatment of lower urinary tract obstruction is beyond the scope of this article, but nursing home residents who are candidates for surgery and who are suspected of having obstruction should be referred for complex urodynamic evaluation. Newer, less invasive techniques are now available for the surgical relief of obstruction in men and may be especially appropriate for consideration in frail, older nursing home residents.

Pads and Undergarments

Highly absorbent launderable and disposable pads and undergarments are the most common method of managing urinary incontinence in the nursing home. This method of management is appropriate for the subgroup of incontinent nursing home residents who remain incontinent despite more specific treatments. Pads and garments may also be helpful at night for residents who are managed by prompted voiding or other interventions during the day and evening. Many factors should be considered when selecting the optimal pad or undergarment for an individual resident, including comfort and fit, absorbency relative to the degree of leakage, availability of laundry services (for reusable products), skin sensitivity, and cost. No adequate well-controlled studies are available on which to base the choice of one specific type of product over another [68]. Pads and undergarments should not be used as the sole solution to urinary incontinence in the nursing home, or in a manner that fosters further dependency. When pads or garments are used, residents should still be regularly checked and given the opportunity to use the toilet or be changed if necessary to avoid skin irritation and breakdown.

Catheters and Catheter Care

Three basic types of catheters and catheterization procedures are used to manage urinary incontinence in nursing homes: external catheters, intermittent straight catheterization, and long-term indwelling catheterization. External catheters consist of some type of condom connected to a drainage system. Improvements in design and observance of proper procedure and skin care when applying the catheter will decrease the risk for skin irritation and the frequency with which the catheter falls off. Few studies have been done on complications associated with the use of these devices, but existing data suggest that male nursing home residents with external catheters are at increased risk for developing symptomatic urinary tract infections [69]. External catheters should therefore only be used to manage intractable incontinence in male residents who do not have urinary retention and who are extremely physically dependent. As with incontinence undergarments and padding, these devices should not be used for convenience because they may foster dependency. The safety and effectiveness of external catheters for women have not been documented in nursing homes.

Intermittent catheterization is used to manage urinary retention and overflow incontinence. The procedure involves straight catheterization two to four times daily to keep the bladder volume below approximately 400 mL. Studies done primarily in younger paraplegic patients have shown that this technique is practical and, compared with long-term catheterization, reduces the risk for symptomatic infection [70]. Intermittent self-catheterization has also been shown to be feasible for elderly female outpatients who are functional, willing, and able to catheterize themselves [71]. However, the results of studies done in young paraplegic patients and elderly female outpatients cannot automatically be extrapolated to the nursing home population. The technique may be useful for certain nursing home residents [72], but the practicality and safety of this procedure in nursing homes have not been well documented. Elderly nursing home residents, especially men, may be difficult to catheterize, and the anatomic abnormalities commonly found in the lower urinary tract of nursing home residents may increase the risk for infection because of repeated straight catheterizations. In addition, using this technique in an institutional setting (which may have an abundance of organisms relatively resistant to many commonly used antimicrobial agents) may yield an unacceptable risk for nosocomial infections. The use of sterile catheter trays for the multiple catheterizations is expensive. Thus, it may be difficult to implement such a program in a typical nursing home.

Long-term indwelling catheterization has been overused in the nursing home [73] and increases the incidence of several other complications, including chronic bacteriuria, symptomatic urinary tract infection, bladder stones, periurethral abscesses, and even bladder cancer [8, 9, 74-78]. Elderly nursing home residents managed by this technique, especially men, are at relatively high risk for developing symptomatic urinary tract infection [9]. Given these risks, it seems appropriate to recommend that long-term indwelling catheterization be limited to certain specific situations, and, when indwelling catheterization is used, sound principles of catheter care should be observed so that complications are minimized (Table 6). Recently implemented federal regulations require that an appropriate indication be documented for nursing home residents with long-term indwelling catheterization [1] (Table 6); similar recommendations were made some 35 years ago [79]. Although few data are available on the most effective routine care regimens for long-term indwelling catheterization, the closed system should be broken as infrequently as possible, and the catheter should be changed every 4 to 8 weeks (in addition to when symptomatic infection occurs Table 6) to help prevent the build-up of encrustations, which can lead to catheter blockage and infection [80, 81].

Fecal Incontinence

Fecal incontinence is less common than urinary incontinence, but a large proportion (50% to 75%) of nursing home residents with frequent urinary incontinence also have episodes of fecal incontinence [5, 6]. Defecation, like urination, is a physiologic process that involves smooth and striated muscles, central and peripheral innervation, coordination of reflex responses, mental awareness, and physical ability to get to a toilet. Disruption of any of these factors can lead to fecal incontinence.

The most common causes of fecal incontinence are problems with constipation and laxative use, hyperosmotic enteral feedings, neurologic disorders, colorectal disorders, and functional dependence [82, 83]. Iatrogenic causes (such as laxatives, hyperosmotic feedings, and some antacids) and dietary factors (such as lactose intolerance) should be excluded. Hypothyroidism and hyperthyroidism can also cause bowel symptoms and should be considered. Constipation is common in the elderly and, when chronic, can lead to fecal impaction and incontinence. The hard stool (or scybalum) of fecal impaction irritates the rectum and results in the production of mucus and fluid. This fluid leaks around the mass of impacted stool and precipitates incontinence. Appropriate management of constipation will help prevent fecal impaction and resultant fecal incontinence. The management of constipation in nursing home residents involves several approaches, including adequate intake of fluid and fiber; the appropriate use of stool softeners; and regular toileting with the use of glycerin suppositories if necessary. Some residents have acquired laxative dependence and require the intermittent use of more potent osmotic or irritant laxatives or suppositories.

Fecal incontinence caused by neurologic disorders is sometimes amenable to biofeedback therapy, although most elderly nursing home residents with dementia cannot cooperate or learn the techniques. Prompted voiding improves bowel and urinary continence in some nursing home residents. For those residents with end-stage dementia, a strategy of alternating constipating agents (if necessary) and laxatives on a routine schedule is generally effective in controlling defecation and preventing fecal incontinence. Such strategies are generally referred to as bowel-training programs. The basic components of most bowel-training programs include the regular use of stool softeners or fiber and adequate fluid intake to prevent fecal impaction; regular toileting after breakfast (to take advantage of the gastrocolic reflex); and the intermittent use (commonly 3 to 4 times per week) of either an oral laxative such as milk of magnesia or sorbitol, or a suppository (glycerin or bisacodyl) to stimulate a bowel movement. Periodic enemas may be necessary for residents who do not respond to these measures.