Over the next 20 years, the number of older adults is projected to increase by 34%. Within the older population, people 85 years of age and older represent the fastest growing age group. In 1997, 14.2% of older adults reported having difficulty performing physical activities of daily living while 21.6% had difficulty performing instrumental activities of daily living. The proportion of elders reporting impairments in activities of daily living increases sharply with age, with 73.6% of those over 79 years of age reporting at least one disability (Administration on Aging, 2002).

The most vulnerable older adults are often referred to as the frail elderly. Although this term is widely used both clinically and in research, the definition of frailty remains elusive (Markle-Reid & Browne, 2003). Frailty can be characterized as multisystem or multidimensional impairments that result in instability, a decline in physical activity, and an increased risk for adverse outcomes (Bortz, 2002;Fried, Ferrucci, Darer, Williamson, & Anderson, 2004;Markle-Reid & Browne, 2003;Rockwood, Hogan, & MacKnight, 2000). There are an estimated 6 million frail elders in the United States (Balducci & Stanta, 2000). These elders reside in nursing homes, other long-term care settings (e.g., personal care homes and residential care homes), and the community. In this presentation, frail older adults are those elders with cognitive or physical impairments, or both, that hinder their ability to live independently or limit their ability to leave their homes.

A variety of factors increase frail older adults' risk of developing urinary incontinence (UI) including age-related changes in bladder function (a decrease in bladder capacity, a decrease in the ability to postpone voiding, diminished detrusor contractility, an increase in the prevalence of detrusor overactivity, a decrease in urethral length and maximal closure pressure in women, and enlargement of the prostate in many men;Fonda et al., 2002). While none of these changes alone cause UI, they predispose frail elders to the effects of additional pathologic, physiologic, and pharmacologic insults and increase their risk for developing UI (Fonda et al., 2002;Lekan-Rutledge & Colling, 2003). It is estimated that more than 50% of nursing home residents are incontinent of urine (Brandeis, Baumann, Hossair, Morris, & Resnick, 1997;Durrant & Snape, 2003;Nelson, Furner, & Jesudason, 2001). Estimates of the prevalence of UI among community-dwelling frail elders vary from 13% to 53% (Baker & Bice, 1995;Landi et al., 2003;Mohide, Pringle, Robertson, & Chambers, 1988;Ouslander, Zarit, Orr, & Muira, 1990;Ruther & Helbing, 1988).

There has been little research on the prevalence of UI in non-nursing home long-term care settings such as assisted living and personal care facilities. One study reported that 37% of personal care home residents (Quinn, Johnson, Andress, McGinnis, & Ramesh, 1999) and one reported that 34% of residential home residents (Georgiou, Potter, Brocklehurst, Lowe, & Pearson, 2001) were incontinent of urine.

Research on UI in frail older adults is limited and most published studies were done in nursing homes. There is much less research to guide nursing care of frail community-dwelling elders and those who reside in long-term care settings other than nursing homes. This presentation will summarize previous research on UI in frail elders in nursing homes, residential care settings, and their homes in the community. It will also suggest directions for future research in each of these settings.

Previous Research

Frail Elders in Nursing Homes

A number of investigators have examined risk factors for UI in nursing home residents (Brandeis et al., 1997;Jirovec & Wells, 1990;Nelson, Furner & Jesudason, 2001;Palmer, German, & Ouslander, 1991). Impaired mobility and cognitive impairment were consistently identified as risk factors for UI in this population. Few studies, however, have addressed the impact of treating identified risk factors on UI severity. Jirovec (1991) examined the impact of a daily exercise program on UI in a sample of 15 nursing home residents and reported significant improvements in incontinence. Schnelle and colleagues (2002) examined the effectiveness of a combined prompted voiding and exercise program in nursing home residence and reported significant decreases in UI frequency as well as improvements in a number of functional measures. The investigators estimate, however, that a nurse aide-to-resident ratio of 1:5 would be required to implement this intervention. For the vast majority of nursing homes this would necessitate an increase in staffing.

Watson, Brink, Zimmer, and Mayer (2003) examined the extent to which 52 nursing homes in upstate New York applied the Agency for Health Care Policy and Research (currently the Agency for Healthcare Research and Quality) guidelines in the evaluation and management of new onset UI in a sample of 200 residents. Overall, they found that only 20% of the standards were met. Only 15% of residents had documentation of their UI history. While 86% had an abdominal exam and 84% were evaluated for edema, only 15% had a rectal exam and only 2% of women had a pelvic examination. When feasible, a postvoid residual volume was assessed in only 6% of residents. The investigators identified potentially reversible causes of UI in 81% of residents. Among those with potentially reversible causes, only 34% had all of them addressed promptly. Twelve weeks after the onset of UI, only 6% of residents were continent, 4% secondary to treatment of a reversible cause and 2% because of a toileting program.

The functional impairments in the vast majority of incontinent nursing home residents make regular toileting assistance an essential component of any intervention for UI. A number of studies have examined the effectiveness of prompted voiding in treating UI in nursing homes (Hu et al., 1989;Ouslander et al., 1995;Palmer, Bennett, Marks, McCormick, & Engel, 1994;Schnelle et al., 1983). In these studies, the mean proportion of wet pad checks was reduced 26% to 50%. The wide variation of effectiveness may have been related to differences in who implemented the prompted voiding protocol (research or nursing home staff) and the frequency of prompting (hourly or every 2 hours). Investigators consistently reported, however, that the sustainability of toileting protocols was limited when nursing home staff were responsible for implementing the protocol (Harke & Richgels, 1992;Lekan-Rutledge, Palmer, & Belyea, 1998;Remsburg, Palmer, Langford, & Mendelson, 1999) or that sustained implementation of the protocol required staffing levels in excess of the levels of the majority of nursing homes (Frantz, Xakellis, Harvey, & Lewis, 2003).

Although an informed patient perspective is important in treatment decisions, little is known about the UI treatment preferences of nursing home residents and their families. In a survey of family members of long-term care residents, investigators (Johnson, Ouslander, Uman, & Schnelle, 2001) used hypothetical scenarios to identify preference choices between two equally efficacious and appropriate treatments for UI. When asked if they preferred an adult diaper or prompted voiding for their family member, 47% of the families preferred the use of diapers, while 44% preferred prompted voiding to manage UI.

Frail Elders in Other Long-Term Care Settings

Almost no published research was found on UI in long-term care settings other than nursing homes. Pfister and Dougherty (1994) attempted to introduce a study examining the effectiveness of bladder training for UI in 6 rural residential care homes in Florida. However, the intervention was not implemented due to limited interest among residents identified as potentially eligible to participate in the study (33%) and the small proportion of interested subjects who were able to complete the 3-day baseline bladder diary (9 of 20 subjects, 45%). In addition to surveying families of nursing home patients about their preferences for treating UI, Johnson and colleagues (2001) also surveyed residents of a residential care facility (n = 70) and cognitively intact nursing home residents (n = 9) about their preferences. Half of the residents (50%) preferred to use diapers, while 41% preferred prompted voiding.

Community-Dwelling Frail Elders

There is also little research related to UI in frail community-dwelling elders. McDowell and colleagues (McDowell, Engberg, Rodriquez, Engberg, & Sereika, 1996) examined the characteristics of frail homebound older adults with UI. Urinary incontinence tended to be severe (M = 3.8 accidents/day) and almost all subjects reported having urge accidents (94.8%). Most subjects (80%) had functional limitations in ambulation and levels of comorbidity were high (M = 8.4 medical problems). Engberg and colleagues (2001) reported that 50.1% of a sample of 345 homebound elders with UI also had significant depressive symptoms.

Two published studies examined the effectiveness of biofeedback-taught pelvic floor muscle exercises (PFME) in treating UI in cognitively intact community-dwelling frail elders (McDowell et al., 1999;Rose, Baigis-Smith, Smith, & Newman, 1990). Rose and colleagues used a single group pretest-posttest design to examine the effectiveness of biofeedback-taught PFME in a sample of 18 home-bound older adults recruited through an urban visiting nurse association. They reported a mean 79% reduction in UI. In a randomized controlled clinical trial, McDowell and colleagues examined the short-term effectiveness of a biofeedback-taught PFME program in 105 homebound older adults with stress, urge, and mixed UI. They reported a median 75% reduction in urinary accidents among subjects randomized to a biofeedback treatment group (n = 48) versus a median 6.4% reduction among those randomized to the control group (n = 45; p < .001).

Three studies examined the effectiveness of a caregiver-initiated toileting program for cognitively impaired, care-giver-dependent, frail community-dwelling older adults (Colling, Owen, McCreedy, & Newman, 2003;Engberg, Sereika, McDowell, Weber, & Brodak, 2002;Jirovec & Templin, 2001). Colling and colleagues examined the effectiveness of Patterned Urge-Response Toileting (a form of habit training) in a sample of 78 caregiver-dependent frail elders and reported a mean 37% reduction in the volume of involuntary urine loss for treatment subjects compared to a 4% decrease for control subjects (p = .02). In a small study (n = 16), Engberg and colleagues reported a mean 60% reduction in daytime UI for subjects completing a prompted voiding protocol compared to a mean 37% reduction for control subjects (p = ns). Jirovec and Templin reported that the mean percentage of incontinent voids decreased from 43% at baseline to 37% at 6 months posttreatment for subjects randomized to an individualized scheduled toileting program compared to an increase from 47% to 49% among control subjects. Significantly more subjects in the treatment group (64%) than in the control group (50%) had a decrease in incontinent episodes at 6 months (p < .05).

Directions for Future Nursing Research

There are many gaps in the knowledge base needed to guide the nursing care of incontinent frail elders across all care settings (Table 1).

TABLE 1.

Frail Elders in Nursing Homes

Most incontinent nursing home residents have severe cognitive impairment and mobility problems (Schnelle, MacRae, Ouslander, Simmons, & Nitta, 1995), making them dependent on staff to meet their toileting needs. A frequently cited barrier to the implementation of toileting programs in nursing homes is the current staff-to-resident ratios in most facilities (Schnelle & Leung, 2004). A treatment (toileting) program that targets residents most likely to benefit may have a greater likelihood of success. Ouslander and colleagues (1995) identified characteristics of incontinent nursing home residents most likely to benefit from prompted voiding; future research should examine the impact of targeted toileting programs based on these characteristics.

When the Cochrane Incontinence Group combined data from three clinical trials comparing prompted voiding to no treatment, the results favored prompted voiding. The difference was not, however, statistically significant. These findings indicate the need for larger, well-conducted trials with longer follow-up of subjects (Eustice et al., 2004).

In frail elders the etiology of incontinence is often multifactorial. Failure to address contributing conditions can limit the success of UI-specific interventions (Fonda et al., 2002). Although a number of studies have identified risk factors associated with UI in nursing home residents (Jirovec & Wells, 1990;Nelson et al., 2001;Ouslander, Kane, & Abrass, 1982;Palmer et al., 1991), there is limited research examining treatment of these factors. Future studies should examine the effects of eliminating reversible factors (e.g., specific medications, physical restraints, and environmental barriers) on the severity of UI.

Comorbidities are common among frail elders with UI and can affect both the prevalence of UI and its management. There is some evidence that addressing mobility impairments may have a positive effects on UI severity (Jirovec, 1991;Schnelle et al., 2002). Additional research is needed to confirm this. Depression, common in this population (Engberg et al., 2001), might increase the risk for UI and can negatively affect an individual's motivation to participate in behavioral interventions for UI (McDowell et al., 1999). Future research should examine response to these therapies after treatment for depression.

Prevention, early identification, and early treatment of those at risk for developing UI are areas for future research in frail elders. Ouslander, Palmer, Rovner, and German (1993) identified male sex, dementia, fecal incontinence (FI), and mobility impairment as risk factors for the onset of UI after nursing home admission. This research should be expanded to examine resident and facility characteristics in relation to the onset and progression of UI in nursing homes.

Future studies should examine the effects that interventions designed to treat FI and maximize the function and mobility of newly admitted long-term care residents have on the onset and progression of UI. Toileting interventions targeted to recently admitted nursing home residents should also be evaluated for their effectiveness in the prevention of UI and FI. Dependency among nursing home residents is often encouraged indirectly in an effort to maximize efficiency given current staffing levels (Rosemond & Mercer, 2002). Future research should examine the impact that this practice has on the onset and progression of incontinence.

Absorbent products play a major role in the management of UI for many nursing home residents, yet there is little research to guide practice decisions about the selection of absorbent pads, the frequency of changing or optimal skin care in relation to resident comfort, cost-effectiveness, and adverse outcomes such as skin problems (Newman, in press) and urinary tract infections (UTIs). Future studies should compare products and changing and skin care routines in relation to these outcomes.

Patient-centered approaches (Lauver et al., 2002) to continence care should be evaluated in long-term settings. Past intervention research for UI in nursing homes has been focused on toileting interventions, predominately prompted voiding. Multicomponent models of continence care should be evaluated in relation to reductions in UI, quality of care indicators (e.g., pressure ulcers and falls), quality of life (QOL), family and resident satisfaction, staff satisfaction and turnover rates, and cost-effectiveness. Multicomponent models could include, for example, scheduled toileting, prompted voiding, PFME, anticholinergic medications, dietary and fluid management, and a regular pad change and skin care regimen. Specific interventions could be individually tailored based on resident or family preference or targeted to characteristics such as cognitive and functional status.

Frail Elders in Other Long-Term Care Settings

There is little evidence to guide continence care for frail older adults living in residential facilities other than nursing homes. Studies should examine the prevalence of UI in this population, its progression, and associated risk factors; researchers could design and test interventions to prevent or delay the progression of UI, perhaps resulting in fewer or later nursing home admissions. Cost-effectiveness should be one of the outcomes examined in relation to the interventions.

In long-term care settings, staff is limited and levels of cognitive impairment are often high, and so the treatment of UI is particularly challenging. A program proposed by Kincade and colleagues (2003) consisted of a series of workshops targeted to (a) operators of the facilities and case managers, (b) professional staff (RNs and LPNs), and (c) personal care assistants. The educational programs were followed by onsite consultation and guidance in dealing with specific resident situations. The bladder management program included recommendations for facility-level changes (e.g., increasing fluid and fiber intake, decreasing caffeine intake, and removing environmental barriers to toileting) and resident-level changes (e.g., toileting for residents likely to benefit or recommendations about the use of appropriate absorbent products). Similarly innovative continence care programs should be examined in long-term care settings.

Technology has the potential to play an important role in the implementation of effective continence programs in long-term care facilities where staffing levels are low. Nurse researchers should work with interdisciplinary teams to develop and test tools designed to promote adherence to toileting regimens and PFME in frail elders with mild to moderate cognitive impairment such as automated reminders, intelligent reminding systems, and robotics.

Community-Dwelling Frail Elders

Limited research exists to guide the care of community-dwelling frail elders with UI and to examine potentially reversible causes of UI in this population. Landi and colleagues (2003) identified UTIs, the use of physical restraints, and environmental barriers as risk factors for UI. Researchers should examine the effects of treating these contributing factors on both the prevention and progression of UI.

There is some evidence to support the short-term effectiveness of biofeedback-assisted PFME in cognitively intact frail community-dwelling older adults. However, in McDowell and colleagues' study (1999) nearly 40% of subjects did not sustain their posttreatment continence levels during the 12 months of follow-up (Engberg, Sereika, & McDowell, 2001). Future research should examine the effectiveness of interventions designed to improve long-term adherence to behavioral therapies for UI in frail elders.

Interventions that allow patients with UI to pick a treatment strategy (Lauver et al., 2002) may be associated with better long-term adherence to treatment recommendations. Dougherty and colleagues (2002) found that using a patient-centered approach to the behavioral management of continence resulted in a 61% reduction in UI compared to an increase of 184% among control participants. Future studies should examine the impact of this type of intervention on UI severity, QOL, patient satisfaction, adherence to treatment recommendations, and cost-effectiveness in frail older adults.

There are an estimated 880,000 community-dwelling older adults with mild dementia, 530,000 with moderate dementia, and 650,000 with severe dementia (Langa et al., 2001). Little is known about how caregivers manage UI in cognitively impaired care recipients. Future studies should examine the techniques caregivers use to manage UI, their perceptions about the effectiveness of interventions they have used, and their satisfaction with them. Previous studies examining the effectiveness of toileting intervention have produced modest improvements in UI (Colling et al., 2003;Engberg et al., 2002;Jirovec & Templin, 2001). Future research should be designed to identify the characteristics of the frail community-dwelling elders most likely to benefit from these interventions. The effectiveness of various toileting interventions should then be tested in this population. Outcome measures should include reductions in UI, caregiver burden, QOL, and cost-effectiveness.

Frail Elders Across All Settings

In all setting s, there is a need for better outcome measures to assess the effectiveness of nursing interventions to treat UI. Bladder diaries are widely used to assess UI severity and to evaluate therapeutic interventions, but their reliability and validity need to be examined in this population. Wet checks are widely used to assess UI severity in research studies in nursing homes. Concerns related to their use are the frequency with which they may need to be performed to reliably assess UI frequency, the consistency with which staff perform and record the checks, and the fact that it is sometimes be difficult to identify whether some currently used pads are wet (Fonda et al., 1998). Pad weight tests may also be a useful method for assessing changes in UI severity in this population; their reliability as a method for quantifying involuntary urine loss needs to be examined in frail elders in long-term care and community settings.

Previous investigators identified UI as a risk factor for a number of negative health outcomes including skin irritation, pressure ulcers, UTIs, and falls (Berlowitz et al., 2001;Brown et al., 2000;Foxman et al., 2001;Girman et al., 2002). Even though Schnelle and colleagues (2003) did not find that a combined prompted voiding and mobility intervention had a significant impact on skin breakdown, UTIs, or falls, these are important and costly outcomes of UI. Future studies of frail older adults should examine nursing interventions in relation to health outcomes. The findings of this research can help to evaluate the cost-effectiveness of nursing intervention for UI and provide a research basis for policy decisions about continence management.

Urinary incontinence has been shown to have a detrimental effect on QOL in other populations (Coyne, Zhou, Thompson, & Versi, 2003) and behavioral therapies have been shown to improve QOL (Burgio et al., 2002). The Standardization Committee of the International Continence Society (Mattiasson et al., 1998) recommends that all investigations of outcomes in patients with lower urinary tract dysfunction include QOL measures. Quality of life is also a recommended outcome measure for investigators conducting research on incontinence is frail older adults (Fonda et al., 1998), though few health-related QOL instruments and no disease-specific measures have been tested for reliability, validity, and responsiveness in frail elders (Fonda et al., 1998). Kane and colleagues (2003) developed and established the psychometric properties of a generic QOL measure for nursing home residents with varying levels of cognitive and physical function. Future nursing home studies could use this QOL questionnaire to examine the results of UI and its treatment in this population.

There is also a need to develop and test UI-specific QOL instruments that can be used to assess response to nursing interventions for UI in cognitively intact frail elders. Since little is known about how UI affects the lives of frail elders, qualitative research may be an appropriate first step in identifying the domains that need to be included in these measures. The findings of this research can be used to modify existing UI-specific QOL instruments (DuBeau, Kiely, & Resnick, 1999;Kelleher, Cardozo, Khullar, & Salvatore, 1997;Lee, Reid, Saltmarche, & Linton, 1995;Shumaker, Wyman, Uebersax, McClish, & Fantl, 1994) developed for use in nonfrail populations or to develop an instrument specifically for this population.

Methodological Challenges of Doing Research with the Frail Elderly

Barriers exist to conducting research in long-term care settings. Administrators may be reluctant to allow research to be conducted in their facilities. Recruitment of residents is difficult and time-consuming (Maas, Kelley, Park, & Specht, 2002). It is often necessary to obtain informed consent from family members of cognitively impaired residents; these family members may be difficult to contact or reluctant to have the resident participate in a research study. Hearing and visual impairment can make obtaining informed consent difficult from those residents who do not have cognitive impairment. Older adults in these settings are often reluctant to participate in research studies (Maas et al.). Subject attrition due to serious illness or death increases sample size needs and can threaten the validity of the study.

Randomization of subjects within a setting may not be feasible because staff may deliver the intervention to control as well as treatment subjects (Ouslander & Schnelle, 1993). If settings, rather than subjects, are randomized to the treatment and control conditions, movement of staff members among facilities increases the likelihood that control subjects will be exposed to the intervention (Maas et al., 2002). On the other hand, using research staff to deliver the intervention increases the cost of study and can also decrease the transferability of findings to clinical practice.

Conducting research with frail elders in the community is also challenging. Recruitment of frail community-dwelling elders is time-consuming and costly (Gill, McGloin, Gahbauer, Shepard, & Bianco, 2001;Ory et al., 2002). Advertisements may be less successful with frail than nonfrail community-dwelling older adults due to sensory deficits (limiting, for example, individuals' ability to read print advertisements or to hear radio ads) and frail elders may be more reluctant to respond to research advertisements.

Many potential subjects may need to be screened in order to meet enrollment targets (Gill et al., 2001;McDowell et al., 1999). The bladder diary is widely used to assess the effectiveness of interventions for UI; however, McDowell and colleagues reported that 29.5% of cognitively intact (Mini Mental Exam Scores of 24 or higher) homebound older adults were excluded from their study because they were unable to provide adequate bladder diary data. Multiple comorbidities among frail community-dwelling elders can result in high rates of attrition during research studies. Finally, interventions generally need to be provided in subjects' homes, increasing costs.

Conclusions

Frail older adults are at increased risk for developing UI and high prevalence rates have been reported in this population, but research on incontinence in frail elders is limited. Most of the studies that have been done were in nursing homes, but even in these settings many questions remain unanswered. Unfortunately, much of the intervention research conducted in nursing homes has not been successfully implemented in clinical practice. There is almost no research to guide the care of incontinent older adults in non-nursing home long-term care settings. Little is known about the prevalence of UI, its progression, or associated risk factors in this growing population of elders. Interventions designed to treat UI have generally not been tested in these settings. There is also a limited research base to guide the care of the large number of cognitively intact and cognitively impaired frail older adults who reside in the community. Finally, there is a need for better outcome measures to assess the effectiveness of nursing interventions for UI in all of these settings. Despite the challenges inherent in doing research with this population, there is an urgent need for research to guide the nursing care of frail elders in all healthcare settings.

References