An Interdisciplinary Approach to the Assessment and Behavioral Treatment of Urinary Incontinence in Geriatric Outpatients B. Joan McDowell, PhD, C:RNP, Kathryn L. Burgio, PhD, Marianne Dombrowski, RN,BSN, Julie L. Locher, MA, and Eric Rodriguez, M D Objective: To test the effectiveness of an interdisciplinary assessment and behavioral treatment of persistent urinary incontinencein geriatric outpatients. Design: Prospective case series in which frequency of incontinence was measured before and after intervention. Setting: We established an interdisciplinarycontinence program within an existing academic center, the Benedum Geriatric Center. Patients: Convenience sample of 70 non-demented outpatients aged 56 to 90 years. Behavioral treatment was provided to 29 patients including marly with multiple medical problems (Mean = 6.0 problems). Intervention: Behavioral treatment consisted of biofeedback, pelvic floor muscle exercise, scheduled voiding, and other strategies for preventing accidental urine loss. Outcome measure: Outcome of treatment was measured by comparing bladder diaries completed in the 2 weeks immediately following treatment to those completed in the pre-
treatment phase. Results: Following an average 5.6 treatment sessions, the mean weekly frequency of accidents was reduced from 16.9 to 2.5 (P < 0.01). Individual reductions ranged from 30.8% to 100%with an average of 81.6%improvement.Ten patients achieved continence. Patients with mixed incontinence had greater improvementthan those with urge incontinence alone (P < 0.05), and patients who reported previous evaluation or treatment had a poorer outcome than those coming for their first evaluation (P = 0.05). Degree of improvement was not significantly related to age, duration of symptoms, baseline frequency of accidents, number of treatment sessions, number of other medical diagnoses, or urodynamic findings. Conclusion: We conclude that older adults who are able and willing to participate in behavioral treatment can benefit significantly despite other health problems or disabilities. J Am Geriatr SOC40370-374,1992
rinary incontinence affects approximately 30% geriatric outpatients is to establish a continence treatof older adults and over 50% of those who live ment program within a multidisciplinary geriatric in nursing homes.', As the second leading risk clinic. In one such program, professionals in nursing factor for institutionali~ation,~ it is of particular concern and urology developed a continence clinic within a for older adults whose other health problems or disa- specialized clinic for ambulatory e l d e r l ~ .In~ similar bilities contribute to their risk of losing independence. fashion, we created an interdisciplinary continence Although most cases can be significantly improved or team, consisting of specialists in nursing, geriatric medcured with the medical, surgical, and behavioral treat- icine, and behavioral psychology, within a multidisciments now available, data on community-dwelling plinary geriatric ambulatory care center noted for its adults indicate that the majority do not seek or receive service to a predominantly frail elderly population. treatment.4 In older adults, concerns about concurrent While we offer treatment of contributory medical probmedical problems may result in hesitancy on the part lems, pharmacologic therapy, and referral to affiliated of both the patients and their health care providers to gynecologic and urologic consultants, our primary address the problem of incontinence. Conventional treatment interventions are behavioral in nature. medical or surgical treatments may be withheld out of Behavioral treatments such as biofeedback, pelvic concern for adverse effects. Referral to specialists in floor muscle exercise, and bladder training have been another setting may be cumbersome or unacceptable shown to be effective for non-demented ambulatory 6-12 Such treatments are especially attracto patients. Unwarranted[pessimism about the ability older adults.4* of elderly patients to tolerate, comply with, and derive tive for use by older adults with other health problems benefit from treatment of incontinence may cause phy- because they avoid the risks of iatrogenic harm assosicians to decide in advance not to offer treatment or ciated with pharmacologic and surgical treatments. By offering behavioral interventions in an interdisreferral. One approach to incontinence that overcomes prob- ciplinary continence program, which is itself a comlems of accessibility and integration of services for ponent of a multidisciplinary geriatric clinic, we enable patients to receive low-risk treatment for urinary incontinence in a single setting in which related medical, From the Benedum Geriahic Cente:r, University of Pittsburgh School of Medfunctional, and social issues can be addressed in a iane, Pittsburgh, Pennsylvania. coordinated manner by appropriate professionals. This Supported in part by a Research Career Development Award from the National Institute on Aging to Dr. Burgio KO4-AG00431. paper reports our experience with this integrated team Address correspondenceto B. JoanIvlcDowell, PhD, CRNP, Benedum Geriatric approach to chronic urinary incontinence in geriatric Center, University of Pittsburgh School of Medicine, 3601 Fifth Avenue, Suite outpatients. 28, Pittsburgh, PA 15213.
U
IAGS 40:370-374, 1992 0 1992 by the American Geriatrics Society
0002-8614/92/$3.50
IAGS-APRIL 1992-VOL. 40, NO. 4
METHODS Subjects Subjects were 70 non-demented community-dwelling older adults who were self-referred to the Continence Program of the University of Pittsburgh’s Benedum Geriatric Center or referred by geriatricians, urologists, or other physicians.
TREATMENT OF URINARY INCONTINENCE
371
asked to record urination and accidents both night and day for a two-week period prior to intervention.
Selection Criteria Men or women with mental status scores greater than 22, who documented urinary accidents on bladder diaries, and who reported involuntary loss of urine of at least 3-months duration were included in the study. Those with stress, urge, or mixed stress and urge inEvaluation continence, as documented in the continence history All patients were evaluated initially by the treatment and corroborated in the bladder diary, were eligible for team which consisted of a nurse practitioner, a behav- behavioral treatment. Cases of overflow or functional ioral psychologist, and a geriatrician. A urologist and a incontinence were not eligible for behavioral treatment gynecologist were also available for consultation. The but were referred or provided treatment by the Contievaluation consisted of history and physical examina- nence team. tion, mental status evaluation, urodynamic testing, labWhen patients were found to have urinary tract oratory tests, and bladder diaries. infection, fecal impaction, severe atrophic vaginitis, or History The history included a continence history a correctable metabolic disorder, entry to the study was and a medical history. The continence history included deferred until medical treatment of those conditions description of the onset, severity, and progression of was delivered by the continence team. Patients with the problem as well as antecedents to incontinent epi- severe uterine prolapse, hematuria, enlarged prostate, sodes such as coughing, sneezing, and urge symptoms. or post-void residual urine greater than 100 mL were Patients were queried about other bladder symptoms excluded and referred to a urologist or gynecologist as as well as frequency of day and nighttime urination appropriate. and bowel habits. The medical history explored disease states that Treatment might cause or contribute to incontinence or that might During treatment, patients continued to document prevent or complicate treatment. The patients‘ current incontinence and voiding habits in their bladder medications were reviewed and evaluated for their diaries. They attended treatment sessions 30 to 90 potential effect on bladder function. minutes in duration at 2 to 4 week intervals. The Physical Examination The physical examination number of treatment sessions was individualized deincluded examination of heart and lungs, abdominal pending on the patient’s progress and abilities. examination for bladder distention, masses, and cosBiofeedback-Assisted Pelvic Floor Muscle Traintovertebral angle tenderness, and a brief neurological examination. Women received a complete pelvic ing Training was accomplished using anorectal or examination, and both men and women had a rectal vaginal biofeedback instruments in order to provide immediate visual and/or auditory feedback of pelvic examination. floor muscle activity. Anorectal manometry was proMental Status Evaluation The Mini-Mental State vided by means of a rectal probe with three small Examination was utilized to evaluate the cognitive balloons attached, one of which measured intra-abstatus of s~bjects.’~ dominal pressure, one held the apparatus in place, and Urodynamic Testing Urodynamic assessment in- one was located at the anal opening to record external cluded measurement of post-void residual urine, su- anal sphincter pressure.l4 The vaginal electromyograpine water cystometry, and provocation. Provocative phy was accomplished using surface electrodes on an maneuvers to elicit stress incontinence were standing acrylic vaginal probe.” Using feedback, patients were up from a sitting position, heel bouncing, and coughing taught to contract and relax pelvic floor muscles while in the supine and standing positions. The sound of keeping abdominal muscles relaxed. running water and hand washing were used to provoke Home Practice After learning to control these musurge incontinence. cles properly, patients were given instructions for pracLaboratory Evaluation Laboratory tests included tice at home. They were instructed to perform 45 pelvic serum glucose, electrolytes, BUN/creatinine, serum floor muscle exercises each day, divided into three calcium, urinalysis and urine culture. Other tests were practice sessions of 15 contractions each. Duration of performed as indicated. contractions was specified based on demonstrated abBladder Diaries Patients were given bladder diary ilities and was increased gradually up to a limit of 10 booklets and asked to record voluntary urination, in- seconds and separated by equal periods of relaxation. continent episodes, approximate volume of urine loss Patients were advised to practice in various positions (large vs small), and the circumstances or antecedents including lying, sitting, and standing. associated with the urinary accidents. These included Patients with stress incontinence were taught to conphysical activities such as coughing or rising from a tract pelvic floor muscles prior to and during activities chair as well as sensations such as an urge to void with that increased intra-abdominal pressure, such as an inability to reach the toilet in time. Patients were coughing or rising from a chair. Those with urge incon-
372
MCDOWELL ET AL
tinence were taught to prevent loss of urine and suppress urgency by contracting pelvic floor muscles. Instead of rushing to the toilet in response to urgency, they were taught to pause and relax, use their muscles to suppress urgency, wait for the urge to pass, and only then to proceed at a normal pace to the toilet. Patients with mixed incontinence were taught to use both strategies. Bladder Retraining Patients who voided more than eight times during the waking day were asked to postpone voiding in spite of a sensation of urgency and instead to void by the clock. Initially a voiding interval was specified that, based on 'bladder diaries, was within a comfortable range. The interval was then increased by 30 minute increments, every 2 weeks, until a voiding schedule of 3-hour intervals was achieved.
RESULTS Seventy men and women underwent screening evaluation. One was referred tafa urologist for hematuria, four were referred for el'evated post-void residual urines, and one for nocturnal enuresis. Five were referred to private or clinic physicians for treatment of unrelated but more urgent medical problems (ie, uncontrolled diabetes mellitus, bladder cancer, congestive heart failure, supranuclear palsy, or vestibular disorder). Three patients with urinary tract infection and one with uncontrolled diabetes were treated by the continence team. They had satisfactory improvement in incontinence and underwent no further treatment. Two with dementia were assisted through education of caretakers regarding prompted voiding. Of the 53 remaining patients, five did not return for treatment because incontinence resolved spontaneously (n = 2) or the patient preferred not to receive treatment at that
JAGS-APRIL 1992-VOL 40, NO.4
time (n = 3). Another patient died prior to entry into treatment. Thus, 47 patients (67%)were considered appropriate and entered behavioral treatment. Although they participated in the treatment, seven patients were either unable or unwilling to complete a meaningful bladder diary. Thus, reliable data on their outcome are unavailable. Another 11 patients dropped out too early for any outcome data to be acquired. Reasons for termination included depression, illness, and family situation. Other patients who terminated prematurely were included in the analysis when bladder diary data was available to document their progress to the point of termination. Characteristics of Patients in Behavioral Treatment Twenty-nine patients, 27 women and 2 men, participated in behavioral treatment and completed bladder diaries adequate to assess the efficacy of the intervention. Sixteen were referred by physicians at the Benedum Geriatric Center which serves predominantly frail community-dwelling elderly with multiple health problems. Seven were self-referred and the remaining six were from other sources. Characteristics of these patients are described in Table 1. The average duration of incontinence was 6.8 years and nine patients had previously received some form of treatment for incontinence. Eighteen subjects (64%) reported that they used some form of protection. Review of the patients' problems revealed 0 to 15 medical problems in addition to incontinence (mean = 6.0 problems). Eleven patients had a cystocele, 10 of which were grade one and one grade two. Upon entry to the study, 12 patients were on diuretics, five were using estrogen, and 11 were taking anticholinergic medications, eight of which were anti-de-
TABLE 1. CH:ARACTERISTICSOF PATIENTS IN BEHAVIORAL TREATMENT (n = 29) Gender 27 women, 2 men Range 56-90 years Mean = 74.6, sd = 8.1 Age Range 1-44 years Mean = 6.8, sd = 8.1 Duration of incontinence Type of incontinence: n (%) Mixed 21(72) Urge only 7 (24) Stress only l ( 3 .4 ) Relevant diagnoses 8 (28) Depression Stroke 4 (14) Diabetes mellitus 3 (10) Parkinsonism 2 (6.9) Prostatectomy l(3.4) Urodynamic findings Positive stress test 3 (10) Detrusor instability/bladder capacity 11 (38) < 350 ml Both 12 (41) 2 (6.9) Normal Test not performed 1 (3.4) Previous treatment Bladder repair only l(3.4) Medications only* 4 (14) Both 4 (14) * Includes estrogen, bethanacol, propantheline, imiprarnine, oxybutynin, flavoxate, and phenylpropanolamine.
TREATMENT OF URINARY INCONTINENCE
JAGS-APRIL 1992-VOL. 40, NO.4
pressants prescribed for treatment of depression. During the protocol, these patients continued to take their medications with no changes in dosage, with the exception of one woman who discontinued imipramine during behavioral treatment as a result of reduced incontinence. Characteristics of Treatment Twenty-eight of the 29 patients underwent biofeedback training to learn sphincter control. One patient was instructed with verbal feedback based on vaginal palpation. For the majority of patients (59%) training was accomplished in one session, but others required up to seven (mean = 1.8, median = 1). In addition, patients attended up to 12 (mean = 3.8) treatment sessions in which biofeedback was not used, but other behavioral techniques such as active use of pelvic floor muscles, other strategies for preventing urine loss, and bladder retraining were discussed and implemented. Thus the average number of biweekly sessions was 5.6 with a range of 2 to 14. Urinary Accidents Outcome of treatment was measured by comparing bladder diaries completed in the 2 weeks immediately following treatment to those completed in the pre-treatment phase (Table 2). Patients recorded a mean of 16.9 accidents per week in the pre-treatment phase (median = 9.2) and 2.5 accidents per week in the post-treatment phase (median = 1.5; t = 2.9, P < 0.01). Each individual experienced some reduction in the frequency of accidents; the poorest response was a 31 % reduction. The mean reduction was 82%. Ten patients had no accidents after treatment. Weekly frequencies of accidents before and after treatment are depicted for each patient in Figure 1. Of the 29 patients, 26 (90%) were at least 50% improved, 20 (69%) were at least 70% improved, and 14 (48%) were at least 90% improved. For patients who did not completely eliminate accidents, a paired t test was performed to examine the volume of urine loss before and after treatment. Prior to treatment, 10% of accidents were recorded as large (as opposed to small). Following treatment this figure dropped to 4.7 percent (P = NS). The urodynamic testing and biofeedback procedures were well tolerated and no adverse effects were noted. Of the total 70 subjects studied, 47 (67%) were considered appropriate and were agreeable' to behavioral treatment. Of this group of eligible subjects, 55% (n = TABLE 2. FREQUENCY OF ACCIDENTS AND URINATION BEFORE AND AFTER TREATMENT Pre-treatment Post-treatment Mean (SO) Mean (SO) Urinary accidents (per week) Daytime urinations (per day) Episodes of nocturia (per night) • P < 0.01. •• P < 0.05.
16.9 (27.7)
2.5 (3.7)*
10.1 (3.4)
8.6 (1.8)**
1.8 (1.1)
1.5 (0.7)
373
m
PRE-TREATMENT
140
•
POST-TREATMENT
50 :::
«
20 10
o
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
PATIENTS FIGURE 1. Mean weekly frequency of accidents before and after treatment for ~clividual patients.
26) participated and improved significantly by at least = 14) dropped out or had a poor response «50% improvement), and the other 15% (n = 7) are unknown because they did not complete useful bladder diaries. Frequency of Urination Before treatment, frequency of urination ranged from 6.1 to 20.5 voidings per day (see Table 2). Following treatment, this was reduced to 5.6 to 10.8 per day (t = 2.53, P < 0.05). Frequency of nocturia ranged from 0.5 to 6.9 per night before treatment and 0 to 2.9 voidings per night after treatment (P = NS). Predictors of Response Correlation coefficients were calculated for the relationship between several independent variables and improvement (percent reduction in frequency of accidents). Outcome was not significantly correlated with age, duration of symptoms, the number of health problems with clinical relevance for bladder control (ie, diabetes mellitus, stroke, parkinsonism, depression, prostatectomy), total number of all medical diagnoses, or the total number of treatment sessions. Outcome was also not related to the results of urodynamic evaluation (ie, bladder instability, reduced bladder capacity, or positive stress test), nor to severity of incontinence as measured by baseline frequency of accidents (r = .11), nor to the frequency of urination (r = -.03) or nocturia (r = -.01). Outcome was negatively correlated with number of biofeedback sessions (r = -.60, P < 0.001), probably reflecting the individualized nature of the treatment in which biofeedback was repeated for patients who were not improving. Patients who were not significantly improved continued to receive biofeedback training in up to seven sessions, depending on their progress. Patients with mixed stress and urge incontinence had significantly greater improvement (mean = 85%) than those with urge accidents alone (mean = 68%; t 50%,30% (n
374
Me DOWELL ET AL
= 2.07, P < 0.05). Patients who reported previous evaluation or treatment for incontinence had a poorer outcome (mean = 72% reduction of incontinence) compared to those coming for first evaluation (mean = 88% reduction; P = 0.05).
DISCUSSION This study demonstrates that a continence team, working in a multidisciplinary geriatric setting, can achieve very satisfactory results treating urinary incontinence in geriatric outpatients. In particular, we have shown that behavioral treatment is an effective and well tolerated treatment for urinary incontinence in the cognitively intact elderly, including those with other health problems and disability. Patients treated behaviorally achieved a mean 82% reduction in frequency of accidents, and 10 became completely dry. Two patients with Parkinson's disease achieved 100% reduction in urinary accidents in spite of tremor, bradykinesia, and rigidity which impaired their toileting skills. Significant improvement was also documented in patients who had a cystocele or history of stroke, had failed previous treatment with anticholinergic medication, were under treatment for depression, or, in one case, with post-prostatectomy incontinence. Patients with significant co-morbidity who required specialized intervention were successfully treated by modifying the approach to accommodate functional impairments. For example, a 79 year old woman, blind since childhood, was able to achieve continence using auditory biofeedback and by maintaining bladder diaries in Braille. This is not to say that behavioral treatment is appropriate for all patients with incontinence. In the present study approximately a third of patients evaluated were considered ineligible based on their mental status, urological findings, or more urgent medical problems. In addition, behavioral treatment requires that the patients be able to learn new skills and participate actively in their own treatment across several weeks. While we found most patients were motivated to cooperate in this way, others will find it unacceptable. In this study all patients who were treated with biofeedback-assisted behavior therapy reduced their incontinence. However, there was wide variation in outcome, and we were not able to predict magnitude of response from the information we gathered in the evaluation. Treatment outcome was not related to age, duration or severity of incontinence, frequency of urination, nocturia, or the number of other health problems. Thus, we have no reliable predictor variables,
JAGS-APRIL 1992-VOL. 40, NO.4
and, therefore, referrals for behavior therapy should be available for any cognitively intact older adult with incontinence. Data on long term maintenance of treatment effects were not collected in this study. Previous studies have examined maintenance, 11, 14 and more work is needed on this issue, particularly with regard to behavioral therapies in which ongoing patient compliance is of utmost importance. One benefit of the team approach was that contributory problems such as urinary tract infections or poorly controlled diabetes mellitus were managed by the continence team, with subsequent improvement in urinary incontinence. Integration with a geriatric center had the advantage that patients found to have related non-medical problems, such as depression or caregiver burden, could be treated by other professionals at the Center and then return to the Continence Program. Finally, a continence team composed of specialists in geriatric health care is experienced in approaching the older adult patient in a comprehensive manner. By integrating behavioral treatment with this philosophy of geriatric health care, it is possible to offer effective and acceptable low-risk treatment, even to multiply impaired older patients.
REFERENCES 1. Diokno AC, Brock BM, Brown MB et al. Prevalence of urinary incontinence and other urological symptoms in the noninstitutionalized elderly. J Urol 1986;136:1022. 2. Ouslander JG. Urinary incontinence in nursing homes. J Am Geriatr Soc 1990;38:289-291. 3. Brazda JF. Washington report. The nation's health. March, 1983. 4. NIH Consensus Conference. Urinary incontinence in adults. JAMA 1989;261:2685-2695. 5. Brink C, Wells T, Diokno A. A continence clinic for the aged. I Gerontol Nurs 1983;9:651-655. 6. Fantl [A, Wyman JF, Harkins SW et al. Bladder training in the management of lower urinary tract dysfunction in women: A review. J Am Geriatr Soc 1990;38:329-332. 7. Wells TI. Pelvic (floor) muscle exercise. J Am Geriatr Soc 1990;38:333337. 8. Burgio KL, Engel BT. Biofeedback-assisted behavioral training for elderly men and women. J Am Geriatr Soc 1990;38:338-340. 9. Bums PA, Pranikoff K, Nochajski T et al. Treatment of stress incontinence with pelvic floor exercises and biofeedback. J Am Geriatr Soc 1990;38:341344. 10. Burton IR, Pearce KL, Burgio KL et al. Behavioral training for urinary incontinence in elderly ambulatory patients. J Am Geriatr Soc 1988;36:693-698. 11. Baigis-Smith I, Smith DAJ, Rose M et al. Managing urinary incontinence in community residing elderly patients. Gerontologist 1989;29:229-233. 12. Fantl [A, Wyman JF, McClish OK et al. Efficacy of bladder training in older women with urinary incontinence. JAMA 1991;265:609-613. 13. Folstein MF, Folstein SE, McHugh PRo "Mini-Mental State": a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-98. 14. Burgio KL, Whitehead WE, Engel BT. Urinary incontinence in the elderly: Bladder-sphincter biofeedback and toileting skills training. Ann Intern Med 1985;103:507-515.
treatment phase. Results: Following an average 5.6 treatment sessions, the mean weekly frequency of accidents was reduced from 16.9 to 2.5 (P < 0.01). Individual reductions ranged from 30.8% to 100%with an average of 81.6%improvement.Ten patients achieved continence. Patients with mixed incontinence had greater improvementthan those with urge incontinence alone (P < 0.05), and patients who reported previous evaluation or treatment had a poorer outcome than those coming for their first evaluation (P = 0.05). Degree of improvement was not significantly related to age, duration of symptoms, baseline frequency of accidents, number of treatment sessions, number of other medical diagnoses, or urodynamic findings. Conclusion: We conclude that older adults who are able and willing to participate in behavioral treatment can benefit significantly despite other health problems or disabilities. J Am Geriatr SOC40370-374,1992
rinary incontinence affects approximately 30% geriatric outpatients is to establish a continence treatof older adults and over 50% of those who live ment program within a multidisciplinary geriatric in nursing homes.', As the second leading risk clinic. In one such program, professionals in nursing factor for institutionali~ation,~ it is of particular concern and urology developed a continence clinic within a for older adults whose other health problems or disa- specialized clinic for ambulatory e l d e r l ~ .In~ similar bilities contribute to their risk of losing independence. fashion, we created an interdisciplinary continence Although most cases can be significantly improved or team, consisting of specialists in nursing, geriatric medcured with the medical, surgical, and behavioral treat- icine, and behavioral psychology, within a multidisciments now available, data on community-dwelling plinary geriatric ambulatory care center noted for its adults indicate that the majority do not seek or receive service to a predominantly frail elderly population. treatment.4 In older adults, concerns about concurrent While we offer treatment of contributory medical probmedical problems may result in hesitancy on the part lems, pharmacologic therapy, and referral to affiliated of both the patients and their health care providers to gynecologic and urologic consultants, our primary address the problem of incontinence. Conventional treatment interventions are behavioral in nature. medical or surgical treatments may be withheld out of Behavioral treatments such as biofeedback, pelvic concern for adverse effects. Referral to specialists in floor muscle exercise, and bladder training have been another setting may be cumbersome or unacceptable shown to be effective for non-demented ambulatory 6-12 Such treatments are especially attracto patients. Unwarranted[pessimism about the ability older adults.4* of elderly patients to tolerate, comply with, and derive tive for use by older adults with other health problems benefit from treatment of incontinence may cause phy- because they avoid the risks of iatrogenic harm assosicians to decide in advance not to offer treatment or ciated with pharmacologic and surgical treatments. By offering behavioral interventions in an interdisreferral. One approach to incontinence that overcomes prob- ciplinary continence program, which is itself a comlems of accessibility and integration of services for ponent of a multidisciplinary geriatric clinic, we enable patients to receive low-risk treatment for urinary incontinence in a single setting in which related medical, From the Benedum Geriahic Cente:r, University of Pittsburgh School of Medfunctional, and social issues can be addressed in a iane, Pittsburgh, Pennsylvania. coordinated manner by appropriate professionals. This Supported in part by a Research Career Development Award from the National Institute on Aging to Dr. Burgio KO4-AG00431. paper reports our experience with this integrated team Address correspondenceto B. JoanIvlcDowell, PhD, CRNP, Benedum Geriatric approach to chronic urinary incontinence in geriatric Center, University of Pittsburgh School of Medicine, 3601 Fifth Avenue, Suite outpatients. 28, Pittsburgh, PA 15213.
U
IAGS 40:370-374, 1992 0 1992 by the American Geriatrics Society
0002-8614/92/$3.50
IAGS-APRIL 1992-VOL. 40, NO. 4
METHODS Subjects Subjects were 70 non-demented community-dwelling older adults who were self-referred to the Continence Program of the University of Pittsburgh’s Benedum Geriatric Center or referred by geriatricians, urologists, or other physicians.
TREATMENT OF URINARY INCONTINENCE
371
asked to record urination and accidents both night and day for a two-week period prior to intervention.
Selection Criteria Men or women with mental status scores greater than 22, who documented urinary accidents on bladder diaries, and who reported involuntary loss of urine of at least 3-months duration were included in the study. Those with stress, urge, or mixed stress and urge inEvaluation continence, as documented in the continence history All patients were evaluated initially by the treatment and corroborated in the bladder diary, were eligible for team which consisted of a nurse practitioner, a behav- behavioral treatment. Cases of overflow or functional ioral psychologist, and a geriatrician. A urologist and a incontinence were not eligible for behavioral treatment gynecologist were also available for consultation. The but were referred or provided treatment by the Contievaluation consisted of history and physical examina- nence team. tion, mental status evaluation, urodynamic testing, labWhen patients were found to have urinary tract oratory tests, and bladder diaries. infection, fecal impaction, severe atrophic vaginitis, or History The history included a continence history a correctable metabolic disorder, entry to the study was and a medical history. The continence history included deferred until medical treatment of those conditions description of the onset, severity, and progression of was delivered by the continence team. Patients with the problem as well as antecedents to incontinent epi- severe uterine prolapse, hematuria, enlarged prostate, sodes such as coughing, sneezing, and urge symptoms. or post-void residual urine greater than 100 mL were Patients were queried about other bladder symptoms excluded and referred to a urologist or gynecologist as as well as frequency of day and nighttime urination appropriate. and bowel habits. The medical history explored disease states that Treatment might cause or contribute to incontinence or that might During treatment, patients continued to document prevent or complicate treatment. The patients‘ current incontinence and voiding habits in their bladder medications were reviewed and evaluated for their diaries. They attended treatment sessions 30 to 90 potential effect on bladder function. minutes in duration at 2 to 4 week intervals. The Physical Examination The physical examination number of treatment sessions was individualized deincluded examination of heart and lungs, abdominal pending on the patient’s progress and abilities. examination for bladder distention, masses, and cosBiofeedback-Assisted Pelvic Floor Muscle Traintovertebral angle tenderness, and a brief neurological examination. Women received a complete pelvic ing Training was accomplished using anorectal or examination, and both men and women had a rectal vaginal biofeedback instruments in order to provide immediate visual and/or auditory feedback of pelvic examination. floor muscle activity. Anorectal manometry was proMental Status Evaluation The Mini-Mental State vided by means of a rectal probe with three small Examination was utilized to evaluate the cognitive balloons attached, one of which measured intra-abstatus of s~bjects.’~ dominal pressure, one held the apparatus in place, and Urodynamic Testing Urodynamic assessment in- one was located at the anal opening to record external cluded measurement of post-void residual urine, su- anal sphincter pressure.l4 The vaginal electromyograpine water cystometry, and provocation. Provocative phy was accomplished using surface electrodes on an maneuvers to elicit stress incontinence were standing acrylic vaginal probe.” Using feedback, patients were up from a sitting position, heel bouncing, and coughing taught to contract and relax pelvic floor muscles while in the supine and standing positions. The sound of keeping abdominal muscles relaxed. running water and hand washing were used to provoke Home Practice After learning to control these musurge incontinence. cles properly, patients were given instructions for pracLaboratory Evaluation Laboratory tests included tice at home. They were instructed to perform 45 pelvic serum glucose, electrolytes, BUN/creatinine, serum floor muscle exercises each day, divided into three calcium, urinalysis and urine culture. Other tests were practice sessions of 15 contractions each. Duration of performed as indicated. contractions was specified based on demonstrated abBladder Diaries Patients were given bladder diary ilities and was increased gradually up to a limit of 10 booklets and asked to record voluntary urination, in- seconds and separated by equal periods of relaxation. continent episodes, approximate volume of urine loss Patients were advised to practice in various positions (large vs small), and the circumstances or antecedents including lying, sitting, and standing. associated with the urinary accidents. These included Patients with stress incontinence were taught to conphysical activities such as coughing or rising from a tract pelvic floor muscles prior to and during activities chair as well as sensations such as an urge to void with that increased intra-abdominal pressure, such as an inability to reach the toilet in time. Patients were coughing or rising from a chair. Those with urge incon-
372
MCDOWELL ET AL
tinence were taught to prevent loss of urine and suppress urgency by contracting pelvic floor muscles. Instead of rushing to the toilet in response to urgency, they were taught to pause and relax, use their muscles to suppress urgency, wait for the urge to pass, and only then to proceed at a normal pace to the toilet. Patients with mixed incontinence were taught to use both strategies. Bladder Retraining Patients who voided more than eight times during the waking day were asked to postpone voiding in spite of a sensation of urgency and instead to void by the clock. Initially a voiding interval was specified that, based on 'bladder diaries, was within a comfortable range. The interval was then increased by 30 minute increments, every 2 weeks, until a voiding schedule of 3-hour intervals was achieved.
RESULTS Seventy men and women underwent screening evaluation. One was referred tafa urologist for hematuria, four were referred for el'evated post-void residual urines, and one for nocturnal enuresis. Five were referred to private or clinic physicians for treatment of unrelated but more urgent medical problems (ie, uncontrolled diabetes mellitus, bladder cancer, congestive heart failure, supranuclear palsy, or vestibular disorder). Three patients with urinary tract infection and one with uncontrolled diabetes were treated by the continence team. They had satisfactory improvement in incontinence and underwent no further treatment. Two with dementia were assisted through education of caretakers regarding prompted voiding. Of the 53 remaining patients, five did not return for treatment because incontinence resolved spontaneously (n = 2) or the patient preferred not to receive treatment at that
JAGS-APRIL 1992-VOL 40, NO.4
time (n = 3). Another patient died prior to entry into treatment. Thus, 47 patients (67%)were considered appropriate and entered behavioral treatment. Although they participated in the treatment, seven patients were either unable or unwilling to complete a meaningful bladder diary. Thus, reliable data on their outcome are unavailable. Another 11 patients dropped out too early for any outcome data to be acquired. Reasons for termination included depression, illness, and family situation. Other patients who terminated prematurely were included in the analysis when bladder diary data was available to document their progress to the point of termination. Characteristics of Patients in Behavioral Treatment Twenty-nine patients, 27 women and 2 men, participated in behavioral treatment and completed bladder diaries adequate to assess the efficacy of the intervention. Sixteen were referred by physicians at the Benedum Geriatric Center which serves predominantly frail community-dwelling elderly with multiple health problems. Seven were self-referred and the remaining six were from other sources. Characteristics of these patients are described in Table 1. The average duration of incontinence was 6.8 years and nine patients had previously received some form of treatment for incontinence. Eighteen subjects (64%) reported that they used some form of protection. Review of the patients' problems revealed 0 to 15 medical problems in addition to incontinence (mean = 6.0 problems). Eleven patients had a cystocele, 10 of which were grade one and one grade two. Upon entry to the study, 12 patients were on diuretics, five were using estrogen, and 11 were taking anticholinergic medications, eight of which were anti-de-
TABLE 1. CH:ARACTERISTICSOF PATIENTS IN BEHAVIORAL TREATMENT (n = 29) Gender 27 women, 2 men Range 56-90 years Mean = 74.6, sd = 8.1 Age Range 1-44 years Mean = 6.8, sd = 8.1 Duration of incontinence Type of incontinence: n (%) Mixed 21(72) Urge only 7 (24) Stress only l ( 3 .4 ) Relevant diagnoses 8 (28) Depression Stroke 4 (14) Diabetes mellitus 3 (10) Parkinsonism 2 (6.9) Prostatectomy l(3.4) Urodynamic findings Positive stress test 3 (10) Detrusor instability/bladder capacity 11 (38) < 350 ml Both 12 (41) 2 (6.9) Normal Test not performed 1 (3.4) Previous treatment Bladder repair only l(3.4) Medications only* 4 (14) Both 4 (14) * Includes estrogen, bethanacol, propantheline, imiprarnine, oxybutynin, flavoxate, and phenylpropanolamine.
TREATMENT OF URINARY INCONTINENCE
JAGS-APRIL 1992-VOL. 40, NO.4
pressants prescribed for treatment of depression. During the protocol, these patients continued to take their medications with no changes in dosage, with the exception of one woman who discontinued imipramine during behavioral treatment as a result of reduced incontinence. Characteristics of Treatment Twenty-eight of the 29 patients underwent biofeedback training to learn sphincter control. One patient was instructed with verbal feedback based on vaginal palpation. For the majority of patients (59%) training was accomplished in one session, but others required up to seven (mean = 1.8, median = 1). In addition, patients attended up to 12 (mean = 3.8) treatment sessions in which biofeedback was not used, but other behavioral techniques such as active use of pelvic floor muscles, other strategies for preventing urine loss, and bladder retraining were discussed and implemented. Thus the average number of biweekly sessions was 5.6 with a range of 2 to 14. Urinary Accidents Outcome of treatment was measured by comparing bladder diaries completed in the 2 weeks immediately following treatment to those completed in the pre-treatment phase (Table 2). Patients recorded a mean of 16.9 accidents per week in the pre-treatment phase (median = 9.2) and 2.5 accidents per week in the post-treatment phase (median = 1.5; t = 2.9, P < 0.01). Each individual experienced some reduction in the frequency of accidents; the poorest response was a 31 % reduction. The mean reduction was 82%. Ten patients had no accidents after treatment. Weekly frequencies of accidents before and after treatment are depicted for each patient in Figure 1. Of the 29 patients, 26 (90%) were at least 50% improved, 20 (69%) were at least 70% improved, and 14 (48%) were at least 90% improved. For patients who did not completely eliminate accidents, a paired t test was performed to examine the volume of urine loss before and after treatment. Prior to treatment, 10% of accidents were recorded as large (as opposed to small). Following treatment this figure dropped to 4.7 percent (P = NS). The urodynamic testing and biofeedback procedures were well tolerated and no adverse effects were noted. Of the total 70 subjects studied, 47 (67%) were considered appropriate and were agreeable' to behavioral treatment. Of this group of eligible subjects, 55% (n = TABLE 2. FREQUENCY OF ACCIDENTS AND URINATION BEFORE AND AFTER TREATMENT Pre-treatment Post-treatment Mean (SO) Mean (SO) Urinary accidents (per week) Daytime urinations (per day) Episodes of nocturia (per night) • P < 0.01. •• P < 0.05.
16.9 (27.7)
2.5 (3.7)*
10.1 (3.4)
8.6 (1.8)**
1.8 (1.1)
1.5 (0.7)
373
m
PRE-TREATMENT
140
•
POST-TREATMENT
50 :::
«
20 10
o
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
PATIENTS FIGURE 1. Mean weekly frequency of accidents before and after treatment for ~clividual patients.
26) participated and improved significantly by at least = 14) dropped out or had a poor response «50% improvement), and the other 15% (n = 7) are unknown because they did not complete useful bladder diaries. Frequency of Urination Before treatment, frequency of urination ranged from 6.1 to 20.5 voidings per day (see Table 2). Following treatment, this was reduced to 5.6 to 10.8 per day (t = 2.53, P < 0.05). Frequency of nocturia ranged from 0.5 to 6.9 per night before treatment and 0 to 2.9 voidings per night after treatment (P = NS). Predictors of Response Correlation coefficients were calculated for the relationship between several independent variables and improvement (percent reduction in frequency of accidents). Outcome was not significantly correlated with age, duration of symptoms, the number of health problems with clinical relevance for bladder control (ie, diabetes mellitus, stroke, parkinsonism, depression, prostatectomy), total number of all medical diagnoses, or the total number of treatment sessions. Outcome was also not related to the results of urodynamic evaluation (ie, bladder instability, reduced bladder capacity, or positive stress test), nor to severity of incontinence as measured by baseline frequency of accidents (r = .11), nor to the frequency of urination (r = -.03) or nocturia (r = -.01). Outcome was negatively correlated with number of biofeedback sessions (r = -.60, P < 0.001), probably reflecting the individualized nature of the treatment in which biofeedback was repeated for patients who were not improving. Patients who were not significantly improved continued to receive biofeedback training in up to seven sessions, depending on their progress. Patients with mixed stress and urge incontinence had significantly greater improvement (mean = 85%) than those with urge accidents alone (mean = 68%; t 50%,30% (n
374
Me DOWELL ET AL
= 2.07, P < 0.05). Patients who reported previous evaluation or treatment for incontinence had a poorer outcome (mean = 72% reduction of incontinence) compared to those coming for first evaluation (mean = 88% reduction; P = 0.05).
DISCUSSION This study demonstrates that a continence team, working in a multidisciplinary geriatric setting, can achieve very satisfactory results treating urinary incontinence in geriatric outpatients. In particular, we have shown that behavioral treatment is an effective and well tolerated treatment for urinary incontinence in the cognitively intact elderly, including those with other health problems and disability. Patients treated behaviorally achieved a mean 82% reduction in frequency of accidents, and 10 became completely dry. Two patients with Parkinson's disease achieved 100% reduction in urinary accidents in spite of tremor, bradykinesia, and rigidity which impaired their toileting skills. Significant improvement was also documented in patients who had a cystocele or history of stroke, had failed previous treatment with anticholinergic medication, were under treatment for depression, or, in one case, with post-prostatectomy incontinence. Patients with significant co-morbidity who required specialized intervention were successfully treated by modifying the approach to accommodate functional impairments. For example, a 79 year old woman, blind since childhood, was able to achieve continence using auditory biofeedback and by maintaining bladder diaries in Braille. This is not to say that behavioral treatment is appropriate for all patients with incontinence. In the present study approximately a third of patients evaluated were considered ineligible based on their mental status, urological findings, or more urgent medical problems. In addition, behavioral treatment requires that the patients be able to learn new skills and participate actively in their own treatment across several weeks. While we found most patients were motivated to cooperate in this way, others will find it unacceptable. In this study all patients who were treated with biofeedback-assisted behavior therapy reduced their incontinence. However, there was wide variation in outcome, and we were not able to predict magnitude of response from the information we gathered in the evaluation. Treatment outcome was not related to age, duration or severity of incontinence, frequency of urination, nocturia, or the number of other health problems. Thus, we have no reliable predictor variables,
JAGS-APRIL 1992-VOL. 40, NO.4
and, therefore, referrals for behavior therapy should be available for any cognitively intact older adult with incontinence. Data on long term maintenance of treatment effects were not collected in this study. Previous studies have examined maintenance, 11, 14 and more work is needed on this issue, particularly with regard to behavioral therapies in which ongoing patient compliance is of utmost importance. One benefit of the team approach was that contributory problems such as urinary tract infections or poorly controlled diabetes mellitus were managed by the continence team, with subsequent improvement in urinary incontinence. Integration with a geriatric center had the advantage that patients found to have related non-medical problems, such as depression or caregiver burden, could be treated by other professionals at the Center and then return to the Continence Program. Finally, a continence team composed of specialists in geriatric health care is experienced in approaching the older adult patient in a comprehensive manner. By integrating behavioral treatment with this philosophy of geriatric health care, it is possible to offer effective and acceptable low-risk treatment, even to multiply impaired older patients.
REFERENCES 1. Diokno AC, Brock BM, Brown MB et al. Prevalence of urinary incontinence and other urological symptoms in the noninstitutionalized elderly. J Urol 1986;136:1022. 2. Ouslander JG. Urinary incontinence in nursing homes. J Am Geriatr Soc 1990;38:289-291. 3. Brazda JF. Washington report. The nation's health. March, 1983. 4. NIH Consensus Conference. Urinary incontinence in adults. JAMA 1989;261:2685-2695. 5. Brink C, Wells T, Diokno A. A continence clinic for the aged. I Gerontol Nurs 1983;9:651-655. 6. Fantl [A, Wyman JF, Harkins SW et al. Bladder training in the management of lower urinary tract dysfunction in women: A review. J Am Geriatr Soc 1990;38:329-332. 7. Wells TI. Pelvic (floor) muscle exercise. J Am Geriatr Soc 1990;38:333337. 8. Burgio KL, Engel BT. Biofeedback-assisted behavioral training for elderly men and women. J Am Geriatr Soc 1990;38:338-340. 9. Bums PA, Pranikoff K, Nochajski T et al. Treatment of stress incontinence with pelvic floor exercises and biofeedback. J Am Geriatr Soc 1990;38:341344. 10. Burton IR, Pearce KL, Burgio KL et al. Behavioral training for urinary incontinence in elderly ambulatory patients. J Am Geriatr Soc 1988;36:693-698. 11. Baigis-Smith I, Smith DAJ, Rose M et al. Managing urinary incontinence in community residing elderly patients. Gerontologist 1989;29:229-233. 12. Fantl [A, Wyman JF, McClish OK et al. Efficacy of bladder training in older women with urinary incontinence. JAMA 1991;265:609-613. 13. Folstein MF, Folstein SE, McHugh PRo "Mini-Mental State": a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-98. 14. Burgio KL, Whitehead WE, Engel BT. Urinary incontinence in the elderly: Bladder-sphincter biofeedback and toileting skills training. Ann Intern Med 1985;103:507-515.
