Fecal incontinence (FI) is a common symptom, not. Symptoms and Quality of Life in Community Women With Fecal Incontinence

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4: Symptoms and Quality of Life in Community Women With Fecal Incontinence ADIL E. BHARUCHA,* ALAN R. ZINSMEISTER, G. RICHARD LOCKE,* CATHY SCHLECK, KIMBERLY MCKEON, and L. JOSEPH MELTON *Division of Gastroenterology and Hepatology, Division of Biostatistics, and Division of Epidemiology, Mayo Clinic College of Medicine, Rochester; and Department of Obstetrics and Gynecology, Olmsted Medical Center, Rochester, Minnesota Background & Aims: Assessments of symptom severity are necessary to guide therapy in fecal incontinence (FI). However, there is no consensus on how self-reported symptom severity in FI should be measured for this purpose, and the relationship between symptoms of FI and its impact on quality of life (QOL) is not known. Methods: A questionnaire was mailed to an age-stratified random sample of 5300 women identified through the Rochester Epidemiology Project. Symptom severity was assessed by a validated scale (Fecal Incontinence and Constipation Assessment [FICA]), and impact on 15 domains of QOL was evaluated for subjects who had any FI during the past year. The scale incorporated the type, frequency, and amount of FI, as well as the circumstances surrounding FI (ie, urge or passive FI). A QOLweighted symptom severity score was derived by weighting the responses for these characteristics, in each subject, by the average QOL impact for all subjects who reported that specific symptom characteristic. The relationship between symptom severity and the QOLweighted symptom severity score was assessed by Spearman rank correlation. Results: Altogether, 2800 of 5300 (53%) women responded to the survey. FI symptoms were mild (45%), moderate (50%), or severe (5%). Among women with FI, the FICA symptom severity scale was strongly correlated (r s 0.92, P <.0001) with the QOL-weighted symptom severity score. Conclusions: Among unselected women in the community, self-reported symptom severity, rated by the FICA scale, was strongly correlated with the impact of FI on QOL, supporting the use of this scale to assess symptom severity in FI. Fecal incontinence (FI) is a common symptom, not only among nursing home residents but also in the general population. 1 Among adult women in Olmsted County, Minnesota, more than 1 of 10 (and 1 in 5 women aged 50 years or older) had FI during the past year, as defined by at least one episode of involuntary leakage of solid or liquid stool from the anus not attributable to a temporary illness (eg, acute gastroenteritis), 2 and the symptom is equally prevalent in men. 3 Patient advocates emphasize that FI, although not a life-threatening symptom, can significantly impair lifestyle. 4 Previous community-based studies in elderly subjects have demonstrated that FI is associated with impaired role functioning as assessed by the SF-36 questionnaire, as well as anxiety, depression, and disability. 5,6 However, the impact of FI on quality of life (QOL), specifically on activities that are likely to be by FI (eg, eating, shopping), in the general community is unknown. Therefore, a recent National Institutes of Health sponsored consensus conference underscored the need to develop standardized scales to assess symptom severity and QOL in FI. 4 Several barriers hinder an assessment of the severity of FI among subjects presenting for care. First, up to 50% of patients with FI do not disclose the symptom to their physician unless asked, perhaps because they are embarrassed to do so. 7 Second, scales for rating symptom severity in FI incorporate the frequency and type but not the amount of leakage Without the latter, FI severity would be identical for 2 subjects, one of whom had minor staining and the other a large liquid bowel movement once a week. We previously developed and validated a scale for rating symptom severity in FI that includes 4 components (frequency of FI, type of FI, amount of FI, and circumstances surrounding FI [ie, urge or passive FI]) derived from a self-report questionnaire (the Fecal Incontinence and Constipation Assessment [FICA]). 12 All 4 components are weighted similarly to derive an overall symptom severity score. However, this method for weighting variables has been criticized because it assumes that different components (eg, amount and frequency) are equally important in determining the severity of FI. 11 Moreover, physician-assigned arbitrary Abbreviations used in this paper: FI, fecal incontinence; FICA, Fecal Incontinence and Constipation Assessment; QOL, quality of life 2006 by the American Gastroenterological Association Institute /06/$32.00 doi: /j.cgh

2 August 2006 SYMPTOM SEVERITY IN FECAL INCONTINENCE 1005 weights might not accurately reflect symptom severity from a patient s perspective. 13 For example, patients assigned a higher severity score to incontinence for flatus than did physicians; conversely, physicians assigned a higher severity score for solid stool incontinence compared with patients. Existing symptom severity scales also do not shed light on the impact of FI on QOL, and separate scales have been devised to accomplish this. 12 However, it is unclear whether symptom severity in FI is correlated to QOL. This is an important consideration because symptom severity is not correlated to its impact on QOL in certain other disorders (eg, urinary incontinence). 14 Thus, it is conceivable that women who are tethered to a toilet have mild FI symptoms but a dismal QOL. To address these issues and to evaluate a scoring system for use in the clinic, we assessed the relationship between FI symptom severity and QOL among a group of unselected women with FI in the community. This study was conducted as part of a large investigation into the epidemiology of FI among community women. 2 Methods The Olmsted County population comprises approximately 100,000 persons, of whom 96% are white; sociodemographically, the community is similar to the white population of the United States. 15 In Olmsted County, 80% of the population resides within 5 miles of Rochester, and residents receive their medical care almost exclusively from 2 group practices, Mayo Medical Center and Olmsted Medical Center. Annually, more than 80% of the entire population is attended by one or both of these 2 practices, and nearly everyone is seen at least once during any given 3-year period. A unique medical records linkage system, the Rochester Epidemiology Project, provides an enumeration of this population from which samples can be drawn. 15 We used this system to draw a random sample of 5200 female residents of Olmsted County stratified by age (10-year intervals between 20 and 80 years). Altogether, 2800 of 5300 (53%) women responded to the survey. Survey Technique Each subject in this random sample was mailed a Fecal Incontinence and Constipation Assessment. The FICA is a reliable and valid measure of gastrointestinal symptoms, particularly FI, constipation, and irritable bowel syndrome. 12 Subjects who responded affirmatively to the question, In the past 12 months have you experienced accidental leakage of liquid or solid stool?, were asked additional questions about the type and frequency of FI, circumstances surrounding FI (ie, urge or passive FI), and the impact of FI on lifestyle. Subjects were asked not to consider leakage during short-term diarrheal illnesses. Incontinence for flatus can be embarrassing but was not considered here because patients find it difficult to quantify the frequency of flatus leakage, and because there is no cutoff to discriminate inadvertent expulsion of gas from incontinence. 16 Severity of Fecal Incontinence Only subjects who had FI during the past year answered more detailed questions. Subjects who reported they often ( 25% of time) or usually ( 75%) experienced an urgent need to empty their bowels making them rush to the toilet were considered to have rectal urgency. Subjects who often ( 25% of time) or usually ( 75%) leaked liquid or solid stool without any warning were considered to have passive incontinence. Patients who did not report symptoms of urge or passive incontinence were classified as neither, whereas those who had symptoms of urge and passive incontinence were classified as combined incontinence. The severity of FI was rated by a previously validated symptom severity scale. 12 A severity score (maximum score, 13) was calculated by summing scores for individual components in this scale (Table 1). Symptom severity scores of 1 6, 7 10, and were categorized as mild, moderate, and severe, respectively. In contrast to our original report, the use of sanitary devices to protect against leakage was not considered in the severity scale used for this study, because this parameter might reflect coping mechanisms rather than the severity of incontinence per se. Impact on Quality of Life The impact of stool leakage on QOL was assessed for 15 domains, adapted from the instrument developed by Rockwood et al 13 ; each was scored as not applicable/not (0), mildly (1), moderately (2), or severely (3). These domains were grouped into 3 categories: (1) activities associated with predictable toilet access (ie, employment, work around house, sex life, visit friends or relatives at their home, stay overnight away from home, and family relationships); (2) activities associated with unpredictable toilet access (ie, going to a movie or church, shopping, recreational activities or sports, ability to leave home, ability to travel by car, and the ability to travel by plane or train); and (3) activities that involve eating (ie, ability to eat before leaving home, ability to go out to eat). By averaging the item values over the domains in each category, 3 scores, corresponding to these 3 categories, were computed for each subject. The scores for activities associated with predictable versus unpredictable toilet access and for predictable toilet access versus eating were compared by a scatter plot. A test for the equiangular line was examined based at an level of Quality of Life Weighted Symptom Severity Score The impact of FI on QOL was quantified by a 2-stage process, ie, a QOL factor was estimated for each symptom attribute in Table 1. Thereafter, this QOL factor was applied to estimate the QOL-weighted symptom severity score for each subject.

3 1006 BHARUCHA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 Table 1. (FICA Symptom) Severity Scale and Symptom Impact on QOL a in FI Symptoms Categories Frequency 1/mo 1/mo several Daily times/wk Symptom severity score QOL factor Composition Mucus/liquid stool Solid stool Liquid and solid stool Symptom severity score QOL factor Amount Small (ie, staining only) Moderate (ie, requiring change of underwear) Large (ie, requiring change of all clothes) Symptom severity score QOL factor Urgency or passive incontinence Neither Passive incontinence Urge incontinence Combined urge and passive incontinence Symptom severity score QOL factor a The symptom severity score was assigned by a physician to grade the severity of FI. The QOL factor represents the average impact of a given symptom on QOL for all subjects who reported that symptom. For each category of every symptom characteristic (eg, small amount of leakage) in Table 1, we calculated the mean QOL impact, separately for all 15 QOL domains, for all subjects who reported that specific symptom characteristic. The sum of these mean impact scores across all 15 domains was considered the QOL impact factor for each cell in Table 1. Subsequently, the QOL factor corresponding to a subject s responses for each of the 4 FI components (ie, frequency, amount, and type of leakage, and circumstances surrounding leakage) was summed to obtain the QOL-weighted symptom severity score. For example, by using the values in Table 1, a subject who had urge FI for a small amount of liquid stool/mucus once per week would have a QOLweighted symptom severity score of (ie, 6.92 [frequency once/week] 2.77 [liquid stool/mucus] 3.26 [small amount] 8.21 [urgency]). The correlation between the original FI severity scale and the QOL-weighted symptom severity score was examined by Spearman rank correlation. Results Symptom Severity in Fecal Incontinence In this study, 507 of 2800 women who responded to the survey had any FI during the past year. A majority of women with FI had infrequent symptoms (54% less than monthly), and 59% reported staining of underwear only (Table 2). The composition of leakage was approximately evenly distributed among leakage of liquid stool/mucus, stool, and both liquid stool/mucus and stool. The symptom severity score was calculated by compiling the 4 attributes (frequency, amount, and type of FI, and circumstances surrounding FI) (Table 1). Thus, 220 women (43%) had mild, 248 (49%) had moderate, and 23 (5%) had severe symptoms, respectively. These results have been presented in detail elsewhere. 14 Impact of Fecal Incontinence on Quality of Life FI had a moderate or severe impact on 1 or more of the 15 QOL domains in 23% of the women with FI Table 2. Symptom Severity in FI Among an Age-Stratified Sample of Olmsted County, Minnesota, Women Aged 20 Years Attribute a n (%) QOL-weighted symptom severity score b Frequency (N 503) Less than once/mo 258 (54%) More than once/mo several 212 (44%) times/wk Daily 8 (2%) Composition (N 502) Liquid stool/mucus 138 (29%) Stool only 153 (32%) Liquid stool/mucus and 187 (39%) solid stool Amount (N 502) Small (underwear stain only) 283 (59%) Moderate (requiring change 161 (34%) of underwear but not clothes) Large (requiring change of 34 (7%) clothes) Incontinence type (N 496) Unspecified 261 (55%) Passive 22 (5%) Urge 166 (35%) Combined 29 (6%) a Numbers represent total number of subjects who answered items pertaining to symptoms and QOL in each category. b This value represents the QOL-weighted symptom severity score (mean standard error of mean), aggregated across all 4 characteristics, for all subjects within a specific category (eg, FI less than once/mo).

4 August 2006 SYMPTOM SEVERITY IN FECAL INCONTINENCE 1007 Table 3. Impact of FI on QOL Among an Age- Stratified Sample of Olmsted County, Minnesota, Women Aged 20 Years Attribute Not Mildly Moderately Severely Not applicable Missing Daily living 333 (66%) 99 (19%) 31 (6%) 6 (1%) 8 (2%) 30 (6%) Predictable toilet access Employment 319 (63%) 45 (9%) 19 (3%) 4 (1%) 80 (16%) 40 (8%) Work around house 380 (75%) 60 (11%) 16 (3%) 2 (0.4%) 9 (2%) 40 (8%) Sex life 328 (65%) 42 (8%) 13 (3%) 5 (1%) 73 (14%) 46 (9%) Visit friends or relatives 369 (73%) 59 (12%) 30 (6%) 10 (2%) 7 (1%) 32 (6%) at their home Stay overnight away from 361 (71%) 55 (11%) 25 (5%) 14 (3%) 21 (4%) 31 (6%) home Family relationships 415 (82%) 36 (7%) 13 (2%) 4 (1%) 14 (3%) 25 (5%) Unpredictable toilet access Movie or church 358 (71%) 61 (12%) 33 (6%) 9 (2%) 10 (2%) 36 (7%) Shopping 326 (64%) 98 (19%) 35 (7%) 12 (2%) 7 (1%) 29 (6%) Recreational activities or 327 (65%) 72 (14%) 30 (6%) 11 (2%) 32 (6%) 35 (7%) sports Leave home 336 (66%) 90 (18%) 39 (8%) 10 (2%) 6 (1%) 26 (5%) Travel by car 325 (64%) 97 (19%) 39 (7%) 9 (2%) 9 (2%) 28 (6%) Travel by plane or train 333 (66%) 58 (11%) 23 (4%) 15 (3%) 50 (10%) 28 (6%) Involve eating Eat before leaving home 345 (68%) 67 (13%) 34 (7%) 17 (3%) 9 (2%) 35 (7%) Out to eat 345 (68%) 66 (13%) 45 (9%) 17 (3%) 8 (2%) 26 (5%) during the past year. The overall impact of FI on QOL (ie, averaged across all 15 domains) was correlated (r 0.3, P.001) to general health rated as excellent, very good, good, fair, or poor. As shown in Table 3, the proportion reporting moderate to severe impact for a given domain ranged from 3% 4 % (eg, for family relationships, employment, sex life) to 12% (for the ability to eat outside home or going out to eat). Figure 1 demonstrates that scores for activities in which toilet access was unpredictable and for activities that involved eating were each higher than scores for activities associated with predictable toilet access (ie, more values were situated to the left of the equiangular line; P.0001). Relationship Between Symptom Severity Scale and Quality of Life Weighted Symptom Severity Score The QOL-weighted symptom severity score was devised to refine the assessment of impact of FI on QOL. Table 1 provides the QOL factor for each symptom (ie, the average impact of a given symptom on QOL for all subjects who reported that symptom). Subsequently, Figure 1. Impact of FI on QOL. Compared with activities associated with predictable toilet access, FI had a greater impact on activities that entailed unpredictable toilet access (A) and on activities that involved eating (B).

5 1008 BHARUCHA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 Figure 2. Relationship between symptom severity rated by the FICA scale and symptom-adjusted QOL score among women with FI in the community. these QOL factors were used to calculate the QOLweighted symptom severity score for every subject. For example, a subject who leaked a large amount of liquid and solid stool daily and who had features of urge and passive FI would have the maximum (worse) QOLweighted symptom severity score, ie, 45. Table 2 demonstrates the distribution of QOL-weighted symptom severity scores by the characteristics of FI. For example, among the subjects who reported FI less than once/ month, the QOL-weighted symptom severity score, derived by aggregating the QOL factor across all 4 categories, was (mean standard error of the mean). This value is greater than the lowest possible QOL-weighted symptom severity score (ie, 11.03), indicating that among some women who had FI less than once/month, the other characteristics (ie, composition, amount, and type of FI) were not in the lowest category. Table 2 suggests that the a priori physician-devised hierarchy within each symptom category (eg, frequency of FI) was correlated with the QOL-weighted symptom severity scores for that category. Moreover, the symptom severity score derived from the FICA symptom severity scale was strongly correlated (r s 0.92, P.0001) with the QOL-weighted symptom severity score. Figure 2 illustrates the distribution of the QOL-weighted symptom severity scores by categories (ie, mild, moderate, or severe) of the original FI severity scale. Discussion The severity of FI can be rated either by evaluating symptoms or by judging its impact on QOL, and generic or disease-specific measures can be used for either approach. Consistent with other scales for rating FI severity, the FICA scale also incorporates the type and frequency of incontinence However, 3 key differences between the FICA scale and other instruments need to be emphasized. First, in contrast to the Fecal Incontinence Severity Index, 11 the FICA does not characterize separately the frequency for every type (ie, solid, liquid, mucus, or gas) of incontinence, because our clinical experience suggests that patients find it difficult to characterize the frequency of incontinence, particularly for flatus. In contrast to leakage of liquid or solid stool, it is difficult to ascertain when leakage of flatus is physiologic versus pathologic. It is also difficult to characterize the odor of flatus, which might be as or more troublesome to patients than the frequency of flatus leakage. Second, the FICA scale includes the symptom of rectal urgency, which can prompt considerable anxiety in incontinent subjects. The St Marks scale also incorporates urgency, assigning a score of 0 4 for patients who can or cannot defer defecation for 15 minutes, respectively. 10 Our clinical experience suggests that this threshold for discriminating normal from excessive rectal urgency is relatively liberal, because a majority of incontinent patients were unable or reluctant to defer defecation for 15 minutes. Third, in contrast to other FI severity scales, the FICA also quantifies the amount of stool leakage. Without the latter, the severity of FI would be identical for 2 subjects who had minor staining or a large liquid bowel movement once a week. Previous studies have demonstrated that FI is associated with anxiety, depression, disability, and impaired role functioning in people aged 65 years or older, 17 and both patient advocates and clinicians recognize that FI can have a devastating impact on QOL. 4 However, this is the first study to document the impact of FI in the community. Only 23% of women with FI reported that the symptom had a moderate to severe impact on 1 or more domains of QOL. Allowing for differences in methods, our results are comparable to the proportion (32%) of incontinent subjects who reported that the symptom had a lot of impact on QOL in the only previous community-based study to address this issue. 3 When responses across all 3 categories of FI (mild, moderate, and severe symptoms) were pooled, the results also suggested that activities that entail eating or unpredictable toilet access were more likely to be than activities associated with predictable toilet access. In this study, 6% of women with mild symptoms, 35% of women with moderate symptoms, and 82% of women with severe symptoms reported a moderate or severe impact on at least 1 of the 15 domains of QOL. These observations support an association between the symptom severity scale and impact of QOL, as quantified by a relatively crude assessment of QOL (ie, moderate or

6 August 2006 SYMPTOM SEVERITY IN FECAL INCONTINENCE 1009 severe impact on 1 domain). Thus, this summary measure of QOL would not recognize differences between subjects who report a moderate-severe impact on 1 domain of QOL but differ in responses to other domains. The QOL-weighted symptom severity score overcomes this limitation and provides a more refined assessment of QOL by incorporating every domain of QOL. However, the QOL factors shown in Table 1 are required to calculate the QOL-weighted symptom severity score. By contrast, the FICA symptom severity score was devised a priori to be more user-friendly by assigning arbitrary weights (ie, 0, 1, 2, 3) for symptoms within each category (eg, frequency of FI). The excellent correlation between the FICA symptom severity score and the QOLweighted symptom severity score indicates that these weights closely approximate the impact of these symptoms on QOL as perceived by patients. This inference is further supported by the correlation between severity rating within each symptom category (eg, frequency of FI) and the QOL-weighted symptom severity score. Conversely, if the physician-assigned weights did not approximate to the relative impact of symptom attributes on QOL, the symptom-adjusted QOL score would have correlated poorly with the symptom severity score. Therefore, the FICA symptom severity score, which is simple to use in the office, is a reasonable indication not only of the physical manifestations of FI (ie, symptom severity) but also its impact on QOL. Although it has been suggested that measures of stool leakage might underestimate the severity of FI in people who avoid FI by staying close to a toilet (eg, by staying at home), 18 the strong correlation between symptom severity and the QOL-weighted symptom severity score suggests that this is not a major pitfall of the FICA symptom severity score. These findings contrast to urinary incontinence, in which the impact on QOL is weakly correlated with symptom severity. 14 Consequently, separate scales are required to assess symptom severity and the impact on QOL in urinary incontinence but not fecal incontinence, where one assessment should suffice in clinical practice. In summary, the FICA symptom severity score is a simple, useful score for quantifying symptom severity in FI and its impact on QOL. We previously documented the concurrent validity and repeatability of the FICA symptom severity scale in patients at a tertiary referral center. Although the response rate to the survey was 53%, the excellent correlation between the FICA symptom severity scale and the more complex QOL-weighted symptom severity scale indicates that the former is a simple and accurate way of characterizing symptom severity in FI. Further studies are required to validate the FICA symptom severity score against QOL in other ethnic groups and to assess the responsiveness of the FICA symptom severity scale to therapy. References 1. Nelson RL. Epidemiology of fecal incontinence. Gastroenterology 2004;126:S3 S7. 2. Bharucha AE, Zinsmeister AR, Locke GR, et al. Prevalence and burden of fecal incontinence: a population based study in women. Gastroenterology 2005;129: Perry S, Shaw C, McGrother C, et al. Prevalence of faecal incontinence in adults aged 40 years or more living in the community. Gut 2002;50: Norton NJ. The perspective of the patient. Gastroenterology 2004;126:S175 S Edwards NI, Jones D. The prevalence of faecal incontinence in older people living at home. Age Ageing 2001;30: O Keefe EA, Talley NJ, Zinsmeister AR, et al. Bowel disorders impair functional status and quality of life in the elderly: a population-based study. J Gerontol A Biol Sci Med Sci 1995;50: M184 M Leigh RJ, Turnberg LA. Faecal incontinence: the unvoiced symptom. Lancet 1982;1: Jorge JM, Wexner SD. Etiology and management of fecal incontinence. Dis Colon Rectum 1993;36: Pescatori M, Anastasio G, Bottini C, et al. New grading and scoring for anal incontinence: evaluation of 335 patients. Dis Colon Rectum 1992;35: Vaizey CJ, Carapeti E, Cahill JA, et al. Prospective comparison of faecal incontinence grading systems. Gut 1999;44: Rockwood TH, Church JM, Fleshman JW, et al. Patient and surgeon ranking of the severity of symptoms associated with fecal incontinence: the fecal incontinence severity index. Dis Colon Rectum 1999;42: Bharucha AE, Locke GR, Seide B, et al. A new questionnaire for constipation and fecal incontinence. Aliment Pharmacol Ther 2004;20: Rockwood TH, Church JM, Fleshman JW, et al. Fecal incontinence quality of life scale: quality of life instrument for patients with fecal incontinence. Dis Colon Rectum 2000;43: Naughton MJ, Donovan J, Badia X, et al. Symptom severity and QOL scales for urinary incontinence. Gastroenterology 2004; 126:S114 S Melton LJ 3rd. History of the Rochester Epidemiology Project. Mayo Clinic Proc 1996;71: Whitehead WE, Wald A, Norton NJ. Treatment options for fecal incontinence. Dis Colon Rectum 2001;44: Miner PB Jr. Economic and personal impact of fecal and urinary incontinence. Gastroenterology 2004;126:S8 S Rockwood TH. Incontinence severity and QOL scales for fecal incontinence. Gastroenterology 2004;126:S106 S113. Address requests for reprints to: Adil E. Bharucha, MD, Clinical and Enteric Neuroscience Translational and Epidemiological Research Program (C.E.N.T.E.R.), Mayo Clinic, 200 First St SW, Rochester, MN bharucha.adil@mayo.edu; fax: Supported in part by USPHS NIH grants R01 HD38666, R01 HD41129, and R01 AR and General Clinical Research Center grant M01 RR00585.