SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Background : The prevalence, severity and risk factors of faecal incontinence in women in the community are incompletely characterized.

Aim : To develop and validate a self-report questionnaire (faecal incontinence and constipation assessment) to address these issues.

Method : Eighty-three women completed the instrument; 20 randomly selected patients answered the faecal incontinence and constipation assessment again 6 weeks later. A gastroenterologist also completed the faecal incontinence and constipation assessment in all 83 subjects after a detailed clinical assessment. Concurrent validity was evaluated by comparing the patient's self-report to a doctor interview for every question. Reproducibility was evaluated by a test–retest approach for every question. The severity of faecal incontinence was rated by incorporating the frequency and type of faecal incontinence, rectal urgency and use of sanitary devices.

Results : The questionnaire was well-understood. Reproducibility [median κ statistic, 0.80 (interquartile range: 0.66–0.90)]; and concurrent validity [0.59 (0.47–0.67)] were acceptable. For the index question on faecal incontinence, the κ for reproducibility and concurrent validity was 0.90 and 0.95, respectively. The faecal incontinence severity score was also valid (κ = 0.5).

Conclusion : The faecal incontinence and constipation assessment has excellent reproducibility and reasonable validity for assessing the presence, risk factors and severity of faecal incontinence and associated bowel disorders in women when compared against clinical assessment.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Functional constipation and faecal incontinence (FI) are relatively common symptoms in the community, particularly in the elderly.1 Functional constipation and FI affect approximately 15% and 2.2–11.3% of the population respectively.2, 3 Constipation and FI may coexist, are often attributable to disordered anorectal sensorimotor function and result from similar risk factors. Our understanding of the clinical features and pathophysiology of FI is derived predominantly from studies in tertiary centres. Community-based studies on FI are particularly important because incontinent patients are often embarrassed to disclose the symptom to their doctor.4 However, previous epidemiological studies have only partially characterized the risk factors for FI, the severity of FI and the association between constipation and FI. For example, findings from tertiary centres suggest that obstetric anal sphincter injury is an important risk factor for FI need to be confirmed in community-based studies.

Since face-to-face interviews are relatively expensive, population-based research is often conducted by questionnaires. Before conducting population-based studies, the measurement properties, primarily reliability (reproducibility) and validity should be evaluated to ensure the data collected are accurate. A number of questionnaires have been developed and tested for use in population-based research of gastrointestinal symptoms, providing insights into the prevalence, risk factors, disease burden and natural history of several functional gastrointestinal disorders. The original Bowel Disease Questionnaire developed by Talley et al. focused on irritable bowel syndrome (IBS) and functional dyspepsia.5 Modified versions of this questionnaire have been validated in an elderly population,6 patients with diabetes mellitus,7 gastro-oesophageal reflux disease8 and FI.9

The previously validated FI questionnaire9 was acceptable for characterizing bowel habits, the nature of, circumstances surrounding, and risk factors for fecal incontinence, and urinary symptoms. However, the question and response format were not always consistent with the Rome II diagnostic criteria for functional GI disorders. Moreover, the questionnaire did not assess for symptoms of IBS, or characterize the severity of FI, or its impact on quality of life. Other FI questionnaires have been designed predominantly for clinic-based studies, focus on disease severity and quality of life, but do not address the risk factors for incontinence, or assess bowel habits.10, 11 Therefore, our objectives were to develop and evaluate a self-report questionnaire to measure three issues: (i) the prevalence of constipation and FI in the community; (ii) assess the risk factors for FI; and (iii) characterize the severity of FI.

Design of the questionnaire

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

The Institutional Review Board of the Mayo Clinic approved this study. The faecal incontinence and constipation assessment (FICA) contained 98 questions and was modelled after a previously validated bowel diseases questionnaire 5, 12 and a focused FI questionnaire.9 These questionnaires have been extensively tested and used to assess the epidemiology of IBS and constipation in the community.13, 14 The questions in the FICA were framed to be consistent with Rome II diagnostic criteria for functional bowel disorders and abdominal pain. Questions were worded to ensure they were understood by the general population. Extra large print was used for maximum legibility. Where medical terminology was necessary, descriptions in layperson's terms were used to clarify the question. For example, a rectocoele was described as a ‘protrusion of the rectum through the opening of the vagina’. This questionnaire was designed to assess the following specific areas:

  • 1
    To carefully characterize bowel habits including stool frequency, stool consistency and the process of defecation.
  • 2
    To ascertain whether subjects had abdominal pain, constipation, or FI. The first question in the sections on abdominal pain, constipation and FI served as a branch point, such that subjects who indicated they did not have the symptom proceeded to the next section. Conversely, patients who answered ‘yes’ to the first question in each section proceeded to answer a series of questions within that section. For FI, the branch question was as follows, ‘Have you ever had problems with leakage of stool (accidents or soiling)’? Respondents were requested not to consider stool leakage during short-term diarrhoeal illnesses such as the flu or a virus. Thereafter, respondents were asked to report all symptoms they had experienced during the past 12 months only. The type of stool leakage was rated as gas only, a small amount of stool, moderate amount of stool (often requiring change of underwear) or a large bowel movement of liquid stool (often requiring a complete change of clothes).
  • 3
    To assess the impact of bowel function on activities, namely: (i) activities of daily living, i.e. daily living, work (occupation), and work around the house; (ii) recreation, e.g. sports and social/entertainment; and (iii) other behaviour, i.e. family relationships, travel and sexual life. The impact of bowel symptoms on these functions were graded as not affected, mildly affected, moderately affected and severely affected as previously described.15
  • 4
    To identify if subjects had urinary incontinence and thereby evaluate the relationship between urinary and fecal incontinence. We included three previously validated questions:9 slow leakage or dribbling of urine throughout the day, urine leakage with coughing or sneezing and awareness of the need to urinate (pass your water) before the leakage occurred.
  • 5
    To assess for anorectal trauma or disorders, e.g. obstetric anal sphincter injury, pelvic/anorectal operations, rectal prolapse or rectocele.
  • 6
    To determine health care utilization, family history of bowel problems, use of laxative/suppositories during childhood and change in appetite or weight over the past year.
  • 7
    To measure frequency and severity of somatic complaints by a somatic symptoms checklist at the end of the survey.16

The severity of FI was rated by using questions within the instrument (Table 1). In addition to the frequency and type of leakage,17, 18 our scale also incorporated the number of perineal protective devices used daily for stool (not urinary) leakage and the severity of urgency, consistent with more recent instruments.19 Urgency was rated by asking subjects whether they were incontinent because they had great urgency and could not reach the toilet on time; responses included never, sometimes (i.e. < 25% of the time), often (i.e. > 25% of the time) and usually (i.e. > 75% of the time). Definitions for never, sometimes, often and usually were specified as indicated above in the questionnaire.

Table 1.  Scale for grading severity of fecal incontinence
 123
  1. Maximum total score = 12. Scores of 1–4, 5–8 and 9–12 were categorized as mild, moderate and severe faecal incontinence, respectively.

Frequency of incontinenceUp to once/month< 2/week≥ 2–3/week
Usual type of bowel incontinenceGas only/only enough to stain your underwear (size of a quarter)Small amount of stoolModerate or large amount of stool
Number of protective pads changed/dayNoneOne> 1
UrgencyNeverSometimesOften/usually

Testing of the questionnaire

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Women being evaluated in a clinic specialized in gastrointestinal motility disorders in the Division of Gastroenterology and Hepatology were asked to participate. The presence or absence of FI was not considered in recruitment. Eighty-three women volunteered to participate in the study, completing the FICA privately without any assistance whatsoever. The doctor (AEB) formulated his clinical impression by a detailed clinical interview, a physical examination and by reviewing clinical features and investigations recorded previously. The doctor, who was blinded to patients responses, completed the FICA thereafter (i.e. not in the patient's presence), completing all questions except for those pertaining to health care utilization and quality of life in FI. To assess the stability of the measure (i.e. test/retest reliability or reproducibility), 20 randomly selected subjects completed an identical FICA a second time by mail within 6 weeks of completing an initial questionnaire.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Reproducibility of the questionnaire was determined by calculating the κ statistic for each individual question using the self-administered test–retest data. Concurrent validity was measured by comparing the doctor's questionnaire responses against the patient's questionnaire responses question by question in a similar manner. In addition to individual questions, reproducibility and concurrent validity for diagnosing IBS, functional constipation and functional diarrhoea were also assessed by combining appropriate questions using established diagnostic (Rome II) criteria.20

For questions with a ‘yes–no’ response variable, intersource (i.e. patient initial vs. patient retest, and patient vs. doctor) agreement was assessed by the unweighted κ statistic, which represents the proportion of agreement beyond that expected by chance alone; it is scaled to vary from −1 to 1. A negative value indicates poorer than chance agreement; values below 0.2 reflect slight agreement; values of 0.21–0.40 fair agreement; values of 0.41–0.6 signify moderate agreement; values between 0.61 and 0.8 represent substantial agreement. A κ of 0.81–1.0 is considered almost perfect agreement.21 Because a low κ-value may reflect a low prevalence of that trait in the cohort and not lack of agreement,22 the proportion of positive agreement was also calculated.23, 24 For questions with multiple categorical, not ordinate responses intersource agreement was assessed by a weighted κ statistic.25 This weighting system assigns more weight to small degrees of disagreement than to larger degrees. For responses rated on an ordinal scale (e.g. number of bowel movements) agreement was assessed by the interclass correlation coefficient (ICC).26

A 95% confidence interval (CI) of κ, which provides a test of the hypothesis that the underlying value of κ differs from zero, was also calculated for each question. In addition, the proportion of overall agreement for each question was tabulated. Values are presented as mean κ (and 95% confidence limits) for individual questions, and median κ (and interquartile range, IQR) for group of questions.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

All participants were women, with a mean age of 38 years (range: 20–85 years). Based on doctor interview, 47 patients (57%) had FI while 34 (64%), 24 (42%) and one (1%) patient had symptom-based (Rome II) criteria for functional constipation, IBS and functional diarrhoea respectively. Similarly, among the 20 randomly selected patients for test–retest agreement, eight patients (40%) had FI, eight had symptoms of IBS, 10 had functional constipation, while one each had functional diarrhoea and abdominal bloating. There was considerable overlap between fecal incontinence and other functional gastrointestinal disorders. Thus, of 49 patients with FI, 15 had IBS, nine had functional constipation, one had functional diarrhoea and three had functional bloating.

Reproducibility for individual questions

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

In the 20 subjects who returned the questionnaire after self-administration of the retest, the overall median κ was 0.80 (IQR: 0.66–0.90). The 95% CI included a value of zero in only six of 98 questions. Overall agreement tended to be high. κ-Values for questions dealing with bowel and bladder symptoms are listed in Table 2. For the branch question on FI, i.e. ‘Have you ever had problems with leakage of stool (accidents or soiling)’? the κ statistic for reproducibility was 0.90. κ-Values for questions dealing with risk factors for FI, impact on quality of life and health care utilization are listed in Table 3.

Table 2.  Reproducibility and concurrent validity for bowel and bladder symptoms
 ReproducibilityConcurrent validity
Observed agreement (%)κ or weighted κ* or ICC†95% CIObserved agreement (%)κ or weighted κ* or ICC†95% CI
  1. The overall median (IQ range) for group of questions incorporated items for which validity and reproducibility were assessed by κ, but not those questions analysed by the ICC.

  2. IQR, interquartile range; ICC, interclass correlation coefficient; CI, confidence interval.

General bowel habits
 Bowel pattern in last year800.52†0.24–0.72770.44*0.30–0.61
 Number of bowel movements (BM)/week850.96†0.90–0.98650.89†0.83–0.93
 Incomplete emptying700.63*0.31–0.74430.32*0.16–0.51
 Straining during defecation950.860.41–0.93840.660.48–0.78
 Duration of straining (years)700.93†0.84–0.97660.84†0.77–0.89
 Presence of hard stools850.670.32–0.82780.560.37–0.70
 Anal blockage950.880.48–0.94830.660.47–0.77
 Presence of mucus950.880.48–0.94790.570.38–0.70
 Abnormal bloating950.770.26–0.89790.590.41–0.71
 Abdominal distention950.860.41–0.93760.520.34–0.66
 Anal digitation during defecation1001.00.59–1.0890.770.58–0.86
 Position change during defecation1001.00.56–1.0780.560.38–0.70
 Time spent on toilet800.92†0.81–0.97660.79†0.77–0.89
 Abdominal pressure during defecation790.88*0.47–0.90660.63*0.42–0.75
 Typical BM680.88*0.49–0.89520.72*0.52–0.78
 Frequent (>3) BM daily930.860.42–0.93840.670.45–0.79
 Loose or watery stools600.820.46–0.83480.630.44–0.72
 Presence of urgency1001.00.55–1.0820.650.44–0.77
Overall above set median (IQ range)89 (79–95)0.86 (0.80–0.88) 78 (66–81)0.63 (0.56–0.66) 
Abdominal pain
 Abdominal pain950.890.48–0.94850.650.4–0.77
 Duration of abdominal pain460.29*−0.19–0.68420.46*0.24–0.63
 Severity of abdominal pain460.61*0.42–0.65340.54*0.42–0.61
 Location of abdominal pain770.62*0.25–0.78610.15*0.02–0.47
 Frequency of abdominal pain690.86†0.62–0.95580.450.23–0.63
 Relief with defecation620.91*0.48–0.92680.62*0.40–0.74
 Number of stools770.63*0.27–0.78680.46*0.23–0.65
 Type of stools770.78*0.39–0.83520.30*0.09–0.54
 Frequency of rectal pain380.60†0.10–0.85520.380.14–0.58
Overall above set median (IQ range)69 (46–77)0.63 (0.62–0.84) 58 (52–68)0.46 (0.38–0.58) 
Treatment of constipation
 Duration of constipation660.89*0.73–0.96710.75*0.59–0.85
 Severity of constipation590.61*0.25–0.80530.71*0.48–0.79
 Any treatment670.78*0.42–0.82780.69*0.40–0.79
 Effectiveness of bran/fibre85−0.01*−0.06–0.61590.60*0.45–0.64
 Effectiveness of laxatives670.74*0.40–0.77560.41*0.22–0.62
 Effectiveness of enemas770.58*0.28–0.72540.54*0.33–0.66
 Effectiveness of suppositories1001.0*0.48–1.0650.67*0.39–0.75
Overall above set median (IQ range)67 (67–81)0.74 (0.60–0.84) 59 (55–68)0.67 (0.57–0.70) 
Faecal incontinence
 Any leakage of stool950.900.51–0.95980.950.78–0.97
 Following questions restricted to subjects with faecal incontinence
  Duration since onset of faecal incontinence870.98†0.91–0.99550.76†0.61–0.85
  Frequency of stool leakage500.55†0.09–0.88400.590.37–0.74
  Type of stool leakage570.73*0.42–0.75600.50*0.31–0.66
  Leakage of liquid stool710.74*0.47–0.74400.50*0.31–0.63
  Leakage of liquid stool related to laxatives380.75*0.34–0.77830.82*0.29–0.86
  Leakage of solid stool250.40*0.33–0.53490.57*0.38–0.67
  Stool leakage related to urgency630.54*0.25–0.68550.58*0.37–0.69
  Duration for which defecation can be delayed750.89†0.61–0.97490.630.44–0.77
  Rectal awareness before incontinent episode880.93*0.43–0.93400.38*0.21–0.56
  Rectal sensation during incontinent episode1001.0*0.45–1.0690.66*0.42–0.75
  Discrimination between gas and stool880.91*0.42–0.92600.59*0.37–0.71
  Leakage during waking or sleeping hours100  96  
  Medications to prevent leakage880.95*0.37–0.95710.80*0.54–0.85
  Use of pad for FI880.77*0.37–0.83720.61*0.67–0.74
  Number of pads used daily660.550.13–0.78630.420.15–0.65
  Use of pad – day/night1001.00.39–1.0710.420.13–0.66
Overall above set median (IQ range)87 (63–88)0.77 (0.73–0.93) 60 (49–71)0.59 (0.50–0.70) 
Urinary symptoms
 Urinary dribbling850.87*0.45–0.89650.66*0.45–0.74
 Frank urinary incontinence950.95*0.28–0.95750.48*0.32–0.63
 Use of pad for urinary incontinence850.94*0.50–0.95730.76*0.54–0.83
 Stress urinary incontinence750.84*0.46–0.86670.68*0.49–0.74
 Awareness of leakage600.39*0.14–0.65430.24*0.12–0.45
Overall above set median (IQ range)85 (75–85)0.87 (0.84–0.94) 67 (65–73)0.66 (0.48–0.68) 
Table 3.  Reproducibility and concurrent validity for risk factors, health care utilization, and impact of faecal incontinence
 ReproducibilityConcurrent validity
Observed agreement (%)κ or weighted κ* or ICC†95% CIObserved agreement (%)κ or weighted κ* or ICC†95% CI
  1. NA indicates items not assessed by doctor; ‡Women who completed the questionnaire twice did not have any Caesarean sections.

  2. The overall median (IQ range) for group of questions incorporated items for which validity and reproducibility were assessed by the κ statistic, but not those questions analysed by the ICC.

  3. IQR, interquartile range; ICC, interclass correlation coefficient; CI, confidence interval.

Anorectal disease, surgical history and other
 History of childbirth950.880.45–0.94990.950.70–0.98
 Number of children1001† 860.98†0.97–0.99
 Number of vaginal deliveries1001† 720.98†0.97–0.99
 Number of Caesarean sections  980.98†0.97–0.99
 Number of forceps deliveries500.790.47–0.93320.73†0.57–0.83
 History of episiotomy900.890.50–0.94840.940.76–0.97
 History of anal injury950.74*0.34–0.87740.30*0.19–0.55
 History of postpartum constipation84−0.08−0.08–0.51940.640.29–0.80
 History of rectocele950.860.41–0.93920.650.36–0.80
 Hysterectomy1001.00.56–1.0930.850.65–0.91
 History of anal abscess/infection850.480.12–0.74860.210.01–0.52
 History of anal fistula75−0.09−0.10–0.4290−0.05−0.05–.40
 History of anal fissure890.780.40–0.88840.430.18–0.64
 History of anorectal surgery750.490.17–0.72890.650.39–0.78
 History of anorectal trauma650.21−0.06–0.56870.13−0.01–0.47
 History of rectal prolapse1001.00.54–1.0890.530.24–0.72
 History of haemorrhoids1001.00.54–1.0780.490.27–0.66
 Reducible or irreducible haemorrhoids710.71*0.30–0.77820.67*0.27–0.82
 History of bowel surgery950.890.50–0.95800.550.13–0.78
 History of anorectal surgery950.900.51–0.95100  
 Confined to bed >1 month940.00.0–0.6088−0.06−0.06–0.42
 History of abdominal radiation1001.00.23–1.0970.650.20–0.83
 Family history of bowel problems850.680.32–0.83670.25−0.08–0.67
 Use of laxatives in childhood1001.00.54–1.0920.800.33–0.90
 History of early satiety850.94*0.51–0.94660.57*0.38–0.69
 Weight change within past year90†0.72†0.42–0.8855†0.75†0.64–0.83
Overall above set median (IQ range)95 (85–100)0.86 (0.49–0.94) 87 (79–91)0.57 (0.30–0.67) 
Health care utilization
 Any MD visit750.70†0.37–0.87NANANA
 Any MD visit for bowel problems700.83†0.62–0.93NANANA
 Any MD visit for FI900.66†0.33–0.85NANANA
 Any MD visit for bladder problems1000.88†0.73–0.95NANANA
 Hospitalized for colon blockage950.770.26–0.89NANANA
Overall above set median (IQ range)90 (75–95)  NANANA
Quality of lifeNANANA 
 Affects sports/recreation630.78*0.45–0.82NANANA
 Affects work around house630.78*0.43–0.82NANANA
 Affects social/entertainment740.83*0.49–0.84NANANA
 Affects family relationships630.76*0.43–0.81NANANA
 Affects travel740.69*0.43–0.75NANANA
 Affects sexual life740.78*0.43–0.86NANANA
 Affects work560.55*0.25–0.72NANANA
 Affects daily living790.83*0.47–0.84NANANA
Overall above set median (IQ range)69 (63–74)0.78 (0.74–0.79) NANANA NA
Current marital status950.72*0.44–0.80NANANA
Current employment status950.98*0.44–0.98NANANA
Educational training1001.0*0.51–1.0NANANA

Concurrent validity for individual questions

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

For the comparison of self-report responses against the criterion standard (doctor interview), the median (κ) value overall was 0.59 (IQR: 0.47–0.67). There were four of 98 questions for which 95% CI value included a value of zero, but overall agreement remained high. The branch question on FI was also valid (κ statistic = 0.95).

Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Responses to questions were combined in accordance with Rome II criteria to assess the κ statistic for reproducibility and concurrent validity for diagnosing IBS, functional constipation, functional diarrhoea (Table 4). Patient and doctor rating of FI severity, graded as mild, moderate or severe, agreed in 77% of patients; the weighted κ for rating severity as mild, moderate or severe was 0.45 (95% CI: 0.3–0.6), while the ICC for the actual score was 0.4 (95% CI: 0.1–0.6).

Table 4.  Agreement for diagnosis of functional gastrointestinal (GI) disorders
GI disordersReproducibilityConcurrent validity
Observed agreement (%)κ95% CIObserved agreement (%)κ95% CI
  1. CI, confidence interval; IBS, irritable bowel syndrome.

IBS850.690.34–0.84690.340.16–0.52
Constipation800.600.26–0.78730.440.25–0.61
Diarrhoea1001.00.23–1.096−0.02−0.02–0.52
Abdominal bloating950.00.0–0.60920.30.08–0.61

The impact of bowel symptoms on quality of life was summarized by adding scores for each of the eight subscales shown in Table 3. The quality of life was correlated with the actual severity score (rs = 0.33, P < 0.01) or severity rated as mild, moderate or severe (rs = 0.31, P < 0.02) by the FI Severity Scale. Thus, increasing severity of FI was associated with more severe impact on quality of life.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

Beginning with a previously validated instrument,9 we developed and validated a new, combined FICA. By framing questions to be consistent with Rome II criteria, assessing for IBS symptoms and obstetric history, and evaluating severity of incontinence and quality of life, the FICA had greater content validity compared with the previous FI questionnaire. For the FICA, the κ statistic for reproducibility and concurrent validity indicate this instrument is acceptable for use. For five of six questions with a κ statistic for reproducibility < 0.4, the observed agreement between the patient's initial and subsequent responses was ≥ 75%, suggesting that the low κ was attributable to a high prevalence of positive responses and not poor agreement.22 The agreement and κ statistic for concurrent validity was lower compared with reproducibility. Agreement between patient and doctor was lowest for questions pertaining to satisfaction after defecation and the efficacy of therapeutic measures for constipation; the doctor also overestimated the severity of abdominal pain perceived by patients.

Clinicians have recognized that incontinent patients have a varying degree of awareness of the desire to defecate before the episode of incontinence, ranging from no awareness to marked urgency before an episode of incontinence. The degree of awareness prior to FI may provide insights into the pathophysiology of FI. Thus, urge and passive incontinence are associated with more severe weakness of the external and internalanal sphincters respectively.27 Moreover, symptoms suggestive of exaggerated rectal sensation (e.g. a sense of incomplete defecation or urgency) were also risk factors for FI in the community.28 The FICA included questions assessing for exaggerated and reduced rectal sensation. We assessed for reduced rectal sensation by asking patients if they were aware of leakage either before or during the episode of incontinence, and also if they could discriminate between the sensation of gas and stool in the rectum. For questions pertaining to awareness before or during an incontinence episode, or the ability to discriminate between the sensation of gas from stool, the κ statistic for reproducibility was ≥0.9. Similarly, patient–doctor agreement for assessing awareness during an episode of incontinence and the ability to discriminate between the sensations of gas from stool was acceptable. However, for the question pertaining to awareness of stool leakage before the episode, agreement was relatively weak (i.e. κ statistic was 0.38), suggesting that the patients and the doctor interpreted this question differently. It is conceivable that the warning prior to an episode of incontinence is attributable to rectal distention by stool. Therefore, patients may have little or no warning when they are incontinent for stool, which has been rapidly propagated from the colon to the rectum by a high-amplitude propagated contraction (HAPC). Under these circumstances, the awareness ‘before’ the episode may in reality reflect perception of stool leakage by the anal mucosa. Therefore, questions on the ability to discriminate between gas and stool in the rectum, or awareness of leakage during the episode of incontinence may be more appropriate questions for assessing for diminished rectal sensation in FI.

We devised a scale for rating the severity of FI by incorporating four key elements (i.e. frequency of incontinence, type of incontinence, severity of urgency and pads used). Patient and doctor ratings of the severity of FI agreed in 77% subjects. Moreover, increasing severity of incontinence was significantly correlated with a greater impact on lifestyle. The κ statistic for the incontinence severity rating scale was 0.45, probably because the κ estimates are affected by the relative prevalence of moderate vs. severe incontinence, i.e. roughly twice as many patients had severe, compared with moderate FI. A perfect scale for rating FI severity has not been devised, as evidenced by the existence of multiple scales.10, 11, 17–19 These scales generally incorporate the frequency and type of leakage. The Rockwood Scale, which was published afterthis study commenced, grades the frequency foreach type (i.e. solid, liquid, mucus or gas) ofincontinence,10, 11, 17–19 but not the amount of stool leaked. Thus, incontinence severity graded by the Rockwood Scale would be identical for a subject who leaked a small amount of stool sufficient to stain underwear once a week, and another subject who was incontinent for a large liquid bowel movement once a week. The FICA Scale incorporated the average frequency of FI, and the most frequent type and frequency of FI. Our clinical experience suggests that patients find it difficult to characterize the frequency of incontinence, which is often variable and influenced by several factors, especially unpredictable alterations in stool consistency. In particular, patients are often reluctant or unable to precisely characterize, the frequency of incontinence for flatus. The FICA Scale also included faecal urgency and the number of perineal protective devices used daily. The St Mark's FI Severity Scale also incorporated urgency, assigning a score of 0 or 4 for patients who could or could not defer defecation for 15 min respectively. Our clinical experience suggested that this threshold for discriminating normal from excessive rectal urgency is relatively liberal, as a majority of incontinent patients were unable or reluctant to defer defecation for 15 min. Moreover, the time required to reach a toilet is influenced by other factors, including the distance from the toilet and a patient's physical status. Therefore, we graded urgency as normal or excessive using the response options to the question, ‘Do you leak stool because you have great urgency and cannot make it to the toilet in time to open your bowels'? The Severity Scale also examined the use of sanitary devices to protect against leakage. Since patients often replace the toilet paper lining the device rather than the device itself, the number of sanitary devices used by incontinent patients may underestimate the frequency/severity of leakage. Conversely, fastidious women may replace a sanitary device more frequently. Since all patients with incontinence in this cohort had moderate or severe FI, the scale needs to be validated in a general population, and compared with bowel diaries in men and women.

Previous studies have demonstrated that maternal recall of obstetric events correlated to a variable degree with the criterion standard, i.e. the medical record.29, 30 In this study, we elicited recall for selected obstetric events, including type of delivery (i.e. vaginal or Caesarean), use of forceps, episiotomy, and anal sphincter trauma during delivery. While these questions were reproducible and valid, these items, and other issues (e.g. birth weight, gestational age) need to be corroborated by the criterion standard, i.e. the medical record in future studies. To summarize, the FICA has excellent reproducibility and reasonable validity for assessing constipation and FI, the risk factors for FI, the severity of FI and its impact on lifestyle. We plan to validate the questionnaire in the general population and also validate questions pertaining to rectal sensitivity against objective assessments of anorectal sensorimotor functions. Additional studies are necessary to validate the FI Severity Scale against a criterion standard (e.g. bowel diaries), in subjects with a broader spectrum of symptom severity and assess its responsiveness to change during therapeutic interventions.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References

This work was supported in part by USPHS NIH grants RO1 HD 38666 and HD 41129. Authors wish to thank Dr J. H. Pemberton and Dr N. J. Talley for permission to use questions from a previous questionnaire and Cathy Schleck for assistance in data analysis and summarization.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Design of the questionnaire
  6. Testing of the questionnaire
  7. Statistical analysis
  8. Results
  9. Reproducibility for individual questions
  10. Concurrent validity for individual questions
  11. Reproducibility and concurrent validity for overall diagnostic criteria and Fecal Incontinence Severity Scale
  12. Discussion
  13. Acknowledgements
  14. References
  • 1
    Bharucha A. Fecal incontinence. Gastroenterology 2003; 124: 167285.
  • 2
    Drossman DA, Li Z, Andruzzi E, et al. US householder survey of functional gastrointestinal disorders. Prevalence, sociodemography, and health impact. Dig Dis Sci 1993; 38: 156980.
  • 3
    Nelson R, Norton N, Cautley E, Furner S. Community-based prevalence of anal incontinence. JAMA 1995; 274: 55961.
  • 4
    Leigh RJ, Turnberg LA. Faecal incontinence: the unvoiced symptom. Lancet 1982; 1: 134951.
  • 5
    Talley NJ, Phillips SF, Melton LJ, Wiltgen CM, Zinsmeister AR. A patient questionnaire to identify bowel disease. Am Intern Med 1989; 111: 6714.
  • 6
    O'Keefe EA, Talley NJ, Tangalos EG, Zinsmeister AR. A bowel symptom questionnaire for the elderly. J Gerontol 1992; 47: M11621.
  • 7
    Quan C, Talley NJ, Cross S, et al. Development and validation of the diabetes bowel symptom questionnaire. Aliment Pharmacol Ther 2003; 17: 117987.
  • 8
    Locke GR, Talley NJ, Weaver AL, Zinsmeister AR. A new questionnaire for gastroesophageal reflux disease. Mayo Clin Proc 1994; 69: 53947.
  • 9
    Reilly WT, Talley NJ, Pemberton JH, Zinsmeister AR. Validation of a questionnaire to assess fecal incontinence and associated risk factors: fecal incontinence questionnaire. Dis Colon Rectum 2000; 43: 14653; discussion 153–4.
  • 10
    Rockwood TH, Church JM, Fleshman JW, et al. Fecal incontinence quality of life scale: quality of life instrumentfor patients with fecal incontinence. Dis Colon Rectum 2000; 43: 916; discussion 16–7.
  • 11
    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: 152532.
  • 12
    Talley NJ, Phillips SF, Wiltgen CM, Zinsmeister AR, Melton LJ. Assessment of functional gastrointestinal disease: the bowel disease questionnaire. Mayo Clin Proc 1990; 65: 145679.
  • 13
    Talley NJ, Zinsmeister AR, Van Dyke C, Melton LJ. Epidemiology of colonic symptoms and the irritable bowel syndrome. Gastroenterology 1991; 101: 927934.
  • 14
    Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ. Functional constipation and outlet delay: a population-based study. Gastroenterology 1993; 105: 78190.
  • 15
    Grotz RL, Pemberton JH, Talley NJ, Rath DM, Zinsmeister AR. Discriminant value of psychological distress, symptom profiles, and segmental colonic dysfunction in outpatients with severe idiopathic constipation. Gut 1994; 35: 798802.
  • 16
    Attanasio V, Andrasik F, Blanchard EB, Arena JG. Psychometric properties of the SUNYA revision of the psychosomatic symptom checklist. J Behav Med 1984; 7: 24757.
  • 17
    Jorge JM, Wexner SD. Etiology and management of fecal incontinence. Dis Colon Rectum 1993; 36: 7797.
  • 18
    Pescatori M, Anastasio G, Bottini C, Mentasti A. New grading and scoring for anal incontinence. Evaluation of 335 patients. Dis Colon Rectum 1992; 35: 4827.
  • 19
    Vaizey CJ, Carapeti E, Cahill JA, Kamm MA. Prospective comparison of faecal incontinence grading systems. Gut 1999; 44: 7780.
  • 20
    Thompson WG, Longstreth GF, Drossman DA, Heaton KW, Irvine EJ, Muller-Lissner SA. Functional bowel disorders and functional abdominal pain. Gut 1999; 45(Suppl. II): 11437.
  • 21
    Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 15974.
  • 22
    Kraemer HC. Estimating false alarms and missed event from interobserver agreement: comment on Kaye. Psychol Bull 1982; 92: 74954.
  • 23
    Feinstein AR, Cicchetti DV. High agreement but low kappa: I. The problems of two paradoxes [comment]. J Clin Epidemiol 1990; 43: 5439.
  • 24
    Cichetti DV, Feinstein AR. High agreement but low kappa: II. Resolving the paradoxes. J Clin Epidemiol 1990; 43: 5518.
  • 25
    Fleiss JL. Statistical Methods for Rates and Proportions. New York, NY: Wiley, 1981.
  • 26
    Fleiss JL. Measuring agreement between two judges on the presence or absence of a trait. Biometrics 1975; 31: 6519.
  • 27
    Engel AF, Kamm MA, Bartram CI, Nicholls RJ. Relationship of symptoms in faecal incontinence to specific sphincter abnormalities. Int J Colorectal Dis 1995; 10: 1525.
  • 28
    Kalantar JS, Howell S, Talley NJ. Prevalence of faecal incontinence and associated risk factors; an underdiagnosed problem in the Australian community [comment]? Med J Aust 2002; 176: 547.
  • 29
    Elkadry E, Kenton K, White P, Creech S, Brubaker L. Do mothers remember key events during labor? Am J Obstet Gynecol 2003; 189: 195200.
  • 30
    Yawn BP, Suman VJ, Jacobsen SJ. Maternal recall of distant pregnancy events. J Clin Epidemiol 1998; 51: 399405.