SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Aliment Pharmacol Ther 2011; 33: 1245–1251

Summary

Background  Anorectal biofeedback therapy (BFT) is a safe and effective treatment in patients with constipation. Given the high prevalence of constipation and therefore high demand for BFT, there is a need to prioritise patients.

Aims  To explore clinical features and anorectal physiology which predict success or failure of BFT and to derive a statistical model which helps to predict the success of BFT.

Methods  A total of 102 patients with constipation referred for BFT were evaluated. All patients underwent comprehensive clinical and anorectal function assessment, including balloon expulsion testing. The BFT protocol consisted of a comprehensive 6-weekly visit programme comprising instruction on toilet behaviour and abdominal breathing, achieving adequate rectal pressure and anal relaxation, and balloon expulsion and rectal sensory retraining. Success of BFT was based on an improvement in global bowel satisfaction.

Results  Harder stool consistency (P = 0.009), greater willingness to participate (P < 0.001), higher resting anal sphincter pressure (P = 0.04) and prolonged balloon expulsion time (P = 0.02) correlated with an improvement in bowel satisfaction score. A longer duration of laxative use (P = 0.049) correlated with no improvement in bowel satisfaction score. Harder stools, shorter duration of laxative use, higher straining rectal pressure and prolonged balloon expulsion independently predicted successful BFT. A model (inline image, where β represents a regression coefficient, X is a given predictive variable and Si is the weighted index score for each individual) incorporating these four variables enabled prediction of successful BFT, with sensitivity and specificity of 0.79 and 0.81, respectively.

Conclusions  Important clinical and anorectal physiological features were found to be associated with outcome of anorectal biofeedback therapy in patients with constipation. This information and the predictive model will assist clinicians to prioritise patients for anorectal biofeedback therapy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Chronic constipation is a common condition with a prevalence of 2–27% in population studies in North America.1, 2 In tertiary care centres, dyssynergic defecation (DD) is seen in up to 50% of patients referred with chronic constipation.3 DD is significantly associated with a poor quality of life, with 74% and 69% of patients reporting an impact on their social life and work respectively.4 Medical management is often ineffective with 47% of patients dissatisfied with laxatives or fibre therapy.5

The efficacy of anorectal biofeedback therapy (BFT) in constipation was first reported in 1987.6 Since then, there has been a number of controlled studies examining the efficacy of BFT with mixed results.7 Recently, three randomised controlled trials have demonstrated a 70–81% success rate of BFT in patients with functional constipation and DD.8–10 In these studies,8–10 BFT was shown to be more effective than laxatives, muscle relaxants and placebo. Its benefit has also been shown to last for at least 12 months.9 Therefore, BFT is now considered one of the safest and effective therapeutic options for patients with chronic constipation and DD.

Given the high prevalence of constipation and DD,2, 3 there is potentially a large pool of patients who could be treated with BFT and there is a need to prioritise patients. A number of previous studies have examined factors, which predict success or failure of BFT in constipation.11–14 Results of these studies, however, have been inconsistent except for willingness to participate which has been shown to correlate with success of BFT.11 The inconsistencies in results may be due to small sample sizes in many of these studies and the differences in techniques used. These conflicting results make it difficult for clinicians to assess which group of patients is likely to have a successful outcome with BFT. Moreover, development of a model to predict success of BFT would be very useful clinically. This is analogous to the Model for End-Stage Liver Disease (MELD) score used to predict mortality risk as well as prioritising organ allocation, in patients with end-stage liver disease.15

The aims of this study were (i) to further explore clinical and physiological factors which predict success or failure of BFT in patients with chronic constipation, using a comprehensive anorectal manometry and biofeedback protocol16–18 and (ii) to derive a statistical model which helps to predict the success of BFT.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Patients

A total of 102 consecutive patients (mean age 48 ± 2 years, 88 females19) with chronic constipation referred for anorectal function testing to the Gastrointestinal Investigation Unit of Royal North Shore Hospital were included in the study. All patients had failed standard management of constipation including dietary fibre, fluid intake and laxative therapy where appropriate. In addition to the Rome Integrative Questionnaire,20 patients completed the validated Knowles Constipation Questionnaire,21 which evaluates constipation symptomatology with a maximum total constipation score of 25, and scores encompassing duration of constipation (score anchors: 0 = 0–18 months, 4 = >20 years or all life), laxative use (0 = none, 3 = long duration and ineffective), stool frequency (0 = 1–2 times/1–2 days, 3 = <once per fortnight) and stool consistency (0 = soft/loose/normal, 3 = always hard and pellet-like). They also completed 10 cm visual analogue scale (VAS) for (i) global bowel satisfaction (score anchors: 0 = completely unsatisfied; 10 = completely satisfied), (ii) impact of bowel dysfunction on quality of life (score anchors: 0 = no impact; 10 = most impact) and (iii) willingness to participate in the therapeutic course (score anchors: 0 = not willing; 10 = very willing) and the Hospital Anxiety and Depression (HAD) scale.22

For entry into the study, all patients were required to have two or more of the following symptoms of disordered defecation for at least 12 months: straining, incomplete evacuation, blockage and digitation (manual manoeuvres to facilitate evacuation), for at least 25% of the time. All patients also fulfilled symptom criteria for either nondiarrhoea irritable bowel syndrome or functional constipation.20 Some of the baseline assessments in the 88 females in this study have been published previously.19 All patients were required to exhibit two or more physiological criteria for functional defecation disorders23 assessed by anorectal manometry: (i) inadequate propulsive forces, (ii) paradoxical contraction of the anal sphincter or less than 20% relaxation of anal resting sphincter pressure on straining and (iii) evidence of impaired evacuation based on balloon expulsion testing.

Anorectal function testing

All patients underwent comprehensive anorectal function studies and this has been previously described in detail.24 Briefly, a seven-lumen water-perfused manometry catheter with 0.5 cm spaced sideholes and a compliant balloon attached to the end was used (Dentsleeve International, Ontario, Canada). The catheter was connected to calibrated pressure transducers and data from the pressure transducers were displayed in digital form on a computer using data conversion software (Neomedix, Sydney, NSW, Australia). Each individual study comprised assessments of the following parameters: (i) resting anal sphincter pressure, (ii) maximum anal sphincter squeeze pressure and duration of maximum anal squeeze pressure, (iii) straining rectal pressure and concomitant anal relaxation or paradoxical contraction, (iv) rectal sensitivity thresholds and (v) balloon expulsion test (time taken to expel a rectal balloon inflated with 50 mL of warm water whilst the patient was seated on a commode). A balloon expulsion time of >60 s25 is considered abnormal.

Biofeedback treatment

The biofeedback training consisted of a 30–60 min session, every week for 6 weeks, with a gastroenterologist and/or nurse specialist. The protocol comprises (i) education regarding the anatomy of normal defecation, (ii) advice on correct toilet positioning, (iii) diaphragmatic breathing, with manometric feedback, to achieve adequate rectal pressure, (iv) manometry-based biofeedback to allow anal relaxation to be synchronised with strain, (v) balloon expulsion retraining and (vi) rectal sensory retraining. Follow-up assessments were made at the end of treatment with anorectal function testing, Knowles Constipation Questionnaire,21 and VAS scores for global bowel satisfaction and impact on quality of life. The protocol was approved by the Human Research Ethics Committee of the Royal North Shore Hospital and all patients gave informed consent.

Statistical analysis

Success of BFT was measured both by degree of improvement and by forming an operational definition of substantial improvement. Degree of improvement was measured through the change in global bowel satisfaction score (assessed using 10 cm VAS with score anchors: 0 = completely unsatisfied and 10 = completely satisfied) from baseline to post-treatment. Clinical and physiological variables that may potentially predict the degree of improvement were correlated with change in global bowel satisfaction score (defined as the difference in scores between baseline and at the end of treatment), and Pearson correlation analysis was used to describe relationships between variables. Independent predictors of change in bowel satisfaction score were identified using a bootstrapped backward elimination procedure with change in bowel satisfaction score as the outcome. Only predictors selected in ≥50% of bootstrap samples were considered. The statistical independence of the predictors indentified in this way was verified.

A clinical predictive model of substantial improvement in bowel satisfaction scores after BFT (defined as ≥5 points increase in 10 cm VAS global bowel satisfaction score from baseline) was obtained via unconditional logistic regression. A stepwise selection process was used to identify variables that significantly predict substantial improvement. For each individual, an index score was calculated from the following model: inline image, where β represents a regression coefficient (log-odds ratio), X is a given predictive variable and Si is the weighted index score for each individual.26 The relationship between Si and the predicted probability of a substantial improvement in bowel satisfaction scores follows a logistic function such that the predicted probability of substantial improvement only reaches either zero or one asymptotically. The prognostic efficacy of the model was judged through examination of the receiver-operator characteristic (ROC) curve and the area under the ROC curve (AUC). An AUC of exactly 1.0 indicates perfect sensitivity and specificity, while an AUC of 0.5 indicates that the model performed no better than a ‘coin toss’. We considered an AUC of ≥0.8 to indicate useful model performance.

A potential problem with the interpretation of diagnostic models is a tendency to perform better in the sample on which they were developed than in subsequent samples.27 In the absence of an independent validation sample, which was not feasible in this case, some insight into stability of the model’s effect size estimates and performance (AUC) can be gained from calculation of shrinkage estimates.28 A shrinkage estimate pertains to a specific statistic such as a log-odds ratio or AUC. Shrinkage estimates are obtained from bootstrap resampling of the study data many times (1000 in this study) and fitting the model/calculating AUC in each bootstrap sample. The difference between the observed log-odds ratio/AUC and each bootstrap sample estimate is calculated and expressed as a percentage. The mean percentage deviation, which can be either positive or negative, is reported as the shrinkage estimate. Positive shrinkage values suggest that the sample values are over-estimates, while negative shrinkage values indicate potential under-estimates.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Of the 102 patients who entered into the study, four patients did not attend assessments at the end of treatment. Therefore, results of 98 patients (mean age 48 ± 2 years, 84 females) were available for analysis.

Baseline clinical features, symptoms and anorectal physiology

The mean scores (Knowles Constipation Questionnaire) for baseline clinical features and symptoms were as follows: duration of symptoms 2.4 ± 0.1; laxative use 1.4 ± 0.1; stool frequency 0.4 ± 0.1; stool consistency 1.5 ± 0.1; and total constipation score 18.1 ± 0.5. The anxiety and depression scores were 7.0 ± 0.5 and 5.5 ± 0.5, respectively. The mean 10 cm VAS scores for global bowel satisfaction, impact on quality of life, and willingness to participate were 2.4 ± 0.2, 7.1 ± 0.2 and 9.7 ± 0.1 respectively.

The mean values for anorectal physiology were as follows: resting anal sphincter pressure 72 ± 3 mmHg; straining rectal pressure 50 ± 3 mmHg; and the median value (interquartile range) for balloon expulsion time was 40 (12, 94) s. Sixty per cent of patients exhibited paradoxical contraction of the anal sphincter on straining, whereas 33% had absent relaxation on straining.

Correlations between clinical features, anorectal physiology and bowel satisfaction scores

The correlations between clinical and anorectal physiological variables and change in global bowel satisfaction score are given in Table 1. Harder stools, greater willingness to participate, higher resting anal sphincter pressure and prolonged balloon expulsion time correlated with an improvement in bowel satisfaction score. On the other hand, long duration of laxative use correlated with no improvement in bowel satisfaction score. The outcome of BFT was not influenced by age, duration of symptoms, stool frequency, HAD scores, compliance with therapy, straining rectal pressure or relaxation of anal sphincter on straining.

Table 1.   Correlations between clinical and anorectal physiological variables and change in bowel satisfaction scores after biofeedback therapy
 Change in global bowel satisfaction scoresP value
  1. N.S., not significant.

  2. Values are Pearson r correlations.

Age−0.02N.S.
Duration of symptoms−0.06N.S.
Compliance with treatment0.06N.S.
Stool frequency0.09N.S.
Stool consistency0.270.009
Laxative use−0.210.049
Willingness to participate0.390.0001
Resting anal sphincter pressure0.210.04
Straining rectal pressure0.08N.S.
Paradoxical contraction with strain−0.09N.S.
Balloon expulsion time0.230.02

Predictors of a change in bowel satisfaction scores

Multiple linear regression analysis for predictors of bowel satisfaction after BFT is shown in Table 2. Harder stool consistency and greater willingness to participate independently predicted an improvement in global bowel satisfaction score. The model including these two variables explained 21% of the variance in treatment outcome.

Table 2.   Multiple linear regression analysis for predictors of bowel satisfaction after biofeedback therapy
PredictorCoefficient95% CIP value
Stool consistency0.640.11–1.170.018
Willingness to participate1.570.79–2.36<0.001

Predictive model of BFT

When a substantial improvement in global bowel satisfaction (defined as ≥5 points increase in 10 cm VAS global bowel satisfaction score from baseline) was used as a measure of successful BFT, four independent predictors were identified (Table 3). Harder stools, shorter duration of laxative use, higher straining rectal pressure and prolonged balloon expulsion time at baseline predicted a substantial improvement in bowel satisfaction. Shrinkage estimators of the effect size (log-odds ratio) of each independent variable suggests that the estimated effect sizes are not over-estimates with negative values for all four independent variables (range −8% to −11%). An index score can be calculated from the following model: inline image, where β represents a regression coefficient (log-odds ratio), X is a given predictive variable and Si is the weighted index score for each individual.

Table 3.   Logistic regression analysis for predictors of substantial improvement in global bowel satisfaction after biofeedback therapy
PredictorLog-odds ratio (β)S.E.P value
  1. Pseudo R2 = 0.21.

  2. See text for explanation of model.

Stool consistency0.890.280.001
Laxative use−0.720.290.01
Straining rectal pressure0.030.010.009
Balloon expulsion time0.01<0.010.004

The relationship between Si and predicted probability of substantial improvement is shown in Figure 1. From the graph, the index score that corresponds to a predicted probability of substantial improvement can be determined. For example, an index score of zero corresponds to a predicted probability of 0.5 (50%), while a score of approximately 1.4 is needed for a predicted probability of 0.8 (80%). The ROC curve for this model is shown in Figure 2. With an area under the ROC curve of approximately 0.8, our model provides a prediction that would generally be considered useful. Bootstrap simulation of the AUC suggests that this value is likely to be reproducible in future samples given a slightly negative shrinkage estimate at −0.26%, which we interpret as being effectively zero. Inspection of the ROC expressed as sensitivity and specificity suggests a threshold probability of approximately 0.4 (Figure 3). At this threshold, the sensitivity is 0.79 and specificity is 0.81.

image

Figure 1.  The relationship between individual’s weighted index score Si and predicted probability of substantial improvement after biofeedback therapy.

Download figure to PowerPoint

image

Figure 2.  Receiver-operator characteristic (ROC) curve reported in Table 3.

Download figure to PowerPoint

image

Figure 3.  Receiver-operator characteristic (ROC) curve for the predictive model reported in Table 3 expressed as sensitivity and specificity with a threshold probability of 0.40.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Using a comprehensive manometric-based anorectal biofeedback protocol, we have determined important factors, which predict outcome of BFT in a large group of patients with chronic constipation and disordered defecation. We have further established a clinically useful model, which helps to predict an individual’s likelihood of success with BFT. This is important as constipation is common, and BFT is a safe and effective, yet labour-intensive treatment for constipation. We believe, this predictive model can assist clinicians in prioritising patients who should undergo BFT.

In this study, a harder stool was predictive of a substantial improvement in bowel satisfaction after BFT. This is the first time stool consistency has been shown to be associated with success of BFT. This finding is not unexpected as hard stool is a common feature of DD4 and BFT improves dyssynergia, which potentially leads to better efficiency in stool evacuation. Our finding of shorter duration of laxative use as a predictor of successful BFT is consistent with Chiarioni et al.12 in their study of 52 patients with chronic constipation. The willingness to participate was associated with improvement in bowel satisfaction, but was not a useful variable for our predictive model, as the data were skewed with most patients having an extremely high score.

The two physiological parameters predictive of a substantial improvement in bowel satisfaction after BFT in our study were high straining rectal pressure and prolonged balloon expulsion time. Again, our findings demonstrate for the first time that high straining rectal pressure is associated with success of BFT. During attempted defecation, contraction of the anterior abdominal wall muscle and diaphragm leads to an increase in intra-abdominal pressure during straining.29 However, an inappropriately high straining rectal pressure may in fact represent a feature of DD, as with the rise in rectal pressure during defecation present in Type I dyssynergia described by Rao et al.30 Prolonged balloon expulsion time was an independent predictor of a successful BFT in this study, again consistent with Chiarioni et al.12 Our results therefore suggest that patients with physiological features of dyssynergia are more likely to respond to BFT. Interestingly, however, paradoxical contraction of the anal sphincter on straining was not predictive of successful BFT in our study.

Our predictive model is a clinically relevant tool, which helps in selecting patients for BFT. This model predicts a ≥5 point increase in 10 cm VAS global bowel satisfaction score after BFT, which equates to a 50% improvement from baseline. Although we chose a 5-point increase in bowel satisfaction, we did not feel that it limits the outcome as in order to determine factors predicting success, having approximately equal numbers across categories maximises statistical power. Furthermore, this is quite a conservative approach in order to not over interpret data. Hence, by simply applying the data regarding laxative use, stool consistency, straining rectal pressure and balloon expulsion time, one can readily calculate an individual patient’s likelihood of a substantial improvement with BFT from the model and graph. For example, if a patient reports hard stool and short duration of laxative use (which equates to a value of 2 and 1, respectively, on the Knowles Constipation Questionnaire21) and has a straining rectal pressure of 60 mmHg and a balloon expulsion time of 70 s at baseline, the predictive score using the model inline image and values of β from Table 3, can be calculated as: Si = [(0.89 × 2) + (−0.72 × 1) + (0.03 × 60) + (0.01 × 70)] which results in Si = 3.56. Using Figure 1, one can determine from the graph the predicted probability of success with BFT. In our hypothetical case, a value of 3.56 corresponds to a predicted probability of success of approximately 0.98 (98%). On the other hand, a patient who reports normal stool consistency, a long duration of laxative use, and has a straining rectal pressure of 30 mmHg, and a balloon expulsion time of 20 s, will have a predictive score of −0.34, which corresponds to a predicted probability of success of only 15%. This model is particularly relevant in a busy tertiary referral centre, which potentially has a large number of patients with DD and constipation that requires treatment with BFT. We propose that those with a higher predicted probability of success with BFT be given priority over those with a lower predicted probability of success.

In conclusion, we have found that important clinical and physiological features are associated with the outcome of BFT in patients with constipation, including stool consistency, laxative use, straining rectal pressure and balloon expulsion time. In addition, we have developed a predictive model to help clinicians prioritise the large number of potential patients referred for BFT, which can be a labour intensive and therefore costly treatment.

Acknowledgement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Declaration of personal interests: Dr Kellow has served as a speaker and an advisory board member for Janssen-Cilag Australia and Mundipharma Australia. Declaration of funding interests: None.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  • 1
    Pare P, Ferrazzi S, Thompson WG, Irvine EJ, Rance L. An epidemiological survey of constipation in Canada: definitions, rates, demographics, and predictors of health care seeking. Am J Gastroenterol 2001; 96: 31307.
    Direct Link:
  • 2
    McCrea GL, Miaskowski C, Stotts NA, Macera L, Varma MG. A review of the literature on gender and age differences in the prevalence and characteristics of constipation in North America. J Pain Symptom Manage 2009; 37: 73745.
  • 3
    Surrenti E, Rath DM, Pemberton JH, et al. Audit of constipation in a tertiary referral gastroenterology practice. Am J Gastroenterol 1995; 90: 14715.
  • 4
    Rao SSC, Tuteja AK, Vellema T, Kempf J, Stessman M. Dyssynergic defecation: demographics, symptoms, stool patterns, and quality of life. J Clin Gastroenterol 2004; 38: 6805.
  • 5
    Johanson JF, Kralstein J. Chronic constipation: a survey of the patient perspective. Aliment Pharmacol Ther 2007; 25: 599608.
  • 6
    Bleijenberg G, Kuijpers HC. Treatment of the spastic pelvic floor syndrome with biofeedback. Dis Colon Rectum 1987; 30: 10811.
  • 7
    Chiarioni G, Heymen S, Whitehead WE. Biofeedback therapy for dyssynergic defecation. World J Gastroenterol 2006; 12: 706974.
  • 8
    Heymen S, Scarlett Y, Jones K, Ringel Y, Drossman D, Whitehead WE. Randomized, controlled trial shows biofeedback to be superior to alternative treatments for patients with pelvic floor dyssynergia-type constipation. Dis Colon Rectum 2007; 50: 42841.
  • 9
    Chiarioni G, Whitehead WE, Pezza V, Morelli A, Basotti G. Biofeedback is superior to laxatives for normal transit constipation due to pelvic floor dyssynergia. Gastroenterology 2006; 130: 65764.
  • 10
    Rao SSC, Seaton K, Miller M, et al. Randomized controlled trial of biofeedback, sham feedback, and standard therapy for dyssynergic defecation. Clin Gastroenterol and Hepatol 2007; 5: 3318.
  • 11
    Gilliland R, Heymen S, Altomare F, Park UC, Vickers D, Wexner SD. Outcome and predictors of success of biofeedback for constipation. Br J Surg 1997; 84: 11236.
  • 12
    Chiarioni G, Salandini L, Whitehead WE. Biofeedback benefits only patients with outlet dysfunction, not patients with isolated slow transit constipation. Gastroenterology 2005; 129: 8697.
  • 13
    Song HJ, Jung SA, Noh YW, et al. Predictable factors of biofeedback therapy compliance in patients with chronic constipation due to pelvic floor dyssynergia. Am J Gastroenterol 2009; 104: A1319.
  • 14
    Fernandez-Fraga X, Azpiroz F, Casaus M, Aparici A, Malagelada J-R. Responses of anal constipation to biofeedback treatment. Scan J Gastroenterol 2005; 40: 207.
  • 15
    Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33: 46470.
  • 16
    Bassotti G, Chistolini F, Sietchiping-Nzepa F, de Roberto G, Morelli A, Chiarioni G. Biofeedback for pelvic floor dysfunction in constipation. BMJ 2004; 328: 3936.
  • 17
    Heymen S, Jones KR, Scarlett Y, Whitehead WE. Biofeedback treatment of constipation: a critical review. Dis Colon Rectum 2003; 46: 120817.
  • 18
    Rao SS, Welcher KD, Pelsang RE. Effects of biofeedback therapy on anorectal function in obstructive defecation. Dig Dis Sci 1997; 42: 2197205.
  • 19
    Shim L, Prott G, Hansen RD, Simmons LE, Kellow JE, Malcolm A. Prolonged balloon expulsion is predictive of abdominal distension in bloating. Am J Gastroenterol 2010; 105: 8837.
  • 20
    Thompson WG. Appendix B: research diagnostic questions for functional gastrointestinal disorders: Rome II Integrative Questionnaire. In: DrossmanDA, ed. The Functional Gastrointestinal Disorders: Diagnosis, Pathophysiology, and Treatment: A Multinational Consensus, 2nd ed. McLean, VA: Degnon Associates, 2000; 35191.
  • 21
    Knowles CH, Eccersley AJ, Scott SM, et al. Linear discriminant analysis of the symptoms in patients with chronic constipation: validation of new scoring system (KESS). Dis Colon Rectum 2000; 43: 141926.
  • 22
    Zigmond A, Snaith R. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67: 36170.
  • 23
    Wald A, Bharucha A, Enck P, et al. Functional anorectal disorders. In: DrossmanDA, ed. Rome III: The Functional Gastrointestinal Disorders, 3rd ed. Virginia, USA: McLean, Degnon Associates, 2006; 66375.
  • 24
    Suttor VP, Prott GM, Hansen RD, Kellow JE, Malcolm A. Evidence for pelvic floor dyssynergia in patients with irritable bowel syndrome. Dis Colon Rectum 2010; 53: 15660.
  • 25
    Sun WM, Read NW, Prior A, et al. Sensory and motor responses to rectal distension vary according to rate and pattern of balloon inflation. Gastroenterology 1990; 99: 100815.
  • 26
    Dobson A, Barnett A. An Introduction to Generalized Linear Models, 3rd ed. Boca Raton: CRC Press, 2008; 126.
  • 27
    Harrell FE Jr, Lee KL, Mark DB. Multivariate prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996; 15: 36187.
  • 28
    Steyerberg ES, Eijkemans MJC, Harrell F, Habbema JDF. Prognostic modeling with logistic regression analysis: a comparison of selection and estimation methods in small data sets. Stat Med 2000; 19: 105979.
  • 29
    Bajwa A, Emmanuel A. The physiology of continence and evacuation. Best Pract Res Clin Gastroenterol 2009; 23: 47785.
  • 30
    Rao SSC, Mudipalli RS, Stessman M, et al. Investigation of the utility of colorectal function tests and Rome II criteria in dyssynergic defecation (Anismus). Neurogastroenterol Motil 2004; 16: 58996.