Exposure based cognitive behavioral group therapy for IBS at a gastroenterological clinic – a clinical effectiveness study

Abstract

Background

Patients with irritable bowel syndrome (IBS) may benefit from psychological treatment when diet changes and medications do not sufficiently reduce symptoms. Our research team has developed an exposure based cognitive behavioral therapy protocol (ECBT), which has been shown to be effective in several randomized controlled trials.

Aim

To investigate the effectiveness of ECBT in clinical routine care at a gastroenterological clinic in Stockholm and to find predictors for treatment outcome.

Method

A ten session ECBT based on our protocol was given face to face by licensed psychologists in groups of 4–6 patients. A total of 129 patients provided information regarding IBS symptoms, quality of life, gastrointestinal symptom-specific anxiety (GSA), and depression pre and post-treatment. We used linear regression analyses to identify patient characteristics that predicted treatment outcome.

Results

The primary outcome was symptom severity measured with The Gastrointestinal Symptom Rating Scale for IBS (GSRS-IBS). Average pre-and post-treatment GSRS-IBS scores were 49.24 (SD = 11.54) and 37.03 (SD = 10.03), corresponding to a 34.0% reduction in symptom severity (p < .001). Reductions were also found in GSA, 43.9% (p < .001) and depression, 38.6% (p < .001). IBS-related quality of life was on average increased by 68.2% (p < .001). The effect sizes were large and varied between (Cohen’s d) 0.95 and 1.84. None of the patients’ pre-treatment characteristics predicted outcome.

Conclusion

We conclude that ECBT for IBS delivered face-to-face in a group-format is very effective, also in a routine care setting. We did not find any reliable predictors for treatment outcome. The trial was registered at Clinicaltrials.gov with ID: NCT04756414.

Keywords:

• Irritable bowel syndrome
• cognitive behavior therapy group treatment
• effectiveness
• exposure
• prediction analysis

Introduction

Irritable bowel syndrome (IBS) is associated with significant suffering and disability [1]. The global prevalence of IBS has been estimated to 9% [2], and the prevalence is higher in women than in men [3]. Gastroenterologists report that IBS is the primary reason for 20%–30% of first time consultancies [4]. Compared with healthy individuals, IBS patients report significantly lower quality of life [5,6]. The societal costs associated with IBS are extensive and have in the USA been found to equal those of the flu [7].

Psychological treatments for IBS such as cognitive behavioral therapy (CBT) and gut-directed hypnotherapy have shown good effects immediately after treatment as well as at short and long term follow up [8]. Our research group has developed an exposure-based CBT (ECBT) protocol for IBS. In ECBT, exposure exercises are used to reduce fear of and hypervigilance towards IBS symptoms and associated avoidance and control behavior. Patients who experience pain and bodily discomfort often develop hypersensitivity towards bodily sensations and a pattern of fear-avoidance that increase the subjective perception of symptoms [9]. ECBT can be delivered in group [10] or over the Internet [11–14] and has been found to reduce IBS symptoms and improve quality of life [13–15].

A treatment that is effective under rigorously controlled trial conditions may be less effective in a clinically naturalistic environment [16]. To aid the clinician it is therefore important to study how a program works in ordinary practice. This line of research has been labeled pragmatic or clinical effectiveness trials [17]. It is a method of testing whether a treatment yields similar effects in routine care as it does in an efficacy trial. Effectiveness studies are usually characterized by a systematic naturalistic design, few or no exclusion criteria and in the use of regular staff, without extra training or supervision.

Another area of clinical importance is to find predictors for positive response based on individual characteristics. A predictor is a pre-treatment variable that affects treatment outcome [18]. Predictors of CBT concern patient factors as gender, age, duration of illness, comorbid diseases, symptom burden and suitability of the program [19]. Specifically, research has indicated predictors for improvement of IBS in CBT to be degree of gastrointestinal anxiety, alexithymia, trait anxiety, anxiety sensitivity and coping styles [20–22]. The patient characteristics with available data pre-treatment in our study have been used for prediction analyses in earlier studies and in some of these found to be significantly associated with outcome. Female gender predicted poorer outcome in one CBT-study [23] and better outcome in one hypnotherapy-study [24], longer duration of symptoms predicted poorer outcome in a study on brief psychodynamic therapy [25] and age over 50 years predicted poorer outcome in a hypnotherapy study [26].

The present study is a case series, based on data collected from routine care at the Gastroenterology section at Sabbatsbergs Hospital in Stockholm, Sweden. The aim of our study was to investigate if the ECBT was effective in clinical routine care when administered in groups of 4–6 patients. Furthermore, we wanted to investigate if patient characteristics were associated with treatment outcome, in order to aid clinicians’ decision making regarding who should be offered the treatment.

Method

Design

This is a pragmatic effectiveness study in the form of a case series evaluation that was performed within the setting of a gastroenterological clinic. Ethical permit was granted for this study by The Regional Ethical Review Committee in Stockholm, DNR 2016/697-31/1 and the study is reported in accordance with the CONSORT statement for pragmatic trials [17]. Clinicaltrials.gov registration ID: NCT04756414. The ethical permit for the study stated that patient consent was not necessary for inclusion in the study because clinical and assessment data were collected from patients’ medical records by staff and not directly from patients. Researchers who analyzed the data (HW, BL) did not have access to identifying information about patients.

Participants

Participants in this study were IBS-patients undergoing treatment at the Gastroenterology unit, Sabbatsbergs hospital in Stockholm, between 2009 and 2014. The context of this clinic was an outpatient, gastroenterology unit within the tax-financed health-care system in Sweden, mainly serving the primary care system with gastroenterological consultations. The staff consisted of gastroenterologists, nurses, dieticians, and psychologists. The psychologists all had formal training and experience in treating IBS with ECBT. One of the main focuses of the clinic was managing patients with functional GI disorders such as IBS.

At the gastroenterological clinic the ROME III-criteria were [1] used to diagnose IBS. All patients were investigated with a full blood-count, C-reactive protein, and transglutaminase antibodies prior to diagnosis. Additional investigation to rule out other causes to the symptoms was based on individual symptoms and medical history. For example, all patients with diarrhea, of age >50 years or a history of rectal bleeding underwent a coloscopy before final IBS-diagnosis. Patients with refractory IBS (previous inadequate response to dietary changes and/or pharmacological treatment) of any subtype were offered ECBT in group after an initial psychological assessment supported that this intervention was suitable for the patient. During the data collection period, 171 patients initiated the treatment, and of those, 129 provided data on the primary outcome measure The Gastrointestinal Symptom Rating Scale for IBS (GSRS-IBS) at pre- and post-treatment and were therefore included in the present study. The mean age was 35 (17–69) years and 80% (107) were female. Approximately 76% of the patients completed the full treatment. See Table 1 for description of the patients.

Table 1. Description of the participants (N = 129).

	N	%
ICD-10 Diagnosis
K589 Irritable bowel without diarrhea	51	39.6
K580 Irritable bowel with diarrhea	78	60.4
Occupation
Part time work	14	10.9
Student	22	17.1
Full time work	76	58.9
Sick leave	6	4.7
Retired	6	4.7
Unemployed	3	2.3
Parental leave	2	1.6
Years with gastrointestinal symptoms at treatment start
1–2 yrs	12	9.3
>2–4 yrs	17	13.2
>4–10 yrs	46	35.7
>10 yrs	54	41.9
Gender
Female	107	82.9
Male	22	17.1
Mean age (SD, min, max)	35.57 (12.9, 17, 69)
Completed treatment sessions
5–7 sessions	11	8.6
≥8 sessions	98	75.9
Missing session amount data	20	15.5
Psychiatric comorbidity
Yes/no/missing data	51/75/3	39.5/58.1/2.3

Forty-two patients did not provide outcome data (GSRS-IBS) pre-treatment start or post-treatment. Most of these patients chose to terminate the treatment before or shortly after start for different personal reasons, unrelated to the treatment per se, for example a lack of time. We used independent samples t-tests to compare the mean scores on pre-treatment characteristics. There were no significant differences between the group that had missing data and the other participants in the variables: age at treatment start, type of IBS or duration of symptoms. In the group with missing data, there were 67% female and in the group with no missing data, 83% were female. The difference was statistically significant (p=.024).

The following inclusion criteria were used: at least 16 years of age, diagnosis of IBS, poor response to dietary change and medication, avoidance behavior/hypervigilance towards bodily signals, interested in taking part in psychological treatment. Patients were excluded if they had a primary and/or severe psychiatric disorder, such as schizophrenia, eating disorder, drug addiction, autism, severe depression, suicidal ideation or OCD. They could also be excluded if they presented with other problems than IBS that made an IBS-focused group treatment unsuitable. Finally, patients were excluded if they expressed a preference for individual therapy over group therapy (in which case they received individual therapy). The exclusion criteria were assessed by psychologists, but the patients did not routinely go through any standardized diagnostic process (i.e., assessment with specific diagnostic tools given to every patient). Pharmacological treatment was allowed during treatment, but the patients had in these cases a stable dose and no dose-corrections were made during the period of ECBT.

Procedure

The physicians at the clinic referred patients to psychological assessment if inclusion criteria were met. The psychological assessment was made during one or two sessions by one of five psychologists working at the clinic. Treatment started continuously as soon as there were enough patients to fill a group. The treatment was delivered in groups of 4–6 patients and followed the ECBT protocol developed by our research group at Karolinska Institutet [10]. All patients who entered treatment were encouraged to fill out self-report questionnaires (with pen and paper) as part of the clinic’s regular quality follow-up. The questionnaires were answered on the first and last day of treatment. Demographical and clinical data and results from questionnaires, completed before and after treatment were collected from the electronic medical record (EMR) at the clinic. The study included all patients that were diagnosed with IBS, initiated group CBT at the clinic, and provided answers to GSRS-IBS at both pre- and post-treatment.

Intervention

The ECBT protocol (Table 2) comprised six different steps administered in ten sessions, once a week, during ten weeks. The modules were (Session 1) information regarding the treatment, IBS and mindfulness, (Session 2) Hypervigilance and anxiety related to IBS, (Session 3) Thoughts and behaviors related to IBS, (Session 4) Behavior change and exposure focusing on toilet habits, (Session 5–9) Exposure with response prevention for feared stimuli such as food, bodily sensations and potentially embarrassing situations, (Session 10) Summary and relapse prevention. The sessions lasted 90–120 min. The treatment was provided by experienced, licensed psychologists, employed at the gastroenterology unit. No specific supervision was provided and no assessment of adherence to protocol was done.

Table 2. Description of the treatment content.

Step	Session	Content
1	1	Introduction of group leaders (two psychologists with experience of CBT for IBS) and group participants. Signing of agreement of secrecy between participants. Presentation of CBT as an active treatment with homework. Presentation of the treatment and its goals, as well as the content of each of the ten sessions. Presentation of IBS causes and different treatments. Introduction of a psychological model of IBS. Two mindfulness exercises: 1. A 10 min exercise during which the patient observes and labels inner experiences of both breathing and gastrointestinal sensations (practiced once daily). 2. A 20 s exercise that brings the patient into immediate awareness of current thoughts, GI symptoms, feelings, and behavioral impulses (practiced several times daily).
2	2	Explanation of how the brain’s monitoring of IBS symptoms increases because of negative emotions, thoughts, and behaviors. The learning of symptom-related fear through negative experiences of symptoms. The effect of anxiety on gastrointestinal functioning and how it increases awareness of threatening stimuli, specifically IBS-related stimuli.
3	3	Explanation of how symptom-related behaviors may maintain focus on and vigilance towards symptoms. The role of negative thoughts in exacerbating IBS-related fear. A mindful and accepting stance towards negative thoughts and feelings is proposed as an alternative to attempts to control these experiences. Explanation of how IBS-related avoidance and control behaviors maintain the awareness and fear of IBS symptoms and decrease quality of life. Participants begin by recording how these behaviors prevent them from living a full life. They also record how they use the toilet to decrease symptoms.
4	4	Presentation of how the digestive system functions, what is normal. Introduction to behavioral change with focus on toilet visits. 1) For constipation-predominant participants: scheduling of toilet visits. 2) For diarrhea-predominant participants: gradual increase in time between onset of urgency and toilet visits. 3) Patients with mixed forms of IBS use both exercises. 4) All patients decrease the time spent on every toilet visit to 2–5 min. Participants are aware that the exercises may increase symptoms and anxiety short term.
5	5	Presentation of exposure; purpose, goals and how to plan the exercises. The exposure exercises include: 1) Exposure to symptoms by engaging in activities that provoke symptoms, such as eating certain foods, physical activity, and stressful situations. 2) Reduction of safety and control behaviors, such as distraction, eating certain foods, resting, and taking unprescribed medications. 3) Exposure to situations where symptoms are unwanted, such as attending a meeting when experiencing abdominal pain or riding the bus with fear of losing control of the bowels. Instructions on how to use mindfulness as an aid to choose more functional ways of behaving in the presence of aversive experiences. By observing and labeling their environment during exposure, i.e. aversive, neutral, and positive internal and external stimuli, patients will counter distraction from and suppression of thoughts and emotions. By attending to any impulses to flee the situation or decrease the intensity of symptoms they will also be less inclined to act on these impulses. Comprehensive examples are given about how IBS patients with different types of IBS symptoms and disability have experienced symptom relief and increased quality of life by changing their focus from avoidance of symptoms to living a valued life. Patients continue behavioral change related to toilet visits.
5	6–9	Continued work with exposure and behavioral change related to toilet visits. Therapists and patients work together to follow up and create new exercises.
6	10	Planning of how to handle relapses into avoidance behaviors and how to maintain a widened behavioral repertoire. The participants also summarize the main lessons learned, what has been gained through the treatment and their future plans.

Outcome assessments

The primary outcome measure was the amount of IBS symptoms measured with GSRS-IBS [27]. The GSRS-IBS consists of 13 items covering severity of gastrointestinal symptoms. The items are scored between 1 (no discomfort at all) and 7 (very severe discomfort), rendering a total score between 13 and 91. The GSRS-IBS has demonstrated good psychometric properties for the different symptoms that are assessed, with an internal consistency (Cronbach’s a) of 0.88 [28] and test-retest reliability between 0.55 and 0.70 [27].

Secondary outcome measures were: (1) The Quality of Life in persons with Irritable Bowel Syndrome (QOL-IBS) [29]. The scale consists of 34 items scored between 1 and 6 and the total score is transformed to a 0–100 scale, where 0 represents minimum quality of life and 100 represents maximum quality of life. The QOL-IBS shows high internal consistency (Cronbach’s a = .95) and test-retest reliability. r = .86 [29]. (2) The Visceral Sensitivity Index (VSI) [30] is a measure of gastrointestinal-specific anxiety (GSA; i.e. distress, avoidance, and safety behaviors in response to IBS-related stimuli). The VSI contains 15 items, which are scored between 0 and 5, rendering a total score between 0 (minimum GSA) and 75 (maximum GSA). The VSI has demonstrated high internal consistency (Cronbach’s a = 0.90–0.92) and has been shown to be associated with symptom severity [31] and diagnostic status of IBS [32] and also to be a mediator of the effect of exposure-based CBT on IBS symptoms [33,34]. (3) The Montgomery-Åsberg Depression Rating Scale (MADRS-S) consists of 9 items that can be rated between 0 and 6, yielding a result between 0 and 54, where a higher number indicates more severe symptoms of depression. The scale was developed in 1993 and has been widely used. It has demonstrated good psychometric properties with an internal consistency between Cronbach’s a = 0.84–0.91[35,36].

Statistical analysis

Statistical analyses were made with SPSS v28 and R. We used paired samples t-test to compare the pre-treatment and post-treatment scores on the assessments. Effect size was calculated as Cohens d, by dividing the pre- to post-treatment average difference with the baseline standard deviation. Confidence intervals (95%) for effect sizes were based on 5,000 bootstrap resamples of data, using the bias-corrected and accelerated method. We also calculated the proportion of responders to treatment. The definition of a responder was a symptom reduction of 30%, measured with GSRS-IBS. This definition has been used in other studies [37,38] and is recommended by the U.S. Food and Drug Administration (FDA) when evaluating pharmacological treatment effectiveness in IBS [39]. The calculation was based on the formula 1- (GSRS-IBS-post − 13)/(GSRS-IBS-pre − 13) followed by a calculation of percentages with the number of patients with a score of ≥0.3 divided by the total number of patients.

If a patient characteristic affects the outcome it can be used to predict the effectiveness of a treatment (on a group level) if it is known before treatment [18]. There are generally three categories of possible predictors for CBT: patient demographics, clinical characteristics (e.g., symptom burden), and treatment-related variables [40]. In this trial we had access to at least one variable within each domain. To test whether our data could be used to predict treatment outcome, we used the GSRS-IBS score post-treatment as the dependent variable and each of our potential predictor variables as independent variables. Normal distribution of the dependent variable was assessed with Shapiro-Wilks test and all variables were checked for multicollinearity and heteroskedasticity. No variables were found to violate assumptions. We then used a multiple linear regression analysis in SPSS controlling for correlation with GSRS-IBS pre-treatment values and tested each variable with the significance level set to p≤.10. The intention was to select the variables that were significantly predicting the dependent variable and in the next step build a model using multiple linear regression with the significance level set to 0.05.

Missing data

The study sample included all patients with complete GSRS-IBS data at both pre- and post-treatment. Additionally, 16 patients provided either pre- or post-treatment and post-treatment data on the GSRS-IBS. We ran a sensitivity analysis including these 16 patients (total N = 145) using multiple imputation to account to for missing data. Results from these analyses produced similar effect sizes compared to analyzing only complete cases (largest deviation 5.7%, data not reported but available upon request).

In two patients’ medical records, we identified incorrect or unreasonable scores on VSI (one case) and IBS-QOL (both cases) and these were not included in the analysis. Scores on all other questionnaires for these patients were however included in the analysis.

Results

The investigation of the effectiveness of the exposure-focused manual for IBS in this case series study showed that of 171 patients who initiated the group-based ECBT at a gastroenterological clinic, 129 provided answers to the GSRS-IBS questionnaire at pre- and post-treatment. There were significant improvements from pre-treatment to post-treatment on all outcome measurements (p < .01) and the effect sizes were large (Cohens d > 0.8). The highest effect size was found in quality of life (d = 1.84) and the lowest in depression (d = 0.95). Pre- and post-treatment scores, and calculated effect sizes are presented in Table 3. The proportion of responders in terms of 30% symptom reduction measured with GSRS-IBS were 52.7% (68 out of 129 patients).

Table 3. Results on outcome measures before and after treatment including effects sizes.

	N	Mean	SD	Improvement	Cohen’s d	95 % CI	t-test p value
GSRS-IBS pre-treatment	129	49.24	11.54	–	–	–	–
GSRS-IBS post-treatment	129	37.03	10.03	34.0%	1.06	[0.86,1.28]	<.001
QOL-IBS pre-treatment	128	42.29	15.71	–	–	–	–
QOL-IBS post-treatment	128	71.12	16.38	68.2%	1.84	[1.56, 2.12]	<.001
VSI pre-treatment	126	50.29	12.64	–	–	–	–
VSI post-treatment	126	28.21	14.23	43.9%	1.75	[1.47, 2.03]	<.001
MADRS-S pre-treatment	127	16.87	6.86	–	–	–	–
MADRS-S post-treatment	127	10.35	6.23	38.6%	0.95	[0.76, 1.15]	<.001

GSRS-IBS: The Gastrointestinal Symptom Rating Scale for IBS; QOL-IBS: The Quality of Life in persons with Irritable Bowel Syndrome; VSI: The Visceral Sensitivity Index; MADRS-S: The Montgomery-Åsberg Depression Rating Scale.

Table 4. Linear regression analysis, variable by variable, controlling for GSRS-IBS score pre-treatment..

Variable	B	t-value	p Value
Diagnosis (IBS-C)	−0.304	−0.248	.805
Years with gastrointestinal symptoms at treatment start	0.699	0.803	.423
Gender (Female)	−2.012	−0.881	.380
Age	−0.116	−1.793	.075
Completed treatment sessions	−0.394	−0.556	.579
VSI	0.044	0.628	.531

GSRS-IBS: Gastrointestinal Symptom Rating Scale for IBS; IBS-C: Irritable Bowel Disorder, constipation predominant; VSI: Visceral Sensitivity Index.

Prediction analysis

The investigated predictors were gender, age at treatment start, IBS-subtype, duration of IBS, number of completed treatment sessions, and GSA. In the first selection step, only one variable, age at treatment start, was associated with the dependent variable (p=.075), controlling for pre-treatment GSRS-IBS scores. We did not perform the second step in the prediction analysis because age would not meet the significance level of 0.05 required in the second step. Thus, a prediction model could not be built. Results from the prediction analysis are presented in Table 4.

Discussion

The present study had two specific aims. Firstly, we wanted to investigate the effectiveness of an ECBT protocol for IBS, in a naturalistic context of a gastroenterological clinic. Secondly, we wanted to find potential predictors for treatment outcome that could be easily obtained and used in clinical decision making.

The treatment protocol we used had earlier been tested face to face in a pilot study [10] and via the internet in several studies [11,12,15] and had yielded promising results, but sometimes the real world provides less than optimal conditions for psychological treatment and effects seen in efficacy studies may be less impressive in clinical practice [16]. The results of this case series study confirmed however that ECBT may significantly reduce symptoms in a group of patients with refractory IBS also when the treatment is offered in gastroenterological routine care and in the format of a group therapy.

The primary outcome measure was GSRS-IBS which is an instrument for measuring self-reported IBS symptoms. The effect size was 1.06 (CI 0.86, 1.28), which is considered a large effect, and 53% of the patients reported reduction in their symptoms with 30% or more, which the U.S. Food and Drug Administration (FDA) proposes as a cut-off for considering a patient a responder in pharmacological trials [39]. The average GSRS-IBS score from pre-treatment to post-treatment was reduced by 34%, indicating that in the present study the patients as an average improved to a clinically significant extent. Of the secondary outcome measures, the QOL-IBS was the most improved with an effect size of 1.84 (CI 1.56, 2.12), indicating a large positive impact on patients’ quality of life. Self-reported depression measured with MADRS was the least affected outcome even though the effect size was 0.95 (CI .76, 1.15) which also is considered a large effect.

In previous randomized controlled trials from our research group on the same treatment-protocol, provided via the internet we have found the controlled effects on IBS-symptoms measured by GSRS-IBS to be 1.21 and 0.77 respectively (compared to waiting list controls) [11,12] which is similar to the effect in the present study.

The study is of particular relevance for clinical practice since it evaluates treatment of regular patients, in a clinical environment at a gastroenterological clinic. We used data from the clinics own electronic medical record and there were no specific adjustments made to the routines at the clinic for study purposes. This gives the study a high ecological validity and shows that ECBT can be integrated in routine care. The format of group therapy with 4–6 participants in each group is of interest since it allows more patients to get access to treatment than individual therapies. The cost per patient is reduced analogously.

The proposed mechanism through which the ECBT works is by reducing gastrointestinal specific anxiety in much the same way as other forms of anxiety is reduced with exposure treatment [41,42]. ECBT involves interoceptive and in-vivo exposure to bodily sensations with the purpose of reducing gastrointestinal specific fear and anxiety and providing coping strategies to patients so that the symptoms associated with IBS can be perceived and handled in a way that reduces harm and hindrance in life. The target of the treatment is therefore not the symptoms themselves but rather how patients react to symptoms. In the present study, GSA measured with VSI, was in average reduced with 43.9% and the effect size was calculated to Cohens d 1.75, indicating that the design of the treatment is precise with regards to the proposed mechanism of action.

The gastroenterological clinic at Sabbatsbergs hospital was at the time of this study run by Aleris on contract with the Stockholm municipality and received patients with referral from primary care units. The clinic employed psychologists in its staff which is less common. Other studies on CBT, and on hypnotherapy as well, have pointed out that access to psychological treatment is low in many areas of the world [43] especially within the field of gastroenterology. Scarcity of psychologists may be an obstacle to overcome if a clinic would want to implement this treatment in its routine care. Although the ECBT manual is very detailed, psychologists who deliver the treatment need to have previous training in CBT and must also be supervised and trained by both psychologists and gastroenterologist to achieve the competence to work with this patient group.

The second aim of the study was to find predictors of treatment outcome. However, we did not find any support for the hypothesis that pre-treatment characteristics can be used to predict the outcome. This is not surprising since there are other studies that have failed to find relevant predictors for psychological treatment outcome in IBS [44] or have been contradictory in their findings [45]. There seems to be some evidence that trait anxiety and anxiety sensitivity can predict treatment outcome in CBT for IBS if it is compared to another treatment (i.e. IBS education) [22], but we did not have access to data regarding these specific pre-treatment factors in our study and we could not compare between groups since it was a case series study. Our research group has in an earlier study showed that baseline avoidance behavior moderates treatment effect [46] and it would therefore be reasonable to believe that the level of pre-treatment GSA could indicate possible responders to the treatment. However, the analytic model employed in our previous study [46] was based on weekly measured changes in symptom level, leading to substantially increased power in the analysis.

A limitation to this study is that we have no follow-up data from a year or more after the treatment. We believe that this is representative of real-world clinical routine, where few clinics collect long-term self-assessment follow up data from patients. Our previous studies of ECBT have however shown that patients’ improvements in IBS symptoms are sustained up to 12, 15, and 18 months after treatment [12,14]. Another limitation was that we had no data on the patients who were not selected for the treatment. It would have been interesting to know for how many patients the treatment was considered appropriate and not appropriate and what factors influenced the patient selection. Additionally, no formal assessment of the therapists’ adherence to treatment protocol was made. However, the large effect sizes (similar to our previous trials) and high treatment adherence suggest that there is a low risk that the therapists made significant deviations from the protocol. Of the total number of 171 patients who were treated at the clinic during the data collection period, 129 (75%) provided complete data on GSRS-IBS and were in included in the analysis. However, we also performed sensitivity analysis using multiple imputation were 145 (85%) of the patients were included, which showed very similar results (data not reported). Nevertheless, there was a significantly larger proportion of women than men who provided complete data and thus the results may not be fully generalizable to male patients with IBS.

Lastly, we did not have any randomized comparison group that could be used to control for the effects of time, expectancy of improvement, and attention from a caregiver. Although this limitation is inherent in a case series study, we cannot conclude that the observed improvements were due to the intervention itself. A follow-up could potentially also have discerned transient placebo effects from true intervention effects. Importantly, we have in previous RCTs shown that ECBT for IBS has treatment effects over and above effects of only expectancy of improvement and attention [13,15], which suggests that a significant proportion of treatment effect in this case series study was due to specific factors included in ECBT.

We conclude that ECBT delivered in a group format is highly effective and can be used in gastroenterological routine care. We also conclude that neither gender, type of IBS, years with IBS, patients age at treatment start nor initial GSA are relevant predictors for treatment outcome. The findings may also further support the underlying theory that avoidance behavior and hypervigilance plays an important role in the maintenance of symptoms and disability for some IBS patients.

Disclosure statement

The author Brjánn Ljótsson has developed the ECBT protocol and licenses it on a commercial basis for digital delivery. Per-Johan Lindfors, Brjánn Ljótsson, Lisa Falk and Sara Rydh were all employed at the gastroenterological clinic at Sabbatsbergs Hospital and took part in the daily work during the data collection period.

Data availability statement

Due to the European Union General Data Protection Regulation (GDPR) any researcher who requires access to raw data must apply to the Swedish authority of ethical approvals (Etikprövningsmyndigheten).

Additional information

Funding

The main author, Hugo Wallén was funded by Sahlgrenska University Hospital with paid work time during the process of writing the paper.