Improving health-related quality of life in patients with endometriosis-related chronic pain: the COGENS trial

A detailed description of the Materials and Methods for this randomized controlled, prospective, single-blind multicenter clinical trial is published elsewhere [ 11 ]. Below, a brief description of the study protocol is presented. Patients were included from one academic tertiary and four secondary hospitals. Patients were eligible for participation when aged 18–50 years, diagnosed with endometriosis that was confirmed by ultrasound, MRI, or surgery, who had an indication for endometriosis surgery due to endometriosis-related chronic pain. Endometriosis-related chronic pain was defined as pain present since at least three months. Participants were also required to have a sufficient understanding of the Dutch language. Exclusion criteria were: currently receiving psychological treatment, current use of psychopharmacological medication prescribed to alter mood, other chronic pain complaints, a current mood, anxiety, or personality disorder, or scalp hair shorter than 4 centimeters. Participants in whom endometriosis was not confirmed during surgery were excluded from the study. All participants gave oral and written informed consent to participate in this study. Table 1 and 5 present demographic variables, as well as seven questions that evaluated patients’ motivation to undergo psychological treatment using a 10-point Likert scale, because motivation to undergo psychological therapy may influence treatment results [ 12 ]. Table 1 Demographic data of the participants  Patient characteristics CBT group Control group N 50 (100.0) 50 (100.0) Division between hospitals (n (%)) Secondary 34 (68.0) 26 (52.0) Tertiary Academic 16 (32.0) 24 (48.0) Age in years (mean (SD)) 34.4 (7.3) 36.4 (8.2) BMI kg/m2 (mean (SD)) 27.1 (5.3) 26.0 (3.9) Educational attainment a (median (25–75 percentile)) 6 (5–6) 5 (5–6) Paid work (n (%)) 39 (78.0) 40 (80.0) Marital status (n (%)) Single 14 (28.0) 4 (8.0) Married/Relationship  33 (66.0) 43 (86.0) Divorced/Separated  3 (6.0) 3 (6.0) Time since diagnosis in years (mean (SD)) 3.9 (5.7) 3.9 (5.4) Tool used for diagnosis (n (%)) Ultrasound only 8 (16.0) 5 (10.0) MRI 29 (58.0) 35 (70.0) Surgery 13 (26.0) 10 (20.0) Type of endometriosis (n (%)) Peritoneal 20 (40.0) 22 (44.0) Ovarian 18 (36.0) 19 (38.0) Deep 17 (34.0) 22 (44.0) Subfertility (n (%)) 8 (16.0) 8 (16.0) Parity (median (25–75 percentile)) 0 (0–2) 1 (0–2) Use of analgesics (n (%)) 44 (88.0) 43 (86.0) Use of hormonal medication (n (%)) 40 (80.0) 36 (72.0) History of DSM-IV diagnosis (n (%)) 10 (20.0) 11 (22.0) Prior psychological treatment (n (%)) 11 (22.0) 11 (22.0) Surgery characteristics rASRM score (median (25–75 percentile)) 4 (2–4) 4 (2–4) Pathology confirmed endometriosis (n (%)) 39 (78.0) 45 (90.0) Complete surgical eradication (n (%)) 48 (96.0) 42 (84.0) Surgery method (n (%)) Conventional laparoscopy 36 (72.0) 38 (76.0) Robot assisted laparoscopy 14 (28.0) 12 (24.0) Conversion to laparotomy 3 (6.0) 0 (0.0) Multidisciplinary surgery (n (%)) 12 (24.0) 15 (30.0) Performed of surgery (n (%)) Removal of peritoneal endometriosis 20 (40.0) 22 (44.0) Adhesiolysis  22 (44.0) 23 (46.0) Ovarian cystectomy  11 (22.0) 10 (20.0) Tubectomy  18 (36.0) 23 (46.0) Unilateral ovariectomy  8 (16.0) 9 (18.0) Bilateral ovariectomy  5 (10.0) 1 (2.0) Hysterectomy  15 (30.0) 22 (44.0) Rectosigmoid shaving  5 (10.0) 4 (8.0) Rectosigmoid resection  4 (8.0) 4 (8.0) Partial cystectomy (bladder)  5 (10.0) 1 (2.0) Removal abdominal wall endometriosis  8 (46.0) 7 (14.0) Removal of vaginal cuff endometriosis  1 (2.0) 1 (2.0) Complications (n (%)) 4 (8.0) 5 (10.0) CBT Cognitive Behavioral Therapy, BMI Body Mass Index, MRI Magnetic Resonance Imaging, DSM-IV Diagnostic and Statistical Manual of Mental Disorders, rASRM revised American Society for Reproductive Medicine. a Patients’ educational attainment was scored using a 7 point rating scale [ 34 ]: [1] unfinished primary school [2], finished primary school [3], unfinished low-level secondary education [4], lower vocational training [5], advanced vocational training or lower professional education [6], finished higher professional education or senior general secondary education, and [7] obtained a university degree. The sum of types of endometriosis and performed surgeries may not equal 100% due to the potential for a single patient to experience multiple outcomes across these variables Demographic data of the participants CBT Cognitive Behavioral Therapy, BMI Body Mass Index, MRI Magnetic Resonance Imaging, DSM-IV Diagnostic and Statistical Manual of Mental Disorders, rASRM revised American Society for Reproductive Medicine. a Patients’ educational attainment was scored using a 7 point rating scale [ 34 ]: [1] unfinished primary school [2], finished primary school [3], unfinished low-level secondary education [4], lower vocational training [5], advanced vocational training or lower professional education [6], finished higher professional education or senior general secondary education, and [7] obtained a university degree. The sum of types of endometriosis and performed surgeries may not equal 100% due to the potential for a single patient to experience multiple outcomes across these variables Patients were randomly allocated by computerized randomization (block 2:2) to surgery and usual care (control group) or to surgery and usual care plus CBT (CBT group). The participants and the psychologists delivering the intervention could not be blinded due to the used intervention. Figure  1 contains an overview of the patient flow throughout the study, including the exclusion criteria. Fig. 1 Patient flow CBT = Cognitive Behavioral Therapy Patient flow CBT = Cognitive Behavioral Therapy Both the control and CBT group received usual endometriosis care. This was defined as pre- and post-surgery care by health care professionals in the hospital and at the out-patient department. All participating hospitals provided similar usual care. Usual care included pre-surgical counseling by the gynecologist about the procedure, possible complications, and expected outcomes. Patients were admitted for surgery on the day of the surgical procedure. Surgical treatment consisted of laparoscopic excision of endometriosis lesions, performed by experienced gynecologic surgeons. Depending on disease extent, surgery could include excision of peritoneal, ovarian, or deep infiltrating endometriosis, with additional procedures such as adhesiolysis, ovarian cystectomy, bowel shaving or resection, or bladder surgery when indicated. If needed, patients consulted a gastrointestinal surgeon and/or urologist. Surgery was performed by a gynecology team, with other specialists if necessary. Patients were asked to use their regular hormonal therapy before and after surgery to reduce menstrual effects. The use of analgesics was permitted at the patient’s discretion; however, participants were asked to refrain from initiating additional psychological treatment during the study period. Medical check-ups took place around six weeks, three months, and six months after surgery, including symptom review and physical exams. Patients could contact the endometriosis nurse anytime by email or phone. Patients in the CBT group additionally received one pre-surgery and six post-surgery individual face-to-face sessions of CBT, either in-person or using a videoconference depending on the preference of the patient. The first session, the pre-surgery one, was always in-person. There were six post-surgery sessions, every two weeks. The post-surgery CBT sessions started four weeks after surgery. The end of CBT was 16 weeks after surgery (T1). T2 was 16 weeks later, 32 weeks post-surgery. All sessions were coordinated by psychologists who were experienced in CBT, had substantial knowledge about endometriosis, and were trained to execute the CBT protocol. The content of the CBT protocol was developed by psychologists together with medical professionals especially for this study. It included pain education and content based on standard CBT interventions for chronic pain. This was supplemented with topics aimed at specific psychological factors present in patients with endometriosis, including pain catastrophizing, hypervigilance, pain anxiety, mood, and intimacy. Table  2 shows an overview of the topics discussed in each session. Patients were asked not to participate in other interventions outside of this study. Table 2 Overview of cognitive behavioural therapy content Session Themes to be discussed Time (weeks, relative to surgery) Duration (min) 1 - Therapeutic compliance and expectation towards therapy - Effects of endometriosis on patient’s life - Expectations towards effect of surgery - General pain-education -2 60 2 - Setting goals for therapy - The biopsychosocial link between pain and behaviour - Relaxation 4 45 3 - The biopsychosocial link between pain and emotion 6 45 4 - The biopsychosocial link between pain and thoughts - Negative automatic thoughts 8 45 5 - Hypervigilance towards pain 10 45 6 - Intimacy and sexuality 12 45 7 - Evaluation of therapy - Relapse prevention 14 45 Overview of cognitive behavioural therapy content - Therapeutic compliance and expectation towards therapy - Effects of endometriosis on patient’s life - Expectations towards effect of surgery - General pain-education - Setting goals for therapy - The biopsychosocial link between pain and behaviour - Relaxation - The biopsychosocial link between pain and thoughts - Negative automatic thoughts - Evaluation of therapy - Relapse prevention Table  3 presents an overview of all variables, detailing the instruments used and the time of measurement. All questionnaires as well as scalp hair were collected four weeks before surgery (T0), and two (T1), and 16 (T2) weeks post-intervention, i.e. post CBT treatment. T1 corresponds with 16 weeks post-surgery and T2 corresponds with 32 weeks post-surgery respectively. Table 3 Measurements Variable T0, T1, T2 Measuring instrument HRQoL ✓ Endometriosis Health Profile 30; Short Form 36 Pain intensity ✓ Numerical Rating Scale Catastrophizing ✓ Pain Catastrophizing Scale Pain anxiety ✓ Pain Anxiety Symptom Scale Fatigue ✓ Checklist Individual Strength Stress ✓ Perceived Stress Scale; Hair cortisol concentration HRQoL Health Related Quality of Life.  T0 = 2–4 weeks prior to surgery, T1 = 2 weeks post-intervention (= 16 weeks post-surgery), T2 = 16 weeks post-intervention (= 32 weeks post-surgery). Measurements HRQoL Health Related Quality of Life. T0 = 2–4 weeks prior to surgery, T1 = 2 weeks post-intervention (= 16 weeks post-surgery), T2 = 16 weeks post-intervention (= 32 weeks post-surgery). The primary outcome of the study was HRQoL, measured with two validated questionnaires: the Endometriosis Health Profile 30 (EHP-30) and the Short Form 36 version 2.0. (SF-36) [ 13 – 15 ]. The Endometriosis Health Profile is a disease-specific HRQoL questionnaire, consisting of two parts. The core questionnaire has five subscales: pain, control, and powerlessness, emotional well-being, social support, and self-image. The second part consists of six subscales: work, relationship with children, sexual intercourse, infertility, medical profession, and treatment. Scores range from 0 to 100 with lower scores indicating better HRQoL. The SF-36 version 2.0 is a survey covering nine different domains: physical functioning, social functioning, role limitations due to physical health, role limitations due to emotional problems, emotional well-being, vitality, pain, general health, and health change. Scores range from 0 to 100 with a higher score corresponding to a better HRQoL. These scores can be averaged into two subscales: the Physical Component Scale, reflecting physical HRQoL, and the Mental Component Scale, reflecting mental HRQoL. The Physical Component Scale captures aspects such as physical functioning and pain, while the Mental Component Scale reflects emotional well-being and social functioning. Together, they provide a broad overview of an individual’s overall health status. Secondary endpoints were pain intensity, pain anxiety, pain catastrophizing, fatigue, perceived stress, and cortisol level as a marker for stress. Pain intensity was measured with a Numerical Rating Scale (NRS) [ 16 ]. Specifically, patients were asked to perform two ratings, one evaluating the overall worst and one evaluating the overall average pain that patients had experienced over the past seven days. Both items were scored on a scale ranging from 0 (no pain) to 10 (worst pain possible). The Pain Anxiety Symptoms Scale (PASS) measures the fear and anxiety individuals experience in relation to pain, including cognitive, physiological, and behavioral responses [ 17 ]. Scores range from 0 to 100, with higher scores indicating greater pain-related anxiety. Pain catastrophizing is measured by the Pain Catastrophizing Scale (PCS) which evaluates feelings of rumination, magnification, and helplessness [ 18 ]. Scores range from 0 to 52 with a higher score corresponding to more catastrophizing. The Checklist Individual Strength (CIS) is a questionnaire designed to assess fatigue and its impact on daily functioning, covering dimensions such as concentration, motivation, and physical activity [ 19 ]. Scores range from 20 to 140, with higher scores indicating greater levels of fatigue. Indicators for stress were measured in two ways. First, the Perceived Stress Scale (PSS) measured perceived stress, ranging from 0 to 40, with higher scores indicating greater levels of perceived stress [ 20 ]. In addition to perceived stress, hair cortisol concentrations (HCC) were measured as an indicator for physical chronic stress [ 21 , 22 ]. A 1 centimeter segment hair fragment nearest to the scalp was cut and selected for analysis, to measure HCC over the preceding month based on the average monthly hair growth rate [ 22 ]. Collection of hair samples was performed by trained researchers. Analysis was conducted using an enzyme-linked immunoassay [ 22 ]. Based on the available literature [ 23 ], a medium to large effect size (Cohen’s d) of 0.75 and 1.35 for the primary outcome HRQoL was expected. Anticipating a 19% dropout rate, a total of 100 patients needed to be included to detect significant effects (α = 0.05, power (1-β) = 0.85). All analyses were performed with SPSS (v.29) and a significance threshold of p ≤ 0.05 was applied. First, univariate linear mixed models (LMM) were used to investigate the differences in outcome variables between the control and intervention group over time. LMM is able to handle multiple observations and handle missing data without compromising power [ 24 ]. For the primary endpoint HRQoL, separate models for the EHP-30 core and modular subscales, and the SF-36 mental component scale and physical component scale were calculated. The models included group (CBT vs. control), time (measurement points at T0, T1, and T2) and the interaction between group and time as fixed effects. In this article we primarily focused on group-level differences to assess the effect of CBT, and explored pairwise comparisons between T0 and T1, and between T0 and T2. A compound symmetry covariance structure was applied to account for correlations among repeated measures, with random intercepts modeling between-subject variability. In addition, we controlled for the difference at T0 measurement between the CBT and the control group by including this in the analysis as a covariate. No other covariates were included. This analysis was repeated for the secondary outcome measures pain intensity, catastrophizing, anxiety, fatigue, perceived stress and HCC. If a questionnaire demonstrated a significant effect, the LMM analysis was repeated for the subscales of the questionnaire to explore which domains contributed to the observed effect. Subsequently, a mediation analysis was conducted (using the PROCESS macro [ 25 ] to answer the question to what extent improvements in pain cognitions, pain intensity, fatigue, and stress levels underlie the anticipated positive effect of CBT on improvement in HRQoL. Therefore, regression models were briefly calculated with the long-term change in HRQoL as dependent variables (difference in HRQoL between T0 and T2), group (CBT versus control group) as independent predictor, and change in pain catastrophizing, anxiety, pain intensity, fatigue, perceived stress, and HCC as mediators. Finally, we performed a paired sample t-test comparing outcomes between T0 and T1, T0 and T2, and T1 and T2 separately for patients in the control and CBT group.

Between November 24, 2020 and December 28, 2023 a total of 291 patients were screened for participation. Of these, 213 (73.2%) met the eligibility criteria, and 100 (34.4%) patients were ultimately included in the study. As presented in Fig.  1 , there were six dropouts in the CBT group and three in the control group. In accordance with the intention-to-treat principle, final analysis included all 100 patients that were initially included. The mean age of the CBT group was 34.4 (SD 7.3) years and of the control group it was 36.4 (SD 8.2) years. The medication use, history of DSM-IV diagnosis and prior psychological treatment were comparable between groups. In the CBT group, five (10.0%) surgical complications occurred, compared to four (8.0%) in the control group. Motivation to undergo CBT was comparable between groups, with mean scores of 7.8 (SD 2.2) in the CBT group and 8.3 (SD 1.8) in the control group. In Table  4 , adherence to the intervention is shown as well as the time in which the intervention sessions were completed. In total 39 (78.0%) patients in the intervention group used the optional, online CBT module and they completed 67.6% of the module on average. Patient characteristics that could affect the HCC measurement are shown in Table S2. Table 4 Compliance to the time planning of the intervention Planning Mean (SD) CBT group Control group Aim T0 -3.5 (5.9) -3.5 (3.5) -4.0 to -2.0 Session 1 -2.6 (5.7) - -4.0 to -2.0 Surgery 0.0 0.0 0.0 Session 2 4.7 (1.7) - 4.0 3 6.8 (1.3) - 6.0 4 9.2 (1.7) - 8.0 5 11.8 (2.3) - 10.0 6 14.6 (2.5) - 12.0 7 17.6 (3.3) - 14.0 T1 19.7 (3.2) 16.8 (1.4) 16.0 T2 34.0 (3.4) 30.5 (1.0) 30.0 Session in reference to time of surgery in weeks.  CBT Cognitive Behavioral Therapy Compliance to the time planning of the intervention Session in reference to time of surgery in weeks. CBT Cognitive Behavioral Therapy For the primary outcome measure HRQoL, an LMM analysis was conducted to examine changes in HRQoL over time and to assess whether these changes differed between the control and intervention group. The results showed that the HRQoL was better when CBT was added, compared to surgery plus usual care alone. Specifically, the LMM analyses revealed that patients in the CBT group showed a significant overall improvement for the modular part of the EHP-30 and physical component scale of the SF-36, compared to the control group. For the modular part of the EHP-30 questionnaire, a significant better HRQoL was found in patients in the CBT group at both T1 and T2, compared to the control group. Exploring the different scales of the modular part in the EHP-30 revealed a significant positive effect of CBT on the subscales for work, relationship with children, feelings towards the medical profession and professionals and feelings towards the treatment itself (presented in Table S3). In addition, patients in the CBT group showed a significant improvement at T1 and T2 on the physical component scale of the SF-36 questionnaire. The most substantial improvement was observed in the subscales for physical functioning and general health. No overall significant effects were found on the core part of the EHP-30 or the mental component scale of the SF-36. However, compared to the control group, the CBT group did report a significantly better HRQoL on the core part of the EHP-30 at T1, but not T2. The mental component scale of the SF-36 questionnaire did not differ between the groups at T1 and T2. The individual patient scores of the primary and secondary outcome measures are shown in Fig.  2 . The questionnaires scores are presented in Table  5 and the results of the LMM analysis are shown in Table  6 . Fig. 2 Individual patient scores of primary and secondary outcomes Fig. 2.a to 2.k present multiple line plots illustrating the trends across all primary and secondary outcome measures. Individual subject data is represented by thin lines, while the bold lines indicate the average scores for each questionnaire. CBT = Cognitive Behavioral Therapy; EHP-30 = Endometriosis Health Profile 30; SF-36 = Short Form 36; NRS = Numerical Rating Scale; PCS = Pain Catastrophizing Scale; PASS = Pain Anxiety Symptom Scale; CIS = Checklist Individual Strength; PSS = Perceived Stress Scale Individual patient scores of primary and secondary outcomes Fig. 2.a to 2.k present multiple line plots illustrating the trends across all primary and secondary outcome measures. Individual subject data is represented by thin lines, while the bold lines indicate the average scores for each questionnaire. CBT = Cognitive Behavioral Therapy; EHP-30 = Endometriosis Health Profile 30; SF-36 = Short Form 36; NRS = Numerical Rating Scale; PCS = Pain Catastrophizing Scale; PASS = Pain Anxiety Symptom Scale; CIS = Checklist Individual Strength; PSS = Perceived Stress Scale Table 5 Means scores of outcome measures Questionnaires Mean (SD) CBT group Control group T0 T1 T2 T0  T1 T2 EHP-30 Core 54.9 (13.3) 26.3 (18.5) 26.0 (19.9) 56.3 (13.0) 34.4 (22.0) 33 (23.7) Modular 42.2 (17.1) 21.2 (16.2) 20.0 (18.8) 41.4 (17.6) 28.7 (22.5) 27.1 (21.7) SF-36 Mental 37.4 (10.7) 45.2 (11.4) 45.3 (11.3) 36.1 (9.7) 42.2 (11.7) 43.0 (13.6) Physical 37.0 (9.1) 49.0 (9.6) 49.6 (10.4) 39.4 (8.8) 47.2 (8.4) 48.0 (8.6) NRS Average 5.7 (2.4) 2.3 (2.4) 2.6 (2.8) 5.2 (2.0) 2.2 (1.9) 2.5 (2.1) Worst 7.4 (2.2) 3.4 (3.3) 3.6 (3.6) 6.9 (2.4) 3.3 (3.0) 3.4 (2.8) PCS Rumination 11.1 (3.5) 6.4 (4.2) 5.8 (4.2) 9.9 (4.1) 7.4 (5.2) 6.4 (4.8) Magnification 3.7 (2.3) 2.5 (2.6) 1.8 (2.3) 3.6 (2.7) 2.5 (2.8) 2.4 (2.6) Helplessness 13.1 (6.1) 6.4 (6.2) 5.3 (5.5) 11.1 (6.0) 7.9 (6.8) 6.6 (6.0) Total 27.9 (10.9) 15.3 (12.2) 12.9 (11.0) 24.6 (11.5) 17.8 (14.0) 15.4 (12.6) PASS 80.1 (28.1) 57.7 (23.7) 51.0 (25.0) 78.5 (27.1) 63.2 (30.3) 60.6 (30.5) CIS Fatigue 45.0 (10.0) 32.8 (13.5) 33.6 (13.7) 44.1 (8.3) 35.7 (12.3) 33.4 (11.5) Concentration 22.0 (6.7) 15.1 (7.8) 16.5 (7.8) 21.2 (7.7) 17.9 (7.6) 17.0 (8.6) Motivation 16.1 (5.8) 10.0 (5.7) 11.6 (6.1) 15.4 (5.5) 13.5 (6.6) 13.9 (6.5) Activity 8.3 (4.0) 5.0 (3.5) 5.4 (3.6) 7.6 (3.4) 6.1 (3.6) 6.0 (3.4) Total 95.9 (21.7) 66.0 (28.6) 70.5 (28.8) 92.5 (19.4) 75.4 (26.4) 73.6 (26.6) PSS 18.9 (7.3) 13.1 (6.6) 13.9 (6.6) 19.5 (6.3) 15.9 (7.2) 15.9 (7.5) Cortisol levels HCC pg/mg hair 2.5 (1.5) 3.0 (2.2) 2.9 (1.7) 2.8 (2.8) 2.1 (2.1) 2.6 (1.9) CBT  Cognitive Behavioral Therapy, EHP-30  Endometriosis Health Profile 30, SF-36  Short Form 36, NRS  Numerical Rating Scale, PCS  Pain Catastrophizing Scale, PASS  Pain Anxiety Symptom Scale, CIS  Checklist Individual Strength, PSS  Perceived Stress Scale, HCC  Hair Cortisol Concentration Means scores of outcome measures CBT  Cognitive Behavioral Therapy, EHP-30  Endometriosis Health Profile 30, SF-36  Short Form 36, NRS  Numerical Rating Scale, PCS  Pain Catastrophizing Scale, PASS  Pain Anxiety Symptom Scale, CIS  Checklist Individual Strength, PSS  Perceived Stress Scale, HCC  Hair Cortisol Concentration Table 6 Linear mixed models Questionnaires Time Mean group differences SE 95% CI p Overall effect ( p ) EHP-30 Core T1 -6.8* 3.4 -13.5 – -0.1 0.047 0.080 T2 -6.4 3.4 -13.1 – -0.3 0.059 Modular T1 -8.2* 3.1 -14.4 – -2.1 0.009 0.021* T2 -7.7* 3.1 -13.8 – -1.6 0.014 SF-36 Mental T1 1.4 1.8 -2.1–4.9 0.434 0.574 T2 0.8 1.8 -2.7–4.3 0.660 Physical T1 4.0* 1.4 1.3–6.7 0.004 0.008* T2 3.9* 1.4 1.2–6.5 0.005 NRS Average T1 0.1 0.4 -0.7–0.9 0.858 0.937 T2 0.0 0.4 -0.8–0.8 0.993 Worst T1 0.0 0.5 -1.1–1.0 0.952 0.951 T2 0.0 0.5 -1.1–1.0 0.944 PCS T1 -4.3* 1.7 -7.7 – -0.9 0.015 0.019* T2 -4.0* 1.7 -7.4 – -0.6 0.021 PASS T1 -5.3 3.8 -12.8–2.1 0.160 0.047* T2 -10.0* 3.8 -17.4 – -2.6 0.008 CIS T1 -10.9* 4.4 -19.5 – -2.3 0.014 0.116 T2 -4.4 4.3 -12.9–4.1 0.307 PSS T1 -2.4* 1.1 -4.6 – -0.3 0.029 0.085 T2 -1.6 1.1 -3.8–0.5 0.134 Cortisol levels HCC T1 0.9* 0.3 0.4–1.5 < 0.001 0.022* T2 0.4 0.3 -0.2–0.9 0.189 * statistically significant ( p  ≤ 0.05), SE  standard error, CI  confidence interval, EHP-30  Endometriosis Health Profile 30, SF-36  Short Form 36, NRS  Numerical Rating Scale, PCS  Pain Catastrophizing Scale, PASS  Pain Anxiety Symptom Scale, CIS  Checklist Individual Strength, PSS  Perceived Stress Scale Linear mixed models * statistically significant ( p  ≤ 0.05), SE  standard error, CI  confidence interval, EHP-30  Endometriosis Health Profile 30, SF-36  Short Form 36, NRS  Numerical Rating Scale, PCS  Pain Catastrophizing Scale, PASS  Pain Anxiety Symptom Scale, CIS  Checklist Individual Strength, PSS  Perceived Stress Scale For the secondary outcome measures, the LMM analysis showed a significant overall effect on pain catastrophizing, pain anxiety, and HCC. In detail, pain catastrophizing was significantly lower in the intervention group at T1 and T2, when compared to the control group, with rumination and helplessness showing the largest effect. The pairwise comparison showed that pain anxiety was only significantly reduced at T2, not T1. In addition, the CBT group displayed significantly higher levels of HCC at T1 only, when compared to the control group. Average and worst pain intensity, fatigue, and perceived stress did not show an overall significant effect. In the CBT group, fatigue and perceived stress improved at T1. However, these effects were no longer evident at T2 when compared with the control group. The mediation analysis demonstrated a statistically significant total indirect effect of all mediators combined on the relationship between CBT and HRQoL. Significant effects were observed for CBT on pain catastrophizing, and for pain intensity and perceived stress on HRQoL. No mediator showed a significant indirect effect when examined individually. Results of the mediation analysis are presented in Fig.  3 and Table S4. Fig. 3 Mediation effects Mediation model illustrating the indirect effect of CBT on HRQoL through multiple mediators. The arrows on the left (path a) represent the effect of the CBT intervention on the mediators, while the arrows on the right (path b) indicate the effect of the mediators on HRQoL. The numbers alongside the arrows show the effect sizes for these paths. Together, they illustrate the total indirect effect (a*b path) of CBT on HRQoL for each mediator. The direct path from CBT to HRQoL represents the effect of CBT independent of the mediators. * = Statistically significant; CBT = Cognitive Behavioral Therapy; HRQoL = Health-Related Quality of Life; HCC = Hair Cortisol Concentration Mediation effects Mediation model illustrating the indirect effect of CBT on HRQoL through multiple mediators. The arrows on the left (path a) represent the effect of the CBT intervention on the mediators, while the arrows on the right (path b) indicate the effect of the mediators on HRQoL. The numbers alongside the arrows show the effect sizes for these paths. Together, they illustrate the total indirect effect (a*b path) of CBT on HRQoL for each mediator. The direct path from CBT to HRQoL represents the effect of CBT independent of the mediators. * = Statistically significant; CBT = Cognitive Behavioral Therapy; HRQoL = Health-Related Quality of Life; HCC = Hair Cortisol Concentration Finally, paired sample T-test showed that patients in both the control group and the CBT group improved from T0 to T1 and from T0 to T2 for all primary and secondary outcome measures, except HCC (presented in Table S5).

Endometriosis affects an estimated 10% of women in their reproductive age, leading to significant health care costs and a high economic burden [ 1 ]. The disease is characterized by the presence of active endometrium-like tissue outside the uterine cavity. In the majority of women, the main symptom is cyclic or chronic pelvic pain, and/or infertility. In this study we focus on endometriosis-related chronic pain, which leads to impaired health related quality of life (HRQoL) [ 1 ]. To improve HRQoL, endometriosis treatment has long been aimed at pharmacological pain reduction which primarily consists of hormonal therapies aimed at suppressing ovulation and menstruation, thereby reducing inflammatory activity and pain symptoms. Although medical therapy can be effective in reducing pain for many patients, its effects are often temporary and symptoms frequently recur after discontinuation. In addition, long-term use may be limited by side effects, contraindications, or insufficient symptom relief. When pharmacological medication provides insufficient pain relief, endometriosis surgery can be necessary. Surgical treatment aims to remove visible endometriotic lesions and can result in significant short-term improvement in pain and HRQoL. However, pain recurrence after surgery is common, and symptom severity does not consistently correlate with the extent of disease observed during surgery [ 1 – 3 ]. Recent evidence therefore supports a more comprehensive treatment approach in which medical and surgical strategies are complemented by interventions targeting central pain mechanisms and pain-related cognitions [ 4 ]. Pain is a multidimensional symptom involving ascending and descending pain pathways in the central nervous system [ 5 , 6 ]. Traditional pain mitigation strategies like treatment with analgesics can inhibit these pathways, while surgery enables removal of the source of the pain [ 5 , 6 ]. When pain becomes chronic, the descending pain pathways that modulate pain intensity may become dysregulated and the pain processing system enters an increased state of responsiveness [ 3 , 6 ]. Therefore, effective reduction of pain intensity in endometriosis necessitates interventions targeting ascending and descending pain pathways simultaneously. Chronic pain has several non-pharmacological treatment possibilities, of which cognitive behavioral therapy (CBT) is among the most effective [ 5 , 7 , 8 ]. In endometriosis, pain catastrophizing and the associated pain cognition, pain anxiety, have been identified as potential targets for CBT [ 9 ]. CBT can improve pain catastrophizing and pain anxiety. Therefore, CBT may be promising for pain treatment and HRQoL improvement in endometriosis patients [ 10 ]. In this randomized controlled trial (RCT) we evaluate the effect of CBT on HRQoL in patients undergoing surgery for endometriosis-related chronic pain. CBT and surgery act on different pathways of chronic pain, resulting in a combined intervention that addresses ascending and descending pain pathways. We hypothesize that surgery and CBT will improve HRQoL more than surgery and usual post-surgical endometriosis care only.

The results of this RCT show that adding CBT to the surgical management of endometriosis improves HRQoL more compared to surgery alone, both two weeks after completing CBT treatment (T1), and 16 weeks after completing CBT treatment (T2). In addition, patients receiving CBT also had stronger reductions in pain catastrophizing and pain anxiety, and an increase in HCC. Pain intensity, fatigue, and perceived stress improved similarly in patients receiving and not receiving CBT. We conclude that CBT can help patients cope with endometriosis-related chronic pain by reducing pain catastrophizing and pain anxiety, resulting in better HRQoL. This RCT is the largest study combining CBT with surgery for the treatment of endometriosis. A previous preliminary study has reported similar findings regarding the efficacy of CBT after surgery for endometriosis [ 26 ]. The results of the current study align with this study and are consistent with the broader literature supporting CBT’s effectiveness for chronic pain in general [ 5 , 27 , 28 ], and endometriosis specifically [ 7 , 8 , 10 ]. Pain intensity scores in the control group as well as in the intervention group decreased significantly after surgery, when comparing T0 to T1 and T2. This reconfirms the effectiveness of surgical treatment for endometriosis-related chronic pain, at least until 32 weeks after surgery [ 1 ]. Interestingly, pain intensity did not differ between the intervention group and the control group, while the HRQoL scores did. Therefore, the additional improvement in HRQoL after endometriosis surgery observed in the CBT group cannot be explained by the decrease in pain intensity. The fear-avoidance model offers an explanation for these findings [ 29 ]. According to this model, factors related to pain intensity, such as pain catastrophizing and anxiety, contribute to functional impairment, emotional distress, isolation, and avoidance behaviors, which sustain and worsen chronic pain. Improving pain catastrophizing and pain anxiety could therefore reduce the negative impact of chronic pain, in terms of decreasing avoidance behaviors, promoting engagement, and alleviating emotional distress, resulting in better HRQoL. The CBT intervention was mainly focused on reducing catastrophizing, anxiety, and reducing avoidance behaviors by promoting exposure. Given that the intervention was designed around key components of the fear-avoidance model, the observed greater reductions in pain catastrophizing and anxiety in the CBT group compared to controls are consistent with the intended mechanisms of the fear-avoidance model. This was also reflected in the mediation analysis that showed that the effect of CBT on pain catastrophizing was significant. While the effects of pain intensity and perceived stress on HRQoL were significant too, these effects were not the direct result of the CBT intervention. We conclude that, while both groups reported similar levels of pain intensity, adding CBT to endometriosis surgery is an effective intervention for enhancing HRQoL. The change in HRQoL in the CBT group was not accompanied by a change in fatigue and perceived stress. In the CBT group, fatigue and perceived stress showed improvements at T1, but this effect disappeared at T2 when compared to the control group. Other studies have demonstrated that CBT can improve these factors as well [ 5 , 27 ]. However, a direct comparison with the current study is challenging, as participants in both groups underwent surgery, which contributed to significant improvements in almost all outcome measures compared to T0. This could have diluted the observed impact of CBT on fatigue and perceived stress. Another explanation is that CBT typically encourages behavioral activation, which helps patients re-engage in daily activities [ 30 ]. The psychological benefits, such as improvements in mood, anxiety, and overall mental well-being, might take longer to manifest. However, this cannot be concluded based on the results of this study as a longer follow-up period than 32 weeks would be necessary. Surprisingly, the HCC showed an opposite effect compared to perceived stress that cannot be explained within the framework of the current study. Results from a previous observational study among patients with endometriosis were comparable to this study, showing higher HCC levels in endometriosis patients compared to controls [ 21 ]. In that study it was speculated that increased levels of cortisol might be an adaptive and protective response against pain and for well-being. Furthermore, from psychophysiological studies it is known that exposure-based psychological interventions can be accompanied by acute increases in cortisol levels, reflecting temporary activation of the hypothalamic–pituitary–adrenal (HPA) axis during therapeutic exposure to stressors. These cortisol responses, that have been observed during exposure-based CBT as well are considered part of an adaptive stress response associated with emotional processing and learning [ 31 ], potentially also explaining the temporary rise in cortisol levels in the CBT group in the current study. In one study however, it was found that people with higher levels of in session-cortisol were predicted to benefit less from exposure therapy [ 32 ]. From a clinical perspective, such temporary increases in cortisol are generally not considered harmful and are interpreted as adaptive physiological responses rather than indicators of pathological stress. These responses are expected to normalize over time as patients habituate to stressors and develop more adaptive coping strategies. This interpretation is supported by our findings, as HCC levels no longer differed between groups at T2. Another explanation for the temporary elevation in cortisol levels may be that HCC represent the full chronic stress response of an individual, which may not be fully captured by a single questionnaire evaluating stress [ 33 ]. Therefore, participants may improve their score on the PSS, but stressors not captured by this questionnaire may still lead to an increased HCC. The results of this study underscore the usefulness of integrated, multidisciplinary endometriosis clinics, in which surgical, medical, and psychological care are closely coordinated. Given the multifactorial nature of endometriosis-related chronic pain, embedding CBT within an integrated care pathway may enhance treatment effectiveness, continuity of care, and timely referral to psychological support. Due to the success of the CBT after endometriosis surgery, some participating centers have implemented the CBT in their clinics and continue to offer the intervention to endometriosis patients. Despite the positive effect that was perceived by patients and caregivers, strategies to further improve effectiveness, accessibility, and affordability can help to implement the intervention even broader. Therefore, future studies should examine the long-term effect of CBT, which patient characteristics are associated with treatment success, and how to overcome practical obstacles in the clinic. These factors impact the cost-effectiveness, and ultimately the implementability of the intervention. First, long-term studies evaluating the sustained effects of CBT on HRQoL in endometriosis chronic pain will add to the evidence on long term effectiveness in endometriosis. Next, the availability of well-trained psychologists to accommodate all endometriosis patients requiring psychological support may be a barrier to implementation. Solutions could involve training specialized nurses to deliver this specific treatment protocol, or group therapy which would also accommodate patients’ need for peer-support. In addition, further research is needed to identify patient characteristics, such as high levels pain catastrophizing, that may predict favorable treatment outcomes. In this context, motivation to initiate CBT should be taken into account; however, in the present study, both the CBT and control groups demonstrated comparable levels of motivation. Lastly, the timing of the CBT should be investigated further. Notably, should CBT be given together with surgery, or could it be provided quicker after the diagnosis of endometriosis is confirmed. The most important limitation of this study is the lack of a placebo group. The CBT group received more contact time with a healthcare professional than the control group. This could potentially contribute to the differences that we measured between the control group and the intervention group. An attentional control group, which receives the same amount of attention but without the CBT intervention, would have been helpful to mitigate the placebo effect. However, a strong indication that the observed effect on HRQoL can be attributed to the CBT specifically, can be found in the significant mediating effect of CBT on pain catastrophizing, as this was specifically targeted during the CBT intervention. Based on the findings of this study, we cannot conclude that CBT is effective as a stand-alone treatment for endometriosis-related chronic pain, as it was always combined with a surgical intervention. This highlights the need for a large-scale study to further investigate the potential use in this context. Currently, such a study is carried out by our research group. The results of this RCT show that CBT is effective to improve HRQoL in patients who are surgically treated for endometriosis-related chronic pain. The decrease in pain catastrophizing may underlie this effect. Although future studies should investigate the long-term effect of CBT on HRQoL and which patients are most likely to benefit from CBT, we suggest to consider adding CBT to the surgical management of endometriosis.

Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.