A Core Outcome Set for Inflammatory Bowel Diseases: Development and Recommendations for Implementation in Clinical Practice Through an International Multi-stakeholder Consensus Process

Abstract

Background and Aims

Standardising health outcome measurements supports delivery of care and enables data-driven learning systems and secondary data use for research. As part of the Health Outcomes Observatory [H2O] initiative, and building on existing knowledge, a core outcome set [COS] for inflammatory bowel diseases [IBD] was defined through an international modified Delphi method.

Methods

Stakeholders rated 90 variables on a 9-point importance scale twice, allowing score modification based on feedback displayed per stakeholder group. Two consecutive consensus meetings were held to discuss results and formulate recommendations for measurement in clinical practice. Variables scoring 7 or higher by ≥80% of the participants, or based on consensus meeting agreement, were included in the final set.

Results

In total, 136 stakeholders (45 IBD patients [advocates], 74 health care professionals/researchers, 13 industry representatives, and four regulators) from 20 different countries participated. The final set includes 18 case-mix variables, three biomarkers [haemoglobin to detect anaemia, C-reactive protein and faecal calprotectin to detect inflammation] for completeness, and 28 outcomes (including 16 patient-reported outcomes [PROs] and one patient-reported experience). The PRO-2 and IBD-Control questionnaires were recommended to collect disease-specific PROs at every contact with an IBD practitioner, and the Subjective Health Experience model questionnaire, PROMIS Global Health and Self-Efficacy short form, to collect generic PROs annually.

Conclusions

A COS for IBD, including a recommendation for use in clinical practice, was defined. Implementation of this set will start in Vienna, Berlin, Barcelona, Leuven, and Rotterdam, empowering patients to better manage their care. Additional centres will follow worldwide.

Quality of life, socioeconomical and psychological end points

1. Introduction

Inflammatory bowel diseases [IBD], including Crohn’s disease [CD], ulcerative colitis [UC], and inflammatory bowel disease type unclassified [IBDU], mostly affect young adults who require lifelong drug treatment and often even surgery.^1–3 Because of the disease’s chronic character with alternating periods of flares and remission, and the associated complex management, patients with IBD require close monitoring and longitudinal follow-up. Consequently, it is important to strive for the highest possible quality of care [QoC], including stratification and treatment personalisation where possible. Patients with IBD often report decreased health-related quality of life [HRQoL], caused by a high impact of the disease on mental wellbeing, fatigue, and social and professional activities.^4–7

Following the recent shift towards value-based health care [VBHC], the interest in measuring health outcomes increased significantly.⁸ Health outcomes are defined as the effect of the disease, a treatment, or an intervention on the wellbeing of a patient.⁹ Usually, this information is collected as patient-reported outcomes [PROs], clinician-reported outcomes, or a combination thereof. However, significant differences have been found between what health outcomes clinicians and patients consider important to monitor.^10–12 A better mutual understanding between patients and clinicians is suggested to directly improve health outcomes as well as patients’ experience in clinical care. Routinely measuring PROs can help identify critical unmet needs and support the dialogue and joint decision making between patient and health care professional [HCP], and therefore help drive towards improved and more personalised care. Furthermore, it has been shown that close monitoring of disease activity and HRQoL leads to improved remission rates.^13–15

In 2012, the International Consortium for Health Outcomes Measurement [ICHOM], a non-profit organisation, was established with the goal of promoting the adoption of VBHC through the development of standardised, patient-centred outcome sets.¹⁶ To date, the International Consortium for Health Outcomes Measurement [ICHOM] has developed nearly 50 core outcome sets [COS] for different diseases, including one for IBD which was published in 2018.^11,17 Besides clinical outcomes, the ICHOM set for IBD also includes case-mix variables. These prognostic factors can be understood as characteristics that identify patient subgroups that are likely to have different outcomes, making them important parameters for risk adjustment, and to contextualise outcome data for clinical decision making and research.¹⁸ In addition to the ICHOM set, outcome sets for use in real-world studies, clinical trials, and for more specific IBD subgroups, exist.^19–22

Besides the clear value for patients and HCPs, standardised health outcome measurement facilitates large-scale research through data interoperability and facilitates QoC improvement projects. The pharmaceutical industry integrates PROs into the development of new drugs in clinical trials.²³ Also health regulators can use outcome measurements to guide decisions to direct health care expenditures.^24,25 Given the broad utility of health outcomes data, but different priorities across stakeholder groups, it is important to involve additional stakeholder groups besides HCPs and patients in the development process of a COS.^26–30 Involvement of researchers, industry representatives, and health regulators could bring insights on which health outcomes should be collected for research, in clinical trials and in national health care system initiatives. Moreover, the involvement of a wide variety of relevant stakeholders in the development of a COS will most likely increase its wider acceptability and multi-purpose implementation, driving standardisation, improved treatment, and care for patients overall.

This study was part of the Innovative Medicines Initiative [IMI]-endorsed Health Outcomes Observatories [H2O] project, designed to drive VBHC in Europe. H2O aims to improve the sustainability of health care systems, and to optimise care delivery and secondary use of health outcomes data in a clinical research context by supporting the collection of outcomes that matter most to patients.³¹

First, we aimed to define 'what to measure' by building on existing initiatives and knowledge, presented as a COS supplemented by biomarkers and case-mix variables for contextualisation [later all referred to as ‘items’], that was developed through an international modified Delphi consensus process. Second, we aimed to formulate a recommendation on 'how to measure' the PROs included in the COS, referring to appropriate instruments and measurement frequencies.

2. Materials and methods

A modified Delphi method, more specifically a modified RAND/UCLA appropriateness method, was used to develop a COS for IBD [Figure 1].²⁶

Figure 1.

Flowchart of the process used to develop the core outcome set.

2.1. Patient involvement

Patients were closely involved from start to finish of the project, and additional efforts were implemented to increase patient engagement. A patient advisory board [PAB] was composed consisting of six IBD patients [advocates] to consult and provide guidance throughout the study. Group discussions and focus groups were organised to collect specific inputs. The PAB was invited to attend all decisive consensus meetings. Four PAB members also settled in the core team, next to other relevant stakeholders, preparing and coordinating the modified Delphi process. Additional IBD patients [advocates] were invited to participate in the Delphi exercise.

2.2. Formation of a core team

A multi-stakeholder core team was established to prepare and coordinate the modified Delphi process, consisting of four IBD patients [advocates], one chair [author MF, gastroenterologist specialised in IBD], nine additional IBD clinicians, three industry representatives, a methodologist with expertise in outcome research and Delphi studies, and three academic researchers [Supplementary Table 1]. The core team was composed based on geographical and stakeholder representation, skills, and experience relevant to the disease or process.

2.3. Preparation of a preliminary list of items

PubMed and other specialised databases of the consensus-based standards for the selection of health measurement instruments [COSMIN] and Core Outcome Measures in Effectiveness Trials [COMET] initiatives were screened in January, 2021, by authors LF and NC, to identify review papers on outcomes for IBD and existing COS for IBD [Supplementary Table 2]. Articles providing an overview of existing case-mix variables, objective markers, or outcomes for patients with IBD were selected for further analysis. Objective markers were incorporated in the process for completeness of the set. Disagreements were discussed until consensus was reached. Biomarkers, case-mix variables, and outcomes were extracted from the selected articles and supplemented by outcomes that were already being collected by core team clinicians in clinical practice, to initiate a preliminary list.

Core team members reviewed and scored the identified items on a 1–9-labelled Likert scale, representing ascending importance. The preliminary list was then refined based on the given scores and feedback: several items were redefined or merged, doubles and items of low importance were removed. Each retained item was categorised as a case-mix variable, biomarker, or outcome [patient- and/or clinician-reported] and was given a definition to avoid misinterpretation. The refined preliminary list was agreed on during a core team meeting in October, 2021, as the preliminary list of items to feed into the broader Delphi scoring rounds.

Licensed translators provided forward-backward translations of the English preliminary list in three additional languages to use in the Delphi scoring rounds: Spanish, German, and Dutch. Native-speaking H2O consortium members, experienced with IBD health outcomes, reviewed and refined the translations to ensure linguistic validity.

2.4. Open round: adding missing items

During a virtual meeting, the four core team IBD patients [advocates] provided feedback on the relevance and completeness of the initial list, to ensure inclusion of items that matter most to patients. The other core team members and three additional industry representatives thoroughly reviewed the preliminary list for completeness and suggested missing items, which could be added based on core team assessment.

2.5. Two consecutive Delphi scoring rounds

A broader, multi-stakeholder group of experts was invited through ‘snowball sampling’ to participate in the Delphi scoring rounds, including additional IBD patients [advocates], HCPs with expertise in IBD [gastroenterologists, nurses, abdominal surgeons, psychologists, dieticians, rheumatologists, and dermatologists], [academic] researchers, industry representatives, and health regulators. The principles of geographical representation, diversity, equity, and inclusiveness [gender, sociodemographic, geographical, cultural background, and ethnicity] were applied to the extent feasible.

The Delphi participants were provided the preliminary list of items, including any added items from the open round, and asked to score the importance of each item on a 1–9-labelled Likert scale, representing ascending importance. Participants could indicate 'unable to rate' when uncomfortable with rating an item. At the end of the survey, participants could suggest additional items that were missing from the provided list. Again, suggested items could be added to the list, based on core team assessment.

In a second scoring round, participants were provided with the results of the first scoring round, reported anonymously and aggregated by stakeholder group per item in respective bar charts. Based on this information, participants were asked to re-evaluate their scores given in the first round, intending to grow towards a consensus.

2.6. Data analysis and cut-off criterion

The DelphiManager 5.0 platform and software, developed and maintained by COMET [http://www.comet-initiative.org/delphimanager/], were used to coordinate the two consecutive Delphi scoring rounds.

The statistical program RStudio PBC [version 2021.09.2] was used to analyse the scores and to create graphical output.³²

A consensus threshold of ≥80% of the participants scoring an item 7 or higher on the importance scale in the second round was applied as the cut-off criterion to include an item in the final set. Items that received a score of 7 or higher on the importance scale by 70–79.99% of the participants in the second scoring round were further discussed during a consensus meeting. Items that did not make the ≥80% cut-off but scored remarkably high in the patient stakeholder group in the second scoring round [score of 7 or higher on the importance scale by ≥80% of the patient participants] were marked as ‘patient critical’ and also further discussed during the consensus meeting, with patients [advocates] present.

2.7. Consensus meeting

To ensure inclusion of the ‘items that matter most to patients’, the four core team IBD patients [advocates], one additional IBD patient and one additional IBD patient advocate of the PAB were provided with the consolidated results of the second scoring round. Based on this information, the patients [advocates] indicated specific items to discuss during the consensus meeting. Likewise, the other core team members, provided with the same result report, could indicate specific items to discuss during the consensus meeting.

A virtual consensus meeting was organised in March, 2022. Items that narrowly missed the inclusion threshold, patient critical items, and items specifically suggested by patients [advocates] or other core team members were discussed. The core team was enriched with interested Delphi participants, to ensure representation of each stakeholder group. The core team chair moderated, highlighted areas of ambiguity for clarification, and encouraged equal participation from all stakeholders during the consensus meeting.

Discussed items that were agreed to include in the final set were re-scored by the consensus meeting participants on the same 9-point importance scale, to measure the final level of agreement. Potential adaptations of selected items were supported by appropriate rationale according to the consensus meeting discussion. The agreed selection of items to include in the core set and potential adaptations were shared within the core team members for final feedback, and the set was finalised.

2.8. Selection of recommended instruments and measurement frequency for the core outcome set

Next, potential instruments to collect the included PROs in the COS, or patient-reported outcome measures [PROMs], were identified through literature by author LF. The core team was asked to review this list and suggest missing instruments. The literature report included details on the development process, validation studies, extensiveness, required licenses, and available translations of the identified instruments. Provided with this report, the core team and additional interested Delphi participants to ensure representation of each stakeholder group, were asked in an online survey to indicate which instrument[s] and minimum frequency they considered appropriate to collect the PROs included in the COS. The collected insights were discussed during a second online consensus meeting in September, 2022, where a final selection of instruments and minimum measurement frequency were agreed on.

2.9. Ethical statement

This study was approved on January 7, 2022, by the Ethics Committee for Research of the University Hospitals Leuven/KU Leuven.

3. Results

3.1. Preliminary list of items

In total, 12 articles were identified through COMET, nine additional articles through COSMIN, and 187 additional articles through a PubMed database search, resulting in 208 unique articles. After screening for eligibility based on title and abstract, 22 articles were selected for further analysis to identify existing case-mix variables, biomarkers, and outcomes for IBD. In total, 410 items were extracted from the 22 selected articles and current clinical practice. Articles not providing an overview of existing case-mix variables, biomarkers, or outcomes for patients with IBD, were excluded.

After removing doubles and refinement based on feedback and importance scores given by the core team members, 85 items including 18 case-mix variables, three biomarkers and 31 PROs were retained to be included in the preliminary list [Figure 2].

Figure 2.

Results of the literature review to prepare the preliminary list of items. COMET: Core Outcome Measure in Effectiveness Trials; COS: core outcome set; COSMIN: consensus-based standards for the selection of health measurement instruments; IBD: inflammatory bowel disease; PRO: patient-reported outcome.

3.2. Open round: added missing items

Five additionally suggested missing items were added to the preliminary list: 'Ethnicity', 'Subjective overall wellbeing/HRQoL', 'Social support', 'Financial impact', and 'Coping and acceptance of the disease', resulting in 90 items to feed into the Delphi scoring rounds [Supplementary Table 3].

3.3. Results of the two Delphi scoring rounds

In total, 179 stakeholders participated in the first scoring round (64 IBD patients [advocates], 90 HCPs/researchers, 20 industry representatives, and five health regulators), of whom 136 stakeholders also participated in the second scoring round (participation ratio of 76.0%, 45 IBD patients [advocates], 74 HCPs, 13 industry representatives and four health regulators; Supplementary Table 4).

The results of the first and second scoring round can be found in Supplementary Table 3. No additional items were included after the first scoring round. In total, 34 items received an importance score of 7 or higher by ≥80% of the participants in the second scoring round, and were immediately included in the final set. Thirteen items that received an importance score of 7 or higher by 70–79.99% of the participants, and three patient critical items (receiving an importance score of 7 or higher by ≥80% of the patient [advocate] participants) were identified for further discussion during the consensus meeting.

3.4. Consensus meeting

In total, 17 members of the core team (four IBD patients [advocates], eight IBD clinicians, three academic researchers, two industry representatives), four additional IBD patients [advocates], one additional IBD clinician, two IBD nurses, one additional academic researcher, two additional industry representatives, and two health regulators attended the online consensus meeting.

In addition to the 13 items that narrowly missed the inclusion threshold and three patient critical items, seven more items indicated by core team patients [advocates], and 16 by core team clinicians were discussed during the consensus meeting [Supplementary Table 3]. It was agreed to additionally include 20 out of the 39 discussed items in the final core set. Further, it was agreed to include the presence of extraintestinal manifestations as both a patient- and as a clinician-reported measure, to include both day and night bowel incontinence, and to include symptoms of both draining and intestinal or perianal fistulae.

The final COS [Table 1] includes 18 case-mix variables, three biomarkers (haemoglobin to detect anaemia, and C-reactive protein [CRP] and faecal calprotectin to detect inflammation), and 28 outcomes.

Table 1

Items included in the H2O core outcome set for IBD

Category	Item	Description	Percentage of participants with an importance score of 7, 8, or 9	Origin of the percentage
Case-mix variables—demographics	Agea	Date of birth [DD/MM/YYYY]	71.43b	After consensus meeting
Case-mix variables—demographics	Sexa	Sex at birth [male/female/diverse/prefer not to say]	80.95	After consensus meeting
Case-mix variables—demographics	Smoking statusa	Current smoking status of cigarettes, cigars, or tobacco [never/ex/current/prefer not to say]	90.48	After consensus meeting
Case-mix variables—demographics	Body weighta	Body weight [kg] to calculate loss of weight	80.95	After consensus meeting
Case-mix variables—baseline clinical factors	Malnutrition	The presence of significant or persistent malnutrition [a condition caused by not getting enough calories in or the right food or substances, such as vitamins and minerals, that are needed for good health]	83.47	Scoring round 2
Case-mix variables—baseline clinical factors	Comorbidities including immune-mediated inflammatory conditions	Presence of any other disease including immune-mediated inflammatory diseases [IMIDs] such as rheumatoid arthritis [inflammatory condition affecting joints], spondyloarthropathy [type of arthritis, inflammatory condition affecting joints] or psoriasis [inflammatory condition affecting skin]	100.00	After consensus meeting
Case-mix variables—baseline clinical factors	Vaccination status	Current vaccination status of the recommended vaccinations for patients with inflammatory bowel disease: vaccine for influenza, pneumococci, hepatitis A, hepatitis B, tetanus, human papillomavirus and COVID19	85.71	After consensus meeting
Case-mix variables—baseline clinical factors	Ostomy presence	Presence of an ostomy and date of placement	89.57	Scoring round 2
Case-mix variables—baseline condition factors	IBD diagnosis*	Diagnosis of inflammatory bowel disease: Crohn’s disease [CD], ulcerative colitis [UC], or inflammatory bowel disease unclassified [IBDU] or other	86.51	Scoring round 2
Case-mix variables—baseline condition factors	Date of diagnosis	Date of diagnosis inflammatory bowel disease	85.71	After consensus meeting
Case-mix variables—baseline condition factors	Location of disease activitya	The location of disease activity tracked via the Montreal Classification	90.55	Scoring round 2
Case-mix variables—baseline condition factors	Extraintestinal manifestationsa	The presence of arthritis [disease causing painful inflammation and stiffness of the joints], erythema nodosum [acute and painful red subcutaneous swelling], or pyoderma gangrenosum [large, painful ulcers on the skin], primary sclerosing cholangitis [PSC, inflammation causing scars within the bile ducts], uveitis or episcleritis [external eyeball inflammation], aphthous ulcers [small, shallow lesions that develop on the soft tissues in mouth or at the base of gums], or other extraintestinal manifestations	90.48	After consensus meeting
		The presence of joint, skin or eye problems	85.71	After consensus meeting
Case-mix variables—treatment factors	Fistula surgery	The performance of a surgical procedure to remove an abnormal connection between two organs or vessels that do not usually connect [e.g. intestines or anal channel and skin] and date of intervention	92.24	Scoring round 2
Case-mix variables—treatment factors	Strictureplasty surgery	The performance of a surgical procedure without removing part of the bowel to alleviate bowel narrowing due to scar tissue that has built up in the intestinal wall, and date of intervention	80.36	Scoring round 2
Case-mix variables—treatment factors	Surgical bowel resection	The performance of a surgical procedure to remove a part of the bowel, the reason why, and date of intervention	95.58	Scoring round 2
Case-mix variables—treatment factors	Abscess drainage	The performance of a surgical procedure to drain an abscess, and date of intervention	88.60	Scoring round 2
Case-mix variables—treatment factors	Stenosis endoscopic dilatation	The performance of an endoscopic procedure to restore a narrowing in the intestines, and date of intervention	83.78	Scoring round 2
Case-mix variables—treatment factors	Medication	Current medication use [IBD and non-IBD specific], dose and duration	83.74	Scoring round 2
Biomarker	Haemoglobin [Hb]	Amount of haemoglobin [Hb] detected in blood, biomarker to detect anaemiaa (anaemia is defined by the WHO as a haemoglobin [at sea level] of <12.0 g/dl in non‐pregnant women, <11.0g/dl in pregnant women, and <13.5 g/dl in men)	85.71	After consensus meeting
Biomarker	C-reactive protein [CRP]	Amount of C-reactive protein [CRP] detected in blood, biomarker to detect inflammation	90.48	After consensus meeting
Biomarker	Faecal calprotectin [fcal]	Amount of faecal calprotectin [fcal] detected in stool, biomarker to detect intestinal inflammation	85.09	Scoring round 2
Outcomes—symptoms, function, and quality of life	Abdominal pain	Patient-reported: unpleasant physical sensation in the abdominal region, which may vary in intensity from mild discomfort to pain that limits activities of daily life, limits self-care, and/or requires medication or hospitalisation	92.00	Scoring round 2
Outcomes—symptoms, function, and quality of life	Diarrhoeaa	Patient-reported: presence of liquid or very soft stools	95.00	After consensus meeting
Outcomes—symptoms, function, and quality of life	Stool frequencya	Patient-reported: the amount of [liquid or very soft] stools per day to measure increase/decrease in number of [liquid or very soft] stools	95.24	After consensus meeting
Outcomes—symptoms, function, and quality of life	Rectal bleedinga	Patient-reported: the presence of blood or stripes of blood in stool, or stool in form of only blood	89.34	Scoring round 2
Outcomes—symptoms, function, and quality of life	Perianal disease symptoms	Patient-reported: symptoms caused by perianal disease such as the presence of pain, abscess, pus, or fistula around the anus [small abnormal tunnel between the end of the bowel and the skin near the anus]	90.76	Scoring round 2
Outcomes—symptoms, function, and quality of life	Ostomy symptomsa	Patient-reported: symptoms caused by an ostomy such as ostomy leakage, nightly emptying, or frequent daily emptying for patients with an ostomy [hole made by surgery to allow stool to leave the body through the belly]	88.18	Scoring round 2
Outcomes—symptoms, function, and quality of life	Bowel urgency [tenesmus]	Patient-reported: presence of urgency with a deferral time under 15 min, e.g. caused by rapid bowel movement after eating	87.90	Scoring round 2
Outcomes—symptoms, function, and quality of life	Bowel incontinence	Patient-reported: lack of voluntary control over stool production during daytime	93.55	Scoring round 2
		Lack of voluntary control over stool production during night-time	95.12	Scoring round 2
Outcomes—symptoms, function, and quality of life	Everyday lifea	Patient-reported: ability to carry out everyday physical activities such as walking, climbing stairs, carrying groceries, or moving a chair, and usual social activities and roles The occurrence of missing planned activities because of IBD	85.00	After consensus meeting
Outcomes—symptoms, function, and quality of life	Fatigue	Patient-reported: level of significant or persistent tiredness that is not proportional to recent physical activity	90.00	After consensus meeting
Outcomes—symptoms, function, and quality of life	Need for psychological supporta	Patient-reported: one requires psychological support, and the cause [IBD-related or not]	76.19b	After consensus meeting
Outcomes—symptoms, function, and quality of life	Having control over disease activity	Patient-reported: level of having overall control over disease activity	92.06	Scoring round 2
Outcomes—symptoms, function, and quality of life	Coping and acceptance of the disease	Patient-reported: level to which one is coping and accepting to live with the disease	80.95	After consensus meeting
Outcomes—symptoms, function, and quality of life	Self-management and autonomy	Patient-reported: level to which one feels confident in managing their disease, such as knowing what to do and how to act in certain situations, e.g. when to contact the IBD care team	85.00	After consensus meeting
Outcomes—symptoms, function, and quality of life	Feeling informed	Patient-reported: level to which one feels informed enough to manage their disease, such as skills acquired through experience or education, received factual information, theoretical and practical understanding of disease	66.67b	After consensus meeting
Outcomes—symptoms, function, and quality of life	Subjective overall wellbeing/health-related quality of life	Patient-reported: perceived overall quality of life	82.40	Scoring round 2
Outcomes—disutility of care	Prednisone use	The use of prednisolone [steroid] or equivalent, dose, and duration	95.24	After consensus meeting
Outcomes—health care utilisation	Hospitalisation	Number of days hospitalised per year, cause [IBD-related or not] and the length of stay	88.71	Scoring round 2
Outcomes—health care utilisation	Emergency unit admissions	Number of admissions to the emergency unit per year and cause [IBD-related or not]	86.18	Scoring round 2
Outcomes—survival and disease control	Behaviour of a new intestinal stenosis	Stricturing [abnormal narrowing] behaviour of a new intestinal stenosis [narrowing of the intestines that makes it difficult to pass stool] tracked with the Montreal classification	85.71	Scoring round 2
Outcomes—survival and disease control	Perianal abscess	Behaviour of a new superficial infection that appears as a tender red lump under the skin near the anus	92.37	Scoring round 2
Outcomes—survival and disease control	[Draining] fistula	Behaviour of new intestinal or perianal fistula [abnormal connection or passageway that connects two organs or vessels that do not usually connect, e.g. intestines and skin or anal channel and skin]	91.38	Scoring round 2
		Behaviour of new draining fistula [when the drainage of pus leads to an abnormal connection between two organs or vessels that do not usually connect, e.g. intestines and skin or anal channel and skin]	89.66	Scoring round 2
Outcomes—survival and disease control	Colorectal cancer	The presence of cancer in colon [longest part of the large intestine] or rectum [last part of the large intestine closest to the anus] and date of diagnosis	98.29	Scoring round 2
Outcomes—survival and disease control	Extraintestinal malignancies	The presence of a malignant tumour outside of intestines and date of diagnosis	96.55	Scoring round 2
Outcomes—survival and disease control	Mortality	Death, date of death, cause [IBD-related or not], and location [in or outside of hospital]	90.52	Scoring round 2
Outcomes—treatment related	Treatment-related or treatment-emergent adverse events	Unexpected medical problem related to treatment such as [skin] infections, leukopenia [low white blood cell count], hepatotoxicity [liver damage], nephrotoxicity [kidney damage], pancreatitis [inflammation of the pancreas], steroid-related side effects, infusion reactions, anaphylaxis [severe allergic reaction] and other	93.50	Scoring round 2
Outcomes—treatment related	IBD medication adherencea	Level to which one is compliant with taking the prescribed IBD medication	83.61	Scoring round 2
Outcomes—patient-reported experience	Therapy satisfaction	Level to which one is satisfied with therapy such as prescribed medication and recommended diet	82.40	Scoring round 2

The percentage of participants with an importance score of 7, 8, or 9 represents the percentage of participants scoring the importance of the item 7, 8, or 9 on an ascending 9-point Likert scale in the second scoring round or after the consensus meeting.

H2O: Health Outcomes Observatory; IBD: inflammatory bowel disease.

^aItems that were adapted based on the consensus meeting discussion and agreement.

^bItems that did not reach the cut off criterion of ≥80% of the participants scoring the importance of the item 7, 8, or 9 on an ascending 9-point Likert scale in the second scoring round or after the consensus meeting, but were included in the final set based on core team agreement.

3.5. Selection of recommended instruments and measurement frequency for the core outcome set

Eleven instruments [of which one was additionally suggested by a core team clinician; Supplementary Table 5] were identified for potential collection of the PROs included in the COS and presented in an online survey, to gain insights on the most appropriate instrument[s] and measurement frequencies.^33–43 In total, 18 stakeholders from ten different countries completed the survey (two core team and two additional patients [advocates], five core team and two additional IBD clinicians, three industry representatives, two IBD nurses, one expert in PRO measurements, and one health regulator).

Equally, 18 stakeholders attended the second consensus meeting (three core team and two additional IBD patients [advocates], four core team and two additional IBD clinicians, two core team and one additional academic researcher, three industry representatives and one IBD nurse, from 11 different countries). The PRO-2 questionnaires [both for CD and UC]^33,34 and the IBD-Control Questionnaire³⁹ were agreed to be the most appropriate instruments to collect the IBD-specific PROs, and the PROMIS Global Health⁴² and PROMIS Self-Efficacy short form for chronic symptoms⁴³ to collect the generic PROs. It was additionally agreed to use the generic Subjective Health Experience [SHE] model questionnaire to measure disease acceptance and control⁴⁴ and to supplement the set with additional self-standing questions to measure bowel incontinence, bowel urgency, the extent to which patients feel informed, IBD medication adherence, and perianal disease and ostomy symptoms if applicable [Figure 3; and Supplementary Table 6]. This instrument set and additional questions provide full coverage of all PROs included in the COS. Last, it was agreed that the IBD-specific PROs should be collected at every contact with an IBD practitioner and minimally every 12 months, and generic PROs minimally every 12 months.

Figure 3.

Recommended instruments and measurement frequencies to collect the patient-reported outcomes included in the final core outcome set. The number of questions is indicated in brackets; questions in grey are only recommended if applicable, namely only for IBD patients who are currently prescribed IBD medication, IBD patients with perianal disease, and IBD patients with an ostomy. CD: Crohn's disease; IBD: inflammatory bowel diseases; SHE: Subjective Health Experience; UC: ulcerative colitis.

4. Discussion

In this study, 18 case-mix variables [four demographics, four baseline clinical factors, four baseline condition factors, six treatment factors], three biomarkers [haemoglobin, C-reactive protein, and faecal calprotectin], and 28 outcomes [16 PROs related to symptoms, function and HRQoL, one patient-reported experience, one outcome related to disutility of care, two related to health care use, six to survival and disease control, and two related to treatment] were agreed to be collected as a minimum among patients with IBD. The PRO-2 and IBD-Control questionnaires were identified as the most appropriate instruments to collect the IBD-specific PROs included in this set. It was agreed to collect these IBD-specific PROs at every contact with an IBD practitioner, and at least every 12 months. The SHE model questionnaire, PROMIS Global Health questionnaire, and Self-Efficacy short form were identified as the most appropriate instruments to collect the generic PROs included in this set, which were agreed to be collected every 12 months at a minimum.

It will take patients approximately five minutes to complete either set of questions, the full set of IBD-specific questions or the full set of generic questions. An Israeli research group recently demonstrated high feasibility of a similar questionnaire set [IBD-Control + PROMIS Global Health questionnaire], reaching up to 57% participation rate during follow-up, without any direct patient-clinician interaction.⁴⁵ These findings suggest even higher questionnaire compliance if the digital PRO collection could be integrated into routine clinical practice, as is intended in H2O.

In this current study, the core team agreed not to repeat previous work but to build on existing initiatives, align, re-evaluate, and supplement with new insights where possible. Many of our findings are in line with the previous work that was conducted by the ICHOM working group,¹¹ underlining the high reliability of this work. In addition, the present study also brought some new insights [Supplementary Table 7]. Where ‘education level’, ‘occurrence of previous infections’, and ‘participation of patients in a colorectal cancer surveillance programme’ were not included in this final set, ‘vaccination status’, ‘malnutrition’, and ‘ostomy presence’ were newly identified case-mix variables in addition to those previously identified in the ICHOM paper.¹¹ Also biomarkers ‘faecal calprotectin’, ‘CRP’, and several clinical outcomes were newly identified, namely ‘extraintestinal malignancies’, and outcomes for more specific conditions: ‘behaviour of a new intestinal stenosis’ and ‘behaviour of a perianal abscess’, both tracked with the Montreal classification. Important newly identified symptom-related PROs are ‘diarrhoea’, ‘stool frequency’, ‘rectal bleeding’, ‘perianal disease symptoms’, ‘ostomy symptoms’, ‘day- and night-time incontinence’, and ‘bowel urgency’. Newly identified PROs related to function and HRQoL are ‘subjective overall wellbeing or HRQoL’, ‘self-management and autonomy’, the extent to which patients ‘feel informed’, and ‘coping and acceptance of the disease’. Also ‘treatment-related or treatment-emergent adverse events’, ‘IBD medication adherence’, and ‘therapy satisfaction’ were found important in this current study. Similar to ICHOM, this study also indicated the IBD-Control questionnaire as the most appropriate instrument to measure the included HRQoL PROs.³⁹ To collect patient-reported information about symptoms, instead of the Manitoba Index, the PRO-2 questionnaires for CD and UC were indicated to be the most appropriate.^33,34 As the PRO-2 questionnaires collect detailed information about stool frequency, the presence of diarrhoea, abdominal pain, and rectal bleeding, they can indicate whether disease is active or if a patient is in clinical remission. This information can be used in research and clinical trials, and also by HCPs to support clinical decision making. Additionally, also three appropriate generic instruments were identified in this current study.^42–44 The collection of generic PROs and the use of generic instruments allows comparison of health outcomes across disease settings, and is important for the scalability of H2O.

Thus, several important PROs were newly identified to describe patient symptoms, function, and HRQoL. This can potentially be explained by the consistent patient involvement throughout the current study from start to finish, which is how this initiative distinguishes itself from others. IBD patients [advocates] were not only largely involved in the scoring rounds, namely 33% of the total amount of Delphi participants, but also consistently represented and consulted at every other step of the process, for example by accounting for 19% of the core team members, identifying which outcomes to discuss, and actively participating in consensus meetings. Whereas we did not include wider patient engagement e.g. through focus groups with a broader patient sample or by organising one-on-one interviews, a follow-up study will evaluate the feasibility of the COS in implementing clinics and will collect feedback from patients and HCPs, to continually improve and adjust the set as needed for patients and their care teams. Nevertheless, the consistent involvement of IBD patient experts in the development of the H2O set allowed us to develop a truly patient-centred COS, which will support the dialogue and joint decision making between patient and HCP, and drive towards improved and more personalised care overall. In addition to HCPs and the close contribution of patients in this project, industry representatives and health regulators were also closely involved. The inclusion of a wide variety of stakeholders enriched the study results and validity of the defined COS.

On the other hand, this study also includes some limitations. The COS includes 16 PROs and one patient-reported experience. By deciding to use existing PROMs in their fully validated form, additional outcomes, beyond the 16 PROs and one patient-reported experience agreed in the COS, will automatically be collected. The downside of using full instruments to preserve their psychometric properties is that the question burden for patients increases. Therefore, it was agreed during the second consensus meeting to collect generic PROs only once a year. IBD-specific PROs, on the other hand, were agreed to collect at every contact with an IBD practitioner. By doing so, each question set will include approximately 20 questions and take patients approximately five minutes each to complete. Another limitation is that we registered a noticeable drop in participants moving on from Delphi round 1 to 2 [participation ratio of 76.0%]. Although we purposely waited to launch the first Delphi round until after the end of year holidays to increase participation, the COVID-19 omicron wave in the beginning of 2022 may have influenced participation. Additionally, only four health regulators participated in both Delphi scoring rounds, resulting in poor representation of this stakeholder group. However, in other studies, similar numbers of participating health regulators or policy makers have been reported, if any.^46,47 To further improve representation of this stakeholder group, the participation of two health regulators was ensured during the consensus meeting and their active participation was highly encouraged.

In previous initiatives, the limited multi-stakeholder involvement, or the approach to patient inclusion in the development process of previous COS, might have affected their broad uptake and implementation. In this project we brought together and built on previous initiatives, resulting in a COS that aligns with existing work. We closely involved patients and several other relevant stakeholder groups, including key opinion leaders, to increase the acceptability and support wide and multi-purpose adoption of the COS. A broad implementation of the H2O COS in clinical practice and in clinical studies will improve treatment development, support policy making decisions, and facilitate multicentre research and [inter]national QoC improvement projects, but also help drive towards improved and more personalised care, support the dialogue and shared decision making between patient and HCP, and care for patients overall. A number of studies report the impact of routine PRO collection on costs, outcomes, acceptance, and satisfaction of patients.^48–52 However, there is a great need for more research in this area. Wong et al. previously demonstrated feasibility of implementing the ICHOM COS for IBD in clinical practice, with little burden on HCPs. In addition, the implementation of the COS and collection of predefined PROs allowed the working group to identify critical target points in current care which are important to patients.⁵³ Although the demonstrated feasibility of implementing a COS is a substantial first step towards value-based delivery of IBD care, subsequent impact studies reporting the added value of implementing a COS in practice should follow. Once the COS described in the current study is successfully implemented with proven feasibility, an impact study describing the effects on health care recourse use, costs, and outcomes will follow. This impact study will show the added value and help fill this literature gap that currently exists.

Within the framework of the H2O initiative, a similar exercise was conducted for diabetes, metastatic breast and lung cancer.^54–56 The methodology to develop the COS was aligned across disease areas, and synergies were found among the three disease areas.³¹ Data integration and comparison of health outcomes between the different disease settings will further be explored in a next phase of the project.

The COS and recommendation for use in practice are designed with flexibility for implementation and data interoperability, as a COS should be fine-tuned and supplemented according to local needs and preferences.⁵⁷ Some centres might for example benefit, or even be governmentally obliged, to use a preference-based instrument that allows cost-utility analysis. In practice this may translate into having several local sets that extend the core outcome domains, while retaining standardisation and comparability. Another probability is that different centres will collect the same outcome domains using various instruments, requiring the development of a common metric model.⁵⁸

To support and facilitate implementation of the full H2O COS, a detailed data dictionary can be obtained through the H2O website [https://health-outcomes-observatory.eu/], where also HCP and patient recruitment material can be found. The H2O COS for IBD is currently being implemented in university hospitals in Vienna, Berlin, Barcelona, Leuven, and Rotterdam. This coordinated implementation will determine the feasibility of the COS and may stimulate additional centres to follow worldwide.

To date there is no uniform and standardised data collection across IBD centres. In the context of the H2O initiative, and in accordance with a harmonised H2O methodology, by building on existing initiatives and reaching multi-stakeholder consensus including consistent involvement of patients [advocates], a COS was defined for IBD that will empower patients to better manage their health care. The multi-stakeholder approach in this project will support broad acceptance and utility of the work. The currently ongoing implementation of the COS will demonstrate its feasibility and impact, stimulating broader implementation worldwide. Furthermore, uniform and standardised data collection could improve the sustainability of health care systems and optimise care delivery and secondary use of health outcomes data in a clinical research context. The defined COS should be interpreted as the minimum data to collect for patients with IBD, with flexibility to fine-tune and supplement according to local needs and preferences. In order to be sustainable, the COS will require revision and optimisation over time according to new findings and improved care in the field of IBD.

Supplementary Data

Supplementary data are available at ECCO-JCC online.

Funding

This work was supported by the Innovative Medicines Initiative [IMI] 2 Joint Undertaking [945345-2]. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme, the European Federation of Pharmaceutical Industries and Associations [EFPIA], Trial Nation and the Juvenile Diabetes Research Foundation [JDRF]. The public grant funding is matched with in-kind contributions of EFPIA partners. About IMI: the Innovative Medicines Initiative is a partnership between the European Union and the European pharmaceutical industry, represented by EFPIA. It is working to improve health by speeding up the next generation of medicines, particularly in areas with unmet medical or social needs. It works by facilitating collaboration between the key players involved in health research, including universities, research centres, pharmaceutical and other industries, small and medium-sized enterprises [SMEs], patient organisations, and medicines regulators. IMI is the world’s most extensive public-private partnership [PPP] in the life sciences. This manuscript reflects only the authors’ views. The European Union and IMI are not responsible for any use that may be made of the information it contains.

Conflicts of Interest

NC and IC are employees and shareholders of F. Hoffmann-La Roche. GN received lecture fees/consultant fees/advisory board member fees from AbbVie, Merck Sharp & Dohme, Takeda, Janssen-Cilag, Pfizer, Gilead, Galapagos, Ferring, Vifor, AstroPharma, Falk, Bristol Myers Squibb, Arena Pharmaceuticals, and Amgen. CJvdW has served on advisory boards for Abbvie, Takeda, Pfizer, Galapagos, and Celltrion; she is supported by research funding from ZonMW, Tramedico, Falk Benelux, and Pfizer. BS served as a consultant for AbbVie, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Galapagos, Gilead, Janssen, Lilly, Pfizer, PredictImmune, and Takeda; and received speaker’s fees from AbbVie, BMS, CED Service GmbH, Falk, Galapagos, Ferring, Janssen, Pfizer, and Takeda [served as representative of the Charité]. FC received research support from AbbVie, Ferring, MSD, Shire, Takeda, and Zambon; speaker fees from AbbVie, Chiesi, Ferring, Gebro, MSD, Shire, Takeda, and Zambon. NB received research support from AbbVie, Janssen, Takeda, Adacyte, Actial; speaker’s fees from AbbVie, Adacyte, Janssen-Cilag, MSD, Pfizer, and Takeda; and consultancy fees from Janssen-Cilag and Abbvie. ES has received speaker’s fees from Biogen, Takeda, BMS, AbbVie, and Janssen; and consultancy fees from Janssen. CH received speaker’s fees from Takeda, Ferring, AbbVie, and Janssen; and consultancy fees from Pfizer; she has acted as local principal investigator for clinical trials for Janssen and GlaxoSmithKline; she is PI on projects at the Karolinska Institutet partly funded by investigator-initiated grants from Takeda and Tillotts. TS reports grants and personal fees from AbbVie, Roche, Sanofi, Takeda, and Novartis. MJ has served as a consultant for Ferring and Takeda Pharma; and received speaker’s fee from MSD, Ferring, and Takeda Pharma; she is the PI on projects at Aarhus University Hospital partly funded by investigator-initiated grants from Takeda Pharma. GF served as a consultant for AbbVie, Takeda, Janssen, Pfizer, Galapagos, Sandoz, Amgen, and Ferring. EG received consultant fees/advisory board member fees from AbbVie, Boehringer Ingelheim RCV, Celgene, Coloplast, Dr. Falk, Eli Lilly, Ferring, Galapagos Biopharma, Gilead, Janssen-Cilag, Merck, Merck Sharp & Dohme, Pfizer, Stada, and Takeda. MF reports research support from AbbVie, Amgen, Biogen, EG, Janssen, Pfizer, Takeda, and Viatris; speaker’s fees from AbbVie, Amgen, Biogen, Boehringer Ingelheim, Falk, Ferring, Janssen-Cilag, Lamepro, MSD, Pfizer, Sandoz, Takeda, Truvion health care and Viatris; and consultancy fees from AbbVie, AgomAb Therapeutics, Boehringer Ingelheim, Celgene, Celltrion, Eli Lilly, Janssen-Cilag, Medtronic, MSD, Pfizer, Regeneron, Samsung Bioepis, Sandoz, Takeda, and Thermo Fisher. None of these activities have any relation to the present study.

Author Contributions

Authors LF and NC contributed to the concept and design of the study, the acquisition, analysis, and interpretation of data, drafted the manuscript, and approved the final version to be submitted. Authors GN, CvdW, BS, FC, NB, CRHH, CvdW, BS, FC, NB, CRHH, MJ, GF, ASH, ES, NG, CP, SR, and MLLZ contributed to the acquisition, analysis, and interpretation of data, critically revised the manuscript for important intellectual content and approved the final version to be submitted. Authors Evelyn Gross, Cornelia Sander, Ingrid Arijs, Vasiliki-Rafaela Vakouftsi, Tunde Koltai and The Health Outcomes Observatory [H2O] Patient Advisory Board for Inflammatory Bowel Diseases contributed to the acquisition of data, analysis and interpretation of data, critically revised the manuscript for important intellectual content, and approved the final version to be submitted. Authors TS, IC, MF and the Health Outcomes Observatory [H2O] Steering Committee contributed to the concept and design of the study, the acquisition, analysis and interpretation of data, critically revised the manuscript for important intellectual content, and approved the final version to be submitted. Conference presentations: United European Gastroenterology Week, Vienna, 2022 [oral poster presentation]; European Crohn’s and Colitis Organisation Congress, Copenhagen, 2023 [poster presentation]; Belgian Week of Gastroenterology, Antwerp, 2023 [oral presentation]; Digestive Disease Week, Chicago, 2023 [poster presentation].

Acknowledgements

We thank the core team, Patient Advisory Board members and all Delphi and consensus meeting participants for their time invested in this study.