Organisation, planning and training

In April 2021, the BSG appointed a content and field expert as a guideline chair through a competitive process. To meet the best evidence guidance for such leadership, a non-voting GRADE and synthesis methodologist was appointed as co-chair.5 Administrative support will be provided from both host higher education institutions of the co-chairs and access to a Cochrane and NICE expert information specialist will be arranged through these institutions. The chairs proposed their approach to the BSG and editor of the journal Gut and received approval. A timeline for the process was prepared with core milestones.

A plan for a mixed approach to training and support will be made. A planned hierarchy of involvement will consist of the core leadership team and the rest of the GDG. GDG members will support technical reviewing as well as guideline recommendation voting, with others supporting voting alone. The core leadership team will be offered bespoke GRADE training workshops through a collaboration with Professor Schünemann and Dr Miranda Lamgendam at the Department Epidemiology and Data Science, Amsterdam University Medical Centers, University of Amsterdam, Netherlands.

The wider GDG will have additional resources made available supplemented with drop-in sessions to allow bespoke advice and education.

GDG membership

The GDG was formed in September 2021 following a general call, followed by targeted invitations through BSG media channels. This was to ensure a balance of professionals from a variety of disciplines, across both primary and secondary care. These included the following: gastroenterology, gastrointestinal surgery, paediatric gastroenterology, radiology, dermatology, rheumatology, pathology, pharmacy, clinical psychology, dietetics, specialist IBD nursing, charities including Crohn’s & Colitis UK and contributions from service user representatives. A national call out from Crohn’s and Colitis UK facilitated identification of people living with IBD who joined the GDG. The group met online to outline the roles and operating procedures.

The GDG members will be directed to general guidance in line with general guidance on established national processes from National Institute for Health and Care Excellence 2012 (NICE).6 This will confirm that as members of the team, all GDG members agree to be coauthors of the full guideline, to maintain the confidentiality of open discussion and debate within the guideline process, as well as to the confidentiality of the content of the guideline prior to publication.

Members will be asked to declare all conflicts of interest in line with BSG guidelines. The BSG will appoint a conflicts of interest chair to oversee the process.

GDG priority setting and identifying target audience

The GDG were invited to take part in an online Delphi process in February 2022. To manage this, the GDG was split into key areas, as per the 2019 BSG guidelines.1 These subgroups included: diagnostics and classification, ulcerative colitis, CD, special considerations and service delivery. The identity and content of the subgroups will be changed based on the nature of the thematic questions identified and brought forward for evidence synthesis.

The subgroups were given a list of all core thematic topics and specific thematic questions from the previous BSG IBD guidelines.1

The results were analysed to identify topics for the new guidelines based on consideration of two major factors:

• Need for update due to emerging or new evidence not apparent when the last guideline was published.

• Priority for stakeholders for specific clinical or patient factors.

A final consensus list of thematic questions within a ‘patient, intervention, comparison, outcome’ (PICO) format has been produced and agreed in the final phase of the Delphi process (online supplemental file 2). An example PICO is given below in box 1.

Final subgroups will be formed based on the thematic question grouping. Each of the subgroups will have a chair who will represent the subgroup in the core leadership team. They, together with collaborators from within each of the subgroups, will work closely with the core leadership team to produce technical reviews for the thematic questions within their subgroup. These would form the basis for consideration of recommendations within subgroup discussion prior to wider GDG voting. Each subgroup leads will be offered bespoke GRADE training. Both core leadership and subgroup meetings will be held throughout the process.

Discussions were held within the wider GDG, within the CSSC and with the editor of Gut regarding the most appropriate outputs from the process for our target audiences. The following was decided:

• The prospective publishing of a guideline operating procedure and technical summary protocol in an open access journal (this manuscript).

• The publishing of a succinct main guideline that summarises key recommendations, the certainty of underpinning evidence and the strength of the recommendations all within the main published journal output.

• An accompanying larger detailed technical evidence review that will include all the underpinning primary evidence, secondary synthesis quality and analysis data, also published within the main journal output.

• A final user-focused guideline that summarises the key recommendations and the evidence justification of these recommendations in a manner that is informed, coproduced and for the intended use by people living with IBD.

These outputs will offer systematic, high-quality and high utility output for all of our audiences.

PICO thematic question generation

The generation of questions has been guided by the GRADE guidelines.7 As this is an update guideline, the process of prioritisation has been based on the thematic and PICO questions from the previous guideline. This allows the process to focus on generation of novel questions in areas that have emerged and refinement of existing questions.

The generation of new questions is a key component of the Delphi prioritisation, asking for any new or uncovered areas to be presented in free text, with justification of the specific question. Analysis of these results has allowed new candidate questions to be considered by each of the GDG subgroups. With new questions added in a standard PICO format.

Key areas of focus for refinement of all questions have been considered by the subgroups, guided by the framing question guidance.7 These core elements of refinement around PICO questions and their specific application have been presented in draft form to the GDG and all feedback considered, with the final list below:

• UK regulatory approval is mandatory for any pharmacological or similar intervention to be considered. However, UK NICE approval is not considered.

• Multiple treatment arms will be considered. To allow consideration of non-placebo comparators and standard therapies, network meta-analysis will be deployed in key targeted areas, as decided by the GDG subgroups and when sufficient volume of homogeneous studies is likely to be included.

• For studies of induction of remission, the key outcomes will be clinical remission (including corticosteroid-free), clinical response, endoscopic remission, endoscopic response, biochemical remission and biochemical response, withdrawals due to adverse events or serious adverse events.

• The magnitude of effect of any intervention (against placebo or a comparator) for induction of clinical remission of importance will be defined at 10%.

• For studies on maintenance of remission, the key outcomes will be clinical relapse, loss of clinical response, endoscopic and biochemical relapse, withdrawals due to adverse events and serious adverse events. Maintenance of remission studies will only be considered if patients were in clinical or endoscopic remission at baseline. Preplanned sensitivity analysis will include studies with patients in response, but this will not be the core analysis.

• The magnitude of effect of any intervention (against placebo or comparator) for maintenance of clinical remission of importance will be defined at 10%.

• For all such analysis, planned subgroup analysis will include whether patients received concomitant therapies (more than 50% in a group) with immunosuppression, corticosteroids or biological therapy (not anti-inflammatory, antibiotics or other classes of therapy), whether patients were biologically naïve and in the case of maintenance studies, as stated above, whether they are in a state of remission or response at baseline.

• For studies not targeting induction or maintenance of remission, outcomes to be considered will include quality of life (using study defined scales), morbidity (including escalation of therapy or hospitalisation), abdominal pain and fatigue.

• For all safety outcomes, no predefined levels for magnitude of effect of safety concerns will be made, as this will be dependent on the specific effect. All events with randomised controlled trials (RCTs) will be recorded and presented descriptively, as well as through meta-analysis. However, specific reporting will be made for all the following adverse events in both trials and observational studies: mortality, malignancy, sepsis, organ targeted inflammatory effects (pancreatitis, nephritis, hepatitis) and thromboembolic events will be specifically reported for all interventions.

• Therapy delivered in primary, secondary or tertiary care, as well as self-administered will be considered but setting will be described in the technical summary to allow any clarifying statements to be made regarding the context of evidence.

GRADE recommendations and good practice/research statements

The GDG determined which PICOs are likely to conducive to evidence synthesis and GRADE assessment of outcomes. This included all forms of evidence (observational and RCT). The GDG does not believe there is enough evidence to make a GRADE assessment for certain areas of interest. The GDG will have a number of options following GRADE guideline production methodology.8

The first will be to make GRADE recommendations recognising these limitations, justifying the factors that led to this (these may be strong or conditional). The second option is when the GDG will find it difficult or impossible to formally summarise and GRADE the evidence, but guideline panel members are confident that there is unequivocal benefit or harm from the proposed PICO response, a good practice statement may, under strict criteria, instead.9

At this stage, the GDG has identified PICOs that are likely to fall under the good practice statement context, due to lack of studies. If through the technical review, evidence becomes apparent that allows synthesis, this will be used, but if not, statements will be produced after GRADE recommendations for all such outcomes and delphi agreement sought as per such GRADE recommendations.

Both elements will be clearly labelled throughout the end guideline.

Considering importance of outcomes and interventions, values, preference or utilities to consumers

The list of these elements that have guided the PICO formation has been presented to our partners in Crohn’s & Colitis UK and people living with IBD members. The key principle has been to balance volume of evidence and completeness with the utility to people living with IBD and stakeholders while aligning the BSG recommendations to those produced by the IBD standards.

Evidence selection

Criteria for considering studies for this review

Types of interventions

Trials using the following interventions for induction or maintenance of remission delivered by any route will be included. The types of interventions for ulcerative colitis are mention in table 2 and the types of intervention for CD are mention in table 3.

For studies focusing on symptom management, outside of induction or maintenance of remission, any intervention will be included.

Search methods for identification of studies

Searching other resources

All studies included in the previous BSG guideline will be considered.

Manual searches of reference lists from potentially relevant publications will be performed to ensure that studies that may have been missed by the computer-assisted search are identified.

We will also perform hand searches of the conference proceedings of major gastroenterology meetings (eg, digestive disease week; the annual meeting of the European Crohn’s and Colitis Organisation; and United European Gastroenterology Week) from 2019 to 2024.

Evidence synthesis

Data extraction and management

Two pairs of reviewers will independently extract data. Any disagreements will be resolved by discussion with a third author. A standard data extraction form will be used. The following data will be retrieved from included studies:

1. General information (title, journal, year, publication, type).

2. Participant characteristics (disease activity at inclusion; number and percentage of advanced medical therapy naive vs advanced medical therapy experienced patients; concomitant corticosteroid usage; number and percentage of patients who previously underwent surgery for the treatment of CD; disease duration; luminal or fistulising disease).

3. Primary and secondary outcomes.

4. Risk of bias information.

Assessment of risk of bias in included studies

The methodological quality of each included study will be independently assessed by two pairs of authors using the Cochrane risk of bias tool.11 Factors to be assessed include:

1. Random sequence generation.

2. Allocation concealment.

3. Blinding of participants, outcome assessors and investigators.

4. Incomplete outcome data.

5. Selective outcome reporting.

6. Other potential sources of bias.

The studies will be judged to be of low, high and unclear risk bias based on these factors. Any disagreements regarding risk of bias assessment between the two authors will be resolved through discussion until consensus is reached. If consensus cannot be reached, a third author will be consulted to resolve the disagreement.

Measures of treatment effect

For the dichotomous outcomes, the treatment effect will be expressed as risk ratios with corresponding 95% CIs. For continuous outcomes, the treatment effect will be expressed as mean differences (MD) with 95% CIs. However, if the studies assessed the same continuous outcome differently, we will estimate the treatment effect using the standardised MD (SMD). We will present SMDs as SD units and interpret them as follows: 0.2 represents a small effect, 0.5 a moderate effect and 0.8 a large effect.

Unit of analysis issues

The participant will be the unit of analysis. For studies comparing more than two intervention groups, we will make multiple pair‐wise comparisons between all possible pairs of intervention groups. To avoid double counts, we will divide shared intervention groups evenly among the comparisons. For dichotomous outcomes, we will divide both the number of events and the total number of participants. For continuous outcomes, we will only divide the total number of participants, and leave the means and SD unchanged.

We will include cross-over studies, but we will only pool their data if they were reported separately before and after cross-over, and we will only use pre-cross-over data. In the case of cluster-RCTs, we only use study data if the trial authors had used appropriate statistical methods in taking the clustering effect into account.

Dealing with missing data

The intention-to-treat principle (ie, all patients lost to follow-up are considered treatment failures) will be applied in the case of missing outcome data.

Assessment of heterogeneity

We will scrutinise studies to ensure that they are clinically homogeneous in terms of participants, interventions, comparators and outcomes. To test for statistical heterogeneity, we will use a χ² test. A p<0.1 gives an indication of the presence of heterogeneity. Inconsistency will be quantified and represented by the I² statistic. We will interpret the thresholds as follows.12

0%–40%: might not be important.

30%–60%: may represent moderate heterogeneity.

50%–90%; may represent substantial heterogeneity.

75%–100%: considerable heterogeneity.

We will examine possible explanations for heterogeneity when sufficient data are available, including factors such as participant characteristics (eg, age, sex), condition severity, healthcare system and country.

In the case of considerable statistical heterogeneity, we will investigate whether this can be explained on clinical grounds or risk of bias, in which case, we will aim to conduct sensitivity analyses. If no reasons for statistical heterogeneity are found, we will present the results narratively and in detail.

Evaluating the assumption underlying network meta-analysis

We will use a design-based decomposition of Cochran Q for assessing the homogeneity on the whole network, the homogeneity within designs and the homogeneity/consistency between designs. This approach also allows for an assessment of the consistency assumption after detaching the effect of single designs. Inconsistency will be located using the net heat plot.13

All analyses will be run with R statistical package (R Development Core Team) and the netmeta library.14

Assessment of reporting biases

To assess the presence of small-study effects in the network meta-analysis, we will use a funnel plot that accounts for the fact that each set of studies estimates a different summary effect.15 We will focus on the comparisons of all active treatments against placebo, which may be more prone to small-study effects.

Publication bias will be investigated using funnel plots for each meta-analysis when there are at least four studies. Funnel plot asymmetry will be tested using rank a correlation test when there are at least 10 studies.16

Data synthesis

Subgroup analysis and investigation of heterogeneity

The presence of heterogeneity among studies will be assessed using the χ² test (a p<0.10 will be regarded as statistically significant) and the I² statistic.23 Data will not be pooled for meta-analysis if a high degree of heterogeneity is observed (eg, I²>75%). Sensitivity analyses will be conducted as appropriate to investigate heterogeneity.

In the presence of statistical and clinical heterogeneity, and if sufficient studies are available, we will consider the following potential sources of heterogeneity:

1. Mistakes and inconsistencies in data extraction and entry.

2. Population.

3. Intervention (ie, dose, frequency or route).

4. Risk of bias.

5. Funding source.

6. Study design (ie, duration of treatment, duration of follow-up and number of participants).

7. Date of publication.

If we find significant inconsistency, we will investigate possible sources. Specifically, we will investigate the distribution of prespecified clinical and methodological variables that we suspect may be potential sources of either heterogeneity or inconsistency in each comparison specific group of trials. If sufficient studies are available we will consider the following potential sources of heterogeneity and network inconsistency:

1. Different drug doses and routes of administration.

2. Patients with exposure to immunosuppressive drugs (ie, thiopurines, methotrexate).

3. Patients with more severe disease activity.

4. Patients who are biological-naive.

5. Disease location and extent.

6. Studies that used clinical and objective measures to determine patient eligibility (C reactive protein, faecal calprotectin, endoscopy).

Sensitivity analysis

If possible, we will perform the following sensitivity analyses to examine the impact of the following variables on the pooled effect estimate:

1. Random-effects versus fixed-effect modelling.

2. Studies with low risk of bias versus studies with unclear or high risk of bias.

3. Relevant lost to follow-up (studies with >25% lost to follow-up).

4. Trials with a sample size of less than 50 participants.

5. Studies for maintenance where patients in response were included to be added.

6. Inclusion of studies defining disease activity, remission or relapse with none described or none validated approaches.

GRADE judgements of certainty

We will present summary of findings tables for all comparisons, which include our primary outcomes.

Based on risk of bias, inconsistency, imprecision, indirectness and publication bias, we will grade the certainty of the evidence for each outcome as high, moderate, low or very low (described below). We will justify all decisions to downgrade the certainty of studies using footnotes, and we will make comments to aid the reader’s understanding of the review where necessary.

GRADE Working Group grades of evidence:

• High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.

• Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

• Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect.

• Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect

Development of recommendations and strength of recommendations

An evidence technical review meeting will be run to allow the GDG members to discuss and explore the technical summaries and GRADE recommendations for each of the PICO outcomes. Particular areas of focus will be those with little to no evidence and the almost certainly lower quality safety data.

For areas where no observational or higher evidence or sufficient to allow appropriate GRADE judgements was found, a small working group will be convened to produce proposed good practice statements. These will be produced based on the previous BSG IBD guideline statements with amendments as needed.

Writing of the guideline

The lead representatives from each subgroup together with the co-chairs of the GDG will form a writing group to prepare the full guidelines in line with the proposals stated in this protocol, with a succinct guideline with the recommendations and a separate but complimentary full technical review summary paper. These will be prepared in line with the appropriate Gut journal guidelines.

The full document will be sent to the GDG for internal peer review and amendment before going to the wider BSG executive for review. All indicated amendments will be made prior to the submission of the final guideline.

In parallel the core leadership team will work with CCUK (Crohn's and Colitis UK organization) representatives and people living with IBD to prepare a user-focused guideline.