Article Text

Download PDFPDF

Original article
Development of an index to define overall disease severity in IBD
  1. Corey A Siegel1,
  2. Cynthia B Whitman2,
  3. Brennan M R Spiegel2,
  4. Brian Feagan3,
  5. Bruce Sands4,
  6. Edward V Loftus Jr5,
  7. Remo Panaccione6,
  8. Geert D'Haens7,
  9. Charles N Bernstein8,
  10. Richard Gearry9,
  11. Siew C Ng10,
  12. Gerassimos J Mantzaris11,
  13. Balfour Sartor12,
  14. Mark S Silverberg13,
  15. Robert Riddell13,
  16. Ioannis E Koutroubakis14,
  17. Colm O'Morain15,
  18. Peter L Lakatos16,
  19. Dermot P B McGovern17,
  20. Jonas Halfvarson18,
  21. Walter Reinisch19,
  22. Gerhard Rogler20,
  23. Wolfgang Kruis21,
  24. Curt Tysk18,
  25. Stefan Schreiber22,
  26. Silvio Danese23,
  27. William Sandborn24,
  28. Anne Griffiths25,
  29. Bjorn Moum26,
  30. Christoph Gasche27,
  31. Francesco Pallone28,
  32. Simon Travis29,
  33. Julian Panes30,
  34. Jean-Frederic Colombel4,
  35. Stephen Hanauer31,
  36. Laurent Peyrin-Biroulet32
  1. 1 Dartmouth-Hitchcock Inflammatory Bowel Disease Center, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, USA
  2. 2 Department of Health Services, Cedars-Sinai Medical Center, Los Angeles, California, USA
  3. 3 Robarts Clinical Trials, London, Ontario, Canada
  4. 4 Icahn School of Medicine at Mount Sinai, New York, New York, USA
  5. 5 Mayo Clinic, Rochester, Minnesota, USA
  6. 6 University of Calgary, Calgary, Alberta, Canada
  7. 7 Academic Medical Center, Amsterdam, Netherlands
  8. 8 University of Manitoba, Winnipeg, Manitoba, Canada
  9. 9 University of Otago, Christchurch, New Zealand
  10. 10 Department of Medicine and Therapeutics, Institute of Digestive Disease, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Hong Kong, Hong Kong
  11. 11 Evangelismos-PolyCliniki-Ophthalmiatreion Hospital, Athens, Greece
  12. 12 University of North Carolina, Chapel Hill, North Carolina, USA
  13. 13 Mount Sinai Hospital, Toronto, Ontario, Canada
  14. 14 University Hospital Heraklion, Crete, Greece
  15. 15 Faculty of Health Sciences Trinity College Dublin, Dublin, Ireland
  16. 16 Semmelweis University, Budapest, Hungary
  17. 17 Widjaja Inflammatory Bowel and Immunobiology Research Institute, Cedars-Sinai Medical Center, Los Angeles, California, USA
  18. 18 Department of Gastroenterology, Faculty of Medicine and Health, Örebro University, Örebro, Sweden
  19. 19 McMaster University, Hamilton, Ontario, Canada
  20. 20 University Hospital of Zuürich, Zurich, Switzerland
  21. 21 University of Cologne, Koln, Germany
  22. 22 University Hopital Schleswig Holstein, Kiel, Germany
  23. 23 Humanitas University, Milan, Italy
  24. 24 UCSD, San Diego, California, USA
  25. 25 Hospital for Sick Children, Toronto, Ontario, Canada
  26. 26 Oslo University Hospital and University Oslo, Oslo, Norway
  27. 27 Medical University and General Hospital Vienna, Vienna, Austria
  28. 28 University of Rome Tor Vergata, Rome, Italy
  29. 29 Oxford University Hospital, Oxford, UK
  30. 30 Hospital Clinic de Barcelona, IDIBAPS, CIBERehd, Barcelona, Spain
  31. 31 Northwestern Feinberg School of Medicine, Chicago, Illinois, USA
  32. 32 Inserm U954 and CHU de Nancy, Lorraine University, Nancy, France
  1. Correspondence to Dr Corey A Siegel, Inflammatory Bowel Disease Center, Section of Gastroenterology and Hepatology, Dartmouth-Hitchcock Medical Center, Lebanon, NH 03756, USA; corey.a.siegel{at}hitchcock.org

Abstract

Background and aim Disease activity for Crohn's disease (CD) and UC is typically defined based on symptoms at a moment in time, and ignores the long-term burden of disease. The aims of this study were to select the attributes determining overall disease severity, to rank the importance of and to score these individual attributes for both CD and UC.

Methods Using a modified Delphi panel, 14 members of the International Organization for the Study of Inflammatory Bowel Diseases (IOIBD) selected the most important attributes related to IBD. Eighteen IOIBD members then completed a statistical exercise (conjoint analysis) to create a relative ranking of these attributes. Adjusted utilities were developed by creating proportions for each level within an attribute.

Results For CD, 15.8% of overall disease severity was attributed to the presence of mucosal lesions, 10.9% to history of a fistula, 9.7% to history of abscess and 7.4% to history of intestinal resection. For UC, 18.1% of overall disease severity was attributed to mucosal lesions, followed by 14.0% for impact on daily activities, 11.2% C reactive protein and 10.1% for prior experience with biologics. Overall disease severity indices were created on a 100-point scale by applying each attribute's average importance to the adjusted utilities.

Conclusions Based on specialist opinion, overall CD severity was associated more with intestinal damage, in contrast to overall UC disease severity, which was more dependent on symptoms and impact on daily life. Once validated, disease severity indices may provide a useful tool for consistent assessment of overall disease severity in patients with IBD.

  • IBD
  • CROHN'S DISEASE
  • ULCERATIVE COLITIS

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Significance of this study

What is already known on this subject?

  • There is accumulating evidence that providers need to treat beyond symptoms of IBD to prevent disease complications.

  • Currently, disease activity of IBD refers to a cross-sectional snapshot in time of symptoms and biological inflammatory activity.

  • There is currently no standard definition of overall disease severity, which takes into account longitudinal and historical factors, which gives a more complete picture of overall burden of disease.

What are the new findings?

  • Based on specialist opinion, the attributes of IBD that are most important for overall disease severity have been defined.

  • These attributes were ranked in order of importance, and then an overall disease severity index was created.

  • Overall Crohn's disease severity was associated more with intestinal damage, in contrast to overall UC disease severity, which was more dependent on symptoms and impact on daily life.

How might it impact on clinical practice in the foreseeable future?

  • When selecting therapy for patients, providers need to focus on the short-term goal of controlling disease activity and use the overall disease severity to guide long-term management plans.

  • We expect this work to help identify those at the higher end of the risk spectrum, so that appropriate intensive treatment can be initiated and optimised in an efficient, precise and cost-effective manner.

Introduction

IBD, Crohn's disease (CD) and UC are chronic and often disabling conditions.1–3 There is accumulating evidence indicating that physicians need to treat beyond symptoms due to the well-documented disconnect between clinical symptoms and inflammatory disease activity, especially in CD. IBD is a destructive disease entity that may be complicated by irreversible bowel damage and the need for surgery. Avoidance of hospitalisation, surgery and the prevention of bowel damage are increasingly considered as important outcomes in disease modification trials.4–9 The premise is to treat inflammatory disease before intestinal damage becomes irreversible and disability ensues.

Disease activity and disease severity refer to two distinct yet overlapping concepts. Disease activity reflects a cross-sectional assessment of biological inflammatory activity, whereas overall disease severity may include longitudinal and historical factors which give a more complete picture of the overall burden of disease. To be included in the registration clinical trials for biologics, patients with IBD had to have moderate to severely active disease according to the clinical activity indices which were developed to measure disease activity at a point in time.10–12 However, basing overall disease severity on symptoms alone does not account for the multitude of additional patient and disease characteristics that drive our decisions towards appropriate medical therapy, including bowel damage and impact on daily activities (disability). Since the labels for biologic drugs are based on the registration trials, these medications are limited to patients with moderate to severe disease activity. Therefore, patients are required to have substantial symptoms at the time of prescribing of these agents for payers to agree to the initiation of treatment. Historically, these medications have not been indicated for those with disease that is currently less active or inactive, even if the past pattern or burden of disease would be likely to benefit from treatment.

A patient's disease course and risk of future complications are important features of IBD that should also guide management in addition to disease activity at present. In an effort to shift the paradigm from stratifying patients based on current disease activity towards a more global course of overall disease severity, researchers conducted a systematic review identifying three main domains relevant to the evaluation of disease severity in IBD: impact of the disease on the patient regarding symptoms and daily activities, inflammatory burden and disease course.10 There are multiple attributes within these domains, some of which have more influence over overall disease severity than others.

The aims of this study were: (1) to define the attributes within these three domains that were most important in determining overall disease severity—‘overall’ accounting for both short-term signs and symptoms, and long-term complications of disease; (2) to rank the importance of and score these individual attributes for both CD and UC and (3) with these rankings and an understanding of how much each attribute contributes to disease severity, to generate an overall disease severity index to place an individual's disease in context.

Methods

Phase 1: Identification of key attributes

Determining the main patient attributes influencing determination of overall IBD severity began with a systematic review process to identify all potential elements of disease activity and severity of IBD. Using a set of predefined MeSH terms, researchers conducted a comprehensive literature search in PubMed from inception until April 2014. During the systematic review, researchers identified three domains instrumental in determining overall IBD severity: impact of the disease on the patient regarding symptoms and daily activities, inflammatory burden and disease course including complications.10

After a list of attributes was identified, 16 IBD specialists met in person to discuss the results using a modified RAND panel. The proposed attributes were discussed and reviewed with the panel. With the intention of reviewing and shortlisting the list of attributes identified through the systematic review, an open discussion ensued. There was an anonymous response process, and participants voted using an audience response system on the attributes on the list by assigning a rating of 1–9 (with 9 being the most preferred) to each of the attributes on the list. Variables that received a score ranging from 7 to 9 by at least 75% of participants were included in the final attribute set. For those attributes not initially receiving a score between 7 and 9, a secondary anonymous vote was initiated using the audience response system. At this point, if an attribute received fewer than 75% of votes with the rating from 7 to 9, it was dropped from the final item set. Output included two defined sets of attributes considered to be markers of severe disease, one for CD and one for UC (see online supplementary tables S1 and S2).

Phase 2: Conjoint analysis

Conjoint analysis in healthcare

Conjoint analysis (CA) has been used as a method to elicit how people make complex decisions when there are many driving factors, by determining the willingness to ‘trade-off’ certain attributes in favour of others. The CA process is a choice-experiment, wherein we assume stakeholder choices are influenced by multiple driving forces that may be underlying. CA uses multinomial logit regression to model the choice process through applying multiple approximations of the utility of an attribute.

A particular form of CA, called adaptive choice-based conjoint (ACBC), takes respondents through a process of evaluating competing profiles of a product (or patient) that are comprised of various attributes or characteristics. Increasingly, CA is being applied to healthcare decision-making processes. Examples of their applications include: conducting cost effectiveness analysis with a consideration for process of care, assessment of willingness to pay, assessing health status, quality-adjusted life-years, health-related quality of life and valuing patient-reported outcome measures to assess perceived risk and preference for various treatment options.13–17 When considering healthcare decisions, the profiles contain a list of patient clinical factors which may influence a decision-making process (eg, frequency of loose stools, disease extent and C reactive protein (CRP) level), each with a set of levels. This technique has been used previously to study patient and providers’ preferences in IBD.18–21

During a CA exercise, there is an iterative process such that each response provided influences subsequent questions until there is ample consistency among respondent clinical factor choices. These results then inform a series of complete hypothetical patient profiles containing varying levels of clinical attributes. In this project, participants were asked to focus on how different clinical factors of IBD influenced their impression of overall disease severity. Respondents indicate which patient profiles meet their evaluation criteria, until finally respondents are shown two side-by-side profiles and asked to select which is more severe. A series of profile comparison continues until the responses achieve internal consistency, and a rank order of the respondent's prioritisation of clinical factors can be determined. The number of specific tasks that each individual answered varied depending on the responses selected by the individual. If the respondent showed internal consistency when answering the questions, they were shown fewer tasks, and if they did not, they were shown more tasks.

Survey design

To conduct a CA exercise, it is essential that attributes are evaluated individually and partitioned into levels with varying degrees of severity. The levels were determined by investigators (CAS, CBW and LP-B) based on the output from phase 1 and the clinical relevance of creating dichotomous versus three-level attributes (table 1, and see online supplementary tables S1 and S2). Attributes and their corresponding levels were adapted for CA from the RAND panel, using the Sawtooth Software (Orem, UT) ACBC module.

Table 1

Clinical factors (and levels) tested in conjoint analysis exercises

The first section of the CA displayed a series of patient vignettes, and respondents were asked to select either ‘Yes, this patient has severe disease’ or ‘No, this patient does not have severe disease’ (figure 1). Each question in the CA was generated according to the respondent's previous selection. The length of the survey varied depending on the internal consistency of the responses. The next set of questions asked respondents to select from a list of attributes whether they believed any specific attributes to always be markers of severe disease, or were never markers of severe disease. Based on the respondents' selections, a series of side-by-side patient profiles was generated. Respondents then selected, among the two patients, which they believed had more severe disease. Some of the attributes remained constant between the two patient profiles. Respondents continued to select between pairs of patients until they established internal consistency and their preferences for certain attributes emerged. Figure 2 displays an example of the side-by-side comparison of two hypothetical patients with CD. Questions for CD and UC were presented separately to allow respondents to rate these diseases independently.

Figure 1

Sample patient vignette.

Figure 2

Side-by side-profiles of two hypothetical Crohn's disease patients generated by the conjoint analysis exercise.

Conjoint analysis administration

In the context of this study, the ACBC form of CA methodology was applied to an in-person panel of specialists to replace a traditional RAND decision panel. The underlying statistics behind ACBC are similar to that of regression modelling, with the main output being a part-worth utility, a number ranging from zero to one that acts as a constant, similar to the role of the coefficient in a regression. The constant shows relative weights between levels within an attribute or clinical factor, with the value of one signifying the most severe and zero the least severe level. The average importance of all clinical factors in the CA profiles adds up to 1.0 (100%). The ACBC analysis uses a hierarchical Bayes logit model, whereby we assume that the individual's part-worth utilities are normally distributed. As each additional individual contributes a new ranking or response, a new iteration of the results is calculated. Eventually, the part-worth utilities for each attribute begin to converge as additional respondents' data are combined with that of the previous respondents. It is through these outputs that we are able to see a rank ordering of attributes and the relative importance of one attribute compared with another, as well as the levels within each attribute.

During the month of April 2015, after an open invitation to the International Organization for the Study of IBD (IOIBD), a selection of these IBD specialists were asked to complete the CA exercise at three different time points to determine the importance of each of the identified attributes. Initially, specialists completed the survey independently without any discussion or influence from others. All CA results were presented and reviewed anonymously at an in-person panel at the Annual IOIBD Conference. After a period of discussion, these specialists then repeated the CA exercise in a group setting using an automated response system (where the majority for question selection was set at 51%). Finally, specialists completed the CA for a third time, independently, after commencement of the group discussion. This last round of ratings was used as the final results.

Phase 3: Overall disease severity index

Following the compilation of the third round of CA results, we worked to translate the average importance, how much of an impact each attribute makes to the overall profile, into a baseline set of IBD severity indices that could be applied to the patient setting and in the future be adapted for use by clinicians in practice. To create the indices, we used the average part-worth utility scores to determine minimum and maximum relative scores of each attribute level. Scores were converted to a 100-point scale, where the scores for the absence of a symptom within each attribute (the lowest average utility) were set to zero. Scores for levels within each attribute were calculated by applying the magnitude of their average utility to the attribute's overall importance, or weight. By applying the average part-worth utilities to the scoring schema, the score within a given attribute became relative to the level of importance that the attribute had in determining overall disease severity across all of the clinical factors included in the CA. Scores within each attribute were rounded to the nearest integer. For example, the Crohn's attribute ‘mucosal lesions’ had three levels of severity, with the lowest level being the absence of the finding. In this case, the average importance for mucosal lesions (15.8) was rounded to the nearest integer (16), and if a patient were to exhibit any presence of this finding, they would receive a score of 16 towards their total potential severity score of 100. If they were to have an absence of mucosal lesions, they would receive a score of zero for that particular attribute.

Results

Sample characteristics

Eighteen IBD specialists completed the final CA exercise. Their ages were equally distributed from 35 to over 65 years, but only one was female. All participants had practiced medicine with a major interest in IBD for more than 10 years and were geographically diverse, representing 12 different countries.

Phase 1: Attribute selection process

Following the initial panel (see online supplementary tables S1 and S2), specialists defined CD and UC overall disease severity using a set of 13 and 12 key attributes, respectively. To meet the parameters of CA methodology, some of these attributes were disassociated with one another: a process that would ensure each attribute would not be influenced by the presence of another. The final CA was comprised of 16 attributes for the CD module and 13 attributes for UC module (table 1).

Phase 2: Premeeting versus postmeeting conjoint analysis results

Following the in-person discussion, there was a shift in the weight of the attributes. Additionally, the rank order of important attributes differed between CD and UC. The presence of mucosal lesions was significantly more important than all other attributes under consideration in both modules. Figures 3 and 4 highlight the relative importance of each of the attributes for the CD and UC CA modules from most to least influential when determining overall patient disease severity according to the postmeeting results.

Figure 3

Premeeting and postmeeting average Crohn's disease (CD) attribute relative importance. Patient characteristics of CD ordered by relative importance based on conjoint analysis part-worth utilities before and after the in-person meeting. Bars are displayed with SEs to highlight significant differences in the rankings between the two time periods. Mucosal lesions accounted for 17.6% and 15.8% of CD overall severity followed by presence of a fistula (5.7% and 10.9%) in premeeting and postmeeting results, respectively. Having a fistula, perianal abscess or stoma, all became more important in determining overall disease severity after the meeting, while having a stricture, use of biologics, impacts on daily activity, albumin level and being anaemic, all became less important.

Figure 4

Premeeting and postmeeting average UC relative importance. Patient characteristics of UC, ordered by relative importance based on conjoint analysis part-worth utilities before and after the in-person meeting. Bars are displayed with SEs to highlight significant differences in the rankings between the two time periods. Mucosal lesions accounted for 18.4% and 18.1% of UC overall severity followed by impacts on daily activity (9.3% and 14.0%) in premeeting and postmeeting results, respectively. The relative rankings of impacts on daily activity, C reactive protein (CRP) level, experience with biologics and recent steroid use all increased after the in-person meeting.

CD module

Overall disease severity in CD was dominated by those attributes measuring intestinal damage caused by disease, with mucosal lesions being the most important clinical factor accounting for 15.8% of the determination of overall CD severity. Presence of a fistula (11.0%), perianal abscess (9.7%) and history of intestinal resection (7.4%) were the next most prominent factors influencing the classification of severe disease.

Comparatively, the presence of daily symptoms (loose stools) and impact on daily activity were less important factors influencing overall disease severity.

There was a marked increase in importance for clinical factors associated with intestinal damage following the in-person meeting, suggesting that the distinction between CD and UC was influenced by the meeting's group discussions.

UC module

In contrast to CD, overall UC severity was influenced by the presence of inflammation and the impact of disease on daily activity. While the most important clinical factor for determining overall UC severity was mucosal lesions (18.1%), the next was the impact of symptoms on daily activity (14.0%). Use and effectiveness of biologics (10.1%) and recent hospitalisation (7.7%) were also influential when determining overall disease severity. The clinical factors measuring intestinal damage were ranked much lower (disease extent, 4.8%; anorectal symptoms, 4.0%) (table 2).

Table 2

Disease attribute contributions to overall disease severity

Comparing results from before and after the in-person meeting, those attributes exhibiting significant shifts in importance ranking included: impacts of daily activity, CRP level, experience with biologics, recent hospitalisation and recent steroid use.

Phase 3: Overall disease severity index

After examining the clinical attribute results for CD and UC separately, we adapted the relative importance of each attribute to two unique 100-point scales (tables 3 and 4). The severity scores calculated within attributes for CD ranged from 16 (large or deep mucosal lesions as confirmed by MRI or endoscopy) to 2 (confirmed steroid use in the past year).

Table 3

Crohn's disease overall disease severity index

Table 4

UC overall disease severity index

Disease severity scores calculated for the clinical attributes of UC ranged from 18 (active ulcers confirmed by endoscopy) down to 4 (presence of rectal bleeding, anorectal symptoms or nocturnal bowel movements).

Discussion

Disease severity has historically been synonymous with disease activity, but this ignores the many different aspects of IBD that impact patients' lives and factors that lead to long-term complications. Disease activity can reflect long-term or short-term activity, but is most often measured by a score depicting only a snapshot of disease status. These measures do not account for the accumulation of hospitalisations, surgeries, complications, associated diseases and impact on lifestyle over time. Intestinal damage begins soon after the onset of inflammation, often in the absence of significant symptoms.22 Although more prominent in CD, this phenomenon also occurs in UC.5 If effective therapy is reserved until moderate–severely active disease is obvious, then the window of opportunity to successfully treat patients to achieve remission is lost for many. To look beyond symptoms and better recognise patients at most risk due to their IBD, we have identified the key attributes that contribute to overall IBD severity. Overall IBD severity accounts for both recent signs and symptoms and long-term disease complications. Furthermore, we have created two overall disease severity indices (DSI) that, when validated, could be used to risk-stratify patients with CD and UC. Since attributes were rated very differently for CD and UC, two separate DSIs were created. The DSI score provides a cross-sectional assessment of severity, and also includes items that accumulate over time, which gives the score a predictive potential.

Based on specialist opinion, mucosal lesions are the most important attributes associated with overall disease severity for both diseases. When comparing CD and UC, overall CD severity is associated more with accumulating intestinal damage, in contrast to overall UC disease severity, which is more dependent on symptoms and impact on daily life. This is consistent with how patients are approached in practice. For example, when patients with CD are seen in clinic, although current symptoms are taken into account, prior surgeries, perianal disease and extent of bowel at risk should typically be weighted higher than day-to-day symptoms when considering a long-term treatment plan. Conversely, since active inflammation is more overt symptomatically with UC and bowel damage less obvious, it is not surprising that this specialist panel treated UC differently than CD. The separate DSIs allow for providers to think about overall disease severity differently between these diseases and create individual treatment plans based on the overall DSI and a patient's specific disease type.

The work of this panel helps redefine overall disease severity for IBD. Currently, disease-scoring indices are used for clinical trials or to track symptoms prospectively.23 These indices are valuable to stratify patients into mild, moderate or severe disease activity at a moment in time,24 but do not take into account overall disease severity in context of both bowel damage and impact on daily life. Other groups have identified risk factors for rapid evolution of disease,25 disabling CD,26 and have defined features of severe, aggressive or complicated CD,27–29 but have not specified and weighted how disease attributes independently contribute to overall disease severity. The disease characteristics identified by the panel, and the overall DSI, will enable the differentiation between disease activity and disease severity, and once validated will offer both further research opportunities and a practical tool by which to classify overall disease severity of patients and offer appropriate treatment without relying on present symptoms alone. In the early 1990s, rheumatologists went through a similar process to identify better outcomes to measure for patients with rheumatoid arthritis and created OMERACT (Outcome Measures in Rheumatoid Arthritis Clinical Trials).30 Since then, this has become a foundation of their field and has led to hundreds of publications. We hope to learn from OMERACT and use this current work to push the field of IBD towards a more meaningful understanding of overall disease severity.

As with any expert panel, there are limitations of this work based on the participants and validity outside of this group. Although the voting group was composed only of members from the IOIBD, this is a globally diverse cohort of specialists possessing varied opinions across a range of years of clinical experience. There was quite a bit of disagreement among the group in the premeeting independent questionnaires, but this gap decreased in the postmeeting voting—demonstrating that the RAND process is a successful method whereby people share opinions, listen to the views of each other and incorporate new points of view into an evolved perspective. Although there was a decrease in disagreement in postmeeting voting, there is still variation in responses as can be appreciated by the SE bars seen in figures 3 and 4, reflecting a broad range of opinions despite this narrow group of participants. With recognition that the voting panel was composed solely of a self-selected group of specialists, validation work with other providers will be important.

CA has been applied in healthcare fields ranging from mental health31 to chronic leukaemia,32 and has been used specifically within the field of IBD.19–21 The methodology of using conjoint trade-off analysis embedded within a RAND panel is to the best of our knowledge a novel approach. Integration of these two methodologies brings along with it the benefit of applying expert knowledge to a systematic process for assigning utility scores to patient disease severity attributes, and then calculating specific values to explain what proportion of overall disease severity is influenced by each attribute.33 This process could not be completed using a standard RAND format. Historically, RAND panels are smaller with nine participants, but conjoint analyses could be much larger. We included 18 participants in the final RAND voting to expand the number of opinions. Arguably, a larger sample would have allowed for broader opinions and ultimately a more precise point estimate for the DSI, but this needed to be taken into consideration while keeping practical control of the in-person meeting.

Future studies using these results should focus on two important aspects: prospective validation of the overall DSI in different patient populations and conducting a similar in-person CA with patients to see how their valuation of attributes compares with providers. In fact, a prospective trial has already been designed to evaluate the DSI by testing for construct validity, internal consistency, test–retest reliability and responsiveness in both referral centres and community practices. Patient opinions will likely differ from those of providers, and this will also be studied as part of the prospective evaluation. After testing the application clinically, it is likely that the overall DSI could be adjusted and later implemented into clinical practice. Furthermore, a validated overall DSI could be used as a stratification factor, inclusion criterion or as a covariate for statistical adjustment in randomised controlled trials.

For patients with IBD, it is important to distinguish disease activity at a point in time from disease severity over a period of time. When selecting therapy for patients, providers need to focus on the short-term goal of controlling disease activity and use the overall disease severity to guide long-term management plans. Short-term symptom control is often the easiest part of IBD management, while preventing the consequences of irreversible bowel damage requires preparation and an understanding of the stratification of risk for an individual patient. We expect this work to begin to address a change in how we think about patients with IBD, and how to identify those at the higher end of the risk spectrum so that appropriate intensive treatment can be initiated and optimised in an efficient, precise and cost-effective manner.

Acknowledgments

The authors thank Lucid for their logistical support during phases 1 and 2, and Kimberly Thompson for assistance with manuscript preparation.

References

View Abstract

Footnotes

  • Contributors CAS (study design, analysis and interpretation, manuscript development and writing). CBW (study design, analysis and interpretation, manuscript development and writing). BMRS (panel moderator, manuscript review). BF, BS, EVL, RP, GD, CNB, RG, SN, GJM, BS, MSS, RR, IK, COM, PLK, DPBM, JH, WR, GR, WK, CT, SS, SD, WS, AG, BM, CG, FP, ST, JP, JFC, SH (panellist, manuscript review). LPB (study design, analysis and interpretation, manuscript development and writing).

  • Funding AbbVie (Unrestricted educational grant), Tillotts (Unrestricted educational grant).

  • Competing interests Phase 1 was funded thanks to an unrestricted educational grant to IOIBD from AbbVie and Tillotts. CAS has served as a consultant on advisory boards for AbbVie, Amgen, Lilly, Janssen, Sandoz, Pfizer, Prometheus, Takeda, Theradiag and UCB; as a speaker for American Regent, AbbVie, Janssen, Pfizer and Takeda; and receives grant support from AbbVie, Janssen and Takeda. BMRS has research grants from Shire and Takeda. BF has served as a consultant for Abbott/AbbVie, Actogenix, Akros, Albireo Pharma, Amgen, Astra Zeneca, Avaxia Biologics, Avir Pharma, Axcan, Baxter Healthcare, Biogen Idec, Boehringer-Ingelheim, Bristol-Myers Squibb, Calypso Biotech, Celgene, Elan/Biogen, EnGene, Ferring Pharma, Roche/Genentech, GiCare Pharma, Gilead, Given Imaging, GSK, Ironwood Pharma, Janssen Biotech (Centocor), JnJ/Janssen, Kyowa Kakko Kirin Co., Lexicon, Lilly, Lycera BioTech, Merck, Mesoblast Pharma, Millennium, Nektar, Nestles, Novonordisk, Pfizer, Prometheus Therapeutics and Diagnostics, Protagonist, Receptos, Salix Pharma, Serono, Shire, Sigmoid Pharma, Synergy Pharma, Takeda, Teva Pharma, TiGenix, Tillotts, UCB Pharma, Vertex Pharma, VHsquared, Warner-Chilcott, Wyeth, Zealand and Zyngenia; has been a speaker for Abbott/AbbVie, JnJ/Janssen, Takeda, Warner-Chilcott and UCB Pharma; has served on advisory boards for Abbott/AbbVie, Amgen, Astra Zeneca, Avaxia Biologics, Bristol-Myers Squibb, Celgene, Centocor, Elan/Biogen, Ferring, JnJ/Janssen, Merck, Nestles, Novartis, Novonordisk, Pfizer, Prometheus Laboratories, Protagonist, Salix Pharma, Takeda, Teva, TiGenix, Tillotts Pharma AG and UCB Pharma; receives research funding from Abbott/AbbVie, Amgen, Astra Zeneca, Bristol-Myers Squibb (BMS), Janssen Biotech (Centocor), JnJ/Janssen, Roche/Genentech, Millennium, Pfizer, Receptos, Santarus, Sanofi, Tillotts and UCB Pharma; and serves on the board of directors for Robarts Clinical Trials. BS has served as a consultant for AbbVie, Celgene, Janssen Biotech, MedImmune, Takeda, Pfizer, TiGenix, Bristol-Myers Squibb, Receptos, Akros Pharma, Arena Pharmaceuticals, Theravance Biopharma R&D, Boehringer-Ingelheim, Synergy Pharmaceuticals, Toplvert Pharma, UCB and Lilly; has served on scientific advisory boards for Celgene, Janssen Biotech, MedImmune, Takeda, Pfizer, TiGenix and Lily; and he receives research funding from Celgene, Janssen Biotech, MedImmune, Takeda and Pfizer. EVL has consulted for UCB Pharma, Takeda, Janssen, AbbVie, Genentech, Celgene, Amgen, Mesoblast, Theradiag, Sun Pharma and Seres Health; has received research support from UCB Pharma, Takeda, Janssen, AbbVie, Genentech, Celgene, Amgen, Pfizer, Receptos, Gilead and Robarts Clinical Trials. RP has served as a consultant for Abbvie/Abbott, Amgen, Aptalis, AstraZeneca, Baxter, Eisai, Ferring, Janssen, Merck, Schering-Plough, Shire, Centocor, Elan, Glaxo-Smith Kline, UCB, Pfizer, Bristol-Myers Squibb, Warmer Chilcot, Takeda, Cubist and Celgene; on speaker's bureaus for Abbvie/Abbott, AstraZeneca, Janssen, Schering-Plough, Shire, Centocor, Elan, Prometheus, Warner Chilcott and Takeda; on advisory boards for Abbvie/Abbott, Amgen, Aptalis, AstraZeneca, Baxter, Eisai, Ferring, Genentech, Janssen, Merck, Schering-Plough, Shire, Centocor, Elan, Glaxo-Smith Kline, UCB, Pfizer, Bristol-Myers Squibb, Warner Chilcott, Takeda, Cubist, Celgene and Salix; and receives research/educational support from Abbvie, Abbott, Ferring, Janssen, Schering-Plough, Centocor, Millenium, Elan, Proctor and Gamble and Bristol-Myers Squibb. GD has served as advisor for Abbvie, Ablynx, Amakem, AM Pharma, Avaxia, Biogen, Bristol-Meyers Squibb, Boerhinger Ingelheim, Celgene, Celltrion, Cosmo, Covidien, Ferring, DrFALK Pharma, Engene, Galapagos, Gilead, Glaxo Smith Kline, Hospira, Johnson & Johnson, Medimetrics, Millenium/Takeda, Mitsubishi Pharma, Merck Sharp Dome, Mundipharma, Novonordisk, Pfizer, Prometheus laboratories/Nestle, Receptos, Robarts Clinical Trials, Salix, Sandoz, Setpoint, Shire, Teva, Tigenix, Tillotts, Topivert, Versant and Vifor; and received speaker fees from Abbvie, Ferring, Johnson & Johnson, Merck Sharp Dome, Mundipharma, Norgine, Pfizer, Shire, Millenium/Takeda, Tillotts and Vifor. CNB is supported in part by the Bingham Chair in Gastroenterology. He has served on advisory boards to Abbvie Canada, Janssen Canada, Shire Canada, Takeda Canada and Pfizer Canada and has consulted to Mylan Pharmaceuticals; he has received unrestricted educational grants from Abbvie Canada, Janssen Canada, Shire Canada and Takeda Canada. RG has received consulting and speaking fees from AbbVie, Janssen, MSD, Ferring, Takeda and Baxter; and research grants from AbbVie. GJM has served as speaker or advisory board member for AbbVie, Angelini, Astellas, Danone, MSD, Falk Pharma, Ferring, Hospira, Janssen, Omega Pharma, Otsuka, Pharmacosmos, Pfizer, Sandoz and Takeda; as consultant for Janssen, MSD, Takeda and Omega Pharma; and has received research grants in the last 3 years from Menarini, AbbVie, and MSD. BS has served on advisory boards for Dann and Yakult North American Probiotic Council, Second Genome, Lilly and Enterome; and receives grant support from Janssen, Salix and GSK. MSS has received consulting and speaker fees from AbbVie, Amgen, Ferring, Janssen, Merck, Pfizer, Prometheus, Shire and Takeda; and research funding from AbbVie, Janssen, Prometheus and Takeda. IEK has served as a consultant and on advisory boards for AbbVie and MSD. CO is the principal investigator for Redhill Pharma. DPBM is a consultant for UCB, Jannsen, Merck and Second Genome. WR has served as a speaker for Abbott Laboratories, Abbvie, Aesca, Aptalis, Centocor, Celltrion, Danone Austria, Elan, Falk Pharma GmbH, Ferring, Immundiagnostik, Mitsubishi Tanabe Pharma Corporation, MSD, Otsuka, PDL, Pharmacosmos, Schering-Plough, Shire, Takeda, Therakos, Vifor and Yakult; as a consultant for Abbott Laboratories, Abbvie, Aesca, Amgen, AM Pharma, Astellas, Astra Zeneca, Avaxia, Bioclinica, Biogen IDEC, Boehringer-Ingelheim, Bristol-Myers Squibb, Cellerix, Chemocentryx, Celgene, Centocor, Celltrion, Covance, Danone Austria, Elan, Falk Pharma GmbH, Ferring, Galapagos, Genentech, Gilead, Grünenthal, ICON, Index Pharma, Inova, Janssen, Johnson & Johnson, Kyowa Hakko Kirin Pharma, Lipid Therapeutics, MedImmune, Millenium, Mitsubishi Tanabe Pharma Corporation, MSD, Nestle, Novartis, Ocera, Otsuka, PDL, Pharmacosmos, Pfizer, Procter & Gamble, Prometheus, Robarts Clinical Trial, Schering-Plough, Second Genome, Setpointmedical, Takeda, Therakos, Tigenix, UCB, Vifor, Zyngenia and 4SC; as an advisory board member for Abbott Laboratories, Abbvie, Aesca, Amgen, AM Pharma, Astellas, Astra Zeneca, Avaxia, Biogen IDEC, Boehringer-Ingelheim, Bristol-Myers Squibb, Cellerix, Chemocentryx, Celgene, Centocor, Celltrion, Danone Austria, Elan, Ferring, Galapagos, Genentech, Grünenthal, Inova, Janssen, Johnson & Johnson, Kyowa Hakko Kirin Pharma, Lipid Therapeutics, MedImmune, Millenium, Mitsubishi Tanabe Pharma Corporation, MSD, Nestle, Novartis, Ocera, Otsuka, PDL, Pharmacosmos, Pfizer, Procter & Gamble, Prometheus, Schering-Plough, Second Genome, Setpointmedical, Takeda, Therakos, Tigenix, UCB, Zyngenia and 4SC; and has received research funding from Abbott Laboratories, Abbvie, Aesca, Centocor, Falk Pharma GmbH, Immundiagnsotik and MSD. GR has consulted to Abbot, Abbvie, Augurix, Boehringer, Calypso, FALK, Ferring, Fisher, Genentech, Essex/MSD, Novartis, Pfizer, Phadia, Roche, UCB, Takeda, Tillots, Vifor, Vital Solutions and Zeller; has received speaker's honoraria from Astra Zeneca, Abbott, Abbvie, FALK, MSD, Phadia, Tillots, UCB and Vifor; has received educational grants and research grants from Abbott/Abbvie, Ardeypharm, Augurix, Calypso, Essex/MSD, FALK, Flamentera, Novartis, Roche, Takeda, Tillots, UCB and Zeller. WK has received fees for advising from Dr. Falk Pharma GmbH, Ferring International, GA-Analysis; support for research from Dr. Falk Pharma GmbH and GA-Analysis; grants for lectures from Abbvie, Ardeypharm, Falk Foundation, Ferring Arzneimittel, GA-Analysis, Institut Allergosan, Nikkiso, Otsuka and Recordati. CT has served as a speaker for Dr. Falk Pharma, Tillotts Pharma, Ferring, MSD and AstraZeneca. SS has received consulting fees from AbbVie, Boehringer; Celltrion/Mundipharma, Jansen, Novartis, Merck, Pfizer/Hospira, Sanofi, Takeda and UCB; and speaking fees from AbbVie Ferring, Falk, Merck, Takeda and Shire. SD has served as a speaker, a consultant and an advisory board member for Abbvie, Ferring, Hospira, Johnson & Johnson, Merck, Millennium Takeda, Mundipharma, Pfizer, Tigenix, UCB Pharma and Vifor. WS reports grant support from Receptos, Exact Sciences, Amgen, the American College of Gastroenterology and the Broad Foundation; grant support and personal fees from Prometheus Laboratories, AbbVie, Boehringer Ingelheim, Takeda, Atlantic Pharmaceuticals, Janssen, Bristol-Myers Squibb, Genentech, Pfizer and Nutrition Science Partners; and personal fees from Kyowa Hakko Kirin, Millennium Pharmaceuticals, Celgene Cellular Therapeutics, Santarus, Salix Pharmaceuticals, Catabasis Pharmaceuticals, Vertex Pharmaceuticals, Warner Chilcott, Gilead Sciences, Cosmo Pharmaceuticals, Ferring Pharmaceuticals, Sigmoid Biotechnologies, Tillotts Pharma, Am Pharma BV, Dr. August Wolff, Avaxia Biologics, Zyngenia, Ironwood Pharmaceuticals, Index Pharmaceuticals, Nestle, Lexicon Pharmaceuticals, UCB Pharma, Orexigen, Luitpold Pharmaceuticals, Baxter Healthcare, Ferring Research Institute, Amgen, Novo Nordisk, Mesoblast, Shire, Ardelyx, Actavis, Seattle Genetics, MedImmune (AstraZeneca), Actogenix NV, Lipid Therapeutics Gmbh, Eisai, Qu Biologics, Toray Industries, Teva Pharmaceuticals, Eli Lilly, Chiasma, TiGenix, Adherion Therapeutics, Immune Pharmaceuticals, Celgene, Arena Pharmaceuticals, Ambrx, Akros Pharma, Vascular Biogenics, Theradiag, Forward Pharma, Regeneron, Galapagos, Seres Health, Ritter Pharmaceuticals, Theravance, Palatin, Biogen and the University of Western Ontario (owner of Robarts Clinical Trials). AG has served as a consultant for AbbVie, Janssen, Merck and Takeda; a speaker for Janssen; and received research and clinical programme support from AbbVie and Janssen. ST has received Grants/Research Support from AbbVie, IOIBD, Lilly, UCB, Vifor and Norman Collison Foundation; consulting fees from AbbVie, Amgen, Biogen, Boehringer Ingelheim, Bristol-Myers Squibb, Celgene, Chemocentryx, Cosmo, Ferring, Giuliani SpA, GlaxoSmithKline, Lilly, MSD, Neovacs, NovoNordisk, Norman Collison Foundation, Novartis, NPS Pharmaceuticals, Pfizer, Proximagen, Receptos, Shire, Sigmoid Pharma, Takeda, Topivert, UCB, VHsquared and Vifor; speaker fees from AbbVie, Ferring and Takeda. JP has received consultant fees from Abbvie, Boehringer Ingelheim, Celltrion, Galapagos, Genentech-Roche, Janssen, Pfizer, Takeda, TiGenix and Topivert; speaker fees from Abbvie, Celltrion, Janssen, MSD and Pfizer. JFC has served as a consultant or advisory board member for AbbVie, Amgen, Boehringer-Ingelheim, Celgene Corporation, Celltrion, Enterome, Ferring, Genentech, Janssen and Janssen, Medimmune, Merck & Co., Pfizer, Protagonist, Second Genome, Seres, Takeda and Theradiag; a speaker for AbbVie, Ferring, Takeda and Shire; receives research support from Abbvie, Janssen and Janssen, Genentech and Takeda; and has stock options for Intestinal Biotech Development and Genfit. SH is a consultant for AbbVie, Actavis, Amgen, Arena, Astellas Pharma Global, Astra Zeneca, Baxter, Boehringer Ingelheim, Bristol Myers Squibb, Catabasis, Cellgene, Celltrion, Cubist, Ferring, Forest Labs, Genentech, Glenmark, GSK, Hospira, Janssen, Lilly, Lutipold/American Regent, Meda, Nestle, Novartis, Novo Nordisk, Pfizer, Prometheus, Receptos, Salix, Sanofi-Avantis, Seattle-Genetics, Seres Health, Shire, Takeda, Theradiag, Tigenex, UCB Pharma and VHsquared; does clinical research with Abbvie, Amgen, Genentech, GSK, Janssen, Lilly, Lutipold/American Regent, Novartis, Novo Nordisk, Pfizer, Prometheus, Receptos, Sanofi-Avantis, Takeda and UCB Pharma; is a speaker for AbbVie, Janssen and Takeda; and serves on a DSMB for Bristol Myers Squibb. LPB reports consulting fees from Merck, Abbvie, Janssen, Genentech, Mitsubishi, Ferring, Norgine, Tillots, Vifor, Therakos, Pharmacosmos, Pilège, BMS, UCB-pharma, Hospira, Celltrion, Takeda, Biogaran, Boerhinger-Ingelheim, Lilly, Pfizer, HAC-Pharma, Index Pharmaceuticals, Amgen, Sandoz, Forward Pharma GmbH, Celgene, Biogen, Lycera and Samsung Bioepis; and lecture fees from Merck, Abbvie, Takeda, Janssen, Takeda, Ferring, Norgine, Tillots, Vifor, Therakos, Mitsubishi and HAC-pharma.

  • Provenance and peer review Not commissioned; externally peer reviewed.