Article Text

Download PDFPDF
From clinical trials to clinical practice: how should we design and evaluate prediction models in the care of IBD?
  1. Ryan William Stidham1,2,
  2. Andrew Vickers3,
  3. Karandeep Singh4,
  4. Akbar K Waljee1,5
  1. 1 Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, USA
  2. 2 Department of Computational Medicine and Bioinformatics, University of Michigan Medical School, Ann Arbor, Michigan, USA
  3. 3 Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York, USA
  4. 4 Department of Learning Health Sciences, University of Michigan, Ann Arbor, Michigan, USA
  5. 5 Center for Clinical Management Research, Department of Veterans Affairs, VA Ann Arbor Healthcare System, Ann Arbor, Michigan, USA
  1. Correspondence to Dr Akbar K Waljee, Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MI 48105, USA; awaljee{at}

Statistics from

As medical treatment options for Crohn’s disease (CD) expand, clinicians need better tools to gauge a patient’s degree of disease burden and ultimately determine whether the therapeutic plan will provide sufficient benefit. Endoscopic scores such as the Simple Endoscopic Score for Crohn Disease (SES-CD) measure the degree of mucosal injury to assess therapeutic change. Though shown in uncontrolled studies to predict some CD clinical outcomes, the SES-CD was not expressly designed to predict the durability of sufficient therapeutic response.1 In Gut, Narula et al present their work optimising the weighting of individual endoscopic features on baseline colonoscopy to predict CD endoscopic remission at 1 year.2 This group demonstrated that the proposed modified multiplier SES-CD (MM-SES-CD) score better predicted endoscopic remission at 1 year compared with using the original SES-CD (area under the curve (AUC) 0.82 vs 0.60, respectively).

Beyond the improvement in model performance, MM-SES-CD feature weighting highlights and quantifies the differential importance of endoscopic feature distribution across anatomic segments. The MM-SES-CD model heavily weights findings in the ileum relative to other sections of the colon, indicating baseline severity at the ileum has a particular influence on outcomes and treatment response. This could be supported by emerging genomic and transcriptional data that suggest ileal or ileocolonic disease behave differently compared with purely colonic disease.3

As the field of gastroenterology prepares to move forward with prediction …

View Full Text


  • KS and AKW are joint senior authors.

  • Twitter @AkbarWaljee

  • Contributors All authors contributed equally.

  • Funding RWS: The work was supported by the National Institute of Health grant NIDDK R01DK124779. KS: The work was supported by Blue Cross Blue Shield of Michigan (for unrelated work). AW: The work was supported by the U.S. Department of Defense and grants from the Department of Veterans Affairs (for unrelated work). AV: The work was supported in part by grants from the National Cancer Institute P30 CA008748.

  • Competing interests R.W.S. has served as a consultant or on advisory boards for AbbVie, Janssen, Takeda, Gilead, Eli Lilly, Exact Sciences, and Corrona.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Commissioned; externally peer reviewed.

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.

Linked Articles