Prognostic tools for identification of high risk in people with Crohn's disease: systematic review and...

Crohn's disease is a lifelong condition that can affect any segment of the gastrointestinal tract. Some people with Crohn's disease may be at higher risk of following a severe course of disease than others and being able to identify the level of risk a patient has could lead to personalised management.

To assess the prognostic test accuracy, clinical impact and cost-effectiveness of two tools for the stratification of people with a diagnosis of Crohn's disease by risk of following a severe course of disease.

The data sources MEDLINE, EMBASE, the Cochrane Database of Systematic Reviews and the Cochrane Central Register of Controlled Trials were searched to inform the systematic reviews on prognostic accuracy, clinical impact of the prognostic tools, and economic evaluations. Additional data sources to inform the review of economic evaluations were NHS Economic Evaluation Database, Database of Abstracts of Reviews of Effects and the Health Technology Assessment Database.

Systematic reviews of electronic databases were carried out from inception to June 2019 for studies assessing the prognostic accuracy and clinical impact of the IBDX^® (Crohn's disease Prognosis Test; Glycominds Ltd, Lod, Israel) biomarker stratification tool and the PredictSURE-IBD™ (PredictImmune Ltd, Cambridge, UK) tool. Systematic reviews of studies reporting on the cost-effectiveness of treatments for Crohn's disease were run from inception to July 2019. Two reviewers independently agreed on studies for inclusion, assessed the quality of included studies and validated the data extracted from studies. Clinical and methodological heterogeneity across studies precluded the synthesis of data for prognostic accuracy. A de novo economic model was developed to compare the costs and consequences of two treatment approaches - the 'top-down' and 'step-up' strategies, with step-up considered standard care - in people at high risk of following a severe course of Crohn's disease. The model comprised a decision tree and a Markov cohort model.

Sixteen publications, including eight original studies (n = 1478), were deemed relevant to the review of prognostic accuracy. Documents supplied by the companies marketing the prognostic tools were also reviewed. No study meeting the eligibility criteria reported on the sensitivity or specificity of the IBDX biomarker stratification tool, whereas one study provided estimates of sensitivity, specificity and negative predictive value for the PredictSURE-IBD tool. All identified studies were observational and were considered to provide weak evidence on the effectiveness of the tools. Owing to the paucity of data on the two tools, in the base-case analysis the accuracy of PredictSURE-IBD was assumed to be 100%. Accuracy of IBDX was assumed to be 100% in a scenario analysis, with the cost of the tests being the only difference between the analyses. The incremental analysis of cost-effectiveness demonstrated that top-down (via the use of PredictSURE-IBD in the model) is more expensive and generates fewer quality-adjusted life-years than step-up (via the standard care arm of the model).

Despite extensive systematic searches of the literature, no robust evidence was identified of the prognostic accuracy of the biomarker stratification tools IBDX and PredictSURE-IBD.

Although the model indicates that standard care dominates the tests, the lack of evidence of prognostic accuracy of the two tests and the uncertainty around the benefits of the top-down and step-up treatment approaches mean that the results should be interpreted as indicative rather than definitive.

This study is registered as PROSPERO CRD42019138737.

This project was funded by the National Institute for Health Research (NIHR) Evidence Synthesis programme and will be published in full in Health Technology Assessment; Vol. 25, No. 23. See the NIHR Journals Library website for further project information.