Diagnostic accuracy for the extent of Crohn's disease: a prospective comparison of magnetic resonance enterography and ultrasound -The METRIC study

Research output: Contribution to journalArticle

Authors

  • Stuart A Taylor
  • Gauraang Bhatnagar
  • Steve Morris
  • Florina Tomini
  • Anne Miles
  • Rachel Baldwin-Cleland
  • Stuart Bloom
  • Arun Gupta
  • Peter J Hamlin
  • Ailsa L Hart
  • Antony Higginson
  • Ilan Jacobs
  • Sara McCartney
  • Charles D Murray
  • Andrew Plumb
  • Richard C Pollok
  • Manuel Rodriguez-Justo
  • Zainib Shabir
  • Andrew Slater
  • Damian Tolan
  • Simon Travis
  • Alastair Windsor
  • Peter Wylie
  • Ian Zealley
  • Steve Halligan

Colleges, School and Institutes

External organisations

  • Applied Health Research, University College London,
  • Centre for Medical Imaging, University College London (UCL), London, UK.
  • University College London
  • St. George's Healthcare National Health Service Trust
  • Portsmouth Hospitals NHS Trust
  • University of Oxford
  • St James' Hospital
  • NHS Foundation Trust
  • Ninewells Hospital
  • UCL

Abstract

Background: Magnetic resonance enterography (MRE) and enteric ultrasound (US) are used to image Crohn’s disease (CD) patients. We compared their diagnostic accuracy for presence, extent and activity of enteric CD.

Objectives: To compare diagnostic accuracy, observer variability, acceptability, diagnostic impact and cost effectiveness of MRE and US in newly diagnosed or relapsing CD.

Design: Prospective multicenter cohort study.

Setting: Eight NHS hospitals.

Participants: Consecutive patients ≥16 years, newly diagnosed with CD, or with established disease and suspected relapse.

Tests: MRE and US.

Main outcome measures: Primary outcome-per patient sensitivity difference between MRE and US for small bowel (SB) CD extent. Secondary outcomes included specificity for SB disease extent; sensitivity and specificity for SB and colonic disease extent, and separately for presence; identification of active disease; interobserver variation; patient acceptability; diagnostic impact; cost effectiveness.

Results: Of 518 patients assessed, 335 entered the trial, with 51 excluded, giving a final cohort of 284 (133 and 151 in new diagnosis and relapse cohorts respectively). Across the whole cohort, MRE sensitivity for SB disease extent (80% [95% Confidence interval [CI] 72 to 86] was significantly greater than US (70% [62 to 78]), a 10% (1 to 18) difference, p=0.027. MRE specificity for SB disease extent was significantly greater than US (95% [85 to 98]) vs. 81% [64 to 91]) respectively, a 14% (1 to 27) difference. MRE sensitivity for SB disease presence was 97% (91 to 99), significantly greater than US (92% [84 to 96]), a 5% (1 to 9) difference. MRE and US specificity for SB disease presence was 96% (86 to 99) and 84% (65 to 94) respectively, a 12% (0 to 25) difference. Test sensitivities for SB disease presence and extent were similar in the two cohorts. US had significantly greater sensitivity for colonic disease presence than MRE in newly diagnosed patients (67% [49 to 81] vs. 47% [31 to 64%]), a 20% (1 to 39) difference. Overall, MRE sensitivity for active SB disease (96% [92 to 99]) was significantly greater than for US (90% [82 to 95]), a 6% (2 to 11) difference. There was some disagreement between readers for both tests. Eight-eight percent of patients rated MRE as very or fairly acceptable, significantly lower than US (99%). Therapeutic decisions based on MRE and US alone agreed with the final decision in 122/158 (77%) and 124//158 (78%) respectively. There were no differences in costs or QALYs between tests.

Limitations: MRE and US were interpreted by practitioners blinded to clinical data (but not patient cohort), which does not reflect use in clinical practice.

Conclusions: MRE has higher accuracy for the presence, extent and activity of SB CD than US. Both tests have variable interobserver agreement and are broadly acceptable to patients, although US produces less patient burden and is preferred. Diagnostic impact and cost-effectiveness are similar.

Future work: Comparative utility of MRE and US for treatment response assessment, and investigation of nonspecific abdominal symptoms to confirm or refute Crohn’s disease.

Study registration: ISRCTN03982913.

Funding details: NIHR Health Technology Assessment Programme

Details

Original languageEnglish
JournalHealth Technology Assessment
Publication statusPublished - 1 Aug 2019