Comparative effectiveness of i-SCAN™ and high-definition white light characterizing small colonic polyps

World J Gastroenterol. 2012 Nov 7;18(41):5905-11. doi: 10.3748/wjg.v18.i41.5905.

Abstract

Aim: To evaluate accuracy of in vivo diagnosis of adenomatous vs non-adenomatous polyps using i-SCAN digital chromoendoscopy compared with high-definition white light.

Methods: This is a single-center comparative effectiveness pilot study. Polyps (n = 103) from 75 average-risk adult outpatients undergoing screening or surveillance colonoscopy between December 1, 2010 and April 1, 2011 were evaluated by two participating endoscopists in an academic outpatient endoscopy center. Polyps were evaluated both with high-definition white light and with i-SCAN to make an in vivo prediction of adenomatous vs non-adenomatous pathology. We determined diagnostic characteristics of i-SCAN and high-definition white light, including sensitivity, specificity, and accuracy, with regards to identifying adenomatous vs non-adenomatous polyps. Histopathologic diagnosis was the gold standard comparison.

Results: One hundred and three small polyps, detected from forty-three patients, were included in the analysis. The average size of the polyps evaluated in the analysis was 3.7 mm (SD 1.3 mm, range 2 mm to 8 mm). Formal histopathology revealed that 54/103 (52.4%) were adenomas, 26/103 (25.2%) were hyperplastic, and 23/103 (22.3%) were other diagnoses include "lymphoid aggregates", "non-specific colitis," and "no pathologic diagnosis." Overall, the combined accuracy of endoscopists for predicting adenomas was identical between i-SCAN (71.8%, 95%CI: 62.1%-80.3%) and high-definition white light (71.8%, 95%CI: 62.1%-80.3%). However, the accuracy of each endoscopist differed substantially, where endoscopist A demonstrated 63.0% overall accuracy (95%CI: 50.9%-74.0%) as compared with endoscopist B demonstrating 93.3% overall accuracy (95%CI: 77.9%-99.2%), irrespective of imaging modality. Neither endoscopist demonstrated a significant learning effect with i-SCAN during the study. Though endoscopist A increased accuracy using i-SCAN from 59% (95%CI: 42.1%-74.4%) in the first half to 67.6% (95%CI: 49.5%-82.6%) in the second half, and endoscopist B decreased accuracy using i-SCAN from 100% (95%CI: 80.5%-100.0%) in the first half to 84.6% (95%CI: 54.6%-98.1%) in the second half, neither of these differences were statistically significant.

Conclusion: i-SCAN and high-definition white light had similar efficacy predicting polyp histology. Endoscopist training likely plays a critical role in diagnostic test characteristics and deserves further study.

Keywords: Adenoma; Colonic polyps; Colonoscopy; Comparative effectiveness; Virtual chromoendoscopy.

Publication types

  • Comparative Study
  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adenoma / pathology*
  • Aged
  • Biopsy
  • Chi-Square Distribution
  • Chromogenic Compounds*
  • Colonic Neoplasms / pathology*
  • Colonic Polyps / pathology*
  • Colonoscopy / methods*
  • Comparative Effectiveness Research
  • Female
  • Humans
  • Light*
  • Linear Models
  • Male
  • Middle Aged
  • North Carolina
  • Pilot Projects
  • Predictive Value of Tests
  • Prospective Studies

Substances

  • Chromogenic Compounds