With an increased emphasis on improving quality and decreasing costs, new tools are needed to improve adherence to evidence-based practices and guidelines in endoscopy.
Dr Timothy Imler and colleagues from Indiana, USA investigated the ability of an automated system that uses natural language processing and clinical decision support to facilitate determination of colonoscopy surveillance intervals.
The research team performed a retrospective study at a single Veterans Administration medical center of patients age 40 years and older who had an index outpatient colonoscopy from 2002 through 2009 for any indication except surveillance of a previous colorectal neoplasia.
The team analyzed data from 10,798 reports, with 6379 linked to pathology results, and 300 randomly selected reports.
NLP-based clinical decision support surveillance intervals were compared with those determined by paired, blinded, manual review.
|An automated system that uses NLP can facilitate guideline-recommended adherence surveillance for colonoscopy|
|Clinical Gastroenterology & Hepatology|
The team's primary outcome was adjusted agreement between manual review and the fully automated system.
κ statistical analysis produced a value of 0.74 for agreement between the full text annotation and the NLP-based clinical decision support system.
The research team found that 55 reports differed between manual review and clinical decision support recommendations.
Of these, NLP error accounted for 55%, incomplete resection of adenomatous tissue accounted for 26%, and masses observed without biopsy findings of cancer accounted for 7%.
NLP-based clinical decision support surveillance intervals had higher levels of agreement with the standard than the level agreement between experts.
Dr Imler's team concludes, "A fully automated system that uses NLP, and a guideline-based clinical decision support system can accurately facilitate guideline-recommended adherence surveillance for colonoscopy."