This software could expedite the quality review process and establish new benchmarks in colonoscopy performance.
Key Features of the Software:
- The software automatically integrates endoscopy and pathology reports across various practice settings.
- Developed by Dr. Todd A. Brenner and colleagues at Johns Hopkins Hospital, Baltimore, the software is designed to be accurate and feasible for real-world usage.
- It gathers procedural and histopathology results following colonoscopy, extracts and classifies relevant data, and outputs ADR along with other metrics like cecal intubation rate, Boston Bowel Preparation Score (BBPS), and withdrawal time.
Evaluation and Performance:
- The software was evaluated using endoscopy and pathology reports from 3,809 colonoscopies performed at six centers over three months.
- A validation cohort of 1,384 colonoscopies conducted over a one-month period was manually reviewed six months later for comparison.
- The automated system showed high congruity with manual review, delivering an ADR of 45.1% compared to 44.3% for manual review.
Advantages and Potential Impact:
- The automated approach could make ADR calculation less resource-intensive, overcoming the limitations of manual collation of data.
- It could be a valuable tool for frequent quality audits at many centers, which is currently impractical with manual methods.
- The software’s ability to accurately identify most ADR-qualifying screening colonoscopies suggests its potential in improving colonoscopy quality and patient outcomes.
Future Development:
- Further programming could address issues like missed identifications due to spelling/syntax errors or pending pathology results.
- The platform can be customized to deliver desired quality metrics and edited to match local practice patterns.
- Additional metrics and novel metrics like size-stratified ADR or location-stratified ADR could be integrated into the platform.
The investigators, who disclosed relationships with various medical companies, believe that automating data collection in this manner could help determine which metrics provide clinically meaningful insights and potentially expand the standard performance benchmarks in colonoscopy.