Can Artificial Intelligence Create an Accurate Colonoscopy Bowel Preparation Prompt?

Area of Science:

• Gastroenterology
• Artificial Intelligence in Medicine
• Medical Communication

Background:

• Colorectal cancer screening relies on colonoscopy, but poor bowel preparation affects procedure efficiency and accuracy.
• Up to 25% of colonoscopies suffer from inadequate preparation, leading to repeat procedures and missed diagnoses.
• Artificial intelligence (AI) is increasingly adopted in healthcare for various applications, including patient education and medical reporting.

Purpose of the Study:

• To evaluate the accuracy and physician perception of an AI-generated (ChatGPT) precolonoscopy bowel preparation prompt.
• To assess the ease of understanding, scientific accuracy, and physician comfort level with using AI-generated patient instructions.

Main Methods:

• A nonrandomized, cross-sectional study surveyed gastroenterologists at a major healthcare institution.
• 208 gastroenterologists with valid email addresses were included in the survey.
• Descriptive statistics were used to analyze survey responses on the AI-generated prompt.

Main Results:

• 91% of physicians found the AI prompt easy to understand.
• 95% considered the prompt scientifically accurate.
• 66% were comfortable providing the AI prompt to patients, though only 32% were confident in identifying it as AI-generated.

Conclusions:

• AI tools like ChatGPT can create effective bowel preparation prompts for colonoscopies.
• Physicians can integrate AI into clinical practice to enhance patient communication regarding bowel preparation.
• AI has the potential to improve the efficiency and ease of patient instruction for colonoscopy preparation.