AI Trained Computer Program Can Monitor Health Forums To Detect Adverse Drug Reactions

MedicalResearch.com Interview with:

Kavita Sarin, M.D., Ph.D.

Dr. Sarin

Kavita Sarin, M.D., Ph.D.
Assistant Professor of Dermatology
Stanford University Medical Center

MedicalResearch.com: What is the background for this study? What are the main findings? 

Response: Drug reactions occur in the majority of patients undergoing cancer therapies. Half of serious drug reactions are detected after market approval which can result in painful complications and interruption in therapy. Post-market drug surveillance platforms such as FDA monitoring rely on medical publications and physician reporting and take time to identify trends. We sought to determine if we could identify trends in patient discussions in internet health forums to more rapidly identify chemotherapeutic drug reactions. We chose skin reactions as a proof-of-principle because patients can more easily describe what they see on their skin.

Julia Ransohoff, a medical student, and Azadeh Nikfarham, an informatics postdoctoral fellow trained a computer to recognize when a patient undergoing anti-cancer treatment with PD-1 antagonists or EGFR-inhibitors described a drug reaction in their internet forum posts.

MedicalResearch.com: What should readers take away from your report?

Response: We were surprised to find that we could detect specific skin drug reactions in internet health forums 6 to 9 months before any published reports and could even identify drug reactions that have not yet been published. This demonstrates the robust potential of internet health forums to provide useful information for doctors and medical researchers.

MedicalResearch.com: What recommendations do you have for future research as a result of this work?

Response: Moving  forward, it will be important to show if this approach can be generalized to survey other types of skin and internal drug reactions. 

MedicalResearch.com: Is there anything else you would like to add?

Response: We thank the Inspire team for allowing us access to their internet health forum for this study.

Citations:

Ransohoff JD, Nikfarjam A, Jones E, Loew B, Kwong BY, Sarin KY, Shah NH. Detecting Chemotherapeutic Skin Adverse Reactions in Social Health Networks Using Deep Learning. JAMA Oncol. Published online March 01, 2018. doi:10.1001/jamaoncol.2017.5688

The information on MedicalResearch.com is provided for educational purposes only, and is in no way intended to diagnose, cure, or treat any medical or other condition. Always seek the advice of your physician or other qualified health and ask your doctor any questions you may have regarding a medical condition. In addition to all other limitations and disclaimers in this agreement, service provider and its third party providers disclaim any liability or loss in connection with the content provided on this website.

 

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.