Abstract

e18304 Background: Watson for Oncology (WFO) is a cognitive computing system developed by the IBM and trained by Memorial Sloan Kettering Cancer Center to provide evidence-based treatment recommendation for cancer patients. This study is to explore the impact on clinical treatment decision and patient communication based on the standard workflow we established of applying WFO in real-world clinical practice. Methods: From May 2018 to January 2019, 57 patients with advanced malignant tumors, including lung, breast, colorectal, gastric, esophageal, liver, ovarian and endometrial cancer, treated at Oncology Department of Beijing Chaoyang Integrative Medicine Emergency Medical Center were included in the study, of which 28 males and 29 females aged 29-89 years old. Treatment plan of each patient generated by WFO was compared with NCCN guidelines and Chinese local CSCO guidelines, and then evaluated and determined by MDT for final decision. Proportion of WFO recommendation fully or partially followed by clinicians was analyzed. Moreover, survey by questionnaire was conducted to collect feedback from these patients on their satisfaction of treatment and communication process. Results: Of all 57 patients enrolled, WFO’s options were highly concordant with NCCN guidelines and CSCO guidelines, and the finally adopted treatment plans after deliberately evaluated and discussed by MDT were 100% in the WFO recommended or for consideration category. Patient satisfaction and the compliance with instruction was significantly improved compared with historical data. Localization was most required in gastric, liver and esophageal cancer due to local drugs and therapies not available in WFO. Conclusions: This study verified that the artificial intelligent clinical decision supporting system can be applied in Chinese clinical practice to achieve high patient satisfaction and good efficacy. WFO meets Chinese and international standards for treatment, which will promote the standardized development of cancer treatment in China. Optimization, iteration, upgrading and localization are required for better practicability of WFO.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call