08/22/23

Assessing the utility of chatgpt throughout the entire clinical workflow: development and usability study.

Rao, Arya; Pang, Michael; Kim, John; Kamineni, Meghana; Lie, Winston; Prasad, Anoop K; Landman, Adam; Dreyer, Keith; Succi, Marc D. J Med Internet Res.

Summary

Interest in testing ChatGPT as a clinical tool is increasing. This study asked ChatGPT to provide a differential diagnosis, diagnostic testing, final diagnosis, and care management for 36 previously published clinical vignettes. ChatGPT had an overall accuracy of 72%, with the highest level of accuracy at the final diagnosis stage (77%). 

The PSNet collection materials were developed with funding under multiple contracts from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services.

Rao A, Pang M, Kim J, et al. Assessing the utility of ChatGPT throughout the entire clinical workflow: development and usability study. J Med Internet Res. 2023;25:e48659.