Friend or Foe? The Interaction Between Human and Artificial Intelligence on Performance in Medical Chart Coding

36 Pages Posted: 24 Jun 2019 Last revised: 21 Aug 2019

See all articles by Weiguang Wang

Weiguang Wang

University of Maryland, Robert H. Smith School of Business

Guodong (Gordon) Gao

University of Maryland - R.H. Smith School of Business

Ritu Agarwal

University of Maryland - Robert H. Smith School of Business

Date Written: June 18, 2019

Abstract

While the impact of artificial intelligence (AI) on jobs and productivity has generated considerable discussion and debate, empirical evidence on how AI interacts with workers at different skill levels is scant. We developed an AI solution for medical chart coding in a publicly traded company and then evaluated its impact on productivity both within and across individual workers. We find evidence that AI improves worker productivity overall, but the productivity gain is mostly associated with human workers’ circadian rhythm. Interestingly, AI is most beneficial in the morning when human performance is also at its peak, rather than afternoon or night when human performance slows down. Results also show that the benefits of AI are dependent on the worker’s skill level: surprisingly, productivity gain does not increase with work experience. Rather, the productivity of junior workers improves significantly more from AI than that of senior workers. While this pattern is inconsistent with skill-biased technological change, further analysis reveals that the performance discrepancy is attributable to senior user resistance. This paper provides new empirical insights into how AI affects knowledge worker productivity, with important implications for wider adoption and use of AI among knowledge workers.

Keywords: artificial intelligence, productivity, worker skills, healthcare, medical coding, circadian rhythm

Suggested Citation

Wang, Weiguang and Gao, Guodong (Gordon) and Agarwal, Ritu, Friend or Foe? The Interaction Between Human and Artificial Intelligence on Performance in Medical Chart Coding (June 18, 2019). Available at SSRN: https://ssrn.com/abstract=3405759 or http://dx.doi.org/10.2139/ssrn.3405759

Weiguang Wang (Contact Author)

University of Maryland, Robert H. Smith School of Business ( email )

College Park, MD
United States

Guodong (Gordon) Gao

University of Maryland - R.H. Smith School of Business ( email )

4325 Van Munching Hall
College Park, MD 20742
United States

HOME PAGE: http://www.rhsmith.umd.edu/faculty/ggao/

Ritu Agarwal

University of Maryland - Robert H. Smith School of Business ( email )

College Park, MD 20742-1815
United States

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
73
Abstract Views
400
rank
344,089
PlumX Metrics