Loading...
Paper Number
1099
Paper Type
short
Description
For many healthcare applications a collaboration of humans and algorithms has been shown to be superior to pure automation in terms of performance. However, the healthcare sector is characterized by shortages in personnel, which can lead to an excessive workload for the employees and thus makes automation highly beneficial to reduce human workload. In our paper, we consider a combination of different work modes and evaluate whether humans have to be involved in every instance of a task or whether they can be replaced by an AI for some instances. We analyze the potential of segmenting tasks based on who is involved in their completion: Either an AI or a human complete the task individually, or they complete the task together. Considering the case of surgery duration predictions and using a dataset from a university hospital, we observe that human effort could be decreased while maintaining a high prediction performance.
Recommended Citation
Walzner, Dominik David; Poreschack, Laura Maria; Fuegener, Andreas; Schiffels, Sebastian; and Denz, Christof, "Division of Labor between Humans and Algorithms in Healthcare: The Case of Surgery Duration Predictions" (2023). ICIS 2023 Proceedings. 9.
https://aisel.aisnet.org/icis2023/ishealthcare/ishealthcare/9
Division of Labor between Humans and Algorithms in Healthcare: The Case of Surgery Duration Predictions
For many healthcare applications a collaboration of humans and algorithms has been shown to be superior to pure automation in terms of performance. However, the healthcare sector is characterized by shortages in personnel, which can lead to an excessive workload for the employees and thus makes automation highly beneficial to reduce human workload. In our paper, we consider a combination of different work modes and evaluate whether humans have to be involved in every instance of a task or whether they can be replaced by an AI for some instances. We analyze the potential of segmenting tasks based on who is involved in their completion: Either an AI or a human complete the task individually, or they complete the task together. Considering the case of surgery duration predictions and using a dataset from a university hospital, we observe that human effort could be decreased while maintaining a high prediction performance.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
16-HealthCare