Authors: Takekoshi M et al.
Cureus, June 16, 2026.
Summary
This study evaluated whether an internally developed computer application could reduce the time anesthesiologists spend preparing preoperative anesthesia informed consent forms.
Documentation requirements can increase anesthesiologists’ workload and cognitive burden. The investigators developed a semi-automated, institution-specific application using Python and generative AI-assisted programming.
The application was designed to work with the hospital’s existing electronic medical record process. It automatically selected relevant checkboxes and inserted required information into anesthesia consent forms. However, it remained a semi-automated tool rather than independently completing or approving the documentation.
The researchers evaluated the application using a randomized crossover study involving 20 anesthesiologists. Each participant prepared informed consent forms for five simulated patients both manually and with the application.
The participants were divided into two groups:
• Group A completed the forms manually first and then used the application.
• Group B used the application first and then completed the forms manually.
This crossover design helped reduce the possibility that the order of testing would influence the findings.
The average preparation time for five consent forms was:
With the application: 9.7 minutes
Without the application: 16.0 minutes
The application saved an average of 6.2 minutes for every five forms prepared. This represented an approximately 40% reduction in preparation time and was statistically significant.
All participants expressed an interest in continuing to use the application:
• Eleven anesthesiologists, or 55%, reported that they very much wanted to continue using it.
• Nine anesthesiologists, or 45%, reported that they wanted to continue using it.
The study provides proof-of-concept evidence that clinicians can use generative AI-assisted programming to build customized tools that address institution-specific documentation problems.
What You Should Know
A customized application reduced the time required to prepare five anesthesia informed consent forms from 16 minutes to approximately 9.7 minutes.
This represented a time savings of approximately 40%.
Every participating anesthesiologist wanted to continue using the application.
The application was developed specifically for the institution’s electronic medical record and consent-form process. Its results may not automatically apply to other hospitals with different documentation systems.
The tool was semi-automated. Clinicians remained responsible for reviewing the information and ensuring that the consent documentation was complete and accurate.
The study involved only 20 anesthesiologists and simulated patient cases. Additional evaluation in routine clinical practice is needed.
The findings demonstrate that relatively small, internally developed applications may produce meaningful efficiency improvements when they are designed around a specific clinical workflow.
Generative AI may be particularly valuable as a programming assistant that helps clinicians create customized administrative tools, even when the final application itself does not directly generate clinical decisions.
Hospitals considering similar tools must address data security, electronic medical record integration, software maintenance, regulatory requirements, and human review of all generated documentation.
Thank you to Cureus for allowing us to summarize this article.