AI Discharge Reports - Futurshealth

Cases of Good Practice

Discharge Reports for
Cataract Surgery Through
Generative AI

A Challenge

Cataract surgery (phacoemulsification) is a very common procedure in healthcare centres. During patient care, healthcare workers can record all information about the care process in the medical record, both in a structured format and as free text.

A draft for the discharge report is created after the surgical procedure based off the previous information that can be found in the medical record, therefore optimising the time spent by consultants to create said report. Not only that, but it also improves the quality of the information shared with the patient and helps codify the episode of care. The consultant validates or corrects the result of the draft, ensuring the information is true for the patient and the events.

Value Added

The business logic was defined to generate the corresponding output report using generative AI based on the input data. The data files were obtained and a map was carried out between input data and the discharge report to be generated.

The project was developed in collaboration with Plain Concepts, Microsoft partner, using agile methodology.

Valor aportado

Following interactions between the development team and end users, and after AI settings, a discharge report draft was generated with the quality criteria expected: summary of health history, clinical diagnosis, procedure performed, secondary diagnosis, and recommendations.

The report is also written in simple language, with no abbreviations throughout the content. In addition, it optimises the work of consultants, improves the quality of reports and, therefore, the coding of episodes, making it easier for patients to understand the report.