Building a Shared, Scalable, and Sustainable Source for the Problem-Oriented Medical Record: Developmental Study.

Area of Science:

• Health Informatics
• Medical Record Management
• Clinical Data Standardization

Background:

• Problem list creation is crucial for medical records but remains a challenge.
• Existing methods struggle to balance clinical language with structured data requirements.

Purpose of the Study:

• To present the development of a shared, multipurpose common problem list at Geneva University Hospitals.
• To integrate free-text clinical language with structured data for secondary uses.

Main Methods:

• Developed a list of 45,946 uniquely identified expressions from clinical documents.
• Encoded expressions across multiple semantic dimensions (e.g., ICD-10, ICPC-2, SNOMED CT).
• Iteratively deployed and refined the list with clinicians, incorporating quality control and real-world usage feedback.

Main Results:

• The problem list was deployed in January 2017 and updated over 4 years.
• Free-text entries decreased from 37.47% to 18.38% during the study period.
• The list became central to the electronic health record, supporting care (e.g., surgical planning) and research (e.g., machine learning predictions).

Conclusions:

• This approach prioritizes clinicians' language and maps it to required structured dimensions.
• It enhances patient problem expression quality and accommodates specific contextual needs.
• The method is scalable, sustainable, and efficient, reducing the burden on clinicians.