Classification of Long COVID from General Practitioner Diagnosis Text

Area of Science:

• Medical Informatics
• Public Health
• Clinical Research

Background:

• Accurate identification of Long COVID patients is crucial for understanding the condition.
• Unstructured diagnostic texts from general practitioners pose a significant challenge for patient identification.

Purpose of the Study:

• To propose a classification framework for identifying Long COVID patients from unstructured diagnostic texts.
• To enable a better understanding of Long COVID risk factors, management, and outcomes in Australia.

Main Methods:

• Development of a classification framework.
• Application to unstructured diagnostic texts.
• Analysis of identified Long COVID patient cohorts.

Main Results:

• A functional classification framework for identifying Long COVID patients was developed.
• The framework facilitates the analysis of patient data from general practitioner records.

Conclusions:

• The proposed framework offers a viable solution for identifying Long COVID patients in primary care settings.
• This approach can support research into Long COVID epidemiology, treatment, and patient outcomes in Australia.