Identifying Long COVID Patients Using General Practice Data: Challenges, Classification and Long COVID Patterns

General practice data, extracted from electronic medical records, holds immense potential to generate a wealth of public health knowledge. But it is not without challenges. Our aim in this study was to identify long COVID patients within a large general practice dataset through data classification. We discuss the classification and its validation, and present initial data patterns for the identified long COVID cohort. We found significant variation in how general practitioners document and describe long COVID presentations. Less than half of the identified long COVID patients had a documented acute COVID infection. The highest proportion of long COVID patients were female and those 40-49 years of age. Overall, this study highlights key lessons for researchers utilizing general practice data, particularly in the context of long COVID, and underscores the vital importance of collaboration between researchers, general practitioners, and data custodians to ensure the robustness of data underpinning knowledge translation.