Document resource
Background Manchester Centre for Clinical Neurosciences is the sole provider of neurological services for 3.3 million people in the Greater Manchester region, with all outpatient clinic letters available on a single electronic system. We used text mining and natural language processing techniques to extract neurological diagnoses from clinic letters.Methods All patients attending a general or specialist neurology clinic in 13 hospitals in Greater Manchester between January 2018 and November 2023 were included. Free-text diagnoses were processed through MedCAT, mapped to SNOMED-CT, categorised into 13 headline groups and linked to demographics.Results A total of 125,273 patients were included in the final dataset, from whom 175,293 diagnoses were mapped. Headaches (n=26631, 15.19%), epilepsy/seizure (n=24880, 14.19%), movement disorders (n= 17468, 9.97%), neuromuscular disorders (n=16454, 9.39%) and functional disorders (n=16074, 9.17%) were the most frequent diagnoses. Compared to the census data, patients were overrepresented in the most deprived postcodes (IMD quintile 1, 37%) and underrepresented in the least deprived postcodes (IMD quintile 5, 13%). Headache, epilepsy and functional disorders were overrepresented in IMD quintile 1.Conclusion We have generated a dataset of neurological diagnoses, the largest to our knowledge, to aid epidemiological study into each disease category and understanding of health inequalities.Rajiv.Mohanraj@nca.nhs.uk