The density of routinely collected neurology data depends on patient visit type: an investigation using the Observational Medical Outcomes Partnership Common Data Model

BackgroundThe Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) is a standardised framework for organising healthcare data. This study investigates the practicality of using the OMOP CDM to analyse data on neurology patients. MethodsAn outpatient neurology patient cohort was defined on the basis of having attended at least one neurology outpatient appointment between 01 April 2022 and 31 March 2023 (n=23,862). All data collected at visits made by this cohort between 01 April 2021 to 31 March 2024 was extracted. The cohort was then divided into 4 sub-cohorts according to appointment types attended: outpatient appointment(s) only (n=15,255); outpatient appointment(s) and inpatient stay(s) (n=2750); outpatient appointment(s) and emergency department attendance(s) (n=1658); outpatient appointment(s), inpatient stay(s) and emergency department attendance(s) (n=4199). ResultsWe found there to be more OMOP-mapped data available for patients who had at least one inpatient stay or emergency department attendance than for those with only outpatient appointments. Notably, an average of 0 out of 100 patients in the outpatient only sub-cohort had a record of a condition, compared to 100 out of 100 patients in the sub-cohort with outpatient appointments, emergency attendances and inpatient stays. ConclusionsNeurology outpatients have far less data recorded than inpatients or patients attending ED. This disparity arises from the lack of outpatient diagnostic coding and impairs the advancement of research in this area. KEY MESSAGESO_ST_ABSWhat is already known on this topicC_ST_ABSthe OMOP common data model (CDM) is being adopted by the NHS to provide a uniform structure to the data within the NHS Secure Data Environments to support research. We know that outpatient coding is not mandated so diagnoses at outpatient appointments are not regularly recorded in EHRs. What this study addswe investigate the variable volume of data available for research through a secondary care dataset that has been converted to the OMOP CDM. We show that outpatients have far less data recorded than inpatients or patients attending ED, in terms of both volume and type of data. How this study might affect research, practice or policythis study highlights the need for data systems such as SDEs to be based on data which is complete. We also highlight the importance of ensuring that data recording for outpatients is as complete as it is for inpatients and ED.