In the UK a large amount of data is collected during the routine treatment of patients. This data, referred to as ‘routinely collected data’ within the health research community is increasingly being used in health research despite not being explicitly collected for this purpose. The fact that data collected for healthcare and administrative purposes is used for research can lead to a number of issues that need to be acknowledged and overcome.

This thesis explores the benefits and limitations of using routinely collected data to research outpatient neurology services in the North-West of England. Neurology services in the UK are under pressure, with large variations in the level of service provided in different geographical areas. We analyse data from an outpatient neurology clinic in the northwest of England covering a diverse population which is dispersed over a large area.

First the current research into neurology services using routinely collected data is explored using a systematic mapping review, identifying gaps and areas for further research. We then use data from the outpatient neurology clinic to explore the issues identified in three separate papers.

Chapter four explores the number of patients using outpatient neurology services, what resources they require and the waiting times they experience. We show that there are a small number of diagnostic categories which account for 60% of new referrals to the clinic, and that waiting times vary by diagnostic category. Chapter five examines referrals for headache patients. Using standardised residuals from a Poisson regression we identify GP practices which have referred an unexpected number of headache patients. Chapter six uses State Sequence Analysis to observe patterns in the types of appointments patients attend over time. We show that there a number of common
patterns of appointments and that these patterns are somewhat related to diagnosis.

Taken together these papers show that routinely collected data can be successfully used to conduct useful and insightful research into outpatient neurology services. Our research also identifies key areas of limitation of using this type of data - such as missing data, the difficulty of accessing data, and the difficulties presented by the lack of diagnostic coding – for which solutions are proposed in the discussion.