Abstract
There have been significant changes in the management of patients with diabetes in the United Kingdom (UK) over the last 20 years, with the introduction of analogue insulins and the drive toward tighter glycaemic control, with its concomitant risk of increased hypoglycaemia. This thesis describes changes in insulin prescribing in the UK between 1999 and 2009, examines the associations between a patient’s characteristics and the type of insulin they are prescribed, and between a patient’s insulin exposure and their risk of serious hypoglycaemic events. Data for the study was obtained from the General Practice Research Database (GPRD), a UK primary-care database providing anonymised patient data for approximately 6% of the UK population. A cohort of insulin-using patients was identified, from which subgroups of patients newly-starting insulin therapy and patients switching insulins were extracted for analysis. There was a rapid uptake of analogue insulins and by 2009 they accounted for more than 80% of insulin prescriptions. While some of a patient’s characteristics, including age and ethnicity, influenced the choice between analogue and human insulin in new insulin users, the largest effect size was the calendar year in which therapy started: clinician (or patient) choice seemed to be the driving factor. The risk of serious hypoglycaemic events under exposure to human or analogue insulin was examined in new users and insulin switchers. After adjusting for patient characteristics, some of which, including ethnicity, were associated with a change in the risk of an event, the study confirmed that new users of long-acting analogue insulins had a lower risk of serious hypoglycaemic events compared with users of human insulin. However, following a switch from human insulin no difference in the risk of serious hypoglycaemic events was found in patients switching to either analogue insulin or to a different combination of human insulins.