Gout affects 2-3 out of every 100 people in the UK. Gout causes episodes of painful joint inflammation called flares. People with gout may be offered medication to reduce the number of flares that they experience which reduces pain, inconvenience and also the chances of joint damage. However, this requires patients to take medication every day and have blood tests to see how well the medication is working. Therefore, this can be a difficult decision for patients. It is important they have good information to help them make this decision.
Although gout is common, the way it will affect an individual varies significantly and we do not understand well what affects the course of the disease. The purpose of this study is to use data collected by GPs in routine consultations to gain a better understanding of what happens to people after they are diagnosed with gout. We will look at how frequently people have flares and whether there are things we can identify that affect this. This information will allow us to provide better information to clinicians and patients about gout which will help patients make decisions about their care.
This is a cohort study comprising patients with incident gout. Following patients from initial diagnosis will allow characteristics to be identified at baseline that may predict, or predispose towards, the severity of disease (measured through the frequency of gout flares).
Patients aged 20 and above will be included in the cohort if they have a new coded diagnosis of gout between 1st January 2010 and 31st December 2018. Previous research has validated the electronic coding of gout in UK electronic medical records to an acceptable level. (6) Included patients will require at least 1 complete year or follow up (starting 30 days after the index date.)
Patients will be excluded if they have a prescription of urate lowering therapy (ULT) prior to, or within 30 days subsequent from, the index date.
The primary outcome variable will be gout flares. Gout flares will be defined according to criteria previously used in the literature. (4)
The cohort will be described. The pattern of gout flares will be described using the number of flares, time to second flare, proportion of people having second flare by the end of each complete year of follow up and the proportion of people having two or more flares in a 12 month period and when this occurs subsequent from diagnosis.
The association between predefined predictor variables and the frequency of flares will be analysed using logistic regression. A time to event analysis considering multiple events will be performed using the Anderson-Gill Cox model.
A second logistic analysis will be undertaken looking at the prescribing of ULT. Number of flares will be included as a predictor variable and the practice ID will be included as a random effect to account for commonalities in prescribing behaviours amongst GP practices.
Gout flares; prescribing of urate lowering therapy
Samuel Finnikin - Chief Investigator - University of Birmingham
Samuel Finnikin - Corresponding Applicant - University of Birmingham
Christian Mallen - Collaborator - Keele University
Edward Roddy - Collaborator - Keele University
Patient Level Index of Multiple Deprivation