A longitudinal study of polymyalgia rheumatica and its treatment: a CPRD cohort study

Polymyalgia rheumatica (PMR) is the commonest inflammatory rheumatological condition in older adults, with a lifetime prevalence of 1.7% for males and 2.4% for females (Crowson et al A&R 2011). The condition is characterised by aching in the shoulder and hip girdles and morning stiffness, often accompanied by raised inflammatory markers (e.g. ESR, CRP) (Dasgupta et al Rheumatology 2010). With the ageing population (Christensen et al Lancet 2009), the number of people with PMR is set to rise. The mainstay treatment for PMR is long-term, low-dose corticosteroids. It is generally accepted that this treatment lasts for around two years, but there is little empirical evidence to either support or refute this, and anecdotal evidence suggests that treatment may be required for much longer periods, at least in some cases.
It is known that there is a range of potential side-effects from the use of corticosteroids, affecting the gastro-intestinal, musculoskeletal, endocrine, neuropsychiatric and ophthalmic systems. Specific, well-known adverse effects include diabetes, osteoporosis (and resultant fractures), proximal myopathy, peptic ulceration and Cushings syndrome (BNF April 2013). It is however possible that PMR disease activity itself is associated with some of these events. For example, it is plausible that the pain and stiffness of PMR result in a deconditioning of the muscles, resulting in myopathy, or increase the rate of falls, resulting in an increased rate of fractures. In order to be able to distinguish the effects of the disease and its treatment, it is necessary to be able to quantify treatment, as well as to define disease. Hence, more information is needed about how to describe and model corticosteroid treatment regimens.
In UK primary care, all prescriptions are issued electronically and thus are recorded in patients’ notes as a matter of routine. This means that information on the corticosteroid dose prescribed for each person with a diagnosis of PMR is available in routinely collected datasets, such as the Clinical Practice Research Datalink (CPRD).
Despite this wealth of information of prescriptions, methods are lacking to process and analyse this data to fully understand the prescribing process. Whilst this may be straightforward for some drugs, which are routinely prescribed at a set dose, it is likely to be less easy to understand the prescribing of corticosteroids. These drugs, which come in a range of preparations (e.g. prednisolone, dexamethasone), are prescribed for a range of conditions with different starting doses and differing amounts of tapering. This means that it is difficult to make assumptions about how to deal with prescriptions for corticosteroids in the context of primary care databases.
This lack of understanding of the prescription process in PMR means that there is a lack of research evidence as to 1) what treatment regimen is used in practice for the condition and whether this is the same for everyone; and 2) the rate of poor outcomes in PMR and how these relate to the corticosteroid regimen received by the individual.

The Arthritis Research UK Primary Care Centre / Research Institute for Primary Care and Health Sciences have funded a PhD studentship. We now seek to appoint a PhD student to conduct a methodological study to explore how to use routinely collected primary care prescription data to further knowledge of the corticosteroid dosing regimen and how to account for this in epidemiological studies.

Expectations of the Candidate
You will benefit from an established infrastructure for research and work alongside an internationally renowned multidisciplinary team. You will join a vibrant group of over 40 postgraduate students, and will play a full part in the Centre, presenting at internal seminars, attending journal clubs, student group meetings and external seminars. You will have access to specialised training courses and the opportunity to present your research at national or international conferences.

With the support of the supervision team, the successful candidate will be expected in the course of the PhD to familiarise themselves with the relevant literature; participate in research training; access data from the Clinical Practice Research Datalink, including gaining relevant permissions; analyse the data with the support of the supervision team; write the thesis and contribute to peer reviewed publications.