Joint replacement is an effective treatment for knee osteoarthritis but it is unclear what the ‘healthcare journey’ is for patients in the years leading up to this. Clinical guidelines recommend a range of drug and non-drug treatments. But what do patients receive in practice? When? And in what order? Describing the types of ‘real world’ healthcare journeys that patients follow before having their knee replacement could give us some important insights and potentially assist in predicting what care patients are likely to receive.
This project is split into a series of studies exploring different aspects of a patient’s journey to joint replacement. We will use a set of methods called “sequence analysis” to find common patterns of healthcare use for patients with osteoarthritis. These methods have been used to understand how people’s lives unfold over time but not to identify healthcare journeys. We will use the Clinical Practice Research Datalink, a large dataset of anonymised GP records, in combination with hospital data, to understand the pathway a patient undertakes towards a joint replacement. By looking back over 10 years before joint replacement we hope to explore when a patient seeks medical advice, what medications patients frequently receive (and in what order), and whether there have been changes in practice over time. We will also describe any possible differences in patterns of care between groups of patients, such as older vs younger, and male vs female. We will then communicate our findings to patients and providers of care.
As a common cause of disability, knee osteoarthritis is managed by non-pharmacological, pharmacological and surgical treatment. Although joint arthroplasty, effective for end-stage knee osteoarthritis, is recommended, it is unclear whether care prior to arthroplasty has changed – i.e. further medical management, other joint procedures, or arthroplasty earlier. Further, variations in the care due to age, sex, and region remain undefined.
We will utilise descriptive analysis, case-control analysis, and sequence analysis of primary care and secondary care data (via linkage to Hospital Episode Statistics
) to explore the care timeline prior to arthroplasty. This will include healthcare consultations, pharmacological and surgical management, as well as referrals to other services.
In the descriptive analysis, a retrospective case-only study among patients who undergo arthroplasty, the absolute frequency of each care strategy will be explored by specific time-windows (i.e. 6, 12months) prior to arthroplasty and stratified by age, sex, region, comorbidities, and socio-economic status (via linkage to Index of Multiple Deprivation).
The nested case-control analysis will be used to estimate the relative frequency of each care strategy outlined above by comparing the frequency of care among cases (patients with osteoarthritis who undergo arthroplasty) with that of risk-set sampled controls (age, sex, practice-matched patients with osteoarthritis who have not undergone arthroplasty) . Odds ratios obtained from conditional logistic regression will suggest when and which care strategies are most strongly associated with future arthroplasty. These building blocks will be classed into channels and then combined in the final stage of multi-channel sequence analysis to produce a small number of discrete care pathways and estimates of the numbers of patients following these and their respective characteristics. We aim to disseminate this work to patients, to provide information of projected care trajectories, as well as to health practitioners and commissioners, to aid in service provision for knee osteoarthritis.
Total Knee Replacement
George Peat - Chief Investigator - Keele University
Dahai Yu - Corresponding Applicant - Keele University
Aleksandra Turkiewicz - Collaborator - Lund University
Elizabeth Cottrell - Collaborator - Keele University
Geraint Thomas - Collaborator - Keele University
John Edwards - Collaborator - Keele University
Martin Englund - Collaborator - Lund University
Thomas Appleyard - Collaborator - Keele University
HES Admitted Patient Care;Patient Level Index of Multiple Deprivation