Rheumatoid arthritis is a common long-term condition, mainly affecting the joints. It usually begins gradually over several months with symptoms including joint pains, joint stiffness in the morning and tiredness.
Because early treatment helps prevent long-term damage to the joints and slows down the progression of rheumatoid arthritis, it is important to diagnose the condition early.
To help doctors diagnose rheumatoid arthritis earlier, a group of researchers is currently analysing a large database of patient records from general practices. They will identify the typical patterns of symptoms patients have before they are diagnosed. This means doctors can be automatically alerted when a patients’ symptoms could be rheumatoid arthritis and undertake further investigations (X-rays or blood tests) to diagnose the condition.
However, we know that people from different ethnic groups and social classes experience symptoms in different ways and describe them using different words. This means the patterns of recorded symptoms may vary between different types of people. To help diagnose rheumatoid arthritis earlier in people from all ethnic groups and social classes, we need to understand how they describe their symptoms before a diagnosis of rheumatoid arthritis.
This study will identify people with rheumatoid arthritis in the same large database of general practice records as the current group of researchers. It will investigate in detail how people from different ethnic groups and social classes reported symptoms in the year before they were diagnosed. This will ensure the automated early warnings can be tailored to work for all types of people.
This project will assess how different demographic groups in England report symptoms in the 12 months preceding a diagnosis of rheumatoid arthritis (RA). It is already known that experiences and reporting of symptoms vary by ethnicity and socioeconomic status. We investigate the hypothesis there are systematic differences in recorded early symptoms of RA between demographic groups, analysed by ethnicity, deprivation, sex, region, and age. This may influence the accuracy of diagnostic tools based on recorded symptoms in different social groups and therefore affect health equity.
This project is part of the NIHR-funded RAPID project, which is an analysis of CPRD to develop a prediction model for early diagnosis of RA in primary care.
In CRPD Aurum from 2004 to 2022, we will identify patients with a diagnosis of RA .
The five articular and extra-articular symptoms most associated with the prodromal phase of RA (as informed by a preceding review), for example hand pain, joint pain, and general fatigue, will be recorded on a patient-level if it occurred 0-12 months prior to the index date of the patient (defined as first diagnosis of RA).
Each symptom will form the dependent for a separate logistic regression model, in which ethnic group, IMD quintile, geographic region, sex, age and covariates will be used as independents.
The results will be presented as odds ratios against category-specific comparators.
The findings will be presented as forest plots depicting the odds ratio for the various symptoms by the various exposures (demographics). The results will feed back into the development of the RAPID tool, providing information about the need to contextualise the prediction model for the population the model is applied in.
The results will aid understanding of potential health-equity effects from data-driven tools derived both from CPRD and large health datasets in general, benefitting health equity.
End-point output is odds ratio by demographic properties for each of the below 10 outcomes, comprising the top-five (by prevalence) articular and top-five non-articular symptoms by odds ratio preceding RA according to a recent CPRD GOLD review (that is, the preceding review compared the odds ratios of having the given symptoms in RA vs non-RA groups). Note that due to the way they are recorded, there is considerable overlap. They will be recorded if the occurred 0-12 months before the index date for the individual patient (defined as first recorded diagnosis of RA).
Articular:
1. Hand pain
2. Joint pain
3. Frozen shoulder
4. Foot pain
5. Shoulder pain
Extra-articular:
1. Morning stiffness
2. Carpal tunnel syndrome
3. Muscle pain
4. Weight loss
5. General weakness
Nicola Adderley - Chief Investigator - University of Birmingham
Alexander d'Elia - Corresponding Applicant - University of Birmingham
Karim Raza - Collaborator - University of Birmingham
Krishnarajah Nirantharakumar - Collaborator - University of Birmingham
Patient Level Index of Multiple Deprivation