Chronic pain and mental illness are common causes of disability and worse quality of life but their interrelationships, impact on health, and access to treatment are unclear. Initial research has suggested that access, treatment, and outcomes for people with chronic pain may be different according to the presence and type of mental illness. However, little information is available on the nature of the relationship between pain and mental illness and treatments received in routine NHS primary care.
The aim of this study is to develop a better understanding of these issues using the CPRD database, i.e., drawing on anonymous GP data from across the United Kingdom. We will analysing information from GP records over time to understand: 1) How common is pain recorded in people with different mental illnesses and what factors influence this? 2) What is the typical direction of the relationship between chronic pain and mental illness (i.e. which comes first) and what influences this? 3) How does having pain and different mental illnesses influence health, social and access to treatment outcomes compared to people with chronic pain but without mental illness?
We have worked with our Patient and Public Involvement group who have told us that this information is important and they would like to understand how pain can be better identified/managed for people with mental illness. More broadly, understanding this will advance the development of a new treatment pathway of chronic pain for patients who experience from both mental illness and chronic pain.
Primary objective is to estimate the incidence of chronic pain across different sites (e.g., neck pain), including type (e.g., musculoskeletal pain) and severity (e.g., moderate) among people with mental illness. Primary exposure is mental illness (e.g., depression, schizophrenia). Primary outcome includes incidence of chronic pain in people with mental illness. Secondary objective is to evaluate the incidence of physical conditions and functional decline among patients with co-existing mental illness and chronic pain. The exposure is co-existing mental illness and chronic pain. The outcomes will include the incidence of physical health disorders (e.g., stroke, diabetes) and functional decline (frailty). Study design: Retrospective, longitudinal cohort study. We will define a cohort of patients with new diagnoses of mental illness between 01st March,2011 and 28th February, 2020 and follow them up to estimate the time to study primary outcome. For the secondary objective, we will select patients with mental illness who developed chronic pain (exposed) during the study period and match (1:2 ratio) them on age, general practice and index date for mental illness) with patients with mental illness only (unexposed). Mental health events include mental health-specific clinical diagnoses or medications (e.g., antidepressants). Chronic pain events will include medical diagnoses and relevant prescriptions (e.g., opioids). Predictors of interest include clinical (e.g., age at diagnosis, severity of condition) and socio-demographic (e.g., ethnicity, deprivation) variables. The start date of the follow-up is the latest of the registration date and index date for mental illness (primary objective) or chronic pain (secondary objective) diagnosis, respectively. The end of the follow-up date is the earliest of the study outcomes, date of death, transferred out date, or study end date. The analyses will adjust for relevant confounders (e.g., comorbidities). Data will be analysed using Cox regression with competing risk analysis. The proportionality assumption will be tested via Schoenfeld residuals.
The primary outcome will be the incidence of specific chronic pain diagnoses across different sites (e.g., lower back pain, neck pain), including type (e.g., musculoskeletal pain and non-cancer pain) and severity (e.g., mild, moderate, severe) in people with different mental health diagnoses (e.g., affective disorders such as bipolar disorder, major depressive disorders, and severe mental illness such as schizophrenia and schizoaffective disorder etc).
Chronic pain diagnoses will be identified via SNOMED-CT and/or Read medical codes using our team codes published in previous research (Dregan et al, 2015, 2019; Stubbs et al, 2015, 2017) and other codes publicly available (e.g., Cambridge University) following validation by clinical experts within the team.
Specific secondary outcomes will be investigated include:
Long term physical health conditions (e.g., diabetes, heart failure): Physical condition diagnoses will be identified from the Read medical codes used by GPs
Functional status: Frailty will be extracted from routinely used scales such as electronic frailty index (eFI). Data permitting, we will also consider extracted clinical events related to mobility (e.g., physically impaired) and/or social functioning (e.g., fit to work).
Brendon Stubbs - Chief Investigator - King's College London (KCL)
Ruimin Ma - Corresponding Applicant - King's College London (KCL)
Alex Dregan - Collaborator - King's College London (KCL)
Eugenia Romano - Collaborator - King's College London (KCL)
Fiona GAughran - Collaborator - King's College London (KCL)
Mark Ashworth - Collaborator - King's College London (KCL)
Robert Stewart - Collaborator - King's College London (KCL)
Patient Level Index of Multiple Deprivation