A break or fracture in a bone of the back can be the first sign of weakened bones (osteoporosis). If they are not treated it is likely the person will break more bones such as their hip. Unfortunately, most people who break a bone in their back in this way are not treated because the broken bone is not picked up by doctors or other healthcare professionals.
We will investigate health records from general practices across England to see how many people had a broken bone in their back detected, and whether this depended on where they live. We will then describe these people – how old they were, whether they were men or women and their ethnic background. We will also describe what healthcare they received before and after their break was picked up, how many times they saw their doctor and what other health problems were diagnosed.
We will learn from those whose broken bone in their back was picked up, to help diagnose more people who have this injury in the future, and to understand what treatment can be given. Our findings will determine which groups of people are more likely to have a broken bone picked up, and which are not, and whether this is different across regions of the country. This will allow doctors to better identify which of their patients may have these broken bones, allowing them to begin medicine to prevent more serious broken bones in the future.
Vertebral fractures (VF) are the most common form of osteoporotic fracture and are associated with future fractures, including serious health events like hip fracture, and increased risk of mortality. Treatment with anti-osteoporotic medications can substantially reduce the risk of future fractures and their negative sequalae. However, 70% of VF do not reach clinical attention due to the non-specific nature of symptoms like back pain and consequent lack of onward referrals for high-risk patients. We aim to estimate the incidence of VF diagnosis across England, and describe the characteristics, pre and post diagnosis health care contact, management and outcomes of these patients.
We will use data from CPRD Aurum for this descriptive observational study to estimate the incidence of VF diagnosis in primary care patients aged 50 and older across the 9 regions of England and describe the cohort of diagnosed patients, the service contact and treatment received before and after diagnosis and their future health outcomes (further fractures, mortality). Age and sex-standardised incidence of VF diagnoses will be calculated by region using the CPRD Aurum population as denominator. Characteristics of newly diagnosed VF patients (age, sex, ethnicity, deprivation, frailty, presenting complaint(s), medical history) and primary care health contact prior to diagnosis (consultation rates and providers, referrals to secondary care, and imaging type leading up to coding) will be summarised using descriptive statistics. Associations between VF management, future fractures and mortality will be modelled using multilevel Poisson and Cox Proportional Hazards models to model variation at the practice and regional levels.
Identifying which patients obtain a diagnosis of VF will help to identify higher risk underdiagnosed patients, while differences in diagnosis rates among high-risk patients across regions will help the development of targeted interventions for poorer performing regions.
Primary outcome
Incidence of VF diagnosis
Secondary outcomes/analyses
Descriptive characteristics of VF patients:
Age; sex; ethnicity; diagnoses of back pain; osteoporosis; previous fractures; diagnoses of chronic conditions (Charlson comorbidity index [1]); electronic Frailty Index (eFI) [2]; smoking status, exercise level (from NHS health check; Index of Multiple Deprivation (IMD).
Care related variables
Consultation frequency (pre and post diagnosis); referrals to secondary care; imaging type leading up to coding; further investigations post diagnosis; further referrals post diagnosis; non-pharmacological management of VF or osteoporosis
Prescriptions
Anti-osteoporotic medications; analgesics; corticosteroids; hormone replacement therapy
Future health outcomes
Further fractures; mortality.
Practice level characteristics:
Size of general practice; case-mix of patients in practice (age, sex, ethnicity); region of England.
See outcomes, exposures and covariates for more detailed specification and description of roles in the analysis.
Matthew O'Connell - Chief Investigator - King's College London (KCL)
Matthew O'Connell - Corresponding Applicant - King's College London (KCL)
Aicha Goubar - Collaborator - King's College London (KCL)
Alex Dregan - Collaborator - King's College London (KCL)
Emma Clark - Collaborator - University of Bristol
Emma Godfrey - Collaborator - King's College London (KCL)
Emma Rezel-Potts - Collaborator - King's College London (KCL)
Frances Williams - Collaborator - King's College London (KCL)
Katie Sheehan - Collaborator - Barts and the London Queen Mary's School of Medicine and Dentistry
Mark Ashworth - Collaborator - King's College London (KCL)
Martin Gulliford - Collaborator - King's College London (KCL)
Sarah Pope - Collaborator - St George's University Hospitals NHS Foundation Trust
HES Admitted Patient Care;Patient Level Index of Multiple Deprivation