Background: To try and reduce the prescribing of unnecessary medications, the NHS in England has introduced routine medication reviews for older people with severe frailty, care home residents and adults across all age groups prescribed high-risk medications. However, there are concerns that some people may not be getting these reviews, for example people from ethnic minority groups or people living in deprived areas. This can cause unfair differences in health outcomes-termed ‘health inequalities’.
Aim: To improve the access to medication reviews to ensure they meet the diverse need of people with severe frailty living in the community, care home residents and adults prescribed high-risk medications across different ethnic groups and in more deprived areas of England.
Methods: We will use anonymous data from CPRD. We will analyse the data to predict which people might be at risk of medication-related side effects. We will also find out if there are differences in how medication reviews are being provided for older people with severe frailty, care home residents and people prescribed high-risk medications (e.g. opioid medications, sedatives, long-term steroids). We will also find out if there are differences in the number of medicines that have been stopped for different ethnic groups and for people living in deprived areas.
For patients prescribed multiple medications, problematic polypharmacy is where the potential for harm outweighs benefits from the medicines, and/or the implications of the medication regimen are not fully understood by patients. Structured Medication Reviews (SMRs) are evidence-based, comprehensive reviews of patients’ medications, which are aimed at decreasing over-prescribing and ultimately reducing adverse drug-related outcomes. From October 2021, all Primary Care Networks (PCNs- groups of typically 5-8 general practices working closely together in priority areas) in England have been required to proactively target patients who would benefit most from an SMR, with five ‘high risk’ groups identified. In this study, we will focus on three of the high-risk groups:
• Those prescribed a high-risk medication: 1) opioids, 2) gabapentinoids, 3) benzodiazepines, 4) z-drugs, 5) steroid medications.
• Patients with severe frailty, as defined by the electronic Frailty Index 2 (eFI2).
• Care home residents, determined using 49 clinical codes, used in a previous CPRD study.
The aim of this study is to address problematic over-prescribing among the three high-risk groups of patients by a) identifying patients at risk of medication-related adverse outcomes; b) examining whether offer and uptake of Structured Medication Reviews (SMRs) varies according to intersectional characteristics (age, gender, ethnicity, socioeconomic status); and c) investigating whether drug-burden and medication-related adverse outcomes decreased following the implementation of the SMR PCN contract.
Clinical prediction models that predict the risk of drug-related adverse outcomes will be developed using Cox regression. The odds of SMR invitation/decline/receipt following PCN SMR contract implementation (1/10/21 onwards) will be modelled according to intersectional characteristics using multilevel logistic regression. Interrupted Time Series (ITS) models will be performed to examine whether there were changes in the level and slope of all medication-related and health and social care outcomes in England in the 2 years following the PCN SMR contract implementation.
Adverse outcomes related to high-risk medications (excluding steroids) (ANALYSIS 1a)
• Hospitalisation with delirium (ICD 10 codes)
• Hospitalisation for a fall or fragility fracture (ICD 10 codes)
Steroid-related adverse outcomes (ANALYSIS 1b)
• New diagnosis of Diabetes (Systematised Nomenclature of Medicine Clinical Terms (SNOMED-CT) codes).
• Hospital admission for a fragility fracture (ICD 10 codes)
• Hospital admissions for infection or mortality with infection listed as cause of death (SNOMED CT code list tbc )
SMR offer/uptake (ANALYSIS 2):
We will investigate inequities at 3 key stages of the SMR process, using the following process measures, all of which are included in the Primary Care Networks (PCN) SMR contract specification using standardised SNOMED-CT coding.
• Invitation of patients for SMR (using SNOMED CT code 1363201000000103, no READ code).
• Decline of SMR by patients (using SNOMED CT code 1363191000000100, READ code 8I3V.00).
• SMR receipt (using SNOMED CT code 1239511000000100, READ code 8B3S200).
Medication-related outcomes (ANALYSIS 3a):
• Number of medications (British National Formulary (BNF) subchapters).
• Number of medications of high-risk drug groups: 1) opioids, 2) gabapentinoid, 3) benzodiazepines, 4) z-drugs, 5) steroid medications). We will refine our operational definition of high-risk drug groups to be considered for inclusion through an initial workshop with stakeholders at the start of the project.
• Anticholinergic Medication Index (ACMI) score, developed and validated by our co-apps Best and West in a Health Data Research UK (HDRUK) funded project led by Chief investigator Clegg.
Health and social care outcomes (ANALYSIS 3b)
• Hospital admission (all-cause; hospitalisation with delirium; hospitalisation with falls).
• Number of primary care consultations.
• Time at home (number of days living at home, taking into account hospitalisation/length of stay and mortality).
• Care home admission (determined using 49 clinical codes, used in previous CPRD study (28)).
• Mortality, using linked ONS data.
Andrew Clegg - Chief Investigator - University of Leeds
Kate Best - Corresponding Applicant - University of Leeds
David Alldred - Collaborator - University of Leeds
Fatima Sabir - Collaborator - University of Leeds
Ifeanyi Chukwu - Collaborator - University of Leeds
Lauren Walker - Collaborator - University of Liverpool
Muhammad Faisal - Collaborator - University of Leeds
Munir Pirmohamed - Collaborator - University of Liverpool
Oliver Todd - Collaborator - University of Leeds
HES Accident and Emergency;HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;CPRD Aurum Ethnicity Record;CPRD GOLD Ethnicity Record;Practice Level Rural-Urban Classification