BACKGROUND: While the safety of new drugs is being evaluated in clinical trials, these studies are restricted in size and duration. Therefore, rare side effects can only be discovered once the drug is used by many people in the community. As part of the safety monitoring process, reports of suspected side effects are issued by healthcare professionals, patients, and carers, usually in the form of so called yellow cards. These remain a major source of information to find new safety signals. However, they do not provide much context on e.g., how many patients have side effects, how many people take the potentially harmful medicine, or if the affected patients were of same age, sex, or had similar conditions.
AIM: Our study will explore how electronic primary care records can support and enhance the preliminary assessment of drug safety signals.
METHOD: The study will be conducted in collaboration with the Uppsala Monitoring Centre, the World Health Organization’s Collaborating Centre for Drug Monitoring, which centrally collects and assesses spontaneous reports from across the world. For a selection of drug-side effect combinations that were seen in spontaneous reports, we will find those people in primary care data who used the respective drug and/or had similar side effects. We can then characterize these people and evaluate if the risk for a particular side effect was higher for one drug than for other drugs.
POTENTIAL IMPORTANCE: The study will evaluate important drug-event combinations as use-cases and inform on methods to improve drug safety monitoring worldwide.
BACKGROUND: Spontaneous reports adverse drug reactions remain a major source of information to identify new, potentially harmful side effects of drugs. However, information at the population level is lacking, including frequencies of a certain drug-event-combination and characteristics of people suffering from the side effect or taking the drug of interest.
AIM: This study will explore to what extent primary care records can support the preliminary assessment of signals identified in the WHO's global medicinal product safety database (Vigibase).
METHODS: The study will be an international network study with CPRD GOLD and AURUM, other real-world databases, and Vigibase. It comprised 3 consecutive work packages, with CPRD data being used in work package 2.
1) Drug-event combinations (DECs) will be identified through statistical signal detection in Vigibase. Of those, DECs that could be reliably identified in primary care data will be selected for work package 2.
2) For each DECs, a characterisation will be conducted in CPRD: Cohorts of (1) drug users, (2) people with the event of interest and (3) drug users with the event of interest will be studied. Additionally, a ‘comparison drug cohort’ with users of other drugs (excl. the drug studied in the DEC) will be generated.
Statistical analyses will comprise: sample size and incidence rate calculation, examination of the distribution of time-to-event onset among drug users, comparison of covariate distributions between different cohorts, visualisation of event occurrence over time using observed-to-expected event ratios.
3) The added value of incorporating primary care data in preliminary signal assessment will be assessed in a semi-structured questionnaire distributed across pharmacovigilance experts at the Uppsala Monitoring Centre. They will answer this after reviewing the results of work package 2.
PUBLIC HELATH BENEFIT: The study will evaluate important drug-event combinations as use-cases and inform on methods to improve drug safety monitoring worldwide.
Selected DECs of specific interest will be identified from Vigibase. Outcomes with at least 100 subjects exposed to any of the drugs of interest will be selected from the following shortlisted ones:
- Angioedema
- anaphylaxis
- acute kidney injury / acute renal failure syndrome
- pancreatitis
- pancytopenia
- stevens-johnson syndrome
- rhabdomyolysis
- agranulocytosis
- erythema multiforme
- blindness
- deafness
- hepatic failure
- pulmonary arterial hypertension
- ischemic stroke
- haemorrhagic stroke
- acute myocardial infarction
- deep venous thrombosis
- pulmonary embolism
- thrombocytopenia
- autoimmune thrombocytopenia
- bell’s palsy
- myocarditis/pericarditis
- narcolepsy
- appendicitis
- encephalomyelitis
- guillain-barre syndrome
- transverse myelitis
Daniel Prieto-Alhambra - Chief Investigator - University of Oxford
Annika Jodicke - Corresponding Applicant - University of Oxford
Daniel Dedman - Collaborator - CPRD
Edward Burn - Collaborator - University of Oxford
Martí Català Sabaté - Collaborator - University of Oxford
Xintong Li - Collaborator - University of Oxford