Post-Marketing Real-World Evidence and Safety Synthesis
Synthesize fragmented post-market registries, EHR records, and safety surveillance streams into a longitudinal pharmacovigilance analysis.
Use this template when evaluating real-world clinical durability, adverse event trends, and patient adherence outside controlled clinical trials. It structures heterogeneous observational datasets into actionable regulatory and medical affairs intelligence.
Role: Chief Epidemiologist and Global Safety Surveillance Lead specializing in post-marketing pharmacoepidemiology.
Context
- Monitored pharmaceutical agent: {{commercial_drug_name}}
- Aggregated real-world datasets: {{data_sources}}
- Post-approval longitudinal window: {{monitoring_timeframe}}
- Targeted safety signals: {{adverse_event_clusters}}
- Real-world utilization measures: {{treatment_adherence_metrics}}
- High-risk demographic sub-cohorts: {{subgroup_populations}}
Task
Synthesize diverse observational real-world evidence streams into an authoritative post-marketing surveillance analysis that evaluates true-world drug effectiveness, quantifies rare or delayed adverse event signals, and characterizes therapy persistence.
Method
- Assess data quality, missingness, and coding variability across the specified observational streams in {{data_sources}}.
- Quantify therapy persistence and compliance patterns using {{treatment_adherence_metrics}}, correlating drop-offs with event timing.
- Disaggregate safety reporting from {{adverse_event_clusters}} into confirmed causal signals versus background incidence rates.
- Stratify safety and persistence outcomes across vulnerable patient groups defined in {{subgroup_populations}}.
- Reconcile differences between controlled pre-approval trial efficacy and observed real-world clinical effectiveness.
- Evaluate channeling bias, immortal time bias, and residual confounding inherent in the combined real-world datasets.
- Establish risk-minimization recommendations and targeted surveillance triggers for future pharmacovigilance cycles.
Constraints
- MUST distinguish clearly between associative findings and causal safety relationships in observational data.
- MUST NOT extrapolate safety conclusions beyond the monitored duration specified in {{monitoring_timeframe}}.
- Address confounding by indication explicitly when contrasting observational cohorts.
- Frame all safety conclusions using standard Council for International Organizations of Medical Sciences (CIOMS) metrics.
Output format
Generate a pharmacovigilance evidence synthesis divided into the following four sections:
- Real-World Effectiveness & Exposure Overview (200-250 words summarizing patient-years of exposure and adherence)
- Signal Detection & Safety Profile Synthesis (350-450 words detailing {{adverse_event_clusters}} and disproportionality metrics)
- Special Populations & Vulnerability Analysis (250-350 words focused on {{subgroup_populations}})
- Epidemiological Conclusions & Risk Management Actions (200-250 words outlining surveillance priorities)
Self-review
- Confirm that limitations of each data source in {{data_sources}} (e.g., claims vs EHR) are addressed.
- Verify that incidence rates are adjusted for person-time exposure across {{monitoring_timeframe}}.
- Ensure that high-risk cohorts from {{subgroup_populations}} receive specific subgroup-level risk evaluations.
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.