Epidemiological analyses are a mandatory component of Module 3 in every AMNOG dossier, covering prevalence, incidence, target population, and a five-year population forecast. In practice, however, this section is far more than a regulatory requirement. It represents a critical methodological exercise in translating the approved therapeutic indication into a transparent and decision-relevant target population for the German healthcare system.

The therapeutic indication described in the Summary of Product Characteristics (SmPC) is only the starting point. The real challenge lies in identifying a target population that reflects routine clinical practice in Germany while meeting the methodological expectations of the AMNOG benefit assessment. This is particularly demanding for complex oncology indications or narrowly defined patient subgroups, where no single data source can answer every epidemiological question.

Rather than asking which data source is best, the more relevant question is: Which data source is best suited to support each step of the target population derivation?

From data availability to data strategy

Target population derivation should be considered a strategic component of dossier planning from the outset. Instead of starting with the available evidence, Market Access teams should begin with the approved indication and define the epidemiological questions that need to be answered.

Only then can the most appropriate evidence sources be selected. Depending on the research question, these may include national disease registries, epidemiological studies, statutory health insurance (SHI) claims data, or data from the German Health Research Data Centre (FDZ Gesundheit). Each source contributes different strengths — and each has its own limitations.

A robust target population is therefore rarely based on a single dataset. It is usually the result of combining complementary evidence sources within a transparent and methodologically sound framework.

Combining evidence in practice

The complexity of this approach is illustrated by the AMNOG benefit assessment for olaparib, where the target population had to be defined using multiple clinical criteria, including tumor type, disease stage, BRCA mutation status, prior treatment history, and treatment response.

The analysis combined data from the German Centre for Cancer Registry Data (ZfKD), regional cancer registries, and published epidemiological studies. Because no individual source captured all relevant eligibility criteria, additional assumptions and complementary evidence were required. Transparent reporting of uncertainties and plausible ranges ensured that the resulting estimates remained robust and reproducible.

Similar challenges can be observed in other AMNOG procedures, ranging from rare diseases such as multidrug-resistant tuberculosis to broad indications such as chronic heart failure, where claims data and epidemiological evidence must be combined to derive reliable target populations.

The growing role of FDZ data

The German Health Research Data Centre (FDZ Gesundheit) has considerably expanded the opportunities for using routine healthcare data in AMNOG dossiers. Unlike analyses based on claims data from individual statutory health insurance funds, FDZ provides access to a comprehensive, cross-payer dataset covering the statutory health insurance system as a whole.

This broader data basis improves the representativeness and generalizability of epidemiological analyses, particularly for rare diseases and small patient populations. At the same time, FDZ projects require substantially more upfront planning. Research questions, cohort definitions, and analysis plans must be clearly specified before analyses begin, making early methodological planning essential.

Three practical takeaways

Successful target population derivation starts long before Module 3 is written. Based on practical experience, three principles are particularly important:

• Start with the indication — not the data. Define the target population first, then identify the evidence needed to support each step.
• Combine complementary evidence. Registries, epidemiological studies, SHI claims data, and FDZ analyses each contribute different pieces of the overall picture. Robust analyses are built by integrating these sources rather than relying on any single dataset.
• Plan for uncertainty. Transparent assumptions, plausible ranges, and sensitivity analyses strengthen both the epidemiological analysis and the overall credibility of the AMNOG dossier.

As access to real-world data continues to expand, deriving target populations is becoming increasingly strategic. Teams that integrate epidemiological planning early, select evidence sources purposefully, and address uncertainty transparently will be better positioned for benefit assessment and subsequent price negotiations.

Interested in learning more?

Watch our recent webinar, “Beyond Traditional Sources: Derivation of Epidemiological Data in the AMNOG Dossier Using Claims Data and FDZ Analyses” on demand:

Share Via