Health Insurance Datasets

When pharma market access teams need to size Continental European market opportunities — and I mean really understand them, not just guess based on extrapolated clinical trial populations — they turn to statutory health insurance claims data. And honestly? That’s because nothing else comes close to capturing what’s actually happening across millions of lives in major Pan-European markets.

Continental European statutory health insurance (SHI) systems cover substantial portions of national populations, in some cases reaching 90% coverage. That’s not a sample. That’s not a model. It’s comprehensive, longitudinal administrative data tracking every diagnosis coded, every prescription filled, every procedure performed, and every euro spent across inpatient, outpatient, and pharmacy settings. For real-world evidence teams, this represents one of the most representative views of treatment patterns, disease burden, and healthcare utilization you can access in the world’s most sophisticated pharma markets.

But here’s what most teams miss: Continental European health insurance claims data isn’t just valuable because it’s big. It’s valuable because it reflects real clinical decisions, real prescribing patterns, real patient journeys — the messy, complex reality that clinical trials deliberately exclude. When you’re preparing HTA submissions, forecasting loss of exclusivity impact, or building budget impact models for payer negotiations, you need data that mirrors the actual treatment landscape.

Why Continental European Health Insurance Datasets Have Become Essential for Pharmaceutical RWE

Over the past decade, reliance on aggregated Pan-European health insurance claims data has grown exponentially across pharma, medtech, and biotech. The reason isn’t hard to understand: regulatory requirements have evolved. National HTA bodies increasingly expect real-world evidence in health technology assessments. Cost-effectiveness analyses must be grounded in local treatment reality, not international averages. The EMA’s 2019 guidance on real-world evidence signaled that administrative claims data — when properly validated and analyzed — can support regulatory decision-making.

Pan-European SHI claims databases have demonstrated their scientific credibility through hundreds of peer-reviewed publications. Research databases maintained by academic institutions cover substantial percentages of national populations with data from multiple statutory health insurance providers. These databases have been extensively validated for mortality outcomes, pregnancy identification, drug utilization patterns, and disease prevalence estimates.¹

What makes these datasets particularly powerful for pharmaceutical analytics is their structure. You’re getting ICD-10 diagnosis codes (with country-specific modifications offering higher granularity), national procedure codes for inpatient and outpatient settings, ATC-classified pharmaceutical prescriptions with dispensing dates, and DRG-based hospital billing information. All longitudinally linked at the patient level through encrypted identifiers, allowing you to track treatment pathways, switching patterns, adherence, and outcomes over time.

Core Applications: Where Health Insurance Claims Data Drives Pharma Decision-Making

The pharmaceutical industry uses Continental European health insurance datasets across virtually every stage of the product lifecycle. Market access teams preparing for launch need accurate prevalence estimates by region, understanding of current standard of care, and insights into prescriber behavior patterns. Commercial analytics directors building forecasting models need historical uptake curves for analogous products, real-world treatment duration data, and switching pattern analysis.

Epidemiology and Disease Burden Assessment

When you need to understand disease incidence, prevalence, or progression in Continental European populations, claims data provides population-level insights that survey data simply can’t match. Studies using validated claims databases have successfully estimated rare disease incidence and case fatality rates,² tracked regional variations in antibiotic prescribing,³ and analyzed trends in hormone therapy prescribing over multi-year periods.⁴

The longitudinal nature of SHI data enables true incidence calculations (new diagnoses among at-risk populations) rather than just prevalence estimates. You can identify treatment-naive patients, track time from diagnosis to treatment initiation, analyze diagnostic pathways, and measure disease progression through escalating treatment patterns or procedure utilization.

Health Economics and Outcomes Research (HEOR)

For HEOR teams supporting HTA dossiers or conducting post-launch value demonstration, Pan-European claims data provides the cost and resource utilization inputs that theoretical models can’t deliver. You get actual hospitalization rates by diagnosis, real-world complication frequencies, genuine specialist consultation patterns, and true medication costs including co-prescriptions and concomitant therapies.

Comparative effectiveness analyses — increasingly requested by national HTA bodies for health technology assessment — require matched cohorts from real-world populations. Claims data enables propensity score matching, inverse probability weighting, or other advanced causal inference methods to compare treatment outcomes when randomized trials aren’t feasible or generalize poorly to local practice patterns.

Market Sizing and Commercial Forecasting

Every pharmaceutical launch starts with a market sizing exercise. But estimates based on literature-derived prevalence rates, applied to census population data, and adjusted by assumed treatment rates… those are educated guesses. Health insurance claims data gives you actual diagnosed patient populations, segmented by severity (through proxy measures like hospitalization frequency or specialty care utilization), age, gender, comorbidities, and geographic distribution.

You can track how new products penetrate the market by analyzing prescription data post-launch. How quickly did uptake occur? Which specialists prescribed first? What patient characteristics predicted early adoption? Which products lost share? These insights inform sales force deployment, KOL engagement strategies, and market access prioritization across regions.

Pharmacovigilance and Post-Authorization Safety Studies

Pan-European SHI claims databases are extensively used for post-authorization safety studies (PASS) required by EMA or national authorities. The ability to identify large exposed populations, matched controls, and outcome events makes claims data ideal for rare adverse event detection or long-term safety monitoring.

Validated databases have been used for identifying pregnancy outcomes,⁵ enabling studies on drug safety during pregnancy — a critical concern for many therapeutic areas. Researchers have investigated associations between drug exposures and birth defects,⁶ analyzed risks of newer treatments in women of childbearing age,⁷ and optimized algorithms for pregnancy identification and classification.⁸

Cardiovascular safety assessments, cancer risk evaluations, and metabolic outcome studies all benefit from the large sample sizes and long follow-up periods available in SHI data. Studies on medication risks and acute cardiac events demonstrate the power of nested case-control designs in Continental European claims data.⁹

Understanding the Data Structure: What You Actually Get

When pharmaceutical companies access Continental European health insurance datasets through research institutes or commercial data providers, they’re working with aggregated, anonymized data that’s been carefully processed to ensure GDPR compliance while maintaining analytical utility.

The core data elements include:

Diagnosis Information: ICD-10 codes from inpatient settings (principal and secondary diagnoses) and outpatient physician encounters. Country-specific modifications include additional specificity useful for pharmaceutical research — for example, more granular diabetes coding or detailed oncology staging information.

Procedure Codes: National procedure classification systems document inpatient procedures with remarkable detail, from surgical interventions to diagnostic imaging to therapeutic procedures. Outpatient physician services are also captured, though sometimes with less granularity than hospital procedures.

Pharmaceutical Prescriptions: ATC-classified medications with dispensing dates, package sizes, and (in some datasets) dosage information. Pharmacy claims capture prescription medications but not over-the-counter products or in-hospital medication administration.

Hospital Admissions: DRG codes, admission and discharge dates, length of stay, discharge disposition, and department-level information. This enables analysis of hospitalization patterns, readmission rates, and acute care utilization.

Demographic Data: Age, gender, and insurance status (member vs. co-insured family member). Regional information may be available but often at aggregated levels to maintain anonymization standards.

Longitudinal Linkage: Encrypted patient identifiers allow tracking over time, essential for treatment pathway analysis, adherence measurement, and outcomes assessment. Individual patients can be followed across care settings — from outpatient diagnosis to specialist referral to hospital admission to prescription fills.

The Continental European Advantage: Why This Market’s Data Infrastructure Matters

Continental European health insurance systems create uniquely valuable data for pharmaceutical research. Unlike fragmented systems with multiple payers using incompatible coding systems, Pan-European SHI operates under standardized coding requirements within countries. National governments mandate specific documentation standards, creating consistency across health insurance providers.

The coverage rates matter enormously. In countries where SHI reaches 70-90% population coverage, data represents nearly entire national populations except for privately insured individuals. This minimizes selection bias that plagues many healthcare databases.

Pan-European SHI data captures both inpatient and outpatient care comprehensively. Many international claims databases struggle with outpatient data quality or completeness. In mature Continental European systems, the integrated nature of SHI coverage means you get the full care continuum — primary care visits, specialist consultations, diagnostic procedures, therapeutic interventions, prescriptions, and hospitalizations — all linked to the same encrypted patient identifier.

For pharmaceutical companies evaluating multi-country strategies, major Continental European markets serve as valuable reference points. Their size, sophisticated healthcare systems, strict regulatory requirements, and powerful payer negotiation processes make Pan-European real-world evidence directly relevant for other markets facing similar value demonstration requirements.

Navigating Common Challenges in Pan-European Claims Data Analysis

Working with Continental European health insurance claims data isn’t without challenges. Understanding these limitations helps pharmaceutical teams design better studies and interpret findings appropriately.

Clinical Depth Limitations

Administrative claims data excels at what it was designed for: documenting billable services. But it lacks the clinical richness of electronic health records or prospective registries. You won’t find lab values, biomarkers, disease severity scores, or patient-reported outcomes. Diagnosis codes indicate that a condition was diagnosed but don’t reveal disease stage or functional status.

This means surrogate measures become important. Hospitalization frequency might proxy for disease severity. Specialty care utilization could indicate complex cases. Prescription patterns may suggest treatment response or disease progression. Sophisticated algorithm development — validated through chart review or registry linkage — becomes essential for outcomes research.

Data Quality and Coding Accuracy

Because claims data originates from billing rather than clinical documentation, coding incentives matter. DRG-based hospital reimbursement creates potential for upcoding to higher-paying diagnosis groups. Outpatient diagnoses may be coded liberally for reimbursement justification. Rule-out diagnoses might be coded similarly to confirmed diagnoses.

Validation studies consistently show that certain data elements are highly accurate (mortality, filled prescriptions, major procedures) while others require careful interpretation (outpatient diagnosis specificity, exact indication for medication use, complication vs. comorbidity coding). Researchers using validated databases have conducted extensive validation, demonstrating strong concordance for mortality outcomes through linkage with disease registries¹⁰ and good performance for core clinical events.

Access Restrictions and GDPR Compliance

Continental European data protection regulations are among the world’s strictest, with GDPR setting baseline privacy standards. Direct access to identifiable patient-level data is prohibited. Pharmaceutical companies work with aggregated, anonymized datasets provided by research institutes or commercial vendors who have data use agreements with SHI providers.

This regulatory framework protects patient privacy but creates workflow challenges. Data requests often require ethics committee approval, detailed study protocols, and specific data use agreements. Turnaround times can extend weeks or months. International data transfers face scrutiny under GDPR, though anonymized data benefits from more flexible transfer rules.

For global pharmaceutical companies, navigating these requirements requires understanding Continental European data protection law (particularly GDPR compliance standards), maintaining appropriate information security infrastructure, and establishing relationships with reputable Pan-European data providers who understand both regulatory requirements and pharmaceutical research needs.

Representativeness Considerations

While Pan-European SHI systems cover substantial portions of national populations, privately insured individuals aren’t randomly distributed. Privately insured populations tend to be higher-income, employed in certain professions, or civil servants — demographic groups with potentially different health profiles and healthcare utilization patterns.

Individual SHI fund datasets may reflect the member characteristics of that fund. Some funds historically enrolled more blue-collar workers, others more white-collar employees. Regional variations exist across countries like Switzerland, Austria, Belgium, and the Netherlands. Aggregated datasets from multiple funds mitigate this through broader inclusion, but researchers should still consider whether their study population generalizes to national populations or specific patient segments of interest.

Methodological Advances: Making Claims Data Work Harder

The pharmaceutical research community has developed increasingly sophisticated methods for extracting maximum value from health insurance claims data.

Algorithm Development and Validation

Identifying specific clinical events from administrative codes requires careful algorithm development. Pregnancy identification in claims data, for instance, involves complex logic considering diagnosis codes, procedure codes, and pharmacy dispensings.¹¹ Mother-baby linkage algorithms enable pediatric outcomes research and maternal medication safety studies.¹²

Disease identification algorithms balance sensitivity (capturing all true cases) against positive predictive value (minimizing false positives). Validation typically involves chart review, registry linkage, or comparison with established cohorts. The pharmaceutical industry benefits from published, validated algorithms that can be adapted to new research questions rather than developing de novo approaches for each study.

Advanced Causal Inference Methods

Observational data can’t randomize patients to treatments, but modern causal inference methods help approximate randomized trial conditions. Propensity score methods balance confounders between treatment groups. Instrumental variable approaches address unmeasured confounding. Parametric g-formula methods enable time-varying exposure analysis, as demonstrated in studies examining medication effects and disease outcomes.¹³

These methods require careful covariate selection, diagnostics for assumption violations, and sensitivity analyses. But when applied rigorously, they enable comparative effectiveness research from claims data that HTA bodies increasingly accept as valid evidence.

Synthetic Data Generation from Real-World Claims

An emerging capability addresses a persistent challenge: sharing granular data for algorithm development, validation, or collaborative research while maintaining strict privacy protections. Synthetic healthcare data generated from real Pan-European claims data using generative adversarial networks (GANs) with differential privacy guarantees enables scenario modeling, method testing, and algorithm development without handling actual patient records.

Synthetic datasets preserve the statistical properties and correlational structures of real claims data while containing no identifiable information and maintaining GDPR compliance principles. This accelerates research workflows, enables broader data access, and supports international collaborations that would be infeasible with anonymized-but-still-regulated real-world data.

Evaluating Data Partnership Options: What to Look For

The challenge many pharmaceutical companies face isn’t finding data sources—it’s determining which combination of capabilities aligns with their strategic priorities. The Continental European claims data landscape includes dozens of providers with overlapping but distinct strengths across several critical dimensions.

Coverage and Scale Considerations

Some sources maintain comprehensive longitudinal records covering hundreds of millions of patient lives across multiple regions and payer types. Others focus on specific geographies or therapeutic areas with greater depth. The trade-off between breadth and depth depends on your use cases.

For market sizing and commercial forecasting, broad population coverage matters most. For rare disease research or specialized therapeutic areas, focused datasets with clinical richness may provide better value than massive databases with limited granularity in your area of interest.

Regional granularity varies significantly. Some databases enable analysis at the national level only, while others support regional breakdowns useful for sales territory planning or understanding geographic treatment pattern variations.

Analytical Sophistication and Tools

Capabilities range from traditional query-based systems to AI-powered platforms offering real-time pattern recognition and predictive analytics. The choice depends on whether you need descriptive intelligence or forward-looking insights.

Self-service platforms enable your internal analysts to explore data directly, accelerating time-to-insight but requiring more internal expertise. Full-service analytical partnerships provide expert interpretation but with longer turnaround times and less flexibility.

API access, integration capabilities, and compatibility with your existing analytical infrastructure influence how easily you can incorporate claims data into broader evidence generation workflows.

Data Integration and Enrichment

Some sources combine claims with complementary data (lab results, specialty pharmacy, provider characteristics, social determinants of health) for richer patient profiles. This vertical integration creates analytical possibilities that pure claims data cannot support.

Linkage to mortality data, cancer registries, or other authoritative sources enhances outcome measurement capabilities. The methodologies used for linkage—deterministic versus probabilistic matching—affect data quality and completeness.

Understanding which enrichments matter for your specific use cases requires careful evaluation. More data isn’t always better if it introduces complexity without improving analytical precision for your questions.

Recency and Update Frequency

Real-time data feeds enable near-current market intelligence but often involve preliminary, unadjudicated claims. Fully adjudicated claims arrive with delays but provide greater accuracy and completeness.

For competitive intelligence and early signal detection, recency matters enormously. For HEOR and regulatory submissions, data quality and validation take priority over currency.

Update frequency varies from weekly refreshes to annual data releases. Your decision cadence should match the data refresh cycle you’re licensing.

Cost Models and Licensing Structures

Pricing approaches span flat annual fees, per-query charges, patient-record-based pricing, and subscription models with usage tiers. Total cost of ownership includes not just licensing fees but also integration, analysis, and ongoing support expenses.

Some providers bundle data with analytical services, while others license data separately from consulting support. The right model depends on your internal analytical capabilities and resource availability.

Multi-year commitments often reduce per-year costs but limit flexibility if your requirements evolve or better options emerge.

Understanding which combination of these capabilities aligns with your specific strategic priorities requires expertise in both pharmaceutical commercialization and the evolving real-world evidence ecosystem. The gap between understanding what claims data can do and identifying which specific data partnership matches your needs is substantial—this is where expert guidance becomes critical.

Strategic Implementation: From Data Access to Actionable Intelligence

Licensing Continental European claims data is just the first step. The gap between accessing data and extracting actionable intelligence is where most pharmaceutical projects struggle.

Building Internal Capabilities vs. Partnering

Some pharmaceutical companies build internal teams with claims data expertise—statisticians, epidemiologists, programmers skilled in working with large administrative datasets. This investment pays off over time if you have consistent, high-volume analytical needs.

Others partner with specialized analytical firms that understand both claims data and pharmaceutical business questions. This approach accelerates time-to-value and provides access to broader expertise without building internal infrastructure.

The right balance depends on your organization’s analytical maturity, budget, and strategic commitment to real-world evidence. Many companies start with external partnerships and gradually build internal capabilities as they understand requirements better.

Defining Requirements Before Evaluating Options

Before engaging with data sources, document your coverage needs (geographies, therapeutic areas, patient populations), analytical capabilities (self-service vs. full-service), integration requirements (API access, data formats), compliance constraints, and budget parameters. This clarity prevents costly mismatches.

Most companies underestimate the importance of this requirements definition phase. Rushing to evaluate data sources before understanding your actual needs leads to suboptimal decisions. Expert consultants who understand pharmaceutical analytics can help structure this requirements definition process to ensure you’re comparing options on the dimensions that actually matter for your use cases.

Pilot Projects and Proof of Value

Start with focused projects that answer specific business questions before committing to large-scale, multi-year licensing agreements. Pilot projects validate that your chosen approach delivers value in your therapeutic areas with your team’s capabilities.

Use pilots to test not just data quality but also analytical workflows, integration processes, and stakeholder engagement. The goal is discovering friction points at small scale before making major commitments.

Well-designed pilots include clear success criteria, defined timelines, and decision frameworks for whether to scale, pivot, or abandon the approach.

Regulatory Acceptance: How Pan-European Authorities View Claims Data Evidence

National HTA bodies across Continental Europe have increasingly signaled openness to real-world evidence in health technology assessments, particularly for rare diseases, pediatric indications, or situations where randomized trials face ethical or practical barriers. Methodological guidelines acknowledge observational data’s role in health economic evaluation, though with careful attention to bias risks and methodological rigor.

EMA’s 2019 guidance on registry-based studies applies principles relevant to claims data analysis: clear study objectives, appropriate design for causal questions, validated outcome measures, transparent methods, and comprehensive sensitivity analyses. Continental European SHI claims data, when analyzed with these principles, supports regulatory decisions.

What matters to regulators isn’t the data source but the evidence quality. Claims data studies with clear research questions, appropriate designs, transparent methods, validated algorithms, and honest discussion of limitations receive serious consideration. Poorly conducted registry studies or claims analyses don’t.

The Future Landscape: Where Pan-European Health Data Infrastructure is Heading

Continental European health data policy is evolving rapidly. National initiatives across countries like Switzerland, Austria, and Norway are testing infrastructure for broader, more efficient data access while maintaining strict privacy protections aligned with GDPR principles.

For pharmaceutical companies, these developments signal expanding opportunities. More standardized data access processes, potentially harmonized data platforms across countries, and clearer regulatory pathways for secondary data use could reduce current friction in accessing Pan-European claims data.

Simultaneously, analytical capabilities continue advancing. Machine learning methods enable more sophisticated outcome prediction, treatment response modeling, and patient segmentation from claims data. Natural language processing applied to diagnosis text fields (where available) could extract additional clinical insights. Linkage with genomic databases opens pharmacogenomic research possibilities.

The integration of claims data with other real-world data sources — electronic health records, pharmacy dispensing data, patient-reported outcomes platforms — creates richer datasets that address current limitations while preserving claims data’s comprehensive population coverage and longitudinal depth.

Navigating this evolving landscape requires staying current with regulatory developments, emerging analytical methods, and shifting data provider capabilities. Expert advisors who understand both your business objectives and the real-world evidence ecosystem can help you anticipate changes and adapt strategies proactively rather than reactively.

Getting Started: Building Your Claims Data Strategy

Pharmaceutical teams new to Pan-European claims data research should approach this strategically rather than tactically.

Start with Business Objectives, Not Data Sources

What decisions will this evidence inform? What level of precision is required? What confounders must be addressed? What outcomes matter? Starting with clear business questions ensures you evaluate data sources against actual needs rather than generic capabilities.

Common mistake: beginning by surveying what data is available and then figuring out what to do with it. Better approach: defining the intelligence gaps in your business strategy and then determining which data can fill those gaps.

Understand the Landscape Dimensions

The claims data market includes dozens of providers with overlapping but distinct capabilities. Key dimensions include data recency (real-time vs. adjudicated), geographic coverage, therapeutic depth, analytical tools, cost models, and integration capabilities.

Understanding which dimensions matter most for your use cases prevents being distracted by impressive but irrelevant features. A provider with the broadest coverage may not be optimal if your therapeutic area requires specialized clinical depth.

Leverage Expert Guidance for Data Partnership Selection

The gap between understanding what claims data can theoretically provide and identifying which specific combination of data sources, analytical approaches, and partnership models best serves your needs is substantial. Most pharmaceutical companies lack the internal expertise to navigate this landscape efficiently.

Expert advisors who understand both pharmaceutical business requirements and the real-world evidence vendor landscape can accelerate your path to value while avoiding common pitfalls. They bring market knowledge about provider capabilities, pricing benchmarks, negotiation leverage, and analytical best practices that internal teams typically lack.

This guidance becomes particularly valuable when evaluating trade-offs between competing options, structuring pilot projects, negotiating licensing terms, and building analytical workflows that actually get used rather than gathering dust.

Plan for Iteration and Learning

First claims data projects in a therapeutic area often reveal unexpected patterns, raise new questions, or identify data limitations that require refined approaches. Successful pharmaceutical companies treat Continental European claims data as a strategic asset, building internal expertise and external partnerships that improve with each project.

Build feedback loops that capture learnings from each analytical project and feed them into future requirements definition. Over time, this iterative approach builds organizational competence in leveraging real-world evidence effectively.

Your Next Steps: Developing Strategic Approaches to Real-World Intelligence

The pharmaceutical decisions you’re making today — launch strategies, market access approaches, commercial investments, pipeline prioritization — deserve evidence grounded in Continental European treatment reality, not theoretical models or borrowed international data.

Pan-European statutory health insurance claims data provides that evidence: comprehensive population coverage, longitudinal patient tracking, validated outcome measures, and proven regulatory acceptance. Whether you’re preparing HTA submissions, building forecasting models, conducting post-market surveillance, or demonstrating value to payers, these datasets enable analysis that other sources simply can’t match.

But the challenges are real — navigating the complex landscape of data providers, understanding which capabilities align with your specific needs, ensuring GDPR compliance, building analytical workflows, and extracting actionable intelligence from raw data. The gap between licensing claims data and generating business value is substantial. This is where expert guidance becomes essential.

Understanding which data sources match specific use cases requires deep market knowledge that most pharmaceutical companies lack internally. The landscape includes dozens of providers with overlapping capabilities, diverse pricing models, varying analytical sophistication, and different strengths across therapeutic areas and geographies. Evaluating these options without insider expertise risks costly mismatches between what you license and what you actually need.

Expert advisors who understand both pharmaceutical commercialization and the evolving real-world evidence ecosystem can help you:

• Define requirements that align with business objectives rather than generic capabilities
• Navigate the provider landscape to identify options worth serious evaluation
• Structure pilot projects that validate value before large commitments
• Negotiate favorable licensing terms based on market benchmarks
• Build analytical workflows that integrate claims data into decision-making processes
• Avoid common implementation pitfalls that derail projects

Ready to explore how Continental European claims data can transform your real-world evidence strategy? The first step isn’t licensing data—it’s understanding what you need and which approaches align with your business objectives.

Contact us to discuss your specific data requirements and explore strategic approaches to real-world intelligence that match your organization’s capabilities, therapeutic focus, and commercial priorities. We help pharmaceutical, medtech, and biotech companies navigate the complex landscape of Pan-European health insurance data to accelerate time-to-insight while avoiding costly mismatches.

Sources & References

1. Ohlmeier C, et al. Mortality in the German Pharmacoepidemiological Research Database (GePaRD) compared to national data in Germany: results from a validation study. BMC Public Health. 2015;15:570. https://pubmed.ncbi.nlm.nih.gov/26087768/
2. Haug U, et al. Estimating Incidence and Case Fatality of Thyroid Storm in Germany Between 2007 and 2017: A Claims Data Analysis. Thyroid. 2022;32(9):1089-1097. https://pubmed.ncbi.nlm.nih.gov/36006371/
3. Kollhorst B, et al. Regional Variations in Outpatient Antibiotic Prescribing in Germany: A Small Area Analysis Based on Claims Data. Antibiotics (Basel). 2022;11(7):947. https://pubmed.ncbi.nlm.nih.gov/35884090/
4. Kostev K, et al. Prescribing of menopausal hormone therapy in Germany: Current status and changes between 2004 and 2016. PLoS One. 2020;15(12):e0243468. https://pubmed.ncbi.nlm.nih.gov/33368726/
5. Mikolajczyk RT, et al. Evaluation of pregnancy outcome records in the German Pharmacoepidemiological Research Database (GePaRD). Pharmacoepidemiol Drug Saf. 2013;22(8):873-880. https://pubmed.ncbi.nlm.nih.gov/23733705/
6. Schink T, Haug U. Investigating drug safety in pregnancy based on the German Pharmacoepidemiological Research Database (GePaRD): A proof-of-concept analysis on the association between valproate and spina bifida. Pharmacoepidemiol Drug Saf. 2023;32(3):354-362. https://pubmed.ncbi.nlm.nih.gov/36401747/
7. Reinold J, et al. Fingolimod, teriflunomide and cladribine for the treatment of multiple sclerosis in women of childbearing age: description of drug utilization and exposed pregnancies in Germany. Pharmacoepidemiol Drug Saf. 2023;32(2):220-229. https://pubmed.ncbi.nlm.nih.gov/36174258/
8. Wentzell N, et al. Optimizing an algorithm for the identification and classification of pregnancy outcomes in German claims data. Pharmacoepidemiol Drug Saf. 2018;27(9):982-992. https://pubmed.ncbi.nlm.nih.gov/30022557/
9. Jobski K, et al. Non-Steroidal Anti-Inflammatory Drug Use and the Risk of Acute Myocardial Infarction in the General German Population: A Nested Case-Control Study. Drugs Real World Outcomes. 2017;4(3):127-137. https://pubmed.ncbi.nlm.nih.gov/28676983/
10. Langner I, et al. Individual mortality information in the German Pharmacoepidemiological Research Database (GePaRD): a validation study using a record linkage with a large cancer registry. BMJ Open. 2019;9(7):e028223. https://pubmed.ncbi.nlm.nih.gov/31270118/
11. Wentzell N, et al. Optimizing an algorithm for the identification and classification of pregnancy outcomes in German claims data. Pharmacoepidemiol Drug Saf. 2018;27(9):982-992. https://pubmed.ncbi.nlm.nih.gov/30022557/
12. Garbe E, et al. Linkage of mother-baby pairs in the German Pharmacoepidemiological Research Database. Pharmacoepidemiol Drug Saf. 2011;20(3):258-264. https://pubmed.ncbi.nlm.nih.gov/21351307/
13. Dathe K, et al. Avoiding Time-Related Biases: A Feasibility Study on Antidiabetic Drugs and Pancreatic Cancer Applying the Parametric g-Formula to a Large German Healthcare Database. Clin Epidemiol. 2021;13:997-1008. https://pubmed.ncbi.nlm.nih.gov/34737647/