Our webinar, Exploring Real World Data in Germany, part of the global Health Data Passport webinar series, provided an in-depth look at the utilization of Real World Data (RWD) assets from Germany for medical and pharmaceutical research.

This comprehensive session highlighted the benefits of RWD, provided examples of data sets, and discussed various research applications, all while maintaining a positive outlook on the potential of these data sources.

German Real World Data and Technology Experts

The webinar featured three distinguished speakers providing insight into German Real World Data, and available technologies. These included Dr. Silvia Dombrowski, Dr. Céline Vetter, and Dr. Agnieszka Wolk, each bringing a wealth of expertise to the discussion.

Dr. Dombrowski, an Engagement Manager, Associated Director and Team Lead at IQVIA Germany, manages health insurance data studies and leads multi-country studies on statutory health insurance data. Her role involves guiding several teams in health data analytics and research, management of projects based on claims data conducting for pharmacology clients.

Dr. Vetter, an Engagement Manager and Team Lead at IQVIA Germany, is a trained epidemiologist with extensive experience in prospective and retrospective RWD analyses. She is an expert scientific communicator with over 85 published peer-reviewed articles and more than 50 invited lectures at scientific meetings.

Dr. Wolk, the Vice President of Data Science and Advanced Analytics at IQVIA Germany, leads the European Data Science and Advanced Analytics team. She has a profound understanding of the pharmaceutical market and broad experience in analytics, including statistical, econometric, and epidemiological modeling, as well as artificial intelligence and machine learning.

Exploring Germany's Healthcare System

The webinar began with an overview of Germany's healthcare system, which provides universal health coverage through statutory health insurance (SHI) and private health insurance (PHI). SHI is financed by contributions from employees, employers, and the government, covering nearly all residents. This system is based on non-profit sickness funds that negotiate contracts with healthcare providers and pay for care. PHI, on the other hand, is an option for self-employed individuals, public officials, and employees above a specific income threshold.

Sources of Real World Data in Germany

The speakers then introduced the various sources of Real World Data available in Germany. These include the Disease Analyzer, which collects data from primary care practices; hospital data, which encompasses information from hospital treatments; prescription data from pharmacies (referred to as LRX Data), which is gathered from pharmacies; and insurance claims data, which is obtained from health insurance companies. Each of these Real World Data sources offers unique insights and has specific strengths and limitations.

How German Real World Data Addresses Crucial Research Questions

German Real World Data can answer a wide range of research questions. For instance, it can provide information on population and disease characteristics, such as the prevalence and incidence of diseases, common comorbidities, disease progression, and diagnostic practices. It can also shed light on treatment outcomes and effectiveness, comparing different treatments in terms of survival rates and other endpoints. Additionally, RWD can reveal treatment patterns, including the timing of treatment initiation, sequences of treatment, and patient adherence to treatment regimens. Furthermore, it can provide insights into resource use and costs, detailing the expenses and resources involved in treating diseases and how these vary with different therapeutic approaches.

Detailed Data Insights

Webinar speakers provided a detailed look at the various available German data sets. The prescription database, or LRx Data, covers approximately 80% of dispensed prescriptions in Germany, making it suitable for analyzing special preparations and niche products. The Disease Analyzer offers a comprehensive view of patient journeys, including demographics, diagnoses, treatments, lab values, and outcomes. Hospital data provides insights into inpatient care, including drug administration, diagnoses, and procedures, Insurance claims data covers both outpatient and inpatient care, demographics, prescriptions, and costs, although there is a delay of one year in data availability.

Prescription Data (LRx Data)

Dr. Céline Vetter elaborated on the LRx Data (prescription database) which is based on prescriptions of patients insured in the statutory health insurance system in Germany. The data is processed by data centers that allow pharmacies to be paid for the dispensed medication by the insurance companies. This database covers about 80% of the dispensed prescriptions in Germany, with coverage varying across different regions. The LRx database is particularly suitable for analyzing special preparations, niche products, and orphan drugs. It provides longitudinal data, enabling the evaluation of patient prescription histories and the calculation of analytics such as compliance parameters and treatment KPIs.

Disease Analyzer

Dr. Silvia Dombrowski discussed the Disease Analyzer, which mirrors the primary care setting and is part of the Darwin network used by the European Medicines Agency. This data source allows for a detailed assessment of patients' medical histories and physicians' behaviors. The Disease Analyzer covers patient data from over 2,800 offices and more than 3,600 physicians, including general practitioners and 13 individual specialties. Data collection occurs through transmission via practice software systems with monthly updates, and it contains over 14 million patient journeys with at least three years of back data. This robust data source provides a holistic view of patients, including practice information, patient demographics, actions, laboratory values, diagnoses, prescriptions, and lab values.

Hospital Data

Dr. Vetter also highlighted hospital data available in the German Real World Data landscape. This data set includes EMR data from about 30 hospitals across Germany, allowing for both retrospective and longitudinal insights from day-to-day care. The hospital patient data covers patient characteristics, drug administration information, diagnoses, and procedures. It provides a comprehensive picture of patient care in the inpatient sector, including outcomes such as hospital stay duration, discharge reasons, and mortality rates. A pilot study on patients with atrial fibrillation demonstrated the power of this data asset, showing how heart insufficiency is associated with longer hospital stays and higher CRP values, a marker of inflammation.

Insurance Claims Data

Dr. Dombrowski also discussed German insurance claims data, which IQVIA accesses through partnerships. German insurance claims data covers outpatient and inpatient care, demographics, medical prescriptions, other medical services, and costs. The claims data is strictly anonymized and includes only treatments that are reimbursed according to German law. Claims data provides valuable insights into healthcare resource utilization and costs, supporting various types of studies, including prevalence and incidence studies, disease burden assessments, and healthcare resource utilization analyses. There is a delay of at least one year in the data availability. German claims data does not include self-paid health services or information about risk factors and clinical severity.

Advanced Analytics and AI/ML Applications

One of the key highlights of the webinar was the discussion on advanced analytics and the application of artificial intelligence and machine learning (AI/ML) to Real World Data. The speakers explained how patient pathways can be analyzed to understand typical pathways prior to diagnosis and disease progression. Predictive analytics can be used to identify undiagnosed patients, study treatment intolerance, and analyze patient pathways. The AutoML engine developed by IQVIA supports multiple AI/ML models for disease detection, indication inferences, and therapy eligibility, providing high-quality, standardized advanced analytics.

The AutoML engine is a sophisticated tool that automates the process of building and deploying machine learning models. It allows researchers to quickly and efficiently analyze large datasets, uncovering patterns and insights that would be difficult to detect using traditional methods. The engine supports a variety of machine learning techniques, including supervised and unsupervised learning, clustering, and predictive modeling. This flexibility enables researchers to tailor their analyses to specific research questions and data characteristics.

In the context of disease detection, the AutoML engine can be used to identify patients who are at high risk of developing a particular condition based on their medical history and other relevant factors. For example, by analyzing patterns in electronic medical records (EMR) and prescription data, the engine can identify patients who exhibit early signs of a disease, allowing for earlier intervention and potentially better outcomes. Similarly, the engine can be used to study treatment intolerance by identifying patients who are likely to experience adverse reactions to a particular medication, enabling healthcare providers to tailor treatments to individual patients' needs.

Use Cases - AI and Machine Learning with Real World Data

Dr. Wolk provided several use cases of AI/ML applications of Real World Data, demonstrating the practical benefits of these advanced analytics techniques.

Find 50X More Patients with EMR and Prescription Data

A notable example involved the identification of undiagnosed patients with idiopathic pulmonary fibrosis (IPF). By building a model on EMR and prescription data, the researchers were able to identify undiagnosed IPF patients 50 times better than random guessing. This model identified an additional 3,000 patients who had a similar profile to diagnosed and treated IPF patients. For each patient, the model provided a clear explanation of why they were considered high risk, based on a list of different events pointing to undiagnosed IPF.

Using AI/ML to Improve Patient Outcomes

Another AI/ML use case focused on studying statin intolerance. Statins are commonly prescribed to reduce cholesterol, but approximately 15% of patients experience intolerance due to various clinical symptoms. The researchers used supervised learning to understand the symptoms related to statin intolerance and unsupervised learning to cluster patients into distinct groups. This approach enabled the identification of patient clusters that could inform diagnosis and optimal treatment pathways, improving patient outcomes and reducing the risk of adverse reactions.

Analyzing Patient Pathways

The third RWD AI/ML use case explored the mutual relationship between epilepsy and depression. The researchers investigated several questions, such as the predictors of depression in patients with epilepsy, the time it takes for patients to develop depression after being diagnosed with epilepsy, and the risk factors leading to epilepsy-related hospitalizations. By analyzing patient pathways, the researchers gained valuable insights into the interplay between these two conditions, which could inform more effective treatment strategies.

A fourth use case involved indication inferences, where the researchers combined EMR and prescription data to enhance the depth of the data. The AI/ML model learned the relationship between prescriptions and indications from the EMR data and then applied this knowledge to the prescription data. This approach boosted the depth of the prescription data.

Explore Real World Data in Germany - Watch the Webinar

The webinar Exploring Real World Data in Germany showed the richness and effectiveness of Real World Data in Germany for advancing medical and pharmaceutical research. The integration of various German data sources combined with advanced analytics capabilities provides comprehensive insights into patient care, treatment outcomes, and healthcare resource utilization. This wealth of information can drive innovations and improve healthcare outcomes, making German Real World Data an invaluable asset for researchers and healthcare professionals alike. Watch the webinar on demand or contact us to connect with our team of experts.

See other webinars in the global Health Data Passport webinar series on demand, or learn more about Real World Data and IQVIA RWD resources.