Planning and Conducting Non-Interventional Studies – Real-World Evidence from ZEG Berlin

Non-interventional studies investigate the use, safety, and effectiveness of therapies under real-world healthcare conditions. Medical treatment is not determined by the study protocol. Treating physicians make their therapeutic decisions independently of a patient’s potential participation in the study.
ZEG Berlin plans and conducts non-interventional studies for companies in the pharmaceutical and healthcare sectors as well as for teams in Medical Affairs, Regulatory Affairs, and Pharmacovigilance. As a Contract Research Organization specializing in pharmacoepidemiology and Real-World Evidence, we support projects from the research question and protocol development through data collection, statistical analysis, reporting, and scientific publication.
Our scientific roots date back to the 1960s, while ZEG Berlin has operated as an independent organization since 1990. Our work is based on transparent methodologies, controlled data processes, and scientifically precise interpretation of results.

What is a Non-Interventional Study?
In a non-interventional study, or NIS, treatment is provided as part of routine medical care. The study protocol does not determine which therapy a patient receives. The prescription of a medicinal product or use of a medical device must also remain independent of participation in the study.
This distinguishes an NIS from an interventional clinical trial. In a clinical trial, the study protocol determines key aspects of treatment and study-related examinations. A non-interventional study, by contrast, focuses on the systematic observation of routine healthcare.
The correct classification of a project influences, among other aspects:
• the appropriate study design,
• the regulatory framework,
• required notifications and submissions,
• the involvement of ethics committees,
• the design of data collection,
• documentation and reporting requirements.
For this reason, we assess at the beginning of each project which design is appropriate for the research question, the product, and the planned approach. The decisive factor is not simply how the study is labeled, but the specific relationship between treatment decisions, study participation, and data collection.
Patient surveys and Patient-Reported Outcomes can also form part of an NIS, provided that the data collection is appropriate for the research question and compatible with the non-interventional nature of the project.

What are Non-Interventional Studies Used For?
Randomized clinical trials generate evidence under clearly defined and controlled conditions. Due to predefined inclusion and exclusion criteria, however, they do not always reflect the full diversity of the population that will later receive treatment.
After approval, medicinal products are used across different age groups, in patients with varying comorbidities, and in heterogeneous healthcare settings. Non-interventional research can complement clinical trial evidence by examining how therapies are used in routine practice and which outcomes can be observed over a defined period.
Typical applications include:
• Post-authorization safety: Post-Authorization Safety Studies investigate defined safety questions concerning authorized medicinal products.
• Effectiveness under real-world conditions: Real-world studies assess treatment outcomes outside a controlled trial setting.
• Long-term or rare events: Large populations and longer observation periods can provide additional insights.
• Product and disease registries: Registries document disease progression, treatment patterns, safety events, and other outcomes.
• Comparative Effectiveness: Comparative analyses assess treatment approaches under real-world healthcare conditions.
• Drug Utilisation: Analyses show how medicinal products are prescribed, used, and switched.
• Adherence and persistence: These analyses examine the duration, interruption, and continuation of treatment.
• Patient-Reported Outcomes: Patients report directly on symptoms, functioning, quality of life, or treatment experiences.
Which method can appropriately answer a research question depends on the target population, endpoints, data source, and regulatory context.

Services Across the NIS Lifecycle
ZEG Berlin supports non-interventional studies throughout the entire project lifecycle. The scope of services can be provided as an integrated study project or as targeted methodological, statistical, or operational support.

Scientific Consulting and Study Design
At the beginning of each project, we define the specific research need together with the client. This includes the scientific or regulatory objective, the relevant population, appropriate comparator groups, and the required endpoints.
When selecting the study design, we consider factors including:
• scientific validity,
• regulatory context,
• availability of suitable data,
• feasibility and recruitment,
• observation period and sample size,
• potential sources of bias,
• time and operational requirements.
Based on these factors, we determine whether prospective data collection, a secondary data analysis, or a hybrid approach is appropriate.

Study Protocol Development
The study protocol forms the scientific and operational basis of the project. It describes, among other aspects:
• background and rationale,
• study objectives and research questions,
• target population and selection criteria,
• exposures and comparator groups,
• primary and secondary endpoints,
• data collection and follow-up,
• statistical framework,
• quality assurance measures,
• organizational responsibilities.
The research question, data source, and statistical analysis are aligned consistently. During protocol development, we also assess whether the required variables and endpoints can actually be captured by the intended data source.

Regulatory Support
Depending on the study type, product, and countries involved, non-interventional studies may be subject to different regulatory requirements.
ZEG Berlin provides support with:
• PRAC- and EMA-related procedures,
• preparation of regulatory study documentation,
• submissions, notifications, and registrations,
• responses to methodological queries,
• coordination of protocol amendments.
The regulatory approach is defined individually for each project.

Operational Study Management
For prospective studies, we provide operational planning and coordination according to the agreed scope of services. This may include the selection of suitable study sites, preparation of study materials, training of participating physicians, setup of data collection, and support with recruitment.
During study conduct, we monitor study progress, incoming data, and agreed quality parameters. Project status and relevant developments are documented in a structured manner and reported to the client.

Data Management and Biostatistics
Robust results require controlled data processes. We develop data structures and workflows that are aligned with the study design and the planned analysis.
Services may include:
• database design,
• electronic data collection tools,
• plausibility and validation rules,
• data cleaning and query management,
• medical coding,
• controlled data transfers,
• preparation of analysis-ready datasets.
Observational data pose specific statistical challenges. Treatment groups are not created through randomization, relevant influencing factors may be unevenly distributed, and data may be incomplete.
Statistical expertise is therefore integrated into study planning at an early stage. The statistical analysis plan defines populations, variables, endpoints, and methods before the main analysis. Depending on the research question, methods may include descriptive analyses, regression models, time-to-event analyses, propensity score methods, and sensitivity analyses.

Reporting and Medical Writing
Results are interpreted in the context of the study design, data quality, and methodological limitations. Interim and final reports undergo internal medical, statistical, and methodological review.
Our Medical Writing services include:
• study protocols,
• statistical analysis plans,
• interim and final reports,
• regulatory documents,
• manuscripts for scientific journals,
• abstracts, posters, and presentations.
Conclusions are formulated with scientific precision. Interpretation remains aligned with the strength of the available data and the limitations of the study design.

Primary and Secondary Data for Real-World Evidence
ZEG Berlin works with prospectively collected primary data and existing secondary data sources. The appropriate basis depends on the research question and the information required.

Primary Data Collection
In primary data collection, information is collected specifically according to a predefined protocol. This includes, among others:
• prospective PASS and PAES,
• product and disease registries,
• hybrid and decentralized studies,
• physician and patient surveys,
• cross-sectional studies,
• active surveillance studies.
Endpoints, covariates, and data collection time points can be aligned with the research question. This also makes it possible to capture symptoms, quality of life, or reasons for treatment decisions that are often not available in routine data.

Secondary Data Analyses
Secondary data analyses use existing datasets to address a new research question. ZEG Berlin works on projects in areas including epidemiology, Comparative Effectiveness, Drug Utilisation, adherence, treatment switching, Real-World Outcomes, and care pathways.
Before starting an analysis, we assess whether the data source adequately captures the population, exposures, endpoints, and relevant influencing factors. A large volume of data alone does not guarantee robust evidence. The origin, completeness, structure, and suitability of the data are decisive.

Addressing Bias and Confounding
Non-interventional studies reflect real-world treatment decisions. These decisions are influenced by medical and healthcare-related factors that may also be associated with the outcomes being investigated.
Potential sources of bias include:
• Confounding by indication,
• selection and information bias,
• misclassification,
• incomplete follow-up,
• missing data,
• time-related biases.
Potential biases are considered as early as the selection of the data source and protocol development. Appropriate methods for adjustment and sensitivity analyses are defined in the analysis plan. Remaining uncertainties are described transparently in the study report.

Patient-Reported Outcomes and Rating Scales
Patient-reported endpoints complement medical findings with the patient perspective. They can systematically capture symptoms, functioning, quality of life, and treatment experiences.
ZEG Berlin has developed several validated Patient-Reported Outcome instruments:
• Menopause Rating Scale,
• Aging Males’ Symptoms Scale,
• Scale for Quality of Sexual Function,
• Short Term Hormone Effect Scale.
The instruments are based on the concept of health-related quality of life and are available in multiple languages. Which scale is suitable for a project depends on the target population, indication, and research question.

Experience in Non-Interventional Research
ZEG Berlin has conducted pharmacoepidemiological studies since 1990 and coordinates research projects in more than 40 countries. The interdisciplinary team includes epidemiologists, medical experts, statisticians, data architects, and medical writers.
As a founding partner of ENCePP, ZEG Berlin follows the network’s methodological principles and standards of transparency. As a CDISC-aligned organization, the company takes established standards for structured data and study documentation into account.
Another area of expertise is the EURAS/INAS design developed by ZEG Berlin. It has been used in Post-Authorization Safety Studies since 2000 and combines direct contact with study participants with the medical validation of relevant events.
ZEG Berlin’s educational workshops also provide knowledge on study planning, operational management, regulatory requirements, statistics, quality management, as well as pharmacovigilance and compliance in non-interventional studies.

Frequently Asked Questions About Non-Interventional Studies

What is the difference between an NIS and a clinical trial?
In an NIS, treatment is not determined by the study protocol. Treatment takes place as part of routine care and independently of study participation. In an interventional clinical trial, by contrast, the protocol determines key aspects of treatment and study-related examinations.

What is a Post-Authorization Safety Study?
A PASS investigates defined questions concerning the safety of an authorized medicinal product. It may be conducted voluntarily or as part of regulatory requirements.

What services does ZEG Berlin provide?
Support may include scientific consulting, study design, protocol development, regulatory coordination, study management, data collection, data management, biostatistics, reporting, and publication support.

Can primary and secondary data be combined?
Yes. Hybrid approaches can combine existing data sources with specifically collected information. The selected data and methods must be appropriate for the research question.

Can Patient-Reported Outcomes be used?
Patient-reported endpoints can be integrated into an NIS if the instrument is suitable for the target population and research question and the data collection is compatible with the study design.

Contact for Your Study Inquiry
Are you planning a non-interventional study, a PASS, or another Real-World Evidence project? ZEG Berlin supports you in selecting the study design, data source, and methodological approach.
ZEG – Berlin Center for Epidemiology and Health Research GmbH
Invalidenstraße 115
10115 Berlin

Phone: +49 30 945 101-20
Email: info@zeg-berlin.de