Patient registries

What is it?

Patient registries are organised systems with a ’predetermined scientific, clinical or public health purpose’ (for further information, see Cross-Border Patient Registries Initiative [PARENT]). They are used to prospectively collect, analyse and disseminate data on a group of patients with specific characteristics in common, for example people with a particular condition, having a certain treatment or using a health-related service. They are cohort studies that are developed with a predetermined health-related aim. Patient registries have evolved from paper-based records in a doctor’s office to data collected electronically, usually in databases. They often contain large amounts of data, such as clinical information or biological samples stored in biobanks.

Patient registries may be classified as:

• disease or condition registries, based on populations with a particular disease or group of diseases
• product-specific registries, based on populations using specific products, often developed by manufacturers to assess long-term safety and adverse effects.

Examples of patient registries

• The European Cystic Fibrosis Society Registry is used to collect data on cystic fibrosis and its treatment, both to improve treatment and provide data for epidemiological research.
• The Cross-Border Patient Registries Initiative (PARENT) has developed a ‘registry of registries’ that lists over 200 patient registries across Europe.

Why is it useful?

• Informing routine clinical practice and disease management: especially in post-marketing surveillance.
• Estimating effectiveness and/or relative effectiveness: by collecting data on:
  • the natural history of disease
  • clinical and/or comparative effectiveness outcomes
  • safety or harm from exposure to health technologies
  • quality of care and quality of life

Registries are being used to:

• support the European Medicines Agency’s (EMA’s) initiatives for early access to medicines, such as the EMA Adaptive Pathways Approach, which strongly builds on real-world data
• monitor orphan drugs, personalised medicines, advanced therapies and biosimilars
• provide insight into the use of medicines under real-life conditions, which inform health technology assessment discussions
• improve healthcare systems, through disease registries that are part of health systems and daily practice and through product-specific registries that provide information on the safety and effectiveness of treatments used in practice.

What are its limitations?

• Quality of data and methods: patient registries are only as good as the underlying data collection and methods. It is critical that the research design is robust and that the relevant outcomes are defined appropriately.
• Incorrect or missing data: problems with remembering information or recalling information correctly can be an issue with patient-reported data.
• Confounding and bias: there may be issues such as channelling bias (where sicker patients are assigned to newer treatments), information bias and selection bias.
• Heavy investment in time and resources: commitment and engagement are needed from participating patients and centres. It can be difficult to recruit patients and it can take time to accumulate data, resulting in delays between data collection and reporting (sometimes years). Maintaining energy and interest over long periods may be challenging. Data can become out of date in fast-changing disease areas.
• Lack of standards and uniformity: lack of standardised data collection across hospitals, regions and countries prevents the pooling of data across registries.
• Lack of comparator: in product-specific registries having no data for comparators prevents assessment of relative effectiveness. 

What do stakeholders say?

Acceptability to stakeholders is dependent on the registry methods and robustness of design.

The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) has published guidance on reliability and validity of data sources, including registries. Many healthcare agencies attach some value to single cohort studies (those without comparator data), as these can be especially useful for rare outcomes and for examining the safety of orphan drugs (ISPOR presentation).

Drug exposure and safety measurements based on registries are well accepted by regulatory bodies and health authorities: these data are often required in post-marketing surveillance.

Advantages of patient registries for patients include:

• outcomes that are relevant to patients, such as health-related quality of life, but might be missed or not appropriately captured in trial data can be collected (see the Agency for Healthcare Research and Quality report)
• patients can be actively involved in clinical research and have input into drug development.

However, as registries contain sensitive health information, data security is a higher priority for patients than other issues.

For findings related to the accessibility of patient registries from GetReal case studies, see GetReal Experience Accessing Registry Data.

Key contributor

Denise Umuhire, Janssen EMEA