<img alt="" src="https://secure.lote1otto.com/219869.png" style="display:none;">
Skip to content

Linked Data Studies: Improving the Way We Do Observational Research in Germany

Asthma affects more than 235 million people worldwide, and due to lacking effective implementation of clinical guidelines, there continue to be high rates of uncontrolled asthma. In order to evaluate differences in healthcare resource utilization and cost among patients with controlled and uncontrolled asthma in Germany, Cytel experts applied a unique approach: a linked data study comparing insurance claims data with primary data. Let’s take a closer look:

What is a linked data study?

A linked data study, as the name implies, combines data from different sources to create a more useful data set.

In this case, as co-authors Thomas Wilke, Sabrina Mueller, and Fraence Hardtstock explain in their recently published paper for Health Services Management Research, they linked primary patient- and physician-reported data and secondary claims data for 492 asthma patients.

“German claims data is a valid and often used source for health-economic analyses . . . . However, claims data generally does not cover any symptom-related or patient reported outcomes. For this reason, we applied a linked data approach, in which primary patient-reported data on asthma control was collected by the research team and subsequently linked to the claims data by an independent trust centre. This linking was based on insurance numbers (deterministic linking), so no probabilistic linking was applied.”

This approach, the authors argue, has the potential to improve the way observational research is done; whereas data linkage is known as a standard approach in the United States, it has hardly been done in Germany.

How can linked data studies improve the way observational research is conducted?

In terms of real-world evidence, there are typically two types: claims (those insurance gets for reimbursement) and medical records (coming from physicians). “The first lack clinical detail; the second typically come from one outpatient or inpatient site only (i.e., you are missing the data from all the other health care facilities),” adds Thomas Wilke, Cytel Real World Principal Scientist. Therefore, “combining both is the best of all worlds, but due to formal hurdles, it is very difficult to accomplish in Europe.”

However, Cytel, with its academic partner IPAM, is uniquely able to link claims with other data on a patient level. (Please contact us for more information.)

To learn more about this study, including its conclusions and a more in-depth discussion of their methods, read “Association between Asthma Control and Healthcare Costs: Results from a German Linked Data Study” here:


Access Publication

Read more from Perspectives on Enquiry and Evidence:

Sorry no results please clear the filters and try again

Why Are There Not More Bayesian Clinical Trials?

Statistical methods have long been fundamental to drug development, and advancements in the last few decades in..
Read more

Topics in Bayesian Statistical Methods

Bayesian methods have been playing a key role in transforming clinical research, and Bayesian topics are frequently..
Read more

Bayesian Strategies in Rare Diseases

When it comes to rare diseases, a handful of major challenges to drug development arise. Bayesians strategies have..
Read more
contact iconSubscribe back to top