Lessons I Learned While Building My First Clinical Registry

Author: Seraina Netzer

Afilliations:

1:Department of Geriatrics, Hôpitaux Universitaires de Genève (HUG), Geneva, Switzerland;

2: Department of Geriatrics, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland;

3: Berner Spitalzentrum für Altersmedizin Siloah, BESAS, Gümligen, Switzerland

---

When I started building a clinical registry for patients with hip fracture, I assumed the main challenge would be the data collection itself. What I didn’t expect was how much work happens long before the first patient is enrolled: defining variables, negotiating scope, and trying to keep the whole structure manageable while still clinically meaningful. These are a few early lessons from the process.

1. Defining variables and scope takes longer than you think.

What looks straightforward quickly becomes surprisingly intricate. Every variable needs a clear definition and a shared understanding across professions. I also learned that feasibility is inseparable from extractability: if something is coded, it can be analysed; if it requires digging through narrative notes, it becomes slow, inconsistent, and open to interpretation. The same principles apply to the scope of the registry. For example, we had to negotiate whether to include patients with periprosthetic fractures and those without an operative indication. Although these groups may not be relevant to every analysis, I argued for including them, as they can easily be excluded later, whereas adding patients retrospectively is much more difficult.

2. Avoid variables that try to do too much.

One thing I encountered often were “composite variables” that mix several concepts into one field. They seem efficient but become a nightmare to extract. A simple yes/no variable followed by a second field that specifies the details is almost always cleaner — and much easier for future you (or the person doing the extraction).

3. REDCap logic is powerful and humbling.

For my registry, I used REDCap, a secure web-based platform commonly used for collecting and managing research data. Branching logic, calculated and repeating fields, events and naming conventions can make a registry elegant, but they also introduce hidden complexity. Testing my own instrument was eye-opening: fields I thought were intuitive weren’t, and tiny inconsistencies created unexpected loops. Investing time in structure early saves time later.

4. Regulatory steps need their own space.

Don’t forget to plan time and resources for understanding local requirements like an ethics committee submission. Every institution has its own expectations about data flow, storage, and follow-up. It’s worth informing yourself early.

5. Collaboration is the heart of the process.

Working with orthopaedics, emergency medicine, anesthesiology, nursing, IT, and research support taught me that registry building is fundamentally interdisciplinary. Each group sees different parts of the patient journey, and their feedback often improved the design in ways I hadn’t anticipated. The process itself became a form of team building.

Looking back, building a registry has shaped my thinking far more than I expected. It taught me to balance ambition with feasibility, to communicate clearly across professions, and to design research tools that are both rigorous and realistic. If you are starting your own registry journey, know that the early stages are challenging, but also incredibly rewarding.