Why research centers are standardizing their patient registries
Patient registries are the hotbed for research in tracking clinical care and outcomes, and leading to better support for care management. We are seeing evidence of a trend in patient registry management that looks set to finally start breaking down some of the biggest barriers to this.
The frustration for researchers is that despite the purpose of the patient registry almost always being pre-determined, the mechanisms to help them fulfil that purpose is challenging. This is because there is rarely any uniformity to the clinical (and other) data, with information stored in and utilizing different, and often bespoke, formats.
It’s a frustration that is inherent whether considering population-based registries, or hospital-based. And with more collective hospital registries collaborating on specific conditions, exposures, or diseases, the challenge is often magnified.
Registry developers often have very little support when it comes to designing registry data elements. While there are some standard terminologies, these may not always be sufficient, and the same concept may be represented in a variety of ways, depending on the setting. Data is collected for different purposes and therefore may be stored in different formats, using different database systems and information models. We would do well to consider here, that research is often not the primary consideration when data is collected. Administrative and clinical purposes are more likely to be the drivers.
What’s the problem with non-standardized data?
Non-standardized data ultimately results in a non-holistic view of a registry. There may also be data integrity issues as a result of inconsistent authoring of data elements, hampering opportunities for coordinated research.
The upshot is that the burden becomes point-loaded on a small collection of individuals or a small team to develop and maintain the knowledge of hugely complex schemas and data dictionaries. This is difficult to impart to researchers in a way that will give either side confidence that observational research is based on holistic and robust evidence.
You can just imagine the scenario: patient registry researchers present their data requirements to database analysts in informatics teams, who then need to interpret the request against the schemas, write the relevant code, and return it to the requestor. That’s a transaction that will happen dozens of times in a single research project, and I’m probably underestimating!
How can standardized data help researchers?
A registry using standardized data elements allows better consistency and a more holistic view of the registry.
There are plenty of good ways to go about standardizing registry data, and that’s comforting, considering there isn’t an option to start from scratch. Using a Common Data Model (CDM) to harmonize datasets is demonstrating that it can produce scalable and repeatable tools for observational research.
Standardization creates great opportunities for interoperability and efficiency within and across clinical registries, helping to bring together data from disparate sources. Along with producing higher quality and more reliable data, standardization also streamlines clinical research by allowing more efficient data collection, not by the analysts, but by the researchers themselves. When the model is standardized, you have removed one of those big blockers to research: the person who can act on the data can now understand how it fits together and how the individual sources and records fit together. Introducing distributed analytical solutions is also more likely to be successful because you have those definitions to refer to.
And don’t worry about the analysts – you researchers will still need them! But the relationship between researcher and analyst becomes less transactional and more collaborative: the analyst imparting their knowledge of the standardized model and advising the researcher how best to harness it within their area of clinical expertise.
We are seeing an increase in the number of research centers looking to standardize their data, and the hope that all registries use standardized data elements in the future would be a great step forward for clinical research, allowing more collaborative research, time and cost savings, and ultimately, more significant discoveries.