The National Radiology Data Registry (NRDR)® is a collection of registries related to a range of radiological procedures and their quality and safety. The data in these registries help organizations with many tasks, such as accurately assessing their performance and marking areas of opportunity for quality improvement. This data can also enable research, including in health equity and disparities where small or incomplete data sets are considerable barriers.1 The Bulletin spoke with Michael Simanowith, MD, director of ACR Quality Registries, about how NRDR’s vast quantities of data can help researchers measure and understand inequities, as well as the equity studies that are already underway.
Why is it especially important right now to advance health equity and examine equity/disparity issues?
The COVID-19 pandemic has brought health equity and disparities front and center. The pandemic has stretched financial and human resources to critical levels, which in turn delays everyday care. We have seen evidence from the data in our cancer screening registries that as waves/variants of the pandemic spread, “less critical” tests such as screenings are delayed and possibly skipped altogether.2 Making sure disadvantaged populations that may already be challenged in getting usual care are not lost to follow-up is likely to be a major challenge. We need to start assessing the impact on these patients now so we can potentially intervene and minimize any long-term consequences.
Why is it important for the ACR to have to resources and partnerships like NRDR that advance health equity through research?
The ACR’s mission is to serve patients and society by advancing the practice and science of radiological care. This is meant to apply to all patients, regardless of background, so advancing health equity is clearly within our mission. NRDR data offers a unique opportunity to explore questions of equity by using our robust data to identify care gaps that should be targeted to reduce disparities. The sheer volume of available data in NRDR — more than 100 million exam records from 5,000 facilities over ten years — can facilitate progress in these population groups that are often small and for whom other data sources aren’t as readily available or in a large enough sample size. NRDR data also contains valuable clinical detail that can be used to do more advanced analysis (such as risk adjustment) that other sources, such as claims data, often do not.
How can researchers use NRDR data fields to look at health equity issues?
Although the primary intent of the registries is for quality improvement at participating hospitals, NRDR registries capture key demographic information, such as patient race/ethnicity, age, gender, and education level. This data can be used to explore disparities, such as exams performed, adherence to screening recommendations, key population health metrics (such as colon, breast, lung cancer detection and stage at detection) and more.
Some demographic fields are optional for participating sites to include, so the ACR quality and safety team works with researchers to determine the most appropriate and complete variables in a registry to achieve their research aims. To that end, we have recently released an online report with counts for each available variable that allows researchers to determine the viability of using the NRDR data for their investigation. This report is currently available for the Lung Cancer Screening Registry (LCSR); the National Mammography Database (NMD) report was made available in February and will be followed by other registry reports in the future.
Are there any studies you’re excited about?
One recent study that is poised to have an important impact, and really highlights the power of the NRDR data was recently published in Radiology. Using five years of data from the Dose Index Registry, researchers developed radiologic diagnostic reference levels (DRLs) and achievable doses (ADs) for the top ten most commonly performed pediatric CT exams. This represents over 1,500 facilities and 1.5 million cases. It is one of the first papers to report DRLs and ADs in a pediatric population and can help guide CT facilities in adjusting pediatric CT protocols and resultant doses.
How can researchers get involved with NRDR?
We are very interested in exploring research topics with any investigator. You don’t need to be a registry participant to submit a request — though we do give preference to participants in the event that multiple requests are made for a similar topic. There is an online portal available to make a submission.
Each registry has a committee that evaluates all requests to ensure the research has merit and there is no conflict with other projects. If the topic is approved, we will assign an analyst to work with you. We have a highly qualified team who will perform the actual data analysis and provide summary tables/figures to the investigator. We do not share raw- or row-level data externally. In general, there is currently no fee to conduct a research project with NRDR and typical projects take about six to nine months to complete, depending on the level of complexity.