Development of a rare disease registry: Valuable lessons learned on how to build a sustainable disease registry

Disease registries/databases in the field of rare diseases are important since basic disease data necessary for clinical study planning can be obtained. However, many investigators using them have difficulties in securing labor and quality in data management and analysis. Our data center (DC) has supported a pediatric leukemia/lymphoma registry in linkage with clinical studies for more than 10 years, and our experience showed that their separation holds the key to effective data management and analysis. Therefore, we clarified known problems of disease registries/databases, and made efforts to optimize labor and quality by taking measures at the construction stage. On the basis of mock tables and figures that were expected to be generated at the time of completion of the paper, the principle investigator, biostatistician, data manager, and system engineer gathered together and undertook substantial discussions (Figure first). As a result, the initial plan to construct a single registry/database was changed, and it was finally separated into 3 parts: a disease registry, prospective observational study, and cross-sectional study. After their simultaneous initiation, cases were accumulated more rapidly than expected. These results suggest that the "Figure first" method allows the design of efficient clinical studies linked with a disease registry.