The Clinical Quality Language as a tool to support data analysis in German clinical cancer registries
Kolja Blohm and David Korfkamp and Joachim Hübner and Florian Oesterling and Stefanie Schulze and Andreas Hein
GMS Medizinische Informatik, Biometrie und Epidemiologie
In this paper we evaluate the suitability of the Clinical Query Language (CQL) for supporting various data analysis tasks in German cancer registries. CQL is a domain-specific query language for clinical data. It is particularly used in the United States, for example to define cohorts and to calculate electronic Clinical Quality Measures. We developed a prototype to execute CQL queries on clinical cancer registry data. For this purpose, we created a unified CQL data model for the analysis of clinical cancer registry data. This model is based on the datasets of the cancer registry of North Rhine-Westphalia and the Agency for Clinical Cancer Data of Lower Saxony, as well as the Oncological Basis Dataset. For the evaluation, we applied CQL to typical questions from the areas of guideline-based quality indicators, data requests from external researchers, plausibility checks, and routine reporting, and compared the results with well-established evaluation methods. We were able to show that CQL is capable of representing the complex criteria and temporal relationships that are often relevant for the analysis of data from clinical cancer registries. We see CQL as a promising method to support cancer registries in the definition of patient cohorts for various internal and external analyses and believe that the use of CQL in combination with a standardized data model can make a significant contribution to the standardization of analyses in cancer registries.
8 / 2024
article
SePaMiM
Sequential Pattern Mining und Pattern Matching von Krankheits- und Behandlungsverläufen für klinische Krebsregister
Sequential Pattern Mining und Pattern Matching von Krankheits- und Behandlungsverläufen für klinische Krebsregister
