Algorithms, broadly speaking, are sets of computational rules or processes, which researchers can use to gain a better understanding of large amounts of data. At HDRN Canada, we are interested in algorithms that measure key health concepts, which can be used to describe study populations or define study outcomes.
Part of our work at HDRN Canada is to collect published algorithms (algorithms that have been investigated in published research) and make them available to other researchers. By making this Inventory as accessible and comprehensive as possible, we’re supporting different researchers or research teams to ensure their research approaches are consistent, reliable and repeatable.
The Algorithm Inventory includes algorithms about measures of population health (e.g. diabetes, depression), health care services use (e.g. costs of care) and social determinants of health (e.g. socioeconomic status, race or ethnicity). Research related to these measures typically require data that are obtained or cross-referenced from more than one data source, so our focus is on algorithms used in multi-site or multi-regional research projects.
The Inventory holds a large number of algorithms, and it is being regularly updated. But we know there are many more we could be including. One of our teams at HDRN Canada – the Algorithms and Harmonized Data (AHD) Working Group – is conducting a study to analyze the Inventory and identify gaps in existing studies. Once we know what types of algorithms might still be needed, we can continue to build an Inventory that is useful and relevant to many different kinds of research and research projects.
We’re looking forward to discussing our findings once the study results are available.