Background: The development and use of health databases have heightened the need for tools, particularly recognized and validated algorithms for identifying hospital stays or patients with cancer. As part of the development of the REDSIAM network (Sniiram Data Network, National Inter-Regime Health Insurance Information System), the "Tumors" group aimed to compile an overview of algorithms published in French and international oncology studies, in order to enable readers seeking algorithms to identify those likely to be of interest to them. Method: A non-systematic literature search was conducted for various cancer sites. The results are presented for lung, breast, colon, and rectal cancer. Publications up to August 2016 were searched in the Medline, Scopus, Public Health Database, and Google Scholar databases, as well as in the abstracts of major French journals in oncology and public health. A data extraction form tailored to oncology was developed and used during the extraction process. Results: A total of 18 publications were selected for lung cancer, 18 for breast cancer, and 12 for colorectal cancer. The performance and choice of algorithms used depend on the context, objective, and site of cancer studied. The description grid developed by the "Tumors" group is more detailed compared to the bibliographic framework shared by other REDSIAM groups, but it remains simple to populate. Conclusions: This study highlights the complexity of detecting cancer cases using health databases and the lack of validated algorithms. It is necessary to promote and develop efforts aimed at standardizing and facilitating the validation of these algorithms.