Topics of interest

The individual randomized clinical trial has often been presented as the only method for finding causality. But this paradigm is waning in proportion to its increasingly accepted limitations: lack of external validity, low power of subgroup analyses, small numbers that do not guarantee comparability, difficulties of random selection, etc. DHSs therefore represent research infrastructures with considerable potential for evidence-based medicine. In this context, two complementary dynamics can be identified within the SCALE-EPI axis of CIC 1402.

The first is aimed at removing the technological barriers to the creation of high-quality EDS for research purposes. It is being driven in particular by the Poitou-Charentes general cancer registry, whether for reliable estimates of public health indicators (1) or for comparisons of therapeutic alternatives in high-definition sub-cohorts (2). Bringing together hospital DHSs and registries will also help build the complementarity of these information systems.

The second dynamic focuses on new statistical methods needed for causality from these real-life data: machine learning to emulate clinical trials from high-dimensional data (3), time-dependent propensity scores to emulate clinical trials with time-dependent treatments (4), multi-state models to model health trajectories and simulate the impact of intervention (5). Thanks to these methods, a new project aims to use historical data such as EDS to increase the performance of future clinical trials and reduce the number of subjects required.

Main publications

1. Defossez G, Uhry Z, Delafosse P, Dantony E, d'Almeida T, Plouvier S, Bossard N, Bouvier AM, Molinié F, Woronoff AS, Colonna M, Grosclaude P, Remontet L, Monnereau A; French Network of Cancer Registries (FRANCIM). Cancer incidence and mortality trends in France over 1990-2018 for solid tumors: the sex gap is narrowing. BMC Cancer. 2021 Jun 24;21(1):726. 
2. Systchenko T, Defossez G, Guidez S, Laurent C, Puyade M, Debiais-Delpech C, Dreyfus B, Machet A, Leleu X, Delwail V, Ingrand P. R-CHOP appears to be the best first-line treatment for second primary diffuse large B cell lymphoma: a cancer registry study. Ann Hematol. 2020 Jul;99(7):1605-1613.
3. Le Borgne F, Chatton A, Léger M, Lenain R, Foucher Y. G-computation and machine learning for estimating the causal effects of binary exposure statuses on binary outcomes. Sci Rep. 2021 Jan 14;11(1):1435.
4. Sabathé C, Casey R, Vukusic S, Leray E, Mathey G, De Sèze J, Ciron J, Wiertlewski S, Ruet A, Pelletier J, Zéphir H, Michel L, Lebrun-Frenay C, Moisset X, Thouvenot E, Camdessanché JP, Bakchine S, Stankoff B, Al Khedr A, Cabre P, Maillart E, Berger E, Heinzlef O, Hankiewicz K, Moreau T, Gout O, Bourre B, Wahab A, Labauge P, Montcuquet A, Defer G, Maurousset A, Maubeuge N, Dimitri Boulos D, Ben Nasr H, Nifle C, Casez O, Laplaud DA, Foucher Y. Improving the decision to switch from first- to second-line therapy in multiple sclerosis: A dynamic scoring system. Mult Scler. 2023 Feb;29(2):236-247.
5. Bonnemains V, Le Borgne F, Savoye E, Legeai C, Pastural M, Bayat-Makoei S, Lenain R, Ragot S, Leffondré K, Couchoud C, Foucher Y. Impact of the Kidney Transplantation Moratorium in France Because of the COVID-19 Pandemic: A Cohort-based Study. Transplantation. 2022 Dec 1;106(12):2416-2425.