Abstract
Immunocompromised patients were excluded most of the time from trials testing corticosteroids in COVID-19. This study aimed to assess the associations between early corticosteroid use and (1) mortality at day 60, and (2) the occurrence of nosocomial infections in immunocompromised patients with severe COVID-19 admitted to the ICU. It was a multicentre retrospective study, achieved in French ICUs of the Outcomerea™ network and medical ICUs of 4 other hospitals in France. This study included immunocompromised patients admitted to an ICU between January 1, 2020, and August 31, 2022, for severe COVID-19, with an ICU stay of more than 2 days. Patients were classified as receiving early corticosteroid therapy if they were given steroids within the first 5 days following ICU admission. Each patient was categorized into one of four immunosuppression subgroups: 'corticosteroid therapy,' 'monocytic alteration,' 'cellular immunosuppression,' or 'humoral immunosuppression.' Survival analyses were performed, and confounding by indication was addressed using propensity score weighting with overlap. 383 patients were included, 50 were into the no-early-corticosteroids group and 333 in the early-corticosteroids group. In the overlap cohort, 118 were included (46 in the non-early-corticosteroids and 72 in the early-corticosteroids group). There was no association with day-60 mortality ((IPTWoverlap)HR = 0.97, 95% CI [0.51; 1.85], p = 0.92). There was also no association with the occurrence of nosocomial infections ((IPTWoverlap)SubHR = 2.59, CI 95% [0.77; 8.7], p = 0.13). We report that steroids had no benefit on mortality in immunocompromised patients admitted to ICU for severe COVID-19.