Endocrine Abstracts

Background: Dexamethasone significantly improved outcomes in patients requiring supplementary oxygen and in ventilated patients with COVID-19 in the RECOVERY trial. Consequently, dexamethasone is now routinely used in these patients. However, dysglycaemia is commonly associated with steroid use and is an established risk factor for poorer outcomes in COVID-19. In this study, we aimed to elucidate the effect of dexamethasone use in patients hospitalised with COVID-19 in a real-world setting.

Methods: : Data from Imperial College Healthcare NHS Trust hospitals were collected from 1372 consecutive patients hospitalised with COVID-19 between 01/11/2020 and 31/01/2021 (Wave 2) and 889 patients admitted between 09/03/2020 to 22/04/2020 (wave 1). The primary outcome was admission to intensive care (ICU) or death within 30 days of COVID-19 diagnosis. Secondary endpoints were post-dexamethasone glycaemic complications. Multivariate logistic regression analyses were performed to determine the factors associated with primary outcome and to determine impact of dexamethasone on the primary outcome.

Results: Mortality alone, without accounting for ICU admission, was significantly lower in wave 2 (wave 27.6%, wave 2 18.8%, 31.8% reduced risk of death, P < 0.01). Male gender, hypertension, increased frailty and lower eGFR were independently associated with the primary outcome. Dexamethasone significantly reduced the risk of death/ICU admission by 56%. In patients with diabetes, dexamethasone use is associated with increased risk of glycaemic complications (OR=22.5, 95% CI 13.98-36.67, P < 0.0001). However, the risk of death/ICU admission was not increased in those with post-dexamethasone complications.

Conclusions: : Dexamethasone reduced the risk of death/ICU admission. There was no difference in ICU admission rates between waves 1 and 2, possibly driven by the dominance of new SARS-Cov2 variants. Patients with diabetes are more likely to develop steroid-induced dysglycaemia, but this did not increase mortality.