The possible association of steroids with fluid accumulation in critically ill patients – a case of a potential bias

Background: Glucocorticoids (GCS) are commonly administered to critically ill patients. Due to their mineralocorticoid effect, GCS might have a substantial influence on a positive fluid balance. We assessed the association between glucocorticoids (GCS) therapy and fluid balance in critically ill patients with sepsis.Methods: This is a retrospective study of patients with sepsis hospitalized during 2006-2018 in a general intensive care unit (ICU) at a 1100-bed tertiary medical center.Results – We considered two definitions of exposure: daily exposure to GCS and GCS treatment at any time in the ICU. Of 945 patients with a diagnosis of sepsis, 375 were treated with GCS at any time and 570 were not. We applied four regression models to assess the association between GCS treatment and fluid balance; in our first model, fluid balance did not differ during days with GCS treatment, between patients who were and were not treated with GCS in the ICU (coefficient estimate 79.5 (-55.4 to 214.4), p=0.25). In our second model, daily fluid balance was increased by 139.8 ml (10.8 to 268.9; p=0.03) in patients who were ever treated with GCS during their ICU stay compared to untreated patients. In the third model, which included only patients treated with GCS during their ICU stay, GCS treatment days were not associated with daily fluid balance (coefficient estimate -190.6 (-485.1 to 103.9), p-value=0.21). In the last model, on "steroid free days", patients who received GCS treatment during their ICU stay had a positive fluid balance compared to those who were never treated with steroids (coefficient estimate 157.7 (-24.6 to 340.1), p-value=0.09).Conclusions – Despite their known mineralocorticoid activity, GCS themselves appear not to contribute substantially to fluid retention. The findings highlight the importance of a clear definition of exposure.