Responder status and long term outcomes of critically ill patients with ARDS receiving high dose steroids as rescue therapy in a specialist acute respiratory failure centre

Abstract

<b>Background:</b> High-dose steroids (HDS) are used in specialist settings for the treatment of ARDS with features suggesting steroid responsiveness or as a last resort.&nbsp;Our previous case-matched cohort&nbsp;demonstrated that&nbsp;HDS&nbsp;increase 6mo. mortality (Remmington et. al, 2019). <b>Aims and objectives:</b> Identify if response to HDS predicted favourable outcomes. <b>Methods:</b> A retrospective observational study from a tertiary specialist respiratory ICU identified 79 patients (5 excluded) between 2009-2017 who received&nbsp;HDS for ARDS (AECC/Berlin criteria). Patients were grouped according to the change in lung injury score (LIS) between days 0 and 10 following HDS initiation. A modified LIS (excluding PF ratios for patients on ECMO) was used. Changes in LIS were grouped as follows; full responders (FR) ≥ 2, partial responders (PR) ≥ 1 and &lt; 2, and non-responders (NR) &lt; 1. <b>Results:</b> This study identified 7 (9.5%) FR, 14 (18.9%) PR, and 53 (71.6%) NR. Figure 1 shows the mortality for each group, which was significantly reduced when comparing the combined PR+FR groups to the NR group (p (log rank test) &lt; 0.0001). Figure 1 – mortality over time in patients with different LIS responses <b>Conclusions:</b> A low proportion (28.4%) of ARDS patients given HDS had a FR or PR by day 10 yet&nbsp;these patients had significantly reduced mortality at 24mo.&nbsp;Further research using modified LIS to predict steroid responsiveness may assist clinicians in selecting ARDS patients for rescue therapy.