Red blood cell distribution width as a marker of hyperinflammation and mortality in COVID-19

Red blood cell distribution width (RDW) could reflect interleukin-6 (IL-6) systemic activity since anisocytosis represents the inhibition of erythropoiesis, leaded by the hyperinflammatory background. Our objective was to analyze RDW performance to predict outcome in coronavirus disease 2019 (COVID-19) acute respiratory distress syndrome (ARDS).

Retrospective observational study including 173 patients with COVID-19-associated ARDS. Data was analyzed at hospital admission, inclusion in the TOCICOV Study (day 0), days 1, 3, 7 and 15 post-inclusion.

Overall, 57% patients received tocilizumab. Overall mortality was 20.8%. RDW was higher in non-survivors compared to survivors at admission (13.53% vs. 14.35, P=0.0016), day 0 (13.60% vs. 14.42, P=0.026), day 3 (13.43% vs. 14.36, P<0.001) and day 7 (13.41% vs. 14.31, P=0.046), presenting better discrimination ability for mortality than other prognostic markers [area under the curve-receiver operating characteristic (AUC-ROC) =0.668 for admission RDW, 0.680 for day 0 RDW, 0.695 for day 3 RDW and 0.666 for day 7 RDW]. RDW values did not vary significantly according to tocilizumab treatment. When adjusted by hemoglobin and tocilizumab treatment, only RDW at admission, day 0, day 3 and C reactive protein (CRP) at day 0 and day 1 were associated with mortality (P<0.05). Only in non-tocilizumab treated patients, IL-6 levels at day 0 were correlated with day 3 RDW (r=0.733, P=0.004) and with day 3 CRP (r=0.727, P=0.022). Both parameters showed significant statistical correlation (r=0.255 for day 1 RDW and CRP in the overall cohort and r=0.358 for day 3 RDW and CRP in patients not treated with tocilizumab, P<0.015).

RDW predicts COVID-19-associated ARDS mortality and reflects the hyperinflammatory background and the effects of cytokines such as IL-6, irrespective of tocilizumab treatment.