https://www.selleckchem.com/products/sgi-110.html in ; Vol. 24, No. 69. See the NIHR Journals Library website for further project information. This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 69. See the NIHR Journals Library website for further project information. To identify predictors of clinical improvement and intubation/death in tocilizumab-treated severe COVID19, focusing on IL6 and CRP longitudinal monitoring. 173 consecutive patients with severe COVID-19 pneumonia receiving tocilizumab in Reggio Emilia province Hospitals between 11 March and 3 June 2020 were enrolled in a prospective cohort study. Clinical improvement was defined as status improvement on a six-category ordinal scale or discharge from the hospital, whichever came first. A composite outcome of intubation/death was also evaluated. CRP and IL-6 levels were determined before TCZ administration (T0) and after 3 (T3), and 7 (T7) days. At multivariate analysis T0 and T3 CRP levels were negatively associated with clinical improvement (OR 0.13, CI 0.03-0.55 and OR 0.11, CI 0.0-0.46) (p=0.006 and p=0.003) and positively associated with intubation/death (OR 17.66, CI 2.47-126.14 and OR 5.34, CI 1.49-19.12) (p=0.01 and p=0.004). No significant associations with IL-6 values were observed. General linear model analyses for repeated measures showed significantly different trends for CRP from day 3 to day 7 between patients who improved and those who did not, and between patients who were intubated or died and those who were not (p<0.0001 for both). ROC analysis identified a baseline CRP level of 15.8 mg/dl as the best cut-off to predict intubation/death (AUC = 0.711, sensitivity = 0.67, specificity = 0.71). CRP serial measurements in the first week of TCZ therapy are useful in identifying patients developing poor outcomes. CRP serial measurements in the first week