The outbreak of coronavirus disease (COVID-19) has brought great challenges to the world . The objectives of this study were to describe the baseline characteristics and changes of biomarkers of these COVID-19 patients and identify predictive value of the above markers for patient death . Using patient death as the observational endpoints, clinical data of inpatients in a special ward for COVID-19 in Wuhan, China were retrospectively collected . Univariate and multivariate Cox regression analyses were used to evaluate prognostic value of baseline characteristics and laboratory data changes . This study included clinical data of 75 patients . Age, c-reactive protein (CRP) and interleukin-6 levels were independent predictors of patient death . Survivors were characterized as having declining neutrophil counts, D-dimer, N-terminal pronatriuretic peptide, troponin I (TnI) and c-reactive protein levels, while counts of lymphocyte gradually came back . Non-survivors were characterized with increasing white blood cell counts (WBC) and neutrophil counts . Changes of WBC, TnI and interleukin-6 were also independently associated with patient death . Older age, baseline CRP and IL-6 levels may be used as meaningful predictors to identify patients with poor prognosis . Changes of biomarkers should be closely monitored in the management of patients with COVID-19, while constantly increasing levels of WBC, TnI and interleukin-6 in the disease course also predict patient death.