ABSTRACT
Background: COVID-19 is declared a global pandemic by the World Health Organization in March 2020 and since then has been significantly challenging health care services. Validated clinical rules are urgently needed to assist clinicians in their decision-making process in identifying admitted COVID-19 patients at risk of in-hospital mortality and mechanical ventilation and appropriately allocating limited health care resources. Methods: A multi-centre observational study was performed on admitted COVID-19 patients between 1 March and 31 May 2020 in Iran. Patient characteristics, clinical manifestations and association with in-hospital mortality and orotracheal intubation were reported. Logistic regression analyses were used to derive predictive clinical tools and the model scoring derived from the coefficients. General bootsrap procedure was used for internal validation. External validation was performed on a randomly selected admitted COVID-19 patients. Findings: 4566 admitted COVID positive patients (2074 females and 2492 males) were enrolled with a median age of 59 years. The most common presentations were dyspnoea (53%), cough (53%), and fever (52%). 64% of patients had O2 saturation below 93% at the time of presentation. Orotracheal intubation was required in 298 of patients (7%) and 628 of patients (14%) died during admission. The features associated with in-hospital mortality were: age more than 65 years, arrival by ambulance, dyspnoea, orotracheal intubation, O2 saturation below 93%, and a history of malignancy. The features associated with orotracheal intubation were: age more than 65 years, arrival by ambulance, dyspnoea, loss of consciousness, prior contact with a confirmed COVID case, and O2 saturation below 93%. External validations confirmed good discrimination with an area under the curve of 0·826 and 0·734 for in-hospital mortality and orotracheal intubation, respectively. Interpretation: Clinical decision tools to predict mortality and requirement for mechanical ventilation will assist clinicians in appropriately making decision of their patients. Funding: No fundingDeclaration of Interests: The authors declare no competing interests.Ethics Approval Statement: Ethical considerations for the study were reviewed and approval granted by the ethics committee of Isfahan University of Medical Sciences with approval number of IR.MUI.MED.REC 1399.423.