Sources of data and definition of variables

We analyzed an integrated database that contained clinical, laboratory and treatment data from all hospitalized patients with a diagnosis of COVID-19 at three Italian referral tertiary centers, two in Lombardy (the “eye of the SARS-COV-2 storm” in Italy) (Bergamo and Pavia) and one in Lazio (Rome): 1) Hospital Papa Giovanni XXIII, Bergamo, Lombardy; 2) Fondazione IRCCS Policlinico San Matteo, Pavia, Lombardy; 3) Fondazione Policlinico Univerisitario A. Gemelli IRCCS, Rome, Lazio. All consecutive patients admitted between February 22^nd and April 7^th, 2020 were enrolled and were followed up until April 30^th, 2020.

Information on the history and physical examination of hospitalized patients with COVID-19 were abstracted from chart reviews by medical officers at each hospital. Variables collected through standardized recording forms included age, sex, comorbidities, smoking status, time of symptoms onset and time of hospital admission. Additional variables were the presence of fever (defined as axillary temperature of at least 37.5°C), dyspnoea, cough and diarrhea. Investigations consisted of chest radiography or computed tomography and hematologic and biochemical blood tests, including complete blood count, coagulation profile, glutamic pyruvic transaminase (GPT), lactate dehydrogenase (LDH), C-reactive protein (CRP), and creatine kinase. Arterial-blood gas analysis (ABG) was performed when clinical signs of oxygen impairment were detected (e.g. tachypnoea and hypoxia). P/F ratio was calculated as the ratio between PaO2 and FiO2.

Laboratory confirmation of the SARS COV-2 infection was defined as positive real-time reverse transcriptase polymerase chain reaction (RT-PCR) from nasal and pharyngeal swab; samples were prospectively collected and analyzed at the Molecular Virology Units of each center according to the WHO guidelines and Corman et al. protocols [4, 5]. More details are provided in Supplementary Materials.

Treatments included the use of antiviral therapy (Lopinavir/ritonavir or darunavir/ritonavir), hydroxychloroquine, enoxaparin, and immunomodulatory/immunosuppressive therapy such as corticosteroids, Tocilizumab and Sarilumab. Lopinavir/ritonavir 400/100 mg was administered orally twice daily for 14 days, while Darunavir/ritonavir 800/100 mg was administered once daily for 14 days. Hydroxychloroquine 600 mg was administered twice on day 1, followed by a dose of 400 mg daily for 7 days. Enoxaparin 1 mg/kg was given once or twice daily. Corticosteroid treatment consisted in dexamethasone 20 mg daily for 5 days or methylprednisolone 1 mg/kg intravenously daily for 5 days. Tocilizumab 8 mg/kg was administered intravenously in 1 or 2 doses. A second dose was administered after 8–12 hours from the first dose in patients with inadequate response. Sarilumab 400 mg was administered intravenously once. Oxygen support consisted of low (cannula and simple masks) and high flow (Venturi and reservoir masks, Nasal High Flow), helmet continuous positive airway pressure (CPAP) or non-invasive ventilation (NIV). The choice for the oxygen support was determined by rapid deterioration of P/F ratio and upgraded if a further worsening after two hours of treatment was detected.

The Institutional Review Boards of the three centres (ASST Papa Giovanni XXIII–Bergamo, Italy; IRCCS Fondazione Policlinico San Matteo, Pavia, Italy; Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy) which comply with the Declaration of Helsinki and its revisions, approved this study. All accessed data were fully anonymized.