Outcomes of intensive care unit patients with COVID‐19: a nationwide analysis in Russia

The COVID-19 pandemic continues to evolve rapidly in many countries and poses a challenge for critical care services. Nevertheless, the outcomes of intensive care unit (ICU) patients with COVID-19 remain ill-defined. In a recent meta-analysis of 24 observational studies that included 10,150 patients, Armstrong et al. reported an ICU mortality rate of 41.6% [1]. The authors suggested that mortality rates have reduced from above 50% to approximately 40% over time. However, only seven studies reported outcome data for all patients, whereas the proportion of patients discharged from ICU at the time of publication varied from 24.5% to 97.2% in the remaining studies. Moreover, six out of the seven studies with known outcomes in all cases were small and included only 101 patients in total. In a nationwide study, we evaluated the mortality rate in 1522 consecutive ICU patients with SARS-CoV-2 pneumonia who had completed their hospital stay (death or recovery) up to 7 July 2020. According to the government decision, medical records were submitted via the internet by COVID-19 hospitals located in 70 regions across Russia to the Federal Center at the Sechenov University, Moscow, that provided advice on critical care of patients. Diagnosis of SARS-CoV-2 pneumonia was established both by polymerase chain reaction (PCR) and CT scanning. In patients with a negative PCR, SARS-CoV-2 pneumonia was defined as severe acute respiratory infection with typical CT scan findings [2] and no other obvious aetiology. Clinical and baseline characteristics of patients with severe COVID-19 admitted to ICU are presented in Table 1. Most patients were > 40 y and had various chronic illnesses, e.g. cardiovascular disease, type-2 diabetes and obesity. Among 1522 patients in this study, 995 (65.4%) died, and 527 (36.4%) recovered. The 14 and 28 day mortality rates were 44.0% and 63.6%, respectively. The most common causes of death were acute respiratory distress syndrome (93.2%), cardiovascular complications (3.7%) and pulmonary embolism (1.0%). The mortality rate was low in patients requiring oxygen therapy (only 10.1%) and significantly higher in patients who required non-invasive (36.8%) or invasive (76.5%) ventilation. The highest mortality rate (86.6%) was reported in patients with septic shock. Median (IQR [range]) duration of mechanical ventilation was 6 (3–12 [1–62]) days in deceased patients and 13 (7–21 [1–40]) days in recovered patients. Mortality rates in Moscow and Moscow province were higher (74.5% and 78.6%, respectively) than in the other regions of Russia (50.2%). However, patients from the regional hospitals had less severe disease and more frequently required only oxygen therapy (24.7% vs. 4.9% in Moscow and 8.2% in Moscow province). Mortality rates were similar in PCR-confirmed and unconfirmed cases (63.5% and 68.9%, respectively). In summary, the average mortality rate was 65.4% in Russian ICU patients with SARS-CoV-2-induced acute respiratory distress syndrome, although it varied widely depending on the level of respiratory support and indications for ICU admission. These factors should be taken into account in future studies to avoid a skewed picture of mortality in ICU patients with COVID-19.