Intensive care unit
The proportion of ICU patients in the SPP was 20%, ranging from 18% in 2016 to 22% in 2019. This was the same scale as described in two case series from Wuhan in the province of Hubei [5,6]. However, the case series from provinces outside of Hubei reported only 11%, 10% and 2% (one case) of ICU patients among COVID-19 cases, which is remarkably lower [9,11,12]. This may be attributed to the preliminary character of the outcome, as more than half of the cases were still hospitalised in these case series (Table and Figure 1).
Figure 1  Reported proportions of patients in intensive care and proportions of patients who were still hospitalised, five COVID-19 case series, China, January–February 2020 (n = 688)
COVID-19: coronavirus disease; ICU: intensive care unit.
Grey line: proportion of ICU in pneumonia sentinel patients. Data from five COVID-19 series [5,6,9,11,12]. Among ICU patients, the proportion of males and the median age were strikingly similar between COVID-19 (61% and 66 years) and SPP (61% and 67 years) [6]. However, the proportions of chronic comorbidities were different: 28% of COVID-19 patients treated on the ICU did not have any reported comorbidity, whereas only 16% of German ICU patients were without comorbidities (Table).

Ventilation
Three case series from the province Hubei described a high rate of cases who needed ventilation (20%, 23% and 25%) [5,6,10]. German pneumonia patients had a two- to threefold lower ventilation rate of 9%. The two case series from Shenzhen and the province Zhejiang also reported much lower ventilation rates of, respectively, 11% and 2% [9,11] (Figure 2). The median duration of ventilation was 9 days (interquartile range: 7–19, n = 13) for non-invasive and 17 days (interquartile range: 12–19, n = 4) for invasive ventilation in the COVID-19 case series described by Chen et al. [5]. In contrast, the median ventilation duration (invasive and non-invasive) in SPP was only 2 days (interquartile range: 1–4, n = 303). The high ventilation rates and the long duration of ventilation (data from one case series) can be attributed to a high rate of acute respiratory distress syndrome (ARDS) among COVID-19 patients. Again, the difference between Hubei (17–20% ARDS) and outside of Hubei (2–4% ARDS) is striking. However, ARDS was observed in only 1% of SPP.
Figure 2  Reported proportions of patients on ventilation and proportions of patients who were still hospitalised, five COVID-19 case series, China, January–February 2020 (n = 734)
COVID-19: coronavirus disease.
Grey line: proportion of ventilation in pneumonia sentinel patients. Data from five COVID-19 series [5,6,9-11].

Case fatality
The observed case fatality ratios in the described case series from Hubei (including the city of Wuhan) ranged from 4% to 12%. Most case series outside Hubei did not report fatalities. One case series had 1% case fatality. However, more than two thirds of the cases reported from outside Hubei were still hospitalised at the reporting dates. The case fatality ratio among SPP was 6%, which is within the range of the reported COVID-19 case fatalities (Figure 3).
Figure 3  Reported proportions of deceased and corresponding proportions of patients that were still hospitalised, seven COVID-19 case series, China, January–February 2020 (n = 1,087)
COVID-19: coronavirus disease; SARS-CoV: severe acute respiratory syndrome coronavirus.
Grey line: proportion of deceased in pneumonia sentinel patients. Data from five COVID-19 series [5,6,8-12]. A study by Yang et al. described clinical courses and outcomes of 52 critically ill patients in a hospital in Wuhan [14]. Among 25 critically ill COVID-19 patients younger than 60 years, 12 died within 28 days after admission to the ICU. Of 31 patients without chronic illnesses, 15 died, which is a comparable proportion of almost half.
In the sentinel hospitals, we identified 462 pneumonia patients who were critically ill (i.e. received intensive care and ventilation). Of those patients, 92 were younger than 60 years, of whom 12 (13%) died. In addition, 18 of the critically ill pneumonia patients were without chronic preconditions and four of them died.