Comparison of Critical Care Occupancy and Outcomes of Critically Ill Patients during the 2020 COVID-19 Winter Surge and 2009 H1N1 Influenza Pandemic in Australia

Abstract

Rationale: Both 2009 pandemic influenza A (H1N1) and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are transmitted by respiratory secretions and in severe cases result in a viral pneumonitis, requiring intensive care unit (ICU) admission. However, no studies have compared the clinical characteristics and outcomes of such patients. Objectives: To report and compare the demographic characteristics, treatments, use of critical care resources, and outcomes of patients admitted to an Australian ICU with H1N1 influenza during the winter of 2009, and SARS-CoV-2 during the winter of 2020. Methods: This was a multicenter project, using national data from previous and ongoing epidemiological studies concerning severe acute respiratory infections in Australia. All ICUs admitting patients with H1N1 or coronavirus disease (COVID-19) were included and contributed data. We compared clinical characteristics and outcomes of patients with H1N1 admitted to ICU in the winter of 2009 versus patients with COVID-19 admitted to ICU in the winter of 2020. The primary outcome was in-hospital mortality. Potential years of life lost (PYLL) were calculated according to sex-adjusted life expectancy in Australia. Results: Across the two epochs, 861 patients were admitted to ICUs; 236 (27.4%) with COVID-19 and 625 (72.6%) with H1N1 influenza. The number of ICU admissions and bed-days occupied were higher with 2009 H1N1 influenza. Patients with COVID-19 were older, more often male and overweight, and had lower Acute Physiology and Chronic Health Evaluation II scores at ICU admission. The highest age-specific incidence of ICU admission was among infants (0-1 yr of age) for H1N1, and among the elderly (≥65 yr) for COVID-19. Unadjusted in-hospital mortality was similar (11.5% in COVID-19 vs. 16.1% in H1N1; odds ratio, 0.68 [95% confidence interval (95% CI), 0.42-1.06]; P = 0.10). The PYLL was greater with H1N1 influenza than with COVID-19 at 154.1 (95% CI, 148.7-159.4) versus 13.6 (95% CI, 12.2-15.1) PYLL per million inhabitants. Conclusions: In comparison with 2009 H1N1 influenza, COVID-19 admissions overwinter in Australia resulted in fewer ICU admissions, and lower bed-day occupancy. Crude in-hospital mortality was similar, but because of demographic differences in affected patients, deaths due to 2009 H1N1 influenza led to an 11-fold increase in the number of PYLL in critically ill patients.

Keywords: Australia; COVID-19; H1N1; outcomes; pandemic.

Figure 1.

Cumulative number of intensive care unit (ICU) admissions during the winter period according to the disease of interest. The blue lines and bars are for H1N1 admission from June 1, 2009, to August 31, 2009. The red lines and bars are for COVID-19 admission from June 1, 2020, to August 31, 2020. The daily number of ICU beds refers to the number of beds occupied by these patients in the ICUs that were contributing data to the study and do not refer to the overall number of beds in Australia. COVID-19 = coronavirus disease; H1N1 = influenza A.

Figure 2.

ICU admission according to age categories. The categories of age according to a previous report (2). The age-specific incidence was calculated by dividing the number of admissions in each age group by the estimated number of persons in that age group in the population of Australia (2009: 295,261 for 0–1 year, 1,130,423 for 1–4 years, 5,776,170 for 5–24 years, 7,724,459 for 25–49 years, 3,874,774 for 50–64 years, and 2,890,566 for 65 years or older; 2020: 302,705 for 0–1 year, 1,264,470 for 1–4 years, 6,435,935 for 5–24 years, 8,854,910 for 25–49 years, 4,467,942 for 50–64 years, and 4,038,345 for 65 years or older). The error bars indicate the 95% confidence intervals, which were calculated assuming a Poisson distribution. COVID-19 = coronavirus disease; H1N1 = influenza A; ICU = intensive care unit.

Figure 3.

Exponentially weighted moving average and variable adjusted life display. The exponentially weighted moving average was constructed considering the weight of 0.005, a target based on the initial predicted risk of death for each disease, and the boundaries of the predicted risk considered 3 SDs. The variable adjusted life display is presented as the cumulative excess of survival over time and was calculated by adding the patient’s probability of death every time he/she survived and by subtracting it if the patient died. The variable adjusted life display analysis was corrected by the total number of patients in each group. The predicted risk of death was derived from the Acute Physiology and Chronic Health Evaluation II score, considering admission due to respiratory infection. COVID-19 = coronavirus disease; H1N1 = influenza A; SD = standard deviation.