Reply: COVID-19 Is Not Comparable to H1N1 Influenza

From the Authors:

We thank Dr. Tiruvoipati and colleagues for their interest in our paper (1). Indeed, as they allude to, the public health response to coronavirus disease (COVID-19) over winter in 2020 in Australia was dramatically different to that for the influenza A (H1N1) pandemic in 2009. As has occurred in many regions, strict public health measures (such as social distancing, border closures, curfew, and mask wearing) were all enacted in 2020 in an attempt to slow the community spread of COVID-19 in Australia and reduce the potential burden on the healthcare system. None of this occurred with the 2009 H1N1 pandemic, a point we reinforced frequently in our manuscript. In this respect, we agree that the differential use of intensive care unit (ICU) resources over winter in 2020 for COVID-19 in Australia (as compared with that for H1N1 influenza over the same seasonal period in 2009) is in part related to the public health measures implemented. However, the inference that Australian hospitals were under more “strain” over the winter of 2009 remains entirely speculative and unproven, particularly in reference to the availability of general ward beds. We would also contend that the older, more comorbid demographic infected with COVID-19 (1) (in the initial stages of the pandemic in Australia, and in comparison with 2009 H1N1 influenza) also impacted ICU usage, as invasive organ support may not have been deemed appropriate for some patients.

Similarly, as detailed in our paper, Australia promoted a highly accessible community testing program for COVID-19, even in mild or minimally symptomatic individuals. Although we agree that this was very useful in planning the broader healthcare and community response, we are not aware of any evidence that this translated into “earlier intervention.” Indeed, during the initial stages of the COVID-19 pandemic in Australia, there were no specific community-based therapies being used to limit the progression to severe disease. Moreover, it is not clear that a positive test for COVID-19 resulted in different healthcare-seeking behavior or outcomes. Similarly, we are not aware of any national policy mandating “early” ICU admission for COVID-19 patients in Australia, with guidelines principally recommending an approach based on clinical need, patient and family preferences, and the likelihood of a meaningful response to treatment (2). In addition, we would also note that illness severity (as determined by acute physiologic assessment and chronic health evaluation II scores) was included in our adjusted analysis, implying that the lower risk of death observed in the 2020 COVID-19 cohort was independent of this parameter.

The implication by Dr Tiruvoipati and colleagues is that our comparison of 2009 H1N1 with 2020 COVID-19, over winter in Australia, was inherently flawed or perhaps even nonsensical; we are uncertain what is meant by “unnatural.” Although we certainly acknowledge the vastly different community responses to each pandemic, and the significant passage of time, we did attempt to limit confounding by using data from the same seasonal period and in the same healthcare setting. Moreover, we would strongly argue that such comparisons are of merit, particularly as a means of placing the COVID-19 healthcare crisis, and the public health response in Australia, into context. Finally, we make no statements as to the generalizability of our data or findings, but rather simply provide these for consideration by the wider scientific community.