Ziyad Al-Aly: Is covid-19 associated with higher risk of hospitalization than seasonal influenza?

Since the early days of the covid-19 outbreak in China, public health officials, scientists, medical professionals, media commentators, and the public at large started drawing comparisons between covid-19 and seasonal influenza. People wanted to know how covid-19 compares with other viral respiratory illnesses that we know very well—for example, seasonal influenza. Is covid-19 more deadly than seasonal influenza? Is covid-19 associated with higher risk of hospitalization than seasonal influenza? Is covid-19 associated with higher risk of mechanical ventilation than seasonal influenza? But there were no answers to these questions. There were speculative media reports, and comparisons made based on disparate data sources and methodologies, but no like for like comparison.

Over the past eleven months of this global pandemic, we have become increasingly aware that covid-19 may be much more than a respiratory virus. We have shown that patients with covid-19 experience high rates of acute kidney injury. Others have shown that covid-19 also leads to neurological, cardiac, and other non-respiratory conditions. But again, these were anecdotal reports, case series, and papers examining one organ system. 

Prior to the pandemic, and before embarking on a project, we thought about a particular research question or a problem for months and even years; we would host multiple face-to-face discussions to develop the aims of a project and hone our analytic approach. 

The pandemic changed everything. We knew the public and medical professionals needed answers fast, and we knew we had to adapt and spring to action. We wanted to do our part in the fight against this global pandemic. We wanted to generate new knowledge to help inform the global discussion around this emerging infectious disease—hoping that the findings would help guide and optimize the response to the pandemic. We assembled our team for several brainstorming sessions over zoom. We specified the questions to answer in this research. Our goal was to design a study to provide a detailed comprehensive comparative evaluation of the clinical manifestations and risks of health resource utilization and death in people hospitalized with of covid-19 vs. seasonal influenza. To do this, we leveraged the breadth and depth of the clinical databases of the US Department of Veterans Affairs—which operates the largest nationally integrated healthcare system in the US—and used advanced statistical methods to even the playing field and provide a comparison.

Our findings suggest that covid-19 affects almost every organ in the body. We observed higher risk of incident acute kidney injury, need for dialysis, incident insulin use, severe septic shock, vasopressor use, pulmonary embolism, deep vein thrombosis, stroke, acute myocarditis, arrythmias and sudden cardiac death, elevated troponin, elevated liver function tests, and rhabdomyolysis. Compared to seasonal influenza, covid-19 was associated with nearly five times the risk of death, and higher risks of mechanical ventilator use and ICU admission, and longer hospital stay.

To better communicate the findings about the increased risk of death, we estimated the absolute risk based on the difference in survival rates in addition to reporting hazard ratios. The absolute risk not only provides information on ratio difference between two groups, but also considers the baseline risk, and provides a more meaningful description of risk to the wider public. The analyses allow us to estimate the number of excess deaths due to covid-19 on an absolute scale; for example, given 100 people hospitalized with the seasonal influenza, and 100 people hospitalized with covid-19, and everything else being equal, what is the excess risk of death that is attributable to covid-19? Our results suggest that compared to seasonal influenza, covid-19 was associated with 16.85 (14.85, 18.99) excess deaths per 100 patients. Excess rates of death per 100 patients in covid-19 vs. seasonal influenza were most pronounced in those >75 years of age with chronic kidney disease or dementia, and patients from black populations, and those with obesity, diabetes, or kidney disease.

Finally, this work may also represent a blueprint into how a comparative framework (covid-19 vs. seasonal influenza) may enhance our understanding of the clinical manifestations and risks of this new emerging infectious disease. We hope that deeper understanding of this disease will inform and guide effort to optimize management of this global pandemic. 

Ziyad Al-Aly, director, Clinical Epidemiology Center, Research and Development Service, VA Saint Louis Health Care System, Saint Louis, MO, USA.

Competing interests: Please see research paper