A Really Good Explanation of the CDC Mortality Data Debacle

One of my favorite sources for Covid-19 data and analysis is the Johns Hopkins Center for Health Security. Their colleagues run one of the most popular dashboards for Covid-19 tracking, and their newsletter is a good distillation of what the latest coronavirus data suggests about case, hospitalization, and death trends. Today the team shared a really good explanation of the inaccurate claim that the CDC was misrepresenting data on the deaths from Covid-19. I am sharing it here, and you can check out their work here.

From the Johns Hopkins Center for Health Security coronavirus update newsletter:

COVID-19 MORTALITY A number of social media posts, including on Facebook and Twitter, are claiming that the CDC “quietly” corrected its Covid-19 mortality data to remove 94% of the reported deaths. This is factually inaccurate, and we want to provide some quick clarification to correct this misinformation.

The posts refer to provisional Covid-19 mortality data published by the CDC’s National Center for Health Statistics, which states in its August 26 update that 6% of the “deaths involving… Covid-19” have Covid-19 listed as the only cause of death. These data come from death certificates, which may list multiple conditions as causing or contributing to death. These conditions could include those resulting directly from Covid-19 (e.g., respiratory failure) or underlying health conditions that compound the effects of Covid-19 (e.g., heart disease).

Assigning a cause of death is complicated, because death is often not attributable to a single, definitive cause. Doctors and medical examiners must judge, based on their expertise, the degree to which various conditions contributed to a patient’s death. It is well understood that numerous underlying health conditions — including heart disease, diabetes, and respiratory conditions — increase the risk of severe Covid-19 disease and death, so it is likely that these conditions would be included on death certificates along with Covid-19. A patient who tests positive for SARS-CoV-2 but dies of a completely unrelated cause (e.g., trauma) should not be listed as a Covid-19 death, but a patient that dies due to complications of Covid-19, whether preexisting comorbidities or conditions caused by Covid-19 itself, will likely have other conditions listed on the death certificate in addition to Covid-19. Covid-19 deaths are not limited to only those caused (or reported as being caused) solely by Covid-19.