madeleineshort

Why is the rate of death going down?

Question: In almost every one of the charts you shared in the Q&A of 1/13 (except TN), every indicator is increasing but one – the rate of death.  Why is the rate of death going down in each chart while everything else is going up?

Answer: The case fatality rate (CFR) that I present in the trend charts is a very simple indicator– the ratio of confirmed deaths to confirmed cases (deaths/cases).  It is impacted by factors that influence the denominator (cases) and numerator (deaths), including:

  • Number of cases captured: As testing becomes more widespread, we expect more asymptomatic and mild cases to be captured through testing, which will lower the CFR.
  • Type of person infected: Case fatality varies dramatically by age and race (see Q&A of 11/18). As more young, otherwise healthy people become infected, the overall CFR will be weighted more heavily by youth and CFR will decline.  The same can be said of race. Indigenous and Black Americans are more likely to die of COVID than White Americans This is a product of racism (see Q&A of 6/7). As the demographics shift and more White Americans become infected, CFR will decline.   
  • Geography of cases: CFR is impacted by the strength of a locality’s health system.  Places with stronger health systems could see lower CFRs. If more cases are occurring in places with stronger health systems or if health systems overall have improved over time, we’d expect to see lower CFR.  
  • Treatment: As health care providers learn more about how to treat patients suffering from COVID-19, their treatments improve, causing health outcomes to improve, leading to fewer deaths and a decline in CFR.   
  • Partial immunity: This could be more important over the coming months, but is still hypothetical.  As more people become vaccinated with the first dose, we can expect an increase in partial immunity, which could reduce the severity of disease.  If vaccination doesn’t impact infection, but does impact disease progression, we would expect to see a decline in case fatality. 

Case fatality rate will also depend on how you decide to measure it.  Age-adjustment is important (see Q&A of 11/24).  So is the recognition that deaths generally lag cases by 2+ weeks (see Q&A of 7/12 and Q&A of 11/22).  Without lag, as of 1/18/21, the crude CFR for the US is 1.6%.  Lagged by 2-weeks, it’s 1.9%. Lagged by 3-weeks, it’s 2.0% (Figure 1).  Even with the lag, case fatality trends are declining over time in the US (see below) and this trend is also revealed in age-adjusted CFR (see Q&A of 11/24).

As I hope you can see from this overview, CFR is not a constant.  Just as it has decreased, it’s also possible that it will increase.  That said, if we can keep the most vulnerable among us from getting infected, if we can keep our health systems buoyed, if we can continue to improve our treatments, and if we continue to expand vaccination, we should expect case fatality rates to continue this downward trend.  Finally, even low case fatality rates result in huge numbers of deaths if many of us are infected.  And that’s what we’re tragically confronting with the US having experienced 399,015 deaths as of this morning.         

Figure 1. Trends in CFR in the USA