Deaths from suicide during lockdown are not higher than COVID-19 deaths.

It is a good thing to have a spirited debate about re-opening after lockdown…but we need to use accurate data.

COVID-19 and suicide

A Google search will show a few stories suggesting that deaths from suicide during the lockdown are higher than the number of deaths from the coronavirus. That may be possible in a single medical clinic or county on a given week or even month. But it is not plausible across larger regions or longer times, and it utterly untrue for the US as a whole.

The annual suicide rate in the US is the highest in the world among wealthy nations at 14 per 100,000 people, almost double the UK number (it’s about 11 per 100,000 in Canada.) In raw numbers, there were 48,344 recorded deaths by suicide in 2018 in the US, the most recent year for which we have data.

At first glance, that seems smaller but possibly comparable to the 98,683 official deaths due to COVID-19 as of May 23. However, those suicide fatalities occur over an entire year, and represent 4,019 deaths per month. On March 23 the US had 689 deaths from the virus, so the monthly death rate since that date is 48,997.

The number of Americans dying per month from the virus since March has been higher than the 2018 suicide deaths per year.

The death rate for suicides in the US has been rising since 1999, when it was about 10/100,000, see chart above. In raw numbers, that means that if we still had the 1999 rate, there would be about 2,900 suicide deaths per month, but tragically we now have over 4,000. The growth in monthly suicide fatalities is therefore about 1,100 over the last two decades. (I’ll explain why this number is important later in this article.)

In order for there to be more suicide deaths than coronavirus deaths so far, the number of suicides would need to have not just doubled compared to the previous level, but grown by more than a dozen times. This has not happened.

It takes time to correctly attribute the cause of death, which is why the suicide numbers for 2019 are not yet public, but the US does report weekly deaths from all causes in (more or less[1]) real time, see chart above. The number of deaths fall into a normal range around the average of the last few years, and anything above an extra ~2,000 deaths per week is called “excess deaths” and is flagged by a little red + symbol on the chart.

The cumulative excess deaths seen so far, adjusting for reporting lag, are in the rough ballpark as official C-19 deaths. If there was a spike in suicides nationally that was bigger than the deaths from the virus, the bars would need to be over twice as high. And they aren’t. Which means that deaths by suicide are not higher than COVID-19 deaths, not even close.

Why does all this matter?

1.  Mental health matters. Suicide matters.

2.  A legitimate argument for returning to normal as rapidly as possible is that it will help reduce mental health problems and the number of suicides over time.

3.  But saying that suicide deaths so far are higher than virus deaths is NOT a legitimate argument. It’s not true. And it causes harm.

4.  The harm it causes is that it sets unreasonable expectations around what a surge in suicides might look like over the next few years. Which will lead to mistakes in public health policy.

5.  The most famous surge in suicides due to economic conditions in the US occurred during the Great Depression: the age-adjusted rate went from 18 per 100,000 in 1928 to 22 per 100,000 in 1932, see chart below. This increase of 4 per 100,000 is the same as the growth seen between 1999 and 2018, or (based on current US population) about 1,100 additional monthly deaths from suicide, or about 13,000 per year.

6.  Assuming we see a roughly similar increase due to the current economic fallout from the pandemic, we should be planning on those kinds of numbers for a year or two, and then declining to previous levels thereafter, as happened in the Great Depression. Suicides went up, but were actually below 1929 levels by 1934.

7.  Governments and mental health professionals should plan accordingly.

[1] “Number of deaths reported on this page are the total number of deaths received and coded as of the date of analysis and do not represent all deaths that occurred in that period. Data are incomplete because of the lag in time between when the death occurred and when the death certificate is completed, submitted to NCHS and processed for reporting purposes. This delay can range from 1 week to 8 weeks or more, depending on the jurisdiction and cause of death.”