A one-week vacation

Tweeting about COVID-19's true risks to kids is risky

Last week, @jack gave me seven days “off” from posting on Twitter. Upside: It gave me a good excuse to start using Substack.

What did I tweet about to earn that one-week vacation?

The actual risks that COVID-19 poses to almost all kids.

The Tweet

The “offending” post was in a thread about the illogical exposure-quarantine protocols in schools.

As parents know too well, students in many states who were able attend in-person school last year were required to stay home for 10 days if a close contact had tested positive for SARS-CoV-2, regardless of whether the exposer or the exposed was actually-sick.

I have kids, and pre-panicdemic, their schools never told me to keep them home after exposure to or close contact with classmate who had a communicable disease. When there was a confirmed case of strep throat, flu, scabies, Fifths Disease, norovirus, etc. in a grade level, the school nurse would notify parents via email and list the indications to monitor. There was no quarantine order or directive to get my child tested for the suspected pathogen.

But it’s probably my claim about COVID-19 presenting lower risk of severe outcomes than flu for 99.99%+ of children that disturbed the Twitter bots & trolls.

Was I exaggerating?

Let’s take a look at data on three “outcomes” that might be considered severe.

Severe Outcome 1: Death

Thankfully, COVID-19 can’t be honestly characterized as a “deadly” disease for children.

From January 2020 through now, there have been 335 pediatric deaths involving COVID-19 in the United States. That’s .0004% of the country’s 74 million children age 0-17, and 0.6% of all 49,725 pediatric deaths that have occurred in the same timeframe. Only 65 (19%) of deaths involved pneumonia - which in countries like Singapore is standard for counting all COVID deaths.

How many U.S. children have had COVID? CDC case demographic data shows nearly 3.5 million positive COVID tests among age 0-17, which gives us a .0097% case fatality ratio (CFR).  “Cases” — which are positive tests, versus actually-sick kids — reflect detected infections, versus all infections.

The agency’s disease burden estimates put the number of infections for children at nearly 27 million between February 2020 - March 2021. By now, we’re at 30 million plus. Either way, we’re still looking at 99.998% survival. (Indeed, in my Chicagoland county of ~1 million residents, there have been zero COVID-related deaths under age 20. As in, n=0.)

Data from the U.S. and England, respectively, show 35% and 60% of pediatric deaths with COVID-19 were not attributable to the virus. In other words, the death totals include children with other severe conditions or adverse events who tested positive for a virus that didn’t play a part in their deaths. Per a November 2020 study from FAIR Health, among COVID patients with no co-morbidities, age 0-18 had 0.00% mortality (no deaths), compared to 0.01 percent mortality in the same age group among all patients. 

So, child deaths resulting from SARS-CoV-2 infection are rare, but are they more rare than seasonal flu deaths have been?

Yep. The CDC’s pediatric flu death estimates over the past ten years show rates per 100,000 from 0.25 to 1.16 or higher, depending on age. Compare to 0.20 and lower for COVID-19 (not accounting for a potential 35% over-ascertainment, which Phil Kerpen illustrates here).

Severe Outcome 2: Hospitalization

What about kids’ risk for the “severe outcome” of hospitalization due to COVID?

The CDC’s mid-range estimate for pediatric COVID hospitalizations since February 2020 is ~200,000. HHS data show ~47,000 COVID-positive pediatric admissions since January 2020, with some states reporting only from July 2020 onward. Assuming 27 million infections, that’s a .07% rate. But there’s a catch.

Neither HHS nor CDC data distinguishes between being hospitalized for COVID per se, and being in the hospital for another reason but testing positive incidentally upon admission or during one’s stay. As Marty Makary said in a recent WSJ op-ed, “[The CDC’s] Covid adolescent hospitalization report, like its death count, doesn’t distinguish on the website whether a child is hospitalized for Covid or with Covid…An asymptomatic child who tests positive after being injured in a bicycle accident would be counted as a ‘Covid hospitalization.’”

He’s right. One recent study at a children’s hospital in northern California found as many as 45% of pediatric hospitalizations were found unlikely to have been caused by SARS-CoV-2.  Similarly, 40% of patients under age 22 classified as “COVID” at another California hospital were incidental positives. Applying these percentages to the HHS data gives us .0086% of infections hospitalized because of COVID.

How does this compare to flu?

We don’t test for flu (or for any virus) like we do for SARS-CoV-2, including automatically upon hospital admission during flu season. But the CDC does maintain estimates. For example, the estimated rate for the 2017-2018 season was .128% for the 0-4 group and .038% for 5-17. Without question, flu carries a higher risk for hospitalization than COVID for younger kids. The risk for older kids and teens appears comparably lower than flu, but also depends on a child's or teen’s overall health.

Severe Outcome 3: “Long COVID”

What about the risk of so-called “long COVID”?

In general, as researchers from the U.K. put it, COVID is short-duration and of low symptom burden in school-aged children. Viral infections of many kinds can have lingering symptoms. (See thread by @heckofaliberal for examples.) Some could be described as “severe”, others merely annoying. But SARS-CoV-2 does not yet appear to be unique, higher prevalence, or greater severity for children in this regard.

Findings from several studies follow:

  • A study of ~1,700 symptomatic children in the U.K. who tested positive for SARS-CoV-2 showed very low rates of prolonged symptoms at 4 weeks post-illness:(4.4%) and <2% beyond 2 months. Severity symptoms was worse in 15 kids in a control group who were not diagnosed with COVID.

  • Another study of 171 Australian children found 95% had very mild or asymptomatic illness, and low rates of mild symptoms at 3-6 month follow-up. By the end of the study, all children were back to baseline health.

  • Swiss researchers found no difference between seropositive and seronegative children with symptoms beyond 4 weeks (~10%), and similarly very low rates of symptoms at 6 months (2% and 4%). The most common long-term symptoms were fatigue, headache, & loss of smell.

  • In Germany, 35% of 1,500+ teens in one study (~1,300 of whom were seronegative for SARS-CoV-2 and ~200 seropositive) reported symptoms associated with long COVID, such as difficulty concentrating, headache, abdominal pain, fatigue, insomnia, and sadness. However, there was no statistical difference between the two groups of students, prompting the researchers to raise the question of whether “Long-COVID19” is really “Long-Pandemic Syndrome”.

Unless new data suggest otherwise, both the very low rate of persistent symptoms - and the low “grade” and duration of those symptoms - further support my claim that COVID illness doesn’t lead to severe outcomes for all but the fewest children.

The Bottom Line

People might quibble with one or two hundredths or the plus sign in the “99.99%+” from my tweet. Fine. The bottom line is that SARS-CoV-2 presents almost every child a near-zero risk for severe outcomes, and has not been shown to be higher risk than flu.

Even mainstream media outlets like New York Magazine are now saying out loud what has been abundantly clear since Spring 2020: The kids were safe from COVID the whole time. Indeed, there’s a case to be made that infection is safer & more desirable for most children than getting a COVID vaccine.

Therefore, I stand by the data—and by the post @jack compelled me to delete.