The needed proof: Volunteers won’t know if they’re getting the real shot or a dummy version. After two doses, scientists will closely track which group experiences more infections as they go about their daily routines, especially in areas where the virus still is spreading unchecked.
I have concerns about the ethicality of the placebo branch in this trial. While the gold standard of clinical trials is a prospective, randomized double-blinded study, in this scenario, you’re asking unprotected people (the placebo branch) to go out there and live normal lives (if I’m understanding “daily routine” correctly) and take a risk on contracting a virus with a significant morbidity and mortality rate.
Wouldn’t they be better off, safety-wise, to use the ambient rate of infection as the placebo branch?
Multiple doctors have told me “if you want to know the direction of the coronavirus, watch what the tech companies are doing.” Today we learn Google will keep its employees home until July 2021
As someone who works in the pharmaceutical and vaccine manufacturing field, mid-2021 isn’t impossible for commercial manufacture of a vaccine but it is unlikely (Note: I don’t have first hand experience in a coronavirus manufacturing process). The fastest to market for a vaccine I’m aware of is ~4 years and 2 years would be lightning fast in my opinion. More likely early to mid-2021 will be the start of a large Phase 3 clinical trial in tandem with vaccinating high risk individuals with a vaccine that showed promise in Phase 1 and 2.
Even if a vaccine is being commercially manufactured and phase 3 clinical results look effective and safe, the likelihood that you personally receive the vaccine at the start is low unless you are high risk.
Here are some rough scenarios that, like many engineering problems, “assume a spherical cow”:
Let’s generously assume 100 million doses can be manufactured per month starting in mid-2021, it would take 50 months to get 5 billion people worldwide vaccinated (which would represent about 70% of the world population to achieve herd immunity).
Let’s assume that (one of) the vaccine(s) is/are made in the US, that company is initially manufacturing 20 million doses per month, and they preferentially distribute to the US. That still means about 10 months to make enough vaccine for herd immunity of the US population (200 million people).
Maybe a more realistic scenario: a US company starts to manufacture about 1 million doses per month, with 100% growth in production per month up to ~50 million (plant capacity), with a mixed distribution that slightly favors US citizens (50%), other developed countries like Europe (30%) and developed countries (20%). In that scenario it would take about 13 months to get herd immunity in the US.
That’s been about the timeline projected as a minimum for all people not named trump. With no fix until a viable vaccine, and timeline of 18 months to (maybe) get to one, let alone actual supply available, that’s been about the minimum since the virus started taking off.
There’s a sort of terrible irony here. The better we are at fighting the disease by public health measures, the harder it will be and the longer it will take to find out whether vaccines work. The current overall Georgia infection rate is somewhere between 0.1% and 1% per day (the lower bound is the confirmed rate). I’m going to go with the lower number because people enrolling in a medical trial ten to be less likely to be stupid yahoos (albeit perhaps more likely to work where there’s a risk of exposure.
So the expected number of infections in each (I’m assuming equal arms) group of 15,000 would be about 10-20 per day. (Note that this is about the same as the number of infections per million in states that have the disease under good control).
Meanwhile, you need a difference of (roughly) 250 people between the two groups to get a number that’s barely statistically significant. What you’d like is a number of infections in the control group that approximates the statewide numbers instead.
With a 100% instantly effective vaccine, you could get to statistical significance in as little as a month. If the immune effect takes a couple of weeks to kick in (not unusual) and the effectiveness is 50%, which is more like what’s expected, then the minimum time is more like 3 months. (And all of that is intensely optimistic, assuming that everyone gets vaccinated at the beginning, instead of the weeks it will likely take to stick 30,000 people, that you don’t have too many dropouts, that behavioral differences don’t confound results, and so forth.)
“President Trump said Wednesday he would take the coronavirus vaccine first — if he was asked to.“
From experience, we all know that this is pure B.S. Trump’s first, last, and only concern is the welfare of Donald Trump. He has but one loyalty, and that is to himself.
Well that’s pretty hard, @emiliano4, since the government is in large part bankrolling the vaccine effort. I hate the idea that Trump will claim whatever vaccine we get as “his,” and he’ll take a victory lap on it, which will be utterly disgusting to see. But that is what’s going to happen if a vaccine proves effective while he’s still in the White House. That’s why, as much as I’d love to see a vaccine tomorrow, I don’t want good news on vaccines till after November 3.
Both the active and placebo groups in the trials have to engage in essentially the same behavior, or else the results will be garbage. So one of the things these studies will have to do is monitor levels of masking and social distancing in participants. (Which will probably lead to safer-on-average behavior because of the reminder effect…)
