A look back at my 2022 predictions
One of the most effective ways to be less full of crap is to evaluate how closely the things that you say track with reality.
And so while the annual review/prediction stuff is a nice content gimmick for those of us in the newsletter business, I think it’s a particularly useful exercise to track how well some of the assumptions that underscore my analysis held up to what actually happened.
Each prediction was made with the idea of being verifiable and describing my level of confidence in its accuracy, so we can see whether I was actually right about the things I was most confident in.
Without further delay, let’s look at my 2022 predictions with some annotations below each one.
Predictions on COVID
COVID will not be eradicated and will still be circulating by the end of 2022 (95% confident)
This was a bit of a layup, but COVID is still around, to no one’s surprise.
The Omicron wave will peak in New York City before the end of January (70% confident)
This one held up pretty well, with COVID cases starting to fall in New York City by the end of January.
Omicron will have varying peaks nationwide throughout the next couple of months, but by the end of April will be on its way out across the United States (60% confident)
It looks like this prediction was pretty good.
There will be at least one new COVID strain identified as a variant of concern by the end of the year (50% confident)
This one doesn’t look like it was right. The subvariant game is tricky, but we’re still working with Omicron subvariants in the US.
The media will still be reporting case numbers by the end of the year (90% confident)
I don’t think I was right about this one. You can still easily find case numbers on major media websites - as well as warnings about the cold and dark winter ahead - but reporting on total numbers of infections has certainly gotten a lot less attention than I expected it to.
Cases will continue to decouple from hospitalizations and deaths (60% confident)
The changing nature of what a COVID infection means has fundamentally altered everything about assessing our risk as the pandemic continues, at least in my eyes.
Not only was this prediction right, but my clinical experience taking care of patients in the hospital has made me think that the official numbers today drastically overstate how scary a disease this is. The number who are hospitalized “with COVID” greatly outnumbers the ones who are hospitalized “from COVID.”
When you hear scary stuff about the winter ahead, keep in mind that the nature of the disease is different today than it was earlier in the pandemic due to societal immunity and widespread vaccination. This is great news, and should be celebrated!
Less than 200 Americans will be dying each day from COVID by the end of 2022 (50% confident)
Not quite, but close.
A 4th COVID vaccine (an additional booster) will be recommended for Americans at high risk of serious illness (60% confident)
This one is half right. The additional booster is definitely recommended, but not just for people at high risk of serious illness. The official recommendations don’t take risk or prior infection into account, so you get a blanket recommendation that applies to everyone, regardless of age or prior COVID infection. Which makes absolutely no sense.
The mRNA boosters will have the same formulation as they current have, rather than a new genetic sequence engineered to specifically combat a new variant (50% confident)
Wildly off base here - new bivalent boosters are available that cover new Omicron subvariants. The evidence we have on their effectiveness really comes down to the fact that they raise antibody levels temporarily, but they’re available for everyone and recommended by the CDC.
Some version of mask mandates will still be in place by the end of the year for air travel and in some (probably blue) states (70% confident)
I was wrong about the air travel mandate, but I certainly still have to wear a mask at work.
Kids will still be masked in school (50% confident)
This one was wrong too.
There will still be absolutely zero good evidence that masks reduce COVID transmission or risk of serious disease in children (80% confident)
I’ve written about masks quite a bit before - and these are the newsletters that have led to some of the most emotional and intense emails that I’ve received.
I believe that this prediction was proven correct, and I stand by the assertion that the strength of evidence on mask mandates for kids is essentially zero. You can quibble about whether reducing COVID transmission in kids is a worthy goal with almost 90% of children seropositive at this point, but the evidence that masks for kids are anything besides COVID theater is essentially non-existent.
Paxlovid will remain in short supply throughout the year (70% confident)
Perhaps the prediction I was most wrong about. Paxlovid has been everywhere. I’ve prescribed it a few dozen times and declined to prescribe it even more.
I wasn’t just wrong about Paxlovid when it comes to this prediction - I was wrong about Paxlovid when it comes to effectiveness and utility.
I interpreted the evidence too generously in favor of using a drug that I didn’t fully understand. I also didn’t put enough weight on the fact that the EPIC HR trial testing Paxlovid was carried out in a majority of unvaccinated people when thinking about the risks and benefits.
Nothing will be done to reform the function of the FDA and CDC with regards to speeding up approvals in an emergency or improving our shoddy data collection capabilities (95% confident)
This one was right. My general impression is that the FDA and CDC under Joe Biden have been bizarro versions of the FDA and CDC under Trump. Which is to say that they’re carrying out political messages rather than scientific ones.
The reason we don’t have high quality data about masks, ventilation, vaccine boosters, social distancing, or anything else is because the people who make policy and who write checks haven’t demanded it. The failure to generate evidence remains the biggest ongoing pandemic sin.
Would strongly recommend this Twitter thread as a look about just how off base the FDA’s recent EUA authorizing bivalent boosters for children is:
I believe the purpose of regulatory agencies is to hold corporations accountable. Yes, you can sell your product for money on the u.s market, but we need proof that Americans are better off as a result. Regulators keep corporations honest. This decision is a failure ....🧵Today, we amended the emergency use authorizations (EUAs) of the updated (bivalent) Moderna and Pfizer-BioNTech COVID-19 vaccines to include use in children down to 6 months of age. https://t.co/dHHwGFJ7fW https://t.co/z4zZNwZZZ0
U.S. FDA @US_FDA
Non-COVID medical predictions
The obesity rate in America will continue to increase (90% confident)
Depressingly correct.
The fastest growing prescription drug class will be SGLT2 inhibitors (50% confident)
This prediction is pretty hard to find data for, but I think that I was wrong here
Although as far as impact on the cultural zeitgeist, when it comes to this prediction I didn’t even pick the right class of diabetic agents. That belongs to the GLP-1 agonists, which are pretty amazing weight loss drugs that are safe and effective.
Don’t just take my word for it - look at New York Magazine. The hashtag #ozempic has over 312 million TikTok views.
At least one drug with a multibillion dollar market will be approved by the FDA without data demonstrating that it is safe and effective (80% confident)
This prediction was right pretty early on. I’m sure there are others, but the first one I saw was bempedoic acid, a cholesterol lowering drug that was approved in February.
Unfortunately, the only data for it still comes in the form of press release rather than paper.
The Roche Alzheimer’s drug gantenerumab will prove ineffective in a clinical trial (50% confident)
This prediction was right, which fits with the fact that the amyloid hypothesis for Alzheimer’s remains in question.
There will be at least one additional high quality clinical trial data published demonstrating the safety and efficacy of psychedelic therapy for depression (50% confident)
This prediction was right, and the psychedelic revolution in medicine remains under-covered.
Telehealth use will increase without any impact on overall health spending, outcomes, or any significant clinical impact (90% confident)
This seems like it’s going to be wrong, even if data for 2022 isn’t available yet.
My old boss, Mehmet Oz, will win his election to the US Senate (50% confident)
This was wrong. And I think I’m done with writing about this story and my experience there.
No professional society will add a polypill for cardiovascular disease prevention or a salt substitute for stroke reduction to their official guidelines even though there’s great evidence for both (90% confident)
Depressingly remains correct, even though the polypill is real and it’s spectacular. And every single time I see a story about approval for a heart disease drug with minimal evidence that it reduces heart attacks or strokes, I’m going to repeat myself about the polypill.
Deaths from opiates will tragically continue to rise (90% confident)
Appears to be right, unfortunately and depressingly.
Narcan will be available without a prescription nationwide by the end of the year (50% confident)
It’s good news that we can say that this one was right.
US healthcare spending will rise (95% confident)
Seems right, again fairly depressingly.
Non-COVID, non-medical predictions
Bitcoin will cross $100,000 at some point, not necessarily ending the year there (40% confident)
Not even close. And I was wrong about this for a million different reasons, from not understanding the technology, to buying into media hype, to being kind of gullible.
The Brooklyn Nets will win the NBA title (20% confident)
Wrong about this one, although 2021-me would be surprised that 2022-me would agree more with Kyrie Irving about vaccines than about Jews.
Tom Brady will not retire in 2022 (70% confident)
I was right, but I’m pretty sure that TB12 regrets this decision.
Pivoting to the metaverse will end up being a good business decision for Facebook AKA Meta and their stock price will end 2022 higher than it ended 2021 (50% confident)
Woof.
There will be zero mass market lab grown meat products available in 2022 (90% confident)
This one was correct. As an aside, my dive into learning about lab grown meat led to perhaps my favorite turn of phrase that I read this year: “And it’s a fractal no,” he told me. “You see the big no, but every big no is made up of a hundred little nos.”
All in all, 16/30 for 2022. Not a great showing. Here’s to a better 2023. You’ll get those predictions over the next few weeks.