Results for “south africa”
264 found

The new Covid equilibrium

Many people have stopped keeping track of where Covid is headed, if only because it is such a stressful and unpleasant topic.  To be clear, under current circumstances I favor complete “Covid laissez-faire,” though with subsidies for new and better vaccines.  Overall, things are not so peachy keen (NYT):

The central problem is that the coronavirus has become more adept at reinfecting people. Already, those infected with the first Omicron variant are reporting second infections with the newer versions of the variant — BA.2 or BA2.12.1 in the United States, or BA.4 and BA.5 in South Africa.

Those people may go on to have third or fourth infections, even within this year, researchers said in interviews. And some small fraction may have symptoms that persist for months or years, a condition known as long Covid.

“It seems likely to me that that’s going to sort of be a long-term pattern,” said Juliet Pulliam, an epidemiologist at Stellenbosch University in South Africa…

“If we manage it the way that we manage it now, then most people will get infected with it at least a couple of times a year,” said Kristian Andersen, a virologist at the Scripps Research Institute in San Diego. “I would be very surprised if that’s not how it’s going to play out.”

I know many of you like to say “No worse than the common cold!”  Well, the thing is…the common cold imposes considerable costs on the world.  Imagine a new common cold, which you catch a few times a year, with some sliver of the population getting some form of Long Covid.  One 2003 estimate suggested that the common cold costs us $40 billion a year, and in a typical year I don’t get a cold even once.  That 2003 estimate also does not include the sheer discomfort of having a cold.

With a pinch of Long Covid in the distribution surely the current virus is a wee bit worse than that?  While many cases of Long Covid are malingerers and hypochondriacs, at this point it is clear that not all of them are.  Toss in some number of immunocompromised individuals (how many?).

Even under mild conceptions of current Covid, it is entirely plausible to believe that the costs of Covid will run into the trillions over the next ten years.

Death rates are not up, but more of the unvaccinated will die off with time and the rest of us will face this steady risk and planning annoyance for — how long?  Plus we’ll get lots of “colds,” some of them considerably worse than a cold.  And with what risk that it might mutate again and get worse? The next generation of vaccines probably will not be directly subsidized.  Which will mean much lower rates of uptake.  The point of maximum Covid immunity may well be behind us.  And you won’t be able to blame it all on lockdowns.

Please keep in mind that when it comes to your reactions I will read many of them as not much better than “I just don’t want to think about this, I am still in denial.”

The equilibrium

The South African drugmaker Aspen Pharmacare earlier this year finalized a deal to bottle and market the Johnson & Johnson vaccine across Africa, a contract that was billed as an early step toward Africa’s development of a robust vaccine production industry. Aspen geared up for production, but no buyers, including the African Union and Covax, have placed orders yet, said Stephen Saad, Aspen’s chief executive.

The Serum Institute of India, the world’s largest vaccine maker, stopped its production of Covid shots in December last year, when its stockpile grew to 200 million doses; Bharat Biotech, another Indian firm that was a major producer, also stopped making vaccines in the face of low demand. The companies say they have no further orders since their contracts with the Indian government ended in March.

Here is more from The New York Times.

Against alcohol, #6437

This paper evaluates the impact of a sudden and unexpected nation-wide alcohol sales ban in South Africa. We find that this policy causally reduced injury-induced mortality in the country by at least 14% during the five weeks of the ban. We argue that this estimate constitutes a lower bound on the true impact of alcohol on injury-induced mortality. We also document a sharp drop in violent crimes, indicating a tight link between alcohol and aggressive behavior in society. Our results underscore the severe harm that alcohol can cause and point towards a role for policy measures that target the heaviest drinkers in society.

That is new research from Kai Barron, Charles D.H. Parry, Debbie Bradshaw, Rob Dorrington, Pam Groenewald, Ria Laubscher, and Richard Matzopoulos.  To be clear, the “policy measure” I favor is absolute individual boycott, not some kind of soporific Pigouvian tax scheme that won’t attract any real extra attention.

Fractional Dosing Trials

My paper Testing fractional doses of COVID-19 Vaccines, co-authored with Kremer et al., has now been published at PNAS. I covered the paper in A Half Dose of Moderna is More Effective than A Full Dose of Astra Zeneca and other posts so I won’t belabor the basic ideas. One new point is that thanks to the indefatigable Michael Kremer and the brilliant Witold Wiecek, clinical trials on fractional dosing on a large scale have begun in Nigeria. Here are a few key points:

WHO SAGE Outreach: The authors have met and presented their work to the World Health Organization (WHO) Strategic Advisory Group of Experts on Immunization (SAGE), with follow-up meetings to present evidence coming from new studies.

DIL Workshop and Updates: In the fall of 2021, the Development Innovation Lab (DIL) at UChicago, led by Professor Kremer, hosted a workshop on fractional dosing, collecting updates from clinical researchers from multiple countries conducting fractional dosing trials for COVID-19 vaccines. The workshop also covered issues relating to trial design and included participants from Belgium, Brazil, Ghana, the Netherlands, Nigeria, Thailand, South Africa, UK and the US. 

CEPI Outreach: Professor Kremer has also presented this research to The Coalition for Epidemic Preparedness Innovations (CEPI), which is now pursuing a platform trial of fractional dosing.

Country Trials – Nigeria: With the support of DIL and the research team and generous support and advice from WAM Foundation, the charitable arm of Weiss Asset Management and Open Philanthropy, a trial is being conducted in Nigeria by the Nigerian Institute of Medical Research, National Institute of Pharmaceutical Research and Development, National Agency for Food and Drug Administration and Control, and the National Primary Health Care Development Agency, in coordination with the Federal Ministry of Health.

A comprehensive list of all the trials on fractional dosing conducted to date is at the link. Fractional dosing may come too late for COVID-19 vaccines but perhaps next time a shortage of a vaccine looms we will be more quick to consider policies to stretch supplies.

52 things Tom Whitwell learned in 2021

Always a good list, here is this year’s edition, and an excerpt:

  1. Beauty livestreamer Li Jiaqi sold $1.9 billion worth of products in one twelve hour show on Taobao. That’s slightly less than the total sales from all four Selfridges stores during 2019. [Jinshan Hong]
  2. 10% of US electricity is generated from old Russian nuclear warheads. [Geoff Brumfiel]
  3. Some South African students sell school Wi-Fi passwords for lunch money. Residents walk up to 6km to connect to schools because 4G data is so expensive. [Kimberly Mutandiro]
  4. Productivity dysmorphia is the inability to see one’s own success, to acknowledge the volume of your own output. [Anna Codrea-Rado]

For the pointer I thank Nabeel.  And here is Whitwell on Get Back and the principles of creativity.

The Covid pandemic is not taking the very best of turns

This was emailed to me, but I am not doing a double indent…in any case I fear the person might be right…

“The prevailing sentiment is that the COVID pandemic is close to over. The vaccines are of course miraculous but we are not currently on a good trajectory.

  • It is increasingly clear that two shots plus a booster of our current vaccines are the least one needs to have effective medium-term protection. Almost nowhere (least of all the US) is on track to reach this kind of coverage. The messaging in the US remains mistaken, where the CDC to this day recommends boosters only for those aged 50 and older. More broadly, the institutional confusion around boosters shows that the adults are not yet in charge.
  • Even though Delta arose in the spring, we are still vaccinating (and boosting) people with the original Wuhan strain. This is insane, and probably meaningfully less effective, and yet nobody is up in arms about it.
  • Severe outbreaks are manifestly possible even in exceptionally vaccinated populations, especially when booster uptake is low. See: Singapore, Gibraltar, Ireland. One should assume that almost every part of the US will see significant waves before COVID “ends”, whatever that turns out to mean. Note that just 60% of the US population is vaccinated today with two doses.
  • There is early suggestive evidence from Israel that boosters may wane.
  • Waning aside, it’s clear that breakthrough infections in boosted individuals are not uncommon. While the vast majority of those infections are not severe, this does mean that there will still be plenty of mutagenesis.
  • It’s unclear that longitudinal cross-immunity is strong. Getting COVID is not enough to confer long-term protection. We probably can’t just “get this over with”, even if we are willing to tolerate a large number of one-time deaths.
  • The currently-breaking news about the South African Nu strain shows that arguments about how the spike protein is running out of mutation search space are almost certainly wrong.
  • While the fog of war is thick right now, the early data on Nu suggests that it may be a big deal. Even if it’s not, however, it has been obvious since we got the vaccines that vaccine escape is a concern. You can debate whether the probability of a vaccine escaping variant is 20% or 80%, but in any case we need effective contingency plans in place. If we fail to respond effectively to Nu, that will be a considerably greater institutional failure than anything that happened at the outset of the pandemic. We’ve had almost two years since the first COVID case and one year from the vaccine approvals to prepare. So I ask: what is the plan for the vaccine-escaping variant?

On current trends, it looks like we will probably need one of two things to effectively end the pandemic: (1) very effective COVID therapeutics (paxlovid, molnupiravir, and fluvoxamine all being candidates but my guess is that none is a silver bullet) or (2) pan-coronavirus vaccines (with broader protection than what is currently available).

It isn’t over yet.

P.S. Has any U.S. health body recommended the clinical use of fluvoxamine (an already-approved drug), or has the FDA given any guidance as to when it might approve paxlovid? If not, can they outline their reasoning? 1,600 Americans died of COVID on Nov 24.”

Nu, a variant of real concern

Here is the Eric Topol thread.  Do read it.  Here is the scary graph, based on preliminary data.  Here is Bloom Lab.  Here is a layperson’s take from the Times of London:

When was the variant first discovered?
South African authorities raised the alarm at 2pm on Tuesday of this week, when they found samples with a significant number of worrying mutations.

The samples dated from tests taken on November 14 and 16. On Wednesday, even as scientists were analysing the genome, other samples were found in Botswana and China, originating from travellers from South Africa.

Why were scientists initially concerned by this variant?
The spike protein is the tool a virus uses to enter cells, and the part of it our vaccines are trained to spot. This variant had 32 mutations in the spike — meaning it would look different to our immune system and behave differently when attacking a body. As a virologist at Imperial College put it, it was a “horrific spike profile”.

Why has worry increased over the course of the week?
When geneticists and virologists looked at the mutations they realised there was a high likelihood they could increase its transmissibility or help it evade immunity. But these concerns were still theoretical. However, today South African scientists spotted a quirk in the testing regimen. PCR tests look for three genes in the coronavirus and amplify them. If, however, the virus was this variant they were only able to amplify two.

In the province of Gauteng, where the proportion of tests coming back positive has rocketed to one in three, they found the proportion in which only two genes were amplified has also rocketed.

What does this mean?
There are three options. It is still possible — though unlikely — this is chance, with the variant’s apparently increased spread relating to an unusual cluster. If it does have a genuine advantage, then it is either better able to spread or better able to infect people who have prior immunity — either from vaccination or infection. Or, it is both.

This might come to nothing, but it is definitely a matter of concern.  One more general point is that even if Nu is a non-event, it seems to show that the space for possible significant mutations is largely than we had thought.

Thursday assorted links

1. “If one wishes to stop the virus, only one goal matters: Getting the reproduction rate below one. e to the 3 t is not a lot less exponential growth than e to the 6 t.”  Is that nowadays a cancellable thought?

2. NPR covers Fast Grants.

3. New Danish data on vaccine effectiveness against delta.  And Bergstrom on Simpson’s paradox.  Based on UK data, Eric Topol has a contrasting view.

4. The economics of Mexican indigenous languages.

5. Austria and Croatia set “expiry dates” for vaccinated travelers — 270 days.

6. “The current landscape sees Economists lobbing impeccably crafted papers into any conceivable area of social science inquiry.

7. Why no Lyme disease vaccine?  (New Yorker)

8. Data fabrication?

9. Thomas Quasthoff update.

10. Has eighty percent of South Africa already had Covid?  (Bloomberg)

What I’ve been reading

1. Richard Zenith, Pessoa: A Biography.  942 pp. of text, yet interesting throughout.  Brings you into Pessoa’s mind and learning to a remarkable degree.  (Have I mentioned that the world is slowly realizing that Pessoa’s The Book of Disquiet is one of the great works of the century?)  His favorite book was Dickens’s Pickwick Papers, and he very much liked Carlyle’s Sartor Resartus.  This biography is also interesting about non-Pessoa topics, such as Durban, South Africa in 1900 (Pessoa did live there for a while).  I am pleased to see Pessoa finally receiving the attention he deserves — definitely one of the books of this year.  Here is a good review of the book.  For a man who never had sex, this book covers his sex life a great deal!  And what a short and lovely title, no long subtitle thank goodness.

2. Nicholas Wapshott, Samuelson Friedman: The Battle over the Free Market.  Quite a good book, though it is interior to my current knowledge set and thus better for others reader than myself.  Contrary to what I have read elsewhere, Wapshott points out that Samuelson did not support Nixon’s wage and price controls, but this LA Times link seems to suggest Samuelson thought they were a good idea?

3. Jamie Mackay, The Invention of Sicily: A Mediterranean History.  While it was less conceptual than I might have preferred, this is perhaps the single best general history of Sicily I know of.  Short and to the point in a good way.

4. N.J. Higham, The Death of Anglo-Saxon England.  In 1066, five different individuals were recognized as de facto King of England — how did that come to pass?  Why was Aethelred the Unready not ready?  (He was only 12 when he assumed the throne, though much of the actual criticism concerned the later part of his reign.)  I find this one of the most intelligible and conceptual treatments of Anglo-Saxon England out there.  I don’t care what the Heritage Foundation says, beware Danish involvement in your politics!

Peter Kinzler, Highway Robbery: The Two-Decade Battle to Reform America’s Automobile Insurance System is a useful look at where that debate stands and how it ended up there.  Here is a good summary of the book.

It does not make sense for me to read Emily Oster’s The Family Firm: A Data-Driven Guide to Better Decision-Making in the Early School Years, but it is very likely more reliable information than you are likely to get from other sources.

A Step Closer to General AI

From Google’s Deep Mind:

In recent years, artificial intelligence agents have succeeded in a range of complex game environments. For instance, AlphaZero beat world-champion programs in chess, shogi, and Go after starting out with knowing no more than the basic rules of how to play. Through reinforcement learning (RL), this single system learnt by playing round after round of games through a repetitive process of trial and error. But AlphaZero still trained separately on each game — unable to simply learn another game or task without repeating the RL process from scratch.

…Today, we published “Open-Ended Learning Leads to Generally Capable Agents,” a preprint detailing our first steps to train an agent capable of playing many different games without needing human interaction data. We created a vast game environment we call XLand, which includes many multiplayer games within consistent, human-relatable 3D worlds. This environment makes it possible to formulate new learning algorithms, which dynamically control how an agent trains and the games on which it trains. The agent’s capabilities improve iteratively as a response to the challenges that arise in training, with the learning process continually refining the training tasks so the agent never stops learning. The result is an agent with the ability to succeed at a wide spectrum of tasks — from simple object-finding problems to complex games like hide and seek and capture the flag, which were not encountered during training. We find the agent exhibits general, heuristic behaviours such as experimentation, behaviours that are widely applicable to many tasks rather than specialised to an individual task. This new approach marks an important step toward creating more general agents with the flexibility to adapt rapidly within constantly changing environments. (Bold added, AT).

Hat tip: Daniel Kokotajlo at Less Wrong who notes “This seems like a somewhat big deal to me. It’s what I would have predicted, but that’s scary, given my timelines.” See also the LW comments.

In other news, South Africa awarded the first ever patent to an AI.