Last week I started with good news from South Africa - HIV and TB co-infection do not drastically increase mortality from covid-19, and case fatality rates from covid-19 in South Africa are so far significantly less than in many places in the world.
This week I'm starting with bad news from South Africa - the epidemic curve still doesn't seem to be flattening in the least. You can see this most clearly in the middle row on the left, showing cases on a log scale. See the trajectory? South Africa looks like it will overtake Italy in number of cumulative cases in about 3 weeks. (Remember that these graphs are plotted per week of the epidemic, starting with the week that the country reached 100 cases, so that on its current trajectory, South Africa will have more cases than Italy by week 17 of the epidemic.) Italy is a useful comparator for South Africa because they have very similar population sizes: 59.3 million in South Africa, 60.4 million in Italy. So when South Africa reaches the same absolute number of cumulative cases as Italy, it will also have a similar number of cases per capita (actually slightly higher). So far most diagnosed cases are in Western Cape, and Western Cape is experiencing the brunt of the epidemic so far.
Even if South Africa's cases outstrip Italy by week 17, this doesn't mean that deaths will also catch up with Italy's at the same time, as South Africa's case fatality rate (CFR) so far is only about 30% of Italy's (thanks to SA's younger population). But deaths, like cases, are also doubling about every 2 weeks in South Africa. This means that at the current trajectory, South Africa's number of deaths will catch up with Italy's by about week 23 (just over 2 months from now). We've discussed before that confirmed cases are not a great measure of an epidemic, as number of cases depends so much on number of tests performed. Yet South Africa is testing far less than Italy, so if SA reaches the same number of cases, it's likely that SA actually has far more infections than Italy. (SA so far has performed 9 tests per diagnosed case, and Italy at this point in its epidemic had tested more than 100 tests per diagnosed case.) But once again, it's really the death count that we should be looking at in assessing the prevalence and severity of the epidemic.
Dr. Michael Osterholm is an American infectious disease epidemiologist who wrote a book 3 years ago mapping out the possibility of just such a pandemic. He's a voice to listen to. This interview was recorded 3 weeks ago and published 2 weeks ago, but I just read it this week, and it pretty much ruined my week (be warned). The rock-in-my-stomach realization for me was that this pandemic is likely not going anywhere soon, that the worst of it is still ahead, and we may be living with infection control measures (masks, physical distancing) for another year or two, even longer. Dr. Osterholm strenuously objects to the term "social distance" and reminds us that the goal is to be physically distanced, but as socially close to each other as possible. He also strongly encourages us all to get outside - it's perfectly safe to go to a park or walk with someone as long as you stay 6 feet apart.
It's a long interview and I can't summarize everything, but some key points:
We are nowhere close to herd immunity; only about 1 in 20 people (globally) has been infected.
Even if we develop a successful vaccine (and this seems to be a particularly tricky vaccine to develop), the huge logistical challenges of producing and distributing the vaccine likely mean it will be 18 months or longer in a best case scenario before there is widespread access to a vaccine. Also, a vaccine may give incomplete or short-term protection.
Wearing a cloth mask does help, but may give you 20 minutes rather than only 10 minutes of close contact with an infected person, before you become infected. Being outdoors is always far safer than indoors.
No one can predict what this virus is going to do. Similar pandemics in the past ebbed and flowed, and the reasons weren't always very clear. BUT in each of the 10 flu pandemics of the last 250 years, there was always a second wave that was usually much more severe.
A strict lockdown is not sustainable for our economies or societies, but in Dr. Osterholm's view a laissez faire approach is also not an option. (He is very critical of the Swedish model.) We need a middle ground between these two extremes in which we can learn to live with the virus, but not see massive casualties.
I'm sure every American who has opened a reputable newspaper or news website in the last week or two knows by now that cases are increasing in about half of U.S. states. It's really alarming. This article from Time magazine has a simple graphic for every state. President Trump and members of his administration are inexplicably arguing with the scientists again (and barreling ahead with a large, indoor rally this weekend) - but to be clear, these increases are not just due to increases in testing, whatever they may say. (There's a good explanation here on why data show that actual infections are increasing, but the short answer is that we are seeing percent of tests which are positive increase in many states, as well as hospitalizations.) It's also true that we are not seeing the 3,000 deaths per day in June that some models were predicting a couple months ago. But increasing cases, in the absence of any game-changing breakthroughs in treatment, are a ticking time bomb that will lead to an explosion of deaths a few weeks later. There was news this week that a widely available steroid may significantly reduce fatality rates for critically ill covid-19 patients - but the full peer-reviewed article hasn't even been published yet, so I think caution (dampened enthusiasm) is warranted until we know more.
Some other interesting research, summarized at STATnews.com, which is a reliable source for info on coronavirus research. A recent large, multi-country study shows that younger school-aged children (ages 5-9) are less likely to show evidence of prior SARS-CoV-2 infection, but older children (ages 10-19) were as likely as adults to have been infected. This is good news regarding the safety of younger pupils, who are most certainly going to have a hard time staying physically distanced and keeping their masks on when they return to school, but also seem to have a natural immunological advantage. However, our data to date come from a time period in which people were largely staying at home. We won’t know how much children become infected and spread the virus while back at school, until they are actually back at school. And we can expect that re-opening schools will have different impacts in different places, judging by different country experiences so far.
And finally, a piece of good news that caught my eye as I prepare to get my hair cut next week. I am fully aware that sitting inches away from a stranger for a prolonged period of time will likely be the most risky thing I have done since the epidemic began (except getting on a plane), but yet nearly 6 months after my last haircut, my vanity has won out. So I was delighted to read that masks (worn by hairdressers and clients) apparently prevented two infected hairdressers in Missouri from infecting any of their clients. I think natural experiments like this are fascinating, and hugely important, as we try to figure out the mysteries of how this virus behaves. Wear your masks!