Is it safe to "open up" the economy?

I’ve been getting this question over and over again. Ending the lockdown is a constant topic of discussion. How do we make a decision about when it’s safe to restart our lives?

To start, my general impression is that the idea of “opening the economy” is something of a false premise. Yes, governmental restrictions play a role in what people do, but if people don’t feel safe going out in public, economic activity is only going to be a fraction of what it was before the pandemic. The individual choices that people for themselves make are going to determine our collective economic path forward more than any mandate from above.

I’m not going to pretend that I’m an expert on modeling the spread of infectious diseases or simulating the death rates without social distancing in place. I’m not an epidemiologist. This isn’t a post about how we model rates of infection and calculate an R0 for SARS-CoV-2 in order to estimate total deaths and intubated patients so that we can compare those numbers to the potential economic damage to guide our path forward.

How I think about the risk here

I’m a doctor. I know how to talk to my patients to help each individual personalize his or her own decision making. This is a post on how I would advise a patient of mine to think about their own personal choices - and those of their family members - with regards to getting back to normal life.

As a physician, it’s usually easy to tell people what to do. When it comes to most of the diseases I treat, I think about the data underlying different possible outcomes in order to make appropriate therapeutic decisions.

For example, a common medical problem I see is atrial fibrillation, an irregular heart beat that increases risk of stroke. To decide on the need for a blood thinner in these patients, I calculate the annual risk of stroke based on well validated data that is applicable to my patient. Once we have the numbers, our discussion can be personalized with an individual’s own approximate risk of a stroke and the approximate amount that a blood thinner reduces that risk.

Everyone has a different method of calculating their own personal risk/benefit analysis. But it’s a lot easier when you have some numbers to help estimate your own risk, even if those numbers are just approximations.

So what do the numbers look like in COVID-19?

The difficult thing with COVID is that we really don’t have the right type of information to help each person make his or her own decision (I’d argue we don’t have that information on a societal level to make policy determinations, but again, I’m not an epidemiologist).

Not to say that we don’t have data in COVID. We have approximate risk of death once someone is sick enough to come to medical attention, a number that is called the case fatality rate, or CFR. Depending on which dataset - and which population - you look at, the number ranges from about 0.3-3.5% (this being an average risk across age ranges, with higher risk in older folks and lower risk in younger ones).

When you think about it like those are the actual numbers, the risk/benefit calculation is straightforward, even if the resulting decision-making for each person may be difficult.

The problem is that we don’t really know how often people who get infected end up with symptoms of disease to warrant seeking medical care because we don’t have widespread testing across the asymptomatic population. If the overall prevalence of infection across our population is already 50%, it means that a lot fewer people who get infected are sick enough to warrant medical care than if only 5% of our population has been infected.

In some diseases, almost everyone who is infected gets sick. That doesn’t seem to be the case in COVID-19. So even if we know quite a bit about the CFR, we know very little about the infected fatality rate, termed the IFR (the number of people who die among all of those who are infected, not just among those sick enough to get tested).

I don’t envision a scenario in which I can have any type of coherent and personalized risk/benefit discussion with a patient when there’s this much of a range in possible risk. We are talking about a potential order of magnitude difference about how sick you might get.

Can you give me an example?

Let’s take a 70 year old man with no medical problems:

• If the IFR is essentially identical to the CFR, our hypothetical patient is looking at an approximately 8% chance of death if he gets infected

• But if the IFR is 1/10 the CFR - meaning that only 1 in 10 patients who are infected have a positive test results and become a “case,” that risk of death for our patient drops to approximately 0.8%.

• That same 70 year old man coming into the hospital with a minor heart attack also has a risk of death of about 8%

Across society, the amount of morbidity and mortality even at a low range of estimates means an almost indescribable amount of pain and suffering. But for each individual trying to decide whether to go back to work, to go to the grocery store, or even to see family, this degree of variation could mean the difference between staying at home in isolation versus going about a quasi-normal life.

So what should I do?

The point of an exercise like this is that there isn’t a blanket recommendation for everyone. This is about understanding your personalized risk - a risk that’s reasonable for me might not be one that’s reasonable for you.

One person can look at a CFR of 8% and think, “there’s no way that I’ll go back to my regular life if there’s a 1 in 12 chance I die if I get infected. Why would take the chance of catching an infection that could be as deadly as a heart attack?”

But you might look at it and say, “I need to get back to the office and I need to see my family, so I’ll take my chances with a 92% likelihood of survival. Plus there’s a chance that the numbers are wrong and my likelihood of dying is much lower than that.”

The more clarity we have with our numbers, the more accurate we can feel about these estimates. As it stands now, there’s still a lot of uncertainty and all of the numbers above must be viewed with a grain of salt.

I’m hopeful that we will get to the point where we aren’t flying so blindly. But until we have more information, I am going to assume that the numbers are on the pessimistic side when I counsel my patients. It’s up to all of us to assess our own personal risk, and the better our testing is, the more comfortable we feel that we’re making those assessments based on reality instead of hope.