New quarantine guidance! Some of it could make sense. Much of it doesn’t.
In brief, CDC is gambling that shortened quarantines will improve adherence, offsetting the higher risk of transmission. Before jumping in, let’s wrangle some jargon: quarantine vs isolation. Both refer to separating and restricting movement of people. Quarantine: People in their usual state of health who have been exposed to COVID-19. Isolation: People infected with COVID-19 (whether symptomatic or not). This entire post refers to people in quarantine, without symptoms -- although of course many people with COVID may be asymptomatic but still infectious, and that’s why there’s a risk of COVID transmission from asymptomatic people coming out of quarantine.
The risk of COVID transmission after 14 days of quarantine is 0.1%. Cutting that down to 10 days, which CDC is now supporting, increases the risk of transmission to 1.4%. There are (at least) three ways to look at this difference, and all can be true simultaneously:
1.4% isn’t a very big number
1.4% is more than 10 times bigger than 0.1%
1.4% is an estimate based largely on mathematical modeling, and: All models are wrong, some are useful. This model was released on Nov. 24 as a pre-print, which means it has not been reviewed by independent scientific peers who would be in the best position to find any errors or flaws in methodology.
The biggest problem we can see with this model is the assumption that the reduced time frame will improve quarantine adherence. The single reference provided for this particular model input is a good review that found the main factors influencing adherence to quarantine go well beyond length of quarantine. The full list: “knowledge people had about the disease and quarantine procedure, social norms, perceived benefits of quarantine and perceived risk of the disease, as well as practical issues such as running out of supplies or the financial consequences of being out of work.” And where length of quarantine was examined, there was no clear evidence with the exception of one study done among college student in Kansas in 2006 (it was a mumps outbreak): “In a college setting, it may be difficult to achieve high compliance with guidelines recommending that persons stay isolated for much longer than 4 days.”
Here’s the public health take-home at the population level: COVID-19 is spreading uncontrolled through communities. CDC has proposed an alternative to reduce quarantine time from 14 to 10 days based on a mathematical model that has not yet been peer-reviewed, and builds its case on an assumption that this change will improve quarantine adherence although the existing evidence is murky at best.
And here’s the take-home for individuals: if you follow all quarantine procedures faithfully for 10 days instead of 14 days, there is some increased risk that you may pass COVID onto others (and again we're just talking about scenarios where you are asymptomatic -- if any signs or symptoms present during quarantine, there's a different pathway to follow). If you’re planning to see people highly vulnerable to severe COVID illness, stick with 14 days. This remains the CDC standard, although this fact seems to be drowned out by the attention to shorter quarantine periods. Otherwise, make your own call based on knowing there’s an increased, but perhaps still acceptable level of risk.
The other option CDC now supports is quarantine that can end after day 7 with a negative COVID test collected on day 5 or later. This recommendation increases the risk of COVID transmission after ending quarantine to … two different figures depending on the type of test you get: 4% with PCR testing and 5.5% with antigen testing. These figures are not only disturbing in context of overwhelming levels of COVID in our communities right now; this approach also further taxes testing resources which are already stretched thin, including further burden on our lab workers and other personnel, likely further worsening of delays in obtaining test results. This alternative is not something we would advise, either from a public health perspective at the population level, or at the clinical level for individuals considering this option.
Going back to the alternative of a 10 day quarantine: what if the CDC model has it right, and this change will result in dramatically more people adhering to quarantine? If we’re at 30% adherence to quarantine now and this new approach results in doubling of adherence, to 60% -- it’s possible this would be a game changer. But the CDC model doesn’t look at potentially realistic scenarios. The model authors point out, “[A]dherence to all measures may be lower in practice than considered” in their projections. And more importantly, as highlighted above, there is no compelling reason to believe this change would result in dramatically higher adherence, as it may not address the root causes of non-adherence. Which brings us to an always-important question: so, who cares? If we’re non-adherent now and non-adherent after writing down some slightly different words in CDC guidance, why does it matter?
First, missed opportunities. There’s a hypothesis to work with here, and interventions to improve adherence to quarantine deserve to be tested. The overly narrow focus on the length of quarantine buries other issues that are likely critical to adherence, and jumps right over much-needed discussions of compensation and support for those who are asked to quarantine. Second, resulting harms. At an individual level, someone seeking to safely visit a vulnerable relative, for example, who would have otherwise quarantined for 14 days -- is now bringing additional risk into the equation. Third, trust. CDC has taken a number of hits to public credibility during this pandemic, and it’s crucially important to establish and maintain trust for an effective public health response. This change has some hallmarks of a shot in the dark fired from the top of a house of cards. Maybe it will work. We hope so. But we find it hard to substantiate that hope.