Warren J. Dew
Forum Master
The analysis in that paper on the effectiveness of various measures is quite good. Their model is detailed enough that it should address the critiques in this thread about the Washington Post "simulation". One could quibble with some of their assumptions, but none are obviously unreasonable. Their division of general strategies into "mitigation" - trying to get to herd immunity with minimal stresses on the health care system - and "suppression" - attempting to suppress all outbreaks - is a good framework for analysis.Here's a paper that summarizes the results of modelling various options. https://www.imperial.ac.uk/media/im...-College-COVID19-NPI-modelling-16-03-2020.pdf
This paper seems to have influenced a lot of governmental responses both in the UK and the US. Is it perfect? I rather doubt it. Are your opinions better than those of this paper? I highly doubt that. But if you have better data and a better model, please share it.
Let's look at some key findings:
- Mitigation alone will cause health care limits to be exceeded. Their most effective mitigation strategy, voluntary isolation of known cases, voluntary quarantine of household members of known cases, and voluntary social distancing of those over 70 - will still result in exceeding the limits of the health care system in the case of the UK. This backs up the discussion upthread on the UK mitigation strategies, which pointed out that they would likely result in age related triage with the NHS denying treatment to the very old.
- Suppression is possible without extreme measures. For example, School closure, case isolation, and voluntary general social distancing effectively suppresses the epidemic until the interventions are removed. Given this example did not even include household quarantine, there are likely other combinations that would be effective; for example, social distancing could likely be eased or eliminated if households of known cases were voluntarily quarantined. Lockdowns are not required. The very large difference between the school closure and no school closure cases suggests that school closure is necessary, though. All of this is exactly consistent with the positions I've taken in the thread.
- Suppression requires years. Suppression measures need to be maintained until an effective vaccine has widespread availability, which is likely to be at least a year or two. I've made this observation myself upthread.
- Intervention measures can have "enormous" social, economic, and health costs. "We do not consider the ethical or economic implications of either strategy here, except to note that there is no easy policy decision to be made. Suppression, while successful to date in China and South Korea, carries with it enormous social and economic costs which may themselves have significant impact on health and well-being in the short and longer-term." This is exactly the point I made in the post you are responding to.
- Bans are not useful. From the paper, "Stopping mass gatherings is predicted to have relatively little impact". This is exactly what I started out arguing: there's little public health value, for example, in shutting down dance competitions.
So basically the paper is backing up every point I've made.
The paper does have a couple of weaknesses. First, they don't do any sensitivity analysis on fatality rates. This is fine in the context of a paper that isn't trying to balance the costs of intervention against the costs of not intervening, but is an issue in discussions that do address such balancing, since reasonable estimates of the fatality rate vary across orders of magnitude. Second, they conclude that suppression is superior to mitigation, a conclusion that isn't reasonably possible to make without addressing the costs of intervention measures, which they "do not consider".
Still, the model seems quite reasonable and the analytical - as opposed to policy - conclusions seem sound.
Last edited: