March 11, 2020

expert reaction to preprint on the impact of non-pharmaceutical interventions carried out by China in the COVID-19 outbreak

A preprint, posted to MedRxiv, reports on the impact of non-pharmaceutical interventions carried out by China in the COVID-19 outbreak.

Prof Mark Woolhouse, Professor of Infectious Disease Epidemiology, University of Edinburgh, said:

“The study by Lai et al. reports a sophisticated, quantitative analysis of the COVID-19 epidemic in China.  The authors estimate the impact of ‘non-pharmaceutical interventions’ – the use of measures including case isolation, contact tracing, self-quarantine and travel bans.  In Wuhan, the worst affected city, these measures have been strictly enforced and the city is described as being in ‘lock down’.  In fitting a mathematical model to the epidemic curve, the authors used large data sets, both historical and real-time, on movement patterns (including location data from mobile phones) and so were able to estimate the impact of reduced rates of movement on the spread of infection.

“The study concludes that reduced rates of movement and contact between people in China has led to the decline the numbers of cases being reported.  This is encouraging in one respect: it tells us that a COVID-19 epidemic can be controlled.  The study also suggests that had measures been implemented just 2 weeks earlier then there would have been significantly less spread within China (and presumably beyond).  What is less clear from this analysis is what should happen next.  One of the authors’ simulations indicates that lifting travel restrictions risks a second epidemic wave unless contact rates are held below normal levels.  Can that be done, and for how much longer?  When can normal life resume?  At present, there are no clear answers to these questions.

“Overall, this type of analysis is invaluable for helping us understand the complex interplay between virus transmission and non-pharmaceutical interventions, but the results need careful interpretation.

“The authors suggest that similar interventions could control COVID-19 epidemics in countries other than China.  However, I would caution that if similar interventions are to be introduced in other countries (as is happening in Italy and could yet happen in the UK) then they should be accompanied by a clear exit strategy that sets out the circumstances under which restrictions on normal life will be lifted.

“The authors also suggest that interventions should be introduced earlier.  This is true if and only if the sole policy objective is to minimise the size of the initial epidemic.  A more comprehensive analysis needs to take into account both i) the very substantial social and economic costs that the interventions impose and ii) the long term dynamics of infection and disease in the population.”

Prof Rowland Kao, Sir Timothy O’Shea Professor of Veterinary Epidemiology and Data Science, University of Edinburgh, said:

“This study exploits a combination of movement data from this year and previous years to estimate the impact of non-pharmaceutical interventions (social distancing, contact tracing, case detection) on the containment of covid-19 in China.  By comparing movement patterns in previous years to this year, and the rate of spread of covid-19 before and after interventions were put in to place, the authors were able to estimate the impact of these controls on disease spread, concluding that early case detection has a greater impact than social distancing (contact reduction and travel restrictions) but that both are valuable.

“This is a clever approach and, if replicated elsewhere, would help authorities to gauge the impact of non-pharmaceutical interventions on covid-19, and aid in disease control planning.  Important consideration of several factors should be made to avoid over-interpretation of these results however.  First, mobility patterns are estimated using near-real time mobile phone data.  These data are likely the best available source to recreate the relevant mobility patterns, but are potentially biased by the mobile phone usage of different groups of individuals.  Older people in particular are less likely to use mobile phones and may have different mobility patterns than average and are also more likely to suffer from severe disease.  Second, there are recent suggestions that the restrictions in place may themselves be changing the pattern of transmission, changes that could not be embedded in the simulation.  Finally the results themselves should not be interpreted as being universal – they will of course depend on the individual circumstances of each country or region, where the controls are put in place.  Therefore any interpretation of the results directly for the UK should be done with caution.  We don’t know how the balance of different restrictions would translate to here – for example the modes of travel themselves will be different, as will the density of travellers and characteristics of travellers.  Also compliance might be different, both to the measures prescribed, but also to the other activities that we would do to reduce transmission – e.g. what is the relative compliance on hand washing, for example?  Because those other factors interact with the measures described, societal differences in those factors will therefore come into play as well.”

Dr Thomas House, Reader in Mathematical Statistics, University of Manchester, said:

“This study makes use of a combination of data on human mobility and behaviour with a model of disease spread to conclude that the dramatic interventions seen in China have been successful in controlling spread of COVID-19.  Overall, this is consistent with various other direct and indirect evidence available to us at this stage, and the modelling and other assumptions seem to be sound.  This study is important, since there is currently a decision to be made about the appropriate level of non-pharmaceutical interventions like self-isolation and quarantine.  Such measures carry economic and health costs that may lead to non-compliance and which must be traded off against their benefits in reduction of transmission – but by providing evidence that they can work to contain the current outbreak, which is not guaranteed and depends on the pathogen in question, this paper contributes to the discussion.”

Press Release – https://www.southampton.ac.uk/news/2020/03/covid-19-china.page

‘Effect of non-pharmaceutical interventions for containing the COVID-19 outbreak: an observational and modelling study’ by Shengjie Lai et al. was posted on the medRxiv preprint server on 9th March.  This work is not peer-reviewed.

DOI – https://doi.org/10.1101/2020.03.03.20029843

All our previous output on this subject can be seen at this weblink:

http://www.sciencemediacentre.org/tag/covid-19

The SMC also produced a Factsheet on COVID-19 which is available here:

https://www.sciencemediacentre.org/smc-novel-coronavirus-factsheet/

Declared interests

None received.