3 ms·

This estimate is broadly consistent with that published last week by a Imai et. al. [1] That previous work looked at foreign cases up to 18th January, and conc

by dzdt 7y ago

This estimate is broadly consistent with that published last week by a Imai et. al. [1] That previous work looked at foreign cases up to 18th January, and concluded the number of contagious cases in Wuhan as of that date is likely in the range of 1000 to 9700. This paper gives estimates of Wuhan-area cases ranging from 7600 to 43600 for January 26. A doubling time of 6 days was estimated for the pre-lockdown phase of the Wuhan epidemic. [2] That gives a factor of 2.5 increase in cases between Jan. 18 and 26. Unfortunately the newer estimated case numbers are somewhat higher than 2.5 times as many; probably the older number were too low. Note the estimated rate of increase is much LOWER than one would naively get from looking at the officially confirmed case numbers. Confirmed cases have been more than doubling every two days. This growth rate of confirmed cases is really the rate of increase of China's capability of carrying out confirmation labwork. There exported case research shows there is still a huge backlog of unconfirmed cases to be identified in the Wuhan area. [1] See "Report 2" at https://www.imperial.ac.uk/mrc-global-infectious-disease-analysis/news--wuhan-coronavirus/ https://www.imperial.ac.uk/mrc-global-infectious-disease-ana... [2] https://news.rthk.hk/rthk/en/component/k2/1505090-20200127.htm https://news.rthk.hk/rthk/en/component/k2/1505090-20200127.h...

cs702 7y ago
2.5 indeed looks too low now. The median estimate of the size of the outbreak by Imai et al as of Jan 18 is 4,000; the new median estimate by Chinazzi et al as of Jan 26 is 21,300 (assuming a catchment population of 20M, their middle case). If both median estimates happen to be spot-on, the size of the outbreak would have grown by a factor of 5.3 between those two dates.
- dzdt 7y ago
  I don't think you've got the right conclusion. The earlier estimate of the total cases was too low. Probably the newer estimate of total cases is ALSO biased too low. The reason for underestimate is they are assuming international surveillance found all exported cases. The paper describes this shortcoming: The estimates contained in this report have been constantly growing with respect to older versions compiled with the data available at previous dates. This is because the number of detected cases at International locations and a travel history from Wuhan city has nearly tripled in the last week. It is worth stressing that this is not implying that the epidemic is growing at the same rate. The estimated size of the outbreak refers to all cases occurred in the area since the beginning of the outbreak, and notification and detection delays may play an important role that, at the moment, do not allow the evaluation of the epidemic growth rate.
  - cs702 7y ago
    Thank you. Yes, you're right. It seems more likely that earlier estimates were low.

by dzdt 7y ago

cs702 7y ago

2.5 indeed looks too low now. The median estimate of the size of the outbreak by Imai et al as of Jan 18 is 4,000; the new median estimate by Chinazzi et al as of Jan 26 is 21,300 (assuming a catchment population of 20M, their middle case). If both median estimates happen to be spot-on, the size of the outbreak would have grown by a factor of 5.3 between those two dates.

dzdt 7y ago
I don't think you've got the right conclusion. The earlier estimate of the total cases was too low. Probably the newer estimate of total cases is ALSO biased too low. The reason for underestimate is they are assuming international surveillance found all exported cases. The paper describes this shortcoming: The estimates contained in this report have been constantly growing with respect to older versions compiled with the data available at previous dates. This is because the number of detected cases at International locations and a travel history from Wuhan city has nearly tripled in the last week. It is worth stressing that this is not implying that the epidemic is growing at the same rate. The estimated size of the outbreak refers to all cases occurred in the area since the beginning of the outbreak, and notification and detection delays may play an important role that, at the moment, do not allow the evaluation of the epidemic growth rate.
- cs702 7y ago
  Thank you. Yes, you're right. It seems more likely that earlier estimates were low.