We use actual case numbers for the Omicron variant in New Zealand to estimate Ro.
We find that case numbers may potentially triple in a two-day period.
To estimate Ro (the average number of people one person may infect without isolation over a 10-day infectious period), we need to estimate (project) the number of people a case may infect if the case had not been isolated.
For background see:
We look at the daily increase (r). We use this to project additional cases.
Below are actual case numbers (first two columns):
Note that on 26 January there are 12 historic cases that have not been included as cases for the day (they are in the total).
“In addition, there is a total of 12 cases linked to the three border workers reported earlier this month with confirmed or suspected Omicron, an increase of six.” See:
On 28 January we had only 15 new cases. The curve appears to be starting to flatten at least until we have some super-spreader events.
We have found that a weighted average works better to estimate r.
We see that r is close to SQRT(3).
This means that case numbers could triple over a two-day period.
Compare the second Total column with the last column.
We use the weighted average column for the factor.
We use 0.732 as the Factor.
We see on 27 January the totals are close.
We look at this table:
We see that for r = 1.732, Ro ~ 12.73.
We also see that for r = 1.66 (Step 2 and Step 3), Ro ~ 11.
1.73 – 1.66 = 0.07
1.73 + 0.07 = 1.8
For r = 1.8, Ro ~ 14.4
These estimates (Ro from 11 to 14.4) are within our previously estimated range for Omicron. See: