We applied models that predicted the pattern of the first peak of COVID-19 cases in late July 2020 but did not predict the second peak in November 2020.
The peak in November 2020 can be explained by now adding to the model: a) data that has become available on negative test numbers; b) an increase in reproductive number (i.e. R(t)) in October/November, which may be related to relaxation of restrictions and increased movement; and c) wider spread of COVID-19 into rural Kenya than in the first wave.
We consider the most plausible effect of schools reopening on 4th January to be that the transmission rate in Kenya will increase the time-varying reproductive number (R(t)) by +25%, and, increase mixing between social clusters that were not in contact whilst schools were closed.
Under the most plausible scenario, we project that the rate of COVID-19 case and death incidence will peak in mid-March 2021.
Under the most plausible scenario, we project 13.7 thousand (10.6k-16.8k) new determined COVID-19 cases and 116 (58-289) new COVID-19 attributed deaths by June 1st 2021 (Fig. 1).
The estimated +25% R(t) increase is conditional on other restrictions that reduce transmission remaining in place, and measures being in place to reduce transmission in the schools setting.
A worst-case scenario would be an increase in R(t) by 50% and resulting in an epidemic of similar magnitude to the second outbreak in the country. We think this is unlikely.
This model needs updating with the latest data on negative test numbers, and to add information on the age distribution on cases and deaths in Kenya. These improvements are also required for medium-term forecasts of vaccine effectiveness in Kenya. This stage of model improvements is expected to be completed by early February.
The impact of other events such as new variants could increase the R(t) by more than +25%, and case numbers and deaths would then exceed our predictions.
Link to the Brief – Projections of COVID-19 cases and deaths following schools reopening