We find that #Omicron is likely more transmissible than #Delta. This finding holds across a wide range of assumptions about immune evasion that are consistent with early neutralization studies.
First, we analyze trends in S-gene target failure (SGTF) at the provincial level. SGTF is a marker of #Omicron sublineage BA.1. Our updated model accounts for SGTF being an imperfect marker, & we now fit the model to data through 6 December.
We estimate a growth advantage for #Omicron of .32 (95% CI: .27 - .39) in Gauteng Province, corresponding to a doubling time of about two days.
We use the estimated proportion of samples with SGTF and the time series of infections to simulate potential growth trajectories for #Omicron and background variants (mostly Delta).
We then estimate the time-varying reproduction numbers based on the simulated infection time series. We find that the relative Rt during mid-Nov was 4-4.7 (medians), depending on what we assume about the time between infections (generation time) of #Omicron.
Next, we calculate mathematical model-based reproduction numbers of #Omicron & #Delta under current conditions. We vary immune evasion for #Omicron, which gives different transmissibility values to match the relative Rt values found in the previous step.
We also consider other assumptions for baseline immune fraction & background immune evasion. They also indicate #Omicron is likely more transmissible than #Delta, BUT do contain more extreme immune evasion regimes where #Omicron would be less transmissible.
We will continue to refine these estimates & look forward to peer review. We hope these estimates will be useful for public health decision making. They also indicate that it is critical to gather detailed epidemiological data on changes to generation time.
Based on these results and emerging global trends, decision makers should assume that infection rates will be very high. Given that, existing policies on self-isolation times, testing, and contact tracing activities may need to be reevaluated.
Code for all analyses is available from github.com/SACEMA/omicron… (n.b., we are still pursuing authorization to make some of the data publicly available).
[1/4] The recent observed increase in COVID-19 cases in Gauteng province is primarily due to a reporting delay for antigen test results. The case plots and metrics on the SACMC Epidemic Explorer are also affected by the reporting delay.
[2/4] Note that the Sustained Increase Monitoring plots on SACMC Epidemic Explorer are useful to interpret trends in cases over time.
[3/4] In Gauteng, the delayed data result in an uptick in cases, but the increase has not occurred consistently over a prolonged period and a “sustained increase” has not been triggered.
[1/5] CITY OF JOHANNESBURG Update: High case incidence across all sub-districts over the last week and extremely high case numbers reported in Regions C, D, E and F; cases increasing by 62%-150% compared to 7 days prior.
[2/5] TSHWANE Update: High case incidence across all sub-districts ranging between 62 and 525 cases per 100k pop. Increases of 80%-144% compared to the 7 days prior.
[3/5] EKURHULENI Update: All sub-districts are experiencing high case incidence, ranging between 76 and 264 cases/100k pop. Percentage change in cases ranges between 25%-124% compared to the week before.
[1/4] In light of the high number of cases reported, we present an update for several provinces which have an increase in cases: Gauteng has had a 12-90% increase in daily new cases, and high case incidence (105-192/100k pop).
[2/4] KwaZulu-Natal, in almost all districts, has had cases increase by 50%-187% compared to 7 days prior. Six districts have a MEDIUM risk incidence (12-24 cases/100k pop). There have been upticks in eThekwini, iLembe and uMgungundlovu.
[3/4] North West province has HIGH risk incidence of 37-87 cases/100k pop, and 2 districts have an increase in cases compared to the last 7 days: Bonjanala Platrium (47%) and Ngaka Modiri Molema (23%).
SACMC released a report on potential 3rd wave scenarios: buff.ly/3aYKW0X. Monitoring trends in cases, eg. on sacmcepidemicexplorer.co.za, is the best indicator of when a 3rd wave is likely to begin. [1/5]
In summary: In the absence of new variant, the peak of the 3rd wave is expected lower than the 2nd wave, and time from initial increase in transmission to peak is on average 2-3 months. [2/5]
Across all ages, hospital admissions are expected lower than in 2nd wave. Admissions in each province depend on seroprevalence after 2nd wave, age distribution and prevalence of comorbidities, individual responses to increasing case numbers and to restrictions. [3/5]