COVID-19 in children and adolescents in Gauteng - changes since December 2020

This Vignette follows up on that released in December 2020, taking another look at the impact of COVID-19 on children and adolescents in Gauteng, relative to other age groups. Over five key periods across the pandemic the patterns have changed markedly, and there is a concerning increase in the proportion of total cases in the 0-19 age bracket since the new school year started in mid-February 2021.

GCRO’s Vignette #38, released in December 2019, showed that cases of COVID-19 amongst those who are 0-19 years old made up a small proportion (8%) of all Gauteng cases between early March and the end of October 2020. However, in December 2020 two significant things happened. First, some coastal cities saw a spike in cases amongst youth attending post-matric ‘rage parties’, followed by heightened concern over renewed spread and a tightening of lockdown measures over the holiday period. Second, various countries became aware of newly dominant COVID-19 variants, with fears that the mutating virus would be more contagious in children. This concern was cited as a reason to delay the opening of schools until mid-February.

This Vignette interrogates the available COVID-19 data again to discern changes in the patterns of infection by age in Gauteng since December 2020. It compares the number and proportion of cases in three age-brackets – children and adolescents (0-19 years old), the working age population (20-64), and the elderly (65+) – over five key periods of the pandemic. These periods, each 30 days long, are: (1) the first peak – 21 June to 20 July; (2) the post-lockdown return to school – 30 August to 28 September; (3) the period of rage events and other post-matric parties – 20 November to 19 December; (4) the second peak, driven by the new South African variant – 22 December to 20 January; and (5) the 2021 back to school period – 21 February to 22 March.

The data shows that in absolute terms, on a per 100 000 population basis, and proportionately, working age adults and the elderly have been far more heavily affected than those aged 0-19 over all stages of the pandemic. This is particularly evident in the number of cases per 100 000 in each age group. For example, over 30 days during the first peak there were 8 314 cases amongst those aged 0-19, slightly more than the 8 056 among the elderly. But 0-19s make up a significantly larger share of the Gauteng population than those 65 years and older – 31% compared to just 5%. So on a per 100 000 population basis there were only 175 cases amongst children and adolescents during that period, compared to 958 cases amongst the elderly.

On the most recent data available to GCRO, downloaded 25 May 2021 and with cases up to 22 May, the percentage of total cases contributed by those aged 0-19 remained at 8%, consistent with the average noted in the December vignette.

However, the pattern has also changed markedly in different periods.

Between late-November and mid-December 2020 the proportion of cases among children and adolescents increased significantly to some 14% on average (see figure 1). On two days, the 5th and 6th of December, the proportion spiked to over 40% of total cases.

Disaggregating cases into five year age cohorts per 100 000, and plotting these over time on a seven day rolling average basis (see figure 2 below), shows clearly that the sudden spike was driven by those in the 15-19 age cohort. This indicates that concern over COVID-19 spread at end of school parties was not limited to post-matric rage events in some coastal cities. It dramatically affected the patterns in Gauteng too, with many more adolescents going to get tested, and testing positive.

The spike in cases among teens was very short lived, and the proportion of 0-19s in the total case numbers rapidly returned to the long run average of 8% over the second peak. In spite of worries that children and youth would be more susceptible to new variants, the data for the second peak does not indicate that 0-19s were relatively more affected than before. The new South African variant, like other variants internationally, may indeed be more contagious than the original virus, and this may have led to more cases in all age groups, including those 0-19. But the evidence does not support the conclusion that there was something about the new variant that makes children specifically more susceptible. 

However, the last few months have seen a gradual but concerning increase in the number and proportion of cases in the 0-19 age bracket. The reopening of schools was delayed by a month to mid-February 2021. And at present children still continue to attend school on alternate days on a rotational basis as a precautionary measure. While the Department of Basic Education has squashed rumours that schools are about to be closed, it has in the last week instructed schools to suspend contact sports and co-curriculum activities such as choir practice. 

The data suggests that this level of caution may have been warranted. In the month following school reopening the 0-19 age group’s share of total cases steadily increased to average 10% for the period, and rising to a 14 day moving average peak of 13% in early April. Though numbers remained low at 19 cases per 100 000, they were slightly above the 15 per 100 000 seen in the return to school period in August and September 2021. The 14 day moving average dropped slightly in April and May as the total Gauteng numbers ticked up, but it has risen again to 13% in the latest data downloaded this week. 

With contributions of thoughts and edits from Christina Culwick Fatti, Julia de Kadt, Gillian Maree, Sthembiso Pollen Mkhize and Alexandra Parker.