COVID-19 – What does the latest “surge” in numbers mean and what is the impact of vaccinations?
COVID-19 is taking over the headlines once again as cases seem to be on the rise. Alongside this, there have been the inevitable discussions among politicians and the media about whether this “surge” in cases will result in restrictions being reintroduced. But are these discussions warranted? Here, PHASTAR’s Vice President for Statistical Research and Consultancy, Jennifer Rogers, takes a look at the latest COVID-19 numbers in more detail and explores the possible impact of vaccinations.
On 21st October, the daily reported number of new COVID-19 cases was 52,009 with the daily reported number of new cases being consistently over 40,000 for the 9 days previous. The last time cases were consistently over 40,000 for this many days in a row was back in January as we were coming out of the last winter wave. The UK Health Secretary, Sajid Javid, has also said that COVID-19 cases could hit a record 100,000 a day. Now, I tend to view these kinds of statements with a sensible amount of scepticism, but there certainly seems to be a suggestion that COVID-19 cases may be on the rise.
However, there is a big fundamental difference between now and January – vaccinations. The COVID-19 vaccination programme has been hugely successful, with the latest data showing that 79% of those aged 12 and over have received two doses of the vaccine, compared with under 1% in January. We know that vaccines reduce the risk of COVID-19 transmission and we also know that vaccinations have been particularly effective in reducing the risk of COVID-19 hospitalisation and deaths.
We can investigate the impact of vaccines by examining what proportion of COVID-19 cases result in associated hospitalisations and death and compare this to this time last year, before any COVID-19 vaccination announcements had even been made.
Table 1 shows a snapshot of the available COVID-19 Daily Summary for the month of October captured on 21st October (https://coronavirus.data.gov.uk/). To allow for noise in the data, we will consider the rolling 7-day averages.
Table 1: COVID-19 Daily Summary Data
To calculate an estimate of the proportion of COVID-19 cases that lead to hospitalisation and death, we assume a 7-day lag between cases and hospital admission and a 14-day lag between cases and death. Figure 1 then shows an estimate of the percentage of COVID-19 cases that have led to hospitalisation and death since the start of September 2020 (which was the start of the academic term and arguably the beginning of the 2020/2021 winter wave).
Figure 1: Percentage of COVID-19 cases that result in hospitalisation and death
We see that the percentages in Figure 1 fluctuate and are noisy. Note that the risk of hospitalisation and death is not constant across all demographics. We know that those who are elderly and/or have comorbidities are at an increased risk of serious illness and death. There is no reason for us to expect the profile of those contracting COVID-19 to stay constant, so it is expected that we may see fluctuations in the percentage of hospitalisations and deaths. Indeed, the start of September saw an increase in COVID-19 cases among school pupils and students, who are at a lower associated risk of serious and illness and death due to their age. This kept COVID-19 deaths slightly lower than may have been expected, but as more and more older people contracted COVID-19 throughout the winter, the proportion of cases resulting in death increased.
What does seem to be clear from Figure 1 is that the proportion of COVID-19 cases that lead to hospitalisations and death have been consistently low since July (when approximately 60% of those over the age of 12 had received both doses of the vaccine) compared with earlier months. Looking at the start of October 2021, around 2% of COVID-19 cases result in hospitalisation and 0.35% result in death. Compare this to the start of October 2020 where around 7.5% of COVID-19 cases resulted in hospitalisation and 1.35% resulted in death. At the peak of the 2020/2021 winter wave, over 10% of COVID-19 cases resulted in hospitalisation and around 3% resulted in death.
Is this reduction in the percentage that we see due to the vaccine? I have already said that the risk of hospitalisation and death is not constant across the whole population and I have said that the COVID-19 case profile changes through time. So maybe this observed decrease in the percentages is due to younger and healthier people contracting COVID-19 now compared to earlier? We can explore this in more detail too by looking at the National Flu and COVID-19 Surveillance Reports (https://www.gov.uk/government/statistics/national-flu-and-covid-19-surveillance-reports-2021-to-2022-season). Looking at the week 42 and week 41 releases, we are able to compare the case rates in different age groups in the week commencing 11th October 2021 to the week commencing 12th October 2020. For these weeks, the weekly confirmed COVID-19 case rates per 100,000 are presented in Table 2.
Table 2: Weekly confirmed COVID-19 case rates per 100,000
To make this data a little easier to digest and to better examine the spread of COVID-19 cases across the different age groups, we can look at the ratios of case rates. Taking the age group “60 to 69” as the baseline, Table 3 shows the ratios of weekly confirmed COVID-19 case rates.
Table 3: Ratios of weekly confirmed COVID-19 case rates per 100,000 with “60 to 69” as baseline
To aid interpretation of this table, consider the age group “10 to 19”. In the week commencing 11th October in 2020, for every 1 COVID-19 case in the “60 to 69” age group there were 2.8 COVID-19 cases in the “10 to 19” age group, so almost three times as many. Comparing this with the week commencing 12th October 2021, we see that for every 1 COVID-19 case in the “60-69” age group there were 5.5 cases in the “10 to 19” age group, so five and a half times as many. Indeed, we do see that the spread of COVID-19 cases in 2021 does seem to be shifted more towards school age children compared with older age groups. Note, however, that we actually see a reduction in COVID-19 case rates in the “20 to 29” age group in 2021 compared with 2020. So, whilst it’s likely that age may have had some impact on the percentage of COVID-19 cases resulting in hospitalisation and death, I’m confident that vaccinations are having a positive impact.
There are a couple of extra points that I think warrant a short discussion. The first being the age-old debate of relative versus absolute risk. As a statistician, I have slapped the wrists of many a journalist for only reporting relative risks and not considering absolute numbers so it would be shameful of me to ignore that here. Yes, vaccinations do seem to be breaking the link between COVID-19 cases and subsequent hospitalisation and death, but if cases continue to rise and rise, so too will hospitalisations and deaths even if the proportions stay low. How do the absolute numbers compare now with October 2020? The latest data we have for 7-day averages is 14th October 2021 where there were 44,532.3 cases, 902.1 hospitalisations, and 115.6 deaths. On 14th October 2020, the 7-day rolling averages were 17,272.6 cases, 962.3 hospitalisations, and 134.7 deaths. So even though the number of cases is vastly higher in 2021 compared to 2020, the number of hospitalisations and deaths do remain lower in comparison, but they’re not far off. Case rates do appear to be increasing, but if we look at the figures for the last few months, this isn’t the first time we have seen case rates increase, followed by a decrease. It’s probably too early to determine if cases will continue to increase, but it’s something that we should perhaps be slightly aware of, especially as we head further into winter.
We can investigate what case numbers might give us the numbers of hospitalisations and deaths that we saw at the peak of the 2020/2021 winter wave given the current percentages of 2% hospitalisations and 0.35% deaths. The highest number of hospital admissions observed was 4,583 on 12th Jan and if 2% of COVID-19 cases now result in hospitalisation, we would need to observe around 230,000 daily cases to see those kinds of hospitalisations numbers again. Considering deaths, the highest number of daily COVID-19 deaths was 1,359 on 19th January and if now 0.35% of COVID-19 cases result in death, we would need to see around 390,000 daily cases to see similar daily case numbers.
The second is not really a comment, but more of a plea for data transparency. We know that vaccinations reduce the risk of COVID-19 and reduce the severity of disease. A booster vaccination programme is currently underway, and this will be vital in maintaining the levels of immunity that we have seen over the last few months, keeping COVID-19 death rates down during the winter months. But we are yet to see any useable, informative data on the booster jab rollout. I was interviewed by National World (https://www.nationalworld.com/health/coronavirus/covid-booster-vaccine-rollout-experts-criticise-lack-of-data-and-say-public-is-not-getting-the-full-story-3420679) where it was highlighted that although NHS England are currently releasing statistics on the number of COVID-19 booster jabs that are being administered (https://www.england.nhs.uk/statistics/statistical-work-areas/covid-19-vaccinations/, 4,269,111 as of 21st October), broken down by age and region, these figures fail to report how many people fall into these groups (i.e. the denominator), so there is no way of assessing the performance of the rollout. It is vital that more detailed data is published so that the public can understand how the rollout is progressing, especially as we enter the 2021/2022 winter wave.