COVID cases in England aren’t rising: here’s why
August 2, 2020
Carl Heneghan
The government has restricted movements on millions of people in England: COVID is apparently on the rise. But what happens when you start digging into the data.
I have used the following data sets to piece together the number of tests, cases and results for Pillar 1* (done in healthcare settings) and Pillar 2* (tests are done in the community).
Looking at the data for July, by the date the PCR test to detect the virus SARs-CoV-2 is reported, shows a trend for an increased number of cases detected – (from about 500 to nearly 750 a day)
If you look at the data by the date the specimen is taken the trend is still apparent (the number of cases varies when assessed by specimen date compared to the date of the reported test)
Now all things being equal, the increase in cases is about 250 per day over a month – not an exponential rise, and no sudden jump. But is this a real increase or could it be down to something else – can an increase in testing explain the rise?
See a time series of positive cases by specimen date: 31 July 2020 to look at the Pillar 1, 2 cases detected.
On the 28th of July in England, Pillar 1 reported 64 cases, and Pillar 2 reports 512 cases (576 in total). Yet www.gov.site reports on the same-day fewer cases – 547 of the two combined. Which one is, therefore right? This inaccuracy makes it difficult to make judgements as to what is happening on the ground.
On first glance it looks like the number of cases in Pillar 2 is trending up and Pillar 1 is trending down. This would suggest that the increase in hospitals – in the sickest (Pillar 1) – is staying the same; while in the community Pillar 2 testing is picking up milder asymptomatic disease.
However, what happens if you adjust for any change in testing over time? On the 1st of July – the seven day moving average of testing was 41,109 for Pillar 1 and 43,161 in Pillar 2. By the 31st July, the Pillar, 1 seven day average for testing had increased to 49,543 (a 20% increase); while the Pillar 2 had risen by much more – by 82% to 78,522 tests.
The next graph shows what happens when you adjust for the number of tests done and then standardise to per 100,000 tests. Pillar 1 is seen to be still trending down, but Pillar 2 is now flatlining. The increase in the number of cases detected, therefore, is likely due to the increase in testing in Pillar 2.
It is essential to adjust for the number of tests being done. Leicester and Oldham have seen significant increases in testing in a short time. Leicester, for example in the first two weeks of July did more tests than anywhere else in England: 15,122 tests completed in the two weeks up to 13th July.
The potential for false-positives (those people without the disease who test positive) to drive the increase in community (Pillar 2) cases is substantial, particularly because the accuracy of the test and the detection of viable viruses within a community setting is unclear.
Standardising cases per tests done, and aligning the counts in different datasets to provide the same numbers will allow a better understanding of whether cases are going up or down.
Inaccuracies in the data and poor interpretation will often lead to errors in decisions about imposing restrictions, particularly if these decisions are done in haste and the interpretation does not account for fluctuations in the rates of testing. The current reporting of the data with its inconsistencies also makes it difficult to provide accurate estimates of the case rates per tests done.
- *Pillar 1: swab (antigen) testing in Public Health England labs and NHS hospitals for those with a clinical need and health and care workers
- *Pillar 2: swab (antigen) testing for the wider population
Carl Heneghan is Professor of Evidence-Based Medicine, Director of the Centre for Evidence-Based Medicine and Director of Studies for the Evidence-Based Health Care Programme. (Full bio and disclosure statement here)
Disclaimer: the article has not been peer-reviewed; it should not replace individual clinical judgement, and the sources cited should be checked. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.
