October 5, 2020

expert reaction to computer error delaying tracing in 16,000 COVID cases

Public Health England (PHE) have announced an error that has resulted in the delay of contact tracing in 16,000 COVID cases.

Prof Paul Hunter, Professor in Medicine at the University of East Anglia, said:

“The news about the latest ‘glitch’ is very disappointing. The very large number of cases that were not included in the daily case number reports or, apparently, not reported to the tracing system is a concern.

“Although we are told that individuals were informed of their results, these were apparently not passed onto the tracing system. For the test track and trace system to have a real impact on reducing transmission of COVID-19 it is essential that test results are communicated rapidly. We now know that people with COVID-19 are most infectious at around the time that they develop symptoms so any delay in following people up will potentially expose a large number of people. Most people know that if they have a positive test they have to self-isolate but contact from the test track and trace system is an important opportunity for this message to be re-enforced.  The test track and trace system is essential to inform contacts that they need to self-isolate and the fact that many people were not adequately informed can only increase the risk of transmission.

“In recent days there has been some talk about whether or not the epidemic in the UK is easing.  This is something that is always difficult to be sure of. We currently monitor the epidemic in several ways: the daily reports of new cases as affected by this glitch, deaths, hospitalisation and the REACT-1 study.  None of these metrics by itself is sufficient and each has advantages and disadvantages.  There have been suggestions over the past week that the epidemic has been slowing.  Clearly if these suggestions were based largely on the daily reports then they were misinformed.  In round 5 of the REACT-1 study there was some suggestion that the rate of increase may have been slowing, though in my view you have to be cautious of this conclusion, based as it is on data from little more than a single week. I do not think that we can yet really say whether or not the epidemic is indeed slowing dramatically.”

Prof Michael Hopkins from the Science Policy Research Unit (SPRU) at the University of Sussex Business School, said:

“NHS Test and Trace operates as a highly centralised system and is vulnerable to system wide IT problems. Indeed this is not the first data-related problem we have seen.

“By contrast the German model for test and trace is much more locally based and would not suffer from this kind of problem.

“Having already accepted that the German model for the Robert Koch Institute is preferable to Public Health England, further lessons could be learned from Germany, with NHS Test and Trace being restructured to provide greater local control of information flows and resources, and more resilience.”

Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:

“Due to a technical issue, which has now been resolved, in the automated process that transfers COVID-19 positive lab results into reporting dashboards and for contact tracing, 15,841 cases between 25 September and 2 October were not included in the reported daily COVID-19 cases. The majority of these cases (11,968, over 75%) occurred in the most recent days.

“It would be helpful if PHE had explained:

i) How the “technical issue” came to light?

ii) When the “technical issue” arose. Is it mere coincidence that the impact of this “technical issue” was first evident  (i.e. affecting the transfer to Test and Trace of new index cases from 25 September) on the day after the government announced, on 24 September, the launch of its new COVID-19 APP across England and Wales. Please see https://www.gov.uk/government/news/nhs-covid-19-app-launches-across-england-and-wales.

iii) Whether PHE and Test & Trace intend to publish, separately, an account of the contact tracing now underway for the contacts of the late-transferred index symptomatic cases; including their number of identified contacts who comprise: a) members of the household of symptomatic index cases (whose self-isolation should not have been impacted); and b) external close contacts (whose quarantine will have been seriously compromised because altered too late. How many days were left out of 14 for external close contacts to be in “self-isolation”?)”

Prof Rowland Kao, Professor of Veterinary Epidemiology and Data Science, University of Edinburgh, said:

“That technical errors have occurred is unfortunate, but is always a risk, especially what a massive, complex system was put into place at speed in response to an emergency.  As situations change and strategies are being altered in response, we can expect that such glitches will only continue. 

“This particular issue has two substantial implications.  All those individuals with positive results that were not entered into the system have contacts who remained an infection risk to others over this period and so we can expect that they will have already contributed extra infections which we shall see over the coming week or so.  While it appears they are now being contacted as a matter of priority, this additional strain on a system already stretched to its limit implies that further delays are likely to occur for other cases where contact tracing is needed.  These knock-on effects may have a substantial influence on the generation of new cases, over a period even longer than that.”

Nigel Marriott FRSS, Independent Statistician, said:

“The big problem with this latest data issue from PHE is that we were misled as to the underlying trend during that period.  Up to Friday it looked like the recent surge in cases had paused and there was hope of a turnaround in some places.  But with the revisions, it is clear that there is still a strong upward trend and more measures may be needed to reverse the trend. At present, the ’50,000 cases by mid-October scenario’ postulated by the CMO last month can’t be ruled out, although I suspect the number will be closer to 25,000.  What hasn’t changed is the sensitivity of the national trend to what is happening in the North.  The sooner the North slows down and reverses, the less likely we are to fulfil the CMO’s scenario”.

All our previous output on this subject can be seen at this weblink:

www.sciencemediacentre.org/tag/covid-19

Declared interests

None received.