An error in an Excel spreadsheet has challenged contact tracing in the UK

A glitch may have contributed to the spread of the Covid-19 pandemic across the Channel. And it is not the first time that Microsoft's spreadsheet has created problems, also thanks to the poor computer knowledge of those who use it

(photo: JUSTIN TALLIS / AFP via Getty Images) The government agency that oversees the response the pandemic in the UK Public Health England missed the figure of 15,841 cases of Covid-19 in the country, which were excluded from the national total due to an Excel system error. Although all those who tested positive for the virus were informed by the British health authorities, the failure to upload the nearly 16,000 patients to the national database could therefore have contributed to the spread of the disease.

In fact, in this way, about 50 thousand potentially infectious people - since they could have come into contact with lost cases - were not included in the tracing of the infections and were not instructed to go into isolation. The missing cases, originally registered between 25 September and 2 October, were made public on Monday 5 October.

According to a Phe spokesperson to BBC journalist Leo Kelion, "the first investigations indicate that the problem was caused by the fact that some files containing positive test results exceeded the maximum size and therefore did not have been uploaded to the central system ”. Although the British health agency has not confirmed (nor, for that matter, denied) the involvement of Excel, newspapers such as the Guardian have traced the error to the famous Microsoft spreadsheet.

What went wrong

Phe is responsible for collecting test results for Covid-19 from public and private laboratories, and for publishing daily updates on case counts and tests performed. Despite significant improvements since the beginning of the pandemic, the system is very far from automation: much of the work is still done manually, with individual laboratories sending files containing their results to the central archives.

In this case, a lab sent its daily report to Phe in the form of a CSV file, a simple comma-separated list of values. The error occurred when, to be added to the central database, the report was uploaded to Microsoft Excel, which has a limit of 1,048,576 rows (or even, when not updated, of 65,536). If the Csv file is longer then the lower lines are cut off. It is unclear which version of the spreadsheet Phe is using, but apparently the row limit has been reached, and nearly 16,000 cases have been rejected by the database.

The agency claimed to have fixed the glitch (although it did not explain how) and to have relayed details of confirmed back cases to UK contact detectors as of 1pm (local time) ) on Saturday 3 October.

Excel errors cost money

According to the University of Connecticut professor James Kwak, there are many problems with Excel: “There is no way to trace the origin of your data, there's no control trail (so you can override the numbers and not know it) and there's no easy way to test your worksheets, to begin with. Because it's so easy to use, creating even important spreadsheets isn't limited to people who understand programming and use it in a methodical and well-documented way. "

It is not the first time that Microsoft's spreadsheet software - one of the most popular business tools in the world - has become implicated in errors that can be costly, if not downright dangerous. In 2013, for example, a mistake in an Excel from the US multinational bank JP Morgan masked the loss of nearly $ 6 billion, due to a cell wrongly divided by the sum of two interest rates instead of the average.