More confirmed Ebola deaths than cases?

Posted onAuthorIan M Mackay, PhD (EIC)

A strange number pattern has appeared in the Democratic Republic of the Congo’s (DRC) Ministry of Health (MOH) Ebola virus disease (EVD) Reports for a couple of the affected health zones. Some health zones (HZs) are reporting more confirmed EVD deaths than cases of people infected.

This is probably just a numbering mistake. I don’t mean to throw shade on the hard work this team has been doing in collating data to provide the world with a daily report of confirmed and fatal EVD cases in the DRC. They’ve done a great job. Keep in mind – we didn’t get regular daily reports from West Africa so the DRC has stepped that level of data up.

Nonetheless from the DRC MOH report dated 14th of April, onwards, there have been more confirmed deaths listed in Butmembo HZ there are confirmed EVD cases. Similarly, from 4th of May onwards, Biena HZ has seen 129% of cases die. Obviously these are errors.

I’ve excerpted the latest table from a MOH report PDF below…

Table from DRC Ministry of Health Ebola update dated 6th May 2019 with data to 5th May 2019.[1] Note that there are more confirmed deaths than confirmed cases in the Butembo and Biena health zones.

As a hobbyist in this area, I am very well aware that mistakes creep in. I fully understand that even official sources can have bad days. Perhaps this is another reason why hobbyists should be able to get access to current and past outbreak numbers; a form of off-site validation.

Alas, for this outbreak, only last 20 days worth of reports are available publicly. I think there is a strong case to be made of the World Health Organization to make public a consistently formatted of line list (see image below; for example, one day’s report data per line) for every outbreak or infectious emergence that it’s already tracking. Those data are probably already in use internally anyway.

A line list. Each line lists all the data from a single report or it could be for a person, country, state – depending on how your organise your list. This is a snippet of the list I personally curate for the current Ebola virus disease outbreak. I use Excel and organise the sheet so that I can easily update the graphs and charts I have on the same sheet.

This seems an important application of the current open access mentality. Scientists want to see it and the public should have access to it.

It’s something the WHO already support for its data and activities and yet it often seems to only extend to producing graphs of the data in lengthy reports rather than openly sharing the data itself for others to help interpret.[2]

But back to the issue of reporting errors. When records don’t make sense or seem to be in error for a while, questions are raised about how well various aspects of a response process is coping with an outbreak.

References

  1. EPIDEMIOLOGICAL SITUATION IN THE PROVINCES OF NORTH KIVU AND ITURI
    https://us13.campaign-archive.com/?u=89e5755d2cca4840b1af93176&id=21a6fc7bce
  2. WHO Policy on Open Access
    https://www.who.int/publishing/openaccess/en/