The complete mess of Bangladesh government Covid-19 statistics. Where are the missing 38,000 positive tests?
What is going on with the Bangladesh government statistics? Yesterday, we wrote about how the World Health Organisation (WHO) documents revealed significant gaps in the data — which could have real life implications for government policy.
Today we look at the wide inconsistency between statistics published by two different parts of the Bangladesh government — from the Directorate General Health Services (DGHS) which provides daily briefings on the numbers of new Covid-19 positive tests and deaths, and the Institute of Epidemiology, Disease Control and Research (IEDCR), the government’s specialist agency in infectious diseases which hosts a website providing detailed breakdown of the government data.
The data, however, simply does not correspond.
On Wednesday, as normal, Dr Nasima Sultana, the additional director general (administration) of DGHS announced the country’s daily Covid-19 statistics.
There had, she said, in the last 24 hours been 3,803 new positive Covid-19 tests, bringing the total to 102,292.
On the previous day, June 17th, she had said there were 4,008 new cases; on June 16th that there had been 3,862 new cases and on June 15th that there were 2,099 new cases. Here this is what this information looks like in table form.
However, none of the figures match with the last three days of data provided on the IECDR website. And not just by a little — by far.
The IEDCR website has a link at the top of the site that provides data on numbers of confirmed tests, set out geographically. On June 15th and 16th, it linked to documents listing the numbers of total positive tests for each of the country’s districts and divisions — with Dhaka City in a separate category. On the 18th, the website contained an undated file which contained the same information, but did not include any information on Dhaka City. The table below extracts the totals from each of these documents.
As you can see there are three serious inconsistencies.
The first — which was touched on in the post yesterday — is that total numbers of positive tests of 52,830 and 55,910 on the IEDCR site for June 15th and 16th are almost half those which are provided by the DGHS, on the same days — 90,619 and 94,481. One might understand a mismatch of a few hundred positive tests or so (if for example they relate to numbers of confirmed cases at different times of the the day) — but not differences of 38,000!
The second is the difference in the total numbers on the IEDCR website between those given on 15th and 16th June — that is to say the difference between 52,830 and 55,910 which should represent the new tests identified in the previous 24 hours. The difference is 3,080. But DGHS announced that there were 3,862 new cases on the 16th — a difference of 782. Where did DGHS’s additional 782 confirmed tests go — and if they did exist, which part of the country were they from?
And the third difference concerns the DGHS announcements on June 17th and 18th of 4,008 and 3,803 new confirmed tests throughout Bangladesh. This is a total of 7,811 new confirmed tests.
As noted above, on 18 June, IEDCR published a document (which excluded data for Dhaka City). This shows an increase of 11,508 tests between June 16th and June 18th. This is 3,697 more confirmed tests than DGHS’s announcement. And this does not even include the number of Dhaka City’s positive tests.
Again, where did DGHS’s additional 3,697 confirmed tests go — and if they did exist, which part of the country were they from?
What is going on? What is the correct number of positive tests? And in which part of Bangladesh are they from?
//David Bergman
