Read on Twitter
1/ Data is critical to the fight against #COVID19. And the most basic data of all are the numbers we are all obsessed with, showing global deaths in real time. Ever wondered where those numbers come from? Read on…. https://ourworldindata.org/covid-deaths ">https://ourworldindata.org/covid-dea...
2/ Someone dies. For family and friends, a tragedy. For health professionals, a question - why? In a hospital, a doctor will record the cause of death, elsewhere it might be a community health worker, or local leader. Either way, it’s not always obvious https://www.nuffieldtrust.org.uk/news-item/measuring-mortality-during-covid-19-a-q-a">https://www.nuffieldtrust.org.uk/news-item...
3/ Social norms and inequalities matter here - e.g., in the 90s some AIDS deaths weren’t recorded as such because of stigma and practical issues such as rules around life insurance. And if there’s a crime involved, the law might decide.
4/ Once the judgement is made about the cause of death information is captured in words on a death certificate & turned into numbers by someone punching code into a computer. This is a skilled job & a country needs thousands of coders for a working system https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4709797/">https://www.ncbi.nlm.nih.gov/pmc/artic...
5/ ICD-10 is an international system of 70,000 codes for diseases and causes of death. COVID-19 has been given emergency codes U07.1 or U07.2, depending on whether a test has been conducted. https://www.who.int/classifications/icd/en/">https://www.who.int/classific...
6/ (ICD-10 is a terrifying litany of all that can go wrong with the human body. There are codes for injuries sustained in a spacecraft (V95.4), five different codes for being poisoned by fish (T61.0,1.1,1.2,1.8,1.9), & tens of thousands of other reasons to never leave the house)
7/ Once coded & in the system the river of data flows from local to regional to national level. Ideally it’s all digitised, but sometimes connections are faulty, hardware goes wrong, or remote areas just aren’t connected to the system. None of this is easy https://devinit.org/publications/digital-civil-registration-and-legal-identity-systems/delivering-digital-crvs-and-legal-identity-systems/">https://devinit.org/publicati...
8/ So far, so good. But here’s the twist - this happens for LESS THAN HALF OF ALL DEATHS in the world. Only 84 countries have what the World Health Organisation classifies as ‘reliable’ data from death registration. https://www.who.int/gho/mortality_burden_disease/registered_deaths/text/en/">https://www.who.int/gho/morta...
9/ How can this be? Don& #39;t we all love data now? It seems no one cares enough to fund the system. Governments don’t get elected for promising death data. Donors fund surveys, not systems, when they want data. So we don’t know why most people die
https://openknowledge.worldbank.org/bitstream/handle/10986/18962/883510WP0CRVS000Box385194B00PUBLIC0.pdf?sequence=1&isAllowed=y">https://openknowledge.worldbank.org/bitstream...
https://openknowledge.worldbank.org/bitstream/handle/10986/18962/883510WP0CRVS000Box385194B00PUBLIC0.pdf?sequence=1&isAllowed=y">https://openknowledge.worldbank.org/bitstream...
10/ Think about that. Billions of dollars have been spent on public health programs over decades but we still don’t know why people die? All the campaigns & appeals & funds to reduce maternal mortality, or under-five deaths, & no one has bothered to properly find out the numbers?
11/ This is a pretty sorry indictment on any so-called commitment to data and evidence. But maybe now we could hope that the new-found appreciation for data translates quickly into a new push to invest in knowing why people die, so that we are better at stopping it from happening
