I am Amanda Glassman is senior fellow and executive price president at cdc. Today we will be talking about death and data in low and middle Income Countries from covid19 and all other causes. Accurate complete and timely data on mortality probably the single most powerful policy tools we have today to mitigate the Economic Impact of covid19 but what is also clear is that we have not collectively built the systems necessary to record peoples death in ways that enable more and to protect livelihoods. This failure does not only spawn individual experiences that creates unnecessary suffering. How is it possible to assess whether Subsaharan African countries are winning the fight against covid if a large share of total deaths and noncovid years go unreported or how to judge the trajectory of pakistan during covid19 if no mortality data of any kind is being ripped ported in the Public Domain according to dub a joes 2019 mortality report. Today will be talking about one strategy and mortality measurement to report on deaths in realtime. Systems cant rely on what deaths of the community are left at the camp are good examples of countries to start off with better data but struggled to produce accurate and timely data and will be focusing on the big picture. What can we do to develop modern systems to accurately record deaths in realtime the weather Innovative Strategies are effective or would he like to see from the global conducive to support low and no income government attack and we visit the issue further as it deserves . We are lucky to be joined today by four world experts in the field the Vice President for Civil Registrations vital strategies the adviser of the brazilian government on these issues for vital strategies semi era osmolik they director general for data and analytics for impact at the World Health Organization was Just Launched their own Data Initiative and Aaron Nichols who leads global Civil Registration and Vital Statistics improvement at the west centers for disease control. Lets get started and i will turn it over to you. Thank you so much all of you. Thanks amanda and if youll give me one moment to share my screen. And put us into live mode here, presentation mode. Okay, good morning and good afternoon or evening good evening and thank you for joining us here. This is an important topic and it really goes to the heart of the adage adage of know adage of no europe but then i can know your response which is a phrase that was coined over a decade ago but how do we know the covid pandemic. We believe every country should know the scope and scale of the pandemic in realtime to the greatest extent possible to shape the datadriven response and here we have our w. H. O. Global dashboard. It does represent deaths and confirmed cases due to covid19 but we need to be careful in terms of understanding and interpreting these numbers and understanding the measures particulate dessa mortalities and data that comes from lowland and sometimes in middleIncome Countries. As amanda indicated this presentation is going to focus on one measure of excess mortality that we fill captures the full scope and scale of the poll of the pandemic. It should be no news that sense the early days of the pandemic there have been shortages and concerns overtesting over its scope and its scale. Since early days there have been worries overtesting capacities throughout Subsaharan Africa not to mention from countries like the united states. Lets look at the implications of this testing context a little more closely when it comes to interpreting big numbers. One point i would like to make is testing a system like it is actually an essential service in the context of the pandemic. In the lancet recently in the absence of the vaccine or highly effective treatment widespread g transmission and death. I think this is something with which we can all concur and yet the reality on the ground particularly in low and middleIncome Countries is highly variable and its actually often inadequate and has been inadequate in relation to the need. Given the scarcity testing has tended to be focused on symptomatic cases of hospitals or to identify cases in clinical settings. The question here then is what are the invocations for that scope and scale for the defection of the total burden of the pandemic and as we have seen from this clipping from the guardian. Com this undercounts cases in the Community Meaning that w. H. O. Warned that it could be the tip of the iceberg in terms of the actual number of cases being detected. But what about death . The key picture that im trying to paint for you today is what a murky picture we have because we do not have the data in hand in many places. The w. H. O. Dashboard tallied over 960,000 confirmed covid deaths as of this past wednesday but again we need to unpack a little bit those numbers that are coming from countries that lack both robust and resilient Vital Statistics and Civil Registration programs and the cause of death data systems usually rooted in the Health Sector that are not yet capable of delivering and providing timely data on the cause of death. In those kinds of context there to issues that arise in particular. The first pertains to getting the covid19 death numbers right themselves. We are just trying to get covid19 debts and how do we get it . Thats going to depend on the adaptation, not adaptation but the adoption of important guidance issued by the World Health Organization in terms of how to correct a certified and code a covid19 death either suspected or confirmed. This guidance which i have looked at and get a little complicated when it concerns comorbidities for example where there is covid president of comorbidity. The upshot is all of this guidance need to be disseminated, adopted and manifest in the practices of coding positions in order to have a robust measure of covid mortality. The second issue however that i want to call attention to is with an exclusive focus on covid deaths and mortalities namely deaths that occur in locations far away from hospitals where they cannot be certified and the stark reality in her opening remarks. We leave out those deaths that arise from disruptions to overextend Health Systems and we leave out death that occurs because people delay seeking hospital care for fear of infection or if they fear they might be in the good are being separated from families and then taken away. It also often excludes emergencyroom death for those broad index of hospitals who are not often counted in hospital tallies. All of these forces are acting in many places where the majority of deaths as amanda pointed out even before the pandemic were occurring at home. In many african contexts for example this really contributes to a murky picture. We only have hypotheses at the moment about the syndication that we have may be comparatively fewer or unexpectedly low cases at the moment of covid19 and covid19 deaths occurring during the pandemic i should say across much of the continent. Is this really a factor of limited Testing Program so we dont have a window into the pandemic . Might the virus had come early already in because Surveillance Systems were not in place we might have missed the arrival of the virus . It may have later attenuated arrival and because less mortality because of competing causes of death at an older age in particular and also there may be a mitigating sense of the residential patterns of the population. Some combination of these factors in the answerer when we need to think in terms of parsing out mortality by cause particularly at the Community Level it becomes enormous and complex. This brings us then to the idea of access to mortality. We think the this complementary measure is really simple and fairly comprehensive way of capturing in a timely manner the full human cost of the pandemic. Im going to explain this mortality in two parts for the first part of measuring is to focus on the enumeration of all deaths now regardless of cause by age sex and location. We are tracking today that current levels of mortality. The second piece of measuring excess mortality is to establish a baseline of expected death for the same at the epidemiologic week and the same location to your go or an average of the Previous Year and a measurement of the gap between todays observed mortality and that is then expected in fact what we call excess mortality. This mortality can be attributed not only to covid direct we but also to the causes of death as the results of the sorts of factors that i outlined a moment ago. These graphs may be somewhat familiar and we have a graph from switzerland by age group on the right and on the left published by the economist. Major deaths are using access mortality visualizations such as those or these two to represent the pandemic. Thats all well and good but again bring us back to our main concern here what about places on the globe that are not being able to routinely and resolutely report the data particularly from immunity. At the moment at least 13 countries and there are more there are more everyday many with the support of philanthropies and other partners are leveraging existing sources of data or creating new ones to measure finality in part they are relying upon the technical package that we have. In partnership with World Health Organization leadership with Cdc Foundation and cdc leadership and regional partners in africa and asia. This technical package is assisting countries along the spectrum of system readiness if you will. I will show up graph from brazil one of the countries thats producing rapid mortality surveillance and the point i would like to say here and youll hear more about brazil surely rapid mortality surveillance was actually an innovative use of existing public data. Brazil and other countries such as colombia and even peru have the availability of data and have had to make fairly minor innovations i could say compared to where else we have been working in order to establish the rapid mortality surveillance and the measures of rapid mortality that rapid mortality surveillance can produce. Visiting a little bit to a discussion of lowIncome Countries what they find is that in these settings systems dulac high coverage and completeness and a lack of timeliness. This makes solutions for coming up with ways of measuring incident deaths more complex slower to implement and more resource intensive. With the notable exception of south africa i think this is generally the case in much of africa and somewhat Southeast Asia where governments have a need to measure the mortality from the community because theres such a such a Big Community mortality burden that exists regardless of the pandemic and eligibility in order to form a complete picture. In such circumstances are early experience has shown that the communitybased surveillance piece of the Technical Work that needs to be accomplished has been a bit more challenging than getting facilitybased surveillance up and running but in that regard strategies have been working with a few countries including rwanda bangladesh and supporting others to leverage routine Health Systems, sorry routine Health Information systems and boost reporting of deaths to a weekly basis and this is mostly for those of you who are familiar with Health Information systems leveraging functionality within the District Health management system. Communitybased rapid surveillance in which deaths occurring are actively detected and reported on we are supporting the government of colombia and the government of bangladesh to undertake this work. In colombia at the intention is to reach remote and harder to access parts of the population in parts of the country and in bangladesh where leveraging an expanding model of vital event notification to be able to identify deaths on a more rapid basis than those previously been the case under the Registration System there. As we are beginning to support countries to produce this data of course the question arises about their youth. In addition to the advocacy and serving as a corrective for misinformation excess mortality data can be viewed in conjunction with other data to assess geographic disparities for example or perhaps even to chart the lagging because mortality is a lagging indicator but a least impact of Public Health and social him matters that ideally have an impact on both a number of cases and hence the number of deaths that are observed. If indeed there is available cause of death data it may be possible to understand the excess mortality more in terms of the specific causes of death that comprise it. Certainly the majority will be covid and suspected covid but the quotient that is left over the additional excess due to other causes may be due to some important system break downs in the Health System and knowing the specific causes can be of help to pinpoint action for dressing those situations. And lastly and this is more significant than it may seem and i think we may have had the opportunity to discuss this further they can shore up death registration during the pandemic as rapid mortality surveillance that can shore up the pandemic and in fact we are working with at least one country intending to undertake rapid mortality surveillance for precisely this reason in addition to getting a handle on the pandemic. Summing up i just want to point out that we know that knowledge is key to the response that a focus on solely covid19 diagnoses and deaths are necessary and we have to unpack it but in a position to understand the true magnitude of the pandemic and measuring mortality is one very familiar and it leaves relatively straightforward thing to do in order to fill the space knowing the epidemic in terms of crucial statistics. I would argue it adds to the urgency of. Covid19 crv s. Improvement. These are i believe the longterm solutions for a resilient system that can indeed meet the needs of future Public Health emergencies and the directors Civil Registration services in kenya summed it up nicely at. And these are the words i will leave you with. She said we have also realized the need for a brazilian crv a system in an emergency such as covid19 to meet the needs not only of the population but vulnerable and marginalized populations in the country. With that i would like to conclude and thank you and hand it back to amanda. Thanks so much. Thank you philip. Its a great overview to get us kicked off. In doing work in brazil looking at this excess mortality measurement and i hope we not only find out that the findings from brazil but also how youve seen it used in policies. Thank you amanda. Good morning or good evening to everyone. Thank you philip for helping with the presentation on excess mortality in brazil. We have talked about methods and how we did this and the Health Department and ongoing work on excess mortality. It is the data source where were we combined the two data sources we have mortality information and get that Historical Data and history in 2015 and 2020. We use Natural Causes of death and crvs data. We compared with the cbr is from 2019 and we had each age group and we applied it to you can see on the right that deaths between the blue line in the red line. This is 2020 in the red line is existing so we had quite a correction when it was needed. Thats how we did this. We used it was our reference and small areas we beat the average. So excess mortality, weeks with values below the projected baseline were disregarded. We have levels of analysis states, sex and age, two age groups and to obtain the excess death we used it the most desegregated age and sex by state or by country. So in brazil we dont have exactly raised could we use the definition in the census and using statistics data. We have a redistribution of missing data by skin color but we work with categories of white lack and pardos, and a mix of a population and analysis by brazil has the consortium of the Health State Department in blue which is the mortality for 2020 and the Historical Data. In red we had the excess mortality. We have the correction applied with death [inaudible] you can see we started in the middle of march peaking at the end of may in beginning of june and it starts to decline because of [inaudible] we dont believe its going down so fast and we are discussing it with the administration. Next, please. It is an example of different phases in brazil. We call it a late pandemic. We just use bolivia and you see it starts to peak in june and the peak of excess mortality and the end of july. This is still going on in the region. On the right you see an early epidemic because they are big surges in the Southeast Region so you see increases in midmarch peaking in the middle of june and now we feel its not so clear here but its starting to increase again access mortality because mortality speaking again. We have following other states closely as well. And the last few numbers and the right resented so members of excess mortality and the most populous states in the country with a higher number and excess mortality death