Hi,

You can email Dr Shahram Sedghi: [email protected]

He is an expert in this field.

Best,

Ehsan

This is his link in Researchgate

https://www.researchgate.net/profile/Shahram_Sedghi

Dear Ehsan, Much appreciated. Rod

Rising numbers of deaths is a subject of discussion in Russia

What does "detailed" mean?

New Zealand provides mortality data, it lags somewhat behind (naturally). The current set is 2012 (http://www.health.govt.nz/publication/mortality-2012-online-tables). It breaks it down against demographics and ICD subsets. I'm not sure if that's enough detail....

http://www.statistik.at/web_de/statistiken/menschen_und_gesellschaft/gesundheit/todesursachen/index.html

I forwarded your question to Mrs. Marian Tervoort M.D., Ph.D. at the Reinier-Haga group of hospitals; in the Netherlands she is involved in related studies. She will contact you. ([email protected])

Hi All, Answers to your replies as follows. 

1. Already have New Zealand data and can confirm that it behaves exactly the same as Australian and UK data.

2. Detailed means monthly deaths (all-cause mortality) by spatial unit (which can be a region, authority, city, or very small area such as LSOA and MSOA in the UK). Data is needed at a minimum from 2006 to date.

3. For those studying the issue of increased deaths in their country need to urgently read the series of publications on my website www.hcaf.biz in the 'Emergency Admissions' folder. Emergency medical admissions and deaths are showing linked behaviour. Especially read the paper 'Are emergency admissions contagious?' The behaviour is consistent with outbreaks of a new type of infectious agent.

Hi All, Forgot to say that is review will be highly relevant. Cheers  Rod

Several more to be published in next few weeks. Easiest if you follow me on RG

This paper has also just been published relating to unexplained periods of higher death in Australia. Whatever is happening is a truly international phenomena.

Dear Rodney

in Italy we have mortality and hospitalization detailed data at municipality level which could allow to apply your interesting approach, but of course it would need intense work.

no  i don't.sorry

Whatever maintained we have access to data. But there is very poorly maintained death statistics in Nepal.

The US Social Security Death Index is searchable by Zip Code, if that is granular enough for your purposes.

Each in turn.

Dear Fabrizio, Monthly deaths and/or medical group admissions for each spatial unit are put into a spreadsheet, running 12 month totals are generated and a moving 12 month difference is calculated. Point of maximum difference is identified along with percentage change. About 200 spatial units can be characterised in around 2 days. Write up paper. Ideal project for junior researchers or students. My aim is to partner with others so they can scrutinise the method and suggest alternative approaches. Ideal is to start with a small subset to establish proof of concept.

Dear Tulsi, Try thr approach suggested to Fabrizio using spatial areas where the counting of deaths is likely to be more accurate.

Dear Ann, Are you able to either send me a link or extract some data to show me?

Much appreciated. Rod

I'm no expert in the use of this data, but I believe the Death Master File can be downloaded from this website: http://www.ntis.gov/products/ssa-dmf/#

Thanks Ann I will give it a go. The CDC have data for 100'ish cities and towns but are not at all helpful in providing the data in an easily accessible time series.

Dear Rodney,

For Norway this can be accessed through the National Cause of Death Registry, albeit in Norwegian langauge. One possibility is standardised mortality rate at municipality level (> 400 municipalities in Norway, 5.1 mio inhabitants). Data are aggregated for periods: 1998-2007 1999-2008 2000-2009 2001-2010 2002-2011 2003-2012, and causes can be broken down til ICD10 classes. Check http://www.fhi.no/eway/default.aspx?pid=240&trg=MainContent_6898&Main_6664=6898:0:25,7524:1:0:0:::0:0&MainContent_6898=6706:0:25,9002:1:0:0:::0:0 or send me a message. 

Kind regards Torben

Dear Torben,

Alas I need monthly data, since I am looking for the equivalent to a moving rectangular wave which appears to be a characteristic of this particular type of infectious event. Using total deaths (all-cause mortality) is usually sufficient to demonstrate the spatial spread. Small area rather than whole country data is preferred. The attached review shows the magnitude of these events at whole country level. Those with neurological disorders do however experience high mortality during these outbreaks. Cheers  Rod

Rod,

Thanks for the interesting paper. I understand the need for more detailed data. The National Cause of Death Registry do of course hold those data, but I'm afraid that you will then have to buy (!) a specified file, which we have been doing for our research on fatal injuries. Those files are not cheap, and usually you need a ethical board review. I would suggest that you contact the National Cause of Death Registry directly if you're interested, by the link in the above communication.

Good luck!

Torben

OK will pursue that option. As you can imagine attempting to do this research across multiple countries could get very expensive, hence my need for local partners. Cheers Rod

Dear Rodney

Brazilian mortality data (and from other health systems) are available on the DATASUS web site, through databases DBF compressed files of the “Declaração de Óbito” (death certificate) from 1979 to (currently) 2013 (ICD-10 since 1996). Among another variables, there are the date of death, municipality and neighborhood (worst registered) of death occurrence and residence of the person. These files are compressed by the Brazilian Ministry of Health (.dbc format) and may be expanded to .dbf (and also directly tabulated and a subset extracted to .dbf in a really efficient way) by the software TabWin, available on the same web site. Related documentation an laws are also available in the same page. I think it is all in portuguese.

Data may also be tabulated by the TabNet, an on-line tool for that, with less options than TabWin, but still a lot of. There you can get the number of deaths by municipality (smallest territorial subset) of death occurrence or subject residence and month, for instance, as you see on the screenshots below. 

I have attached a compressed (.rar) DBF file with one of the database files (for the State of Santa Catarina, 2013), and a PDF file informing its structure.

Hope this help, and sorry for the many english mistakes  (probably, really don't know :) ).

Cheers, Fúlvio

Here the links (in portuguese):

Health systems databases and other information available to download: http://www2.datasus.gov.br/DATASUS/index.php?area=0901

Mortality data and related documentation (if you click on “SIM”, Mortality Information System): http://www2.datasus.gov.br/DATASUS/index.php?area=0901&item=1&acao=26&pad=31655

Database on mortality (since the ICD-10 classification) structure: ftp://ftp.datasus.gov.br/dissemin/publicos/SIM/CID10/Docs/Estrutura_SIM_para_CD.pdf

In order to download the TabWin application and its definitions: http://www2.datasus.gov.br/DATASUS/index.php?area=060805

On-line tabulation (TABNET)

Mortality data:

1. http://www2.datasus.gov.br/DATASUS/index.php?area=0205&VObj=http://tabnet.datasus.gov.br/cgi/deftohtm.exe?sim/cnv/obt10

You have to choose an option to show the map

2. http://www2.datasus.gov.br/DATASUS/index.php?area=0205&VObj=http://tabnet.datasus.gov.br/cgi/deftohtm.exe?sim/cnv/obt10

3. http://tabnet.datasus.gov.br/cgi/deftohtm.exe?sim/cnv/obt10sc.def

4. http://tabnet.datasus.gov.br/cgi/tabcgi.exe?sim/cnv/obt10sc.def

Dear Fulvio, Can you do me an enormous favour and see if you can find someone in Brazil who would be willing to analyse the data. When you know what to do the analysis is very simple spreadsheet manipulation. I suspect someone could do a macro to automate what I do manually. The problem is that I have a huge backlog of partly written papers and have to do the work in my spare time. Had a look at Brazilian hospital admission trends some time ago and they seemed to indicate that the same thing was happening in Brazil. Could even share one of the part written papers to give people an idea of likely layout, etc. Much appreciated  Rod

The real value of death statistics is related to their causes.  This is retrieved from  death certificates . What could be the accuracy of death certificates filled up with no control provided by at least  - autopsy  based  samples ? 

Dear Arthur, I have demonstrated that the increase in both deaths and medical admissions associated with these infectious-like events is diagnosis specific. Hence while analysis of all-cause mortality picks up the timing of the events the obvious next step is to analyse trends for specific conditions/diagnoses. I have done this in a preliminary way in several of the papers listed in the Emergency Admissions folder at www.hcaf.biz which can be downloaded. Have a quick read. Cheers Rod

The attached book chapter may be helpful. Type in 299174 to open a printable version. I have purchased the right to distribute this chapter.