Covid_19: Improve data quality by harmonizing the reporting - "date of test" vs "date of result"

Created on 26 Mar 2020  路  11Comments  路  Source: openZH/covid_19

It looks like GE is reporting based on the "date of test" whereas other cantons use the "date of result". At least it should be clarified, which canton uses which standard. And in a second step e.g. for GE the complete report should be saved on daily bases, so that numbers based both methods can be displayed.

documentation

Most helpful comment

"date of result" is used in BL, GR, SZ, UR, ZG, ZH and FL. BE uses the "date of result", but all cases are collected until 17:00 are reported next day at 8:00.

"date of test" ("date du frottis") is used in VS.

Additional I have requested if the numbers are domicile based or based on hospital bases / current location.

"domicile" is used in BE, BL, GR, SZ, NW, VS, UR, ZG and FL. GL, LU is mainly domicile based.

All 11 comments

The Informationsdienst Kantonaler Krisenstab of BL has confirmed via email to use "date of result".

"date of result" is used in BL, GR, SZ, UR, ZG, ZH and FL. BE uses the "date of result", but all cases are collected until 17:00 are reported next day at 8:00.

"date of test" ("date du frottis") is used in VS.

Additional I have requested if the numbers are domicile based or based on hospital bases / current location.

"domicile" is used in BE, BL, GR, SZ, NW, VS, UR, ZG and FL. GL, LU is mainly domicile based.

Could you make a commit / PR that summarizes our understanding of this data for each canton? And what BAG is recommending for reporting?

TIMING.md maybe?

Thanks @zukunft , your research is truly valuable. I think its a good idea to create an overview over these differences @baryluk, we'll look into this shortly.

ZH relies on date of result and on domicile for confirmed cases. For hospitalized patients however, this differentiation is not relevant. Patients are counted irrespective of domicile.

@jb3-2

BS uses "date of result", as I see from the press releases. Regarding domicile, I have created a pull request that clarifies which columns cover residents of a canton and which cover all cases irrespective of residence, see https://github.com/openZH/covid_19/pull/355, to be reviewed / merged by @tlorusso

@baryluk @zukunft Please let me know if I can be of any help in clarifying and documenting which cantons use which type of date (date of test vs. date of confirmation).

I just receive from SO

Diese Angaben sind schweizweit geregelt und sind 眉berall gleichen.
Alle Antworten zu Ihren Fragen finden Sie in den einschl盲gigen Meldeverordnungen und den dazugeh枚rigen Dokumenten beim BAG.

At least on some Kantons it seems that at least at the beginning this was not fully clear to all. And the question is, why the BAG reports with the "date of sample taking".
I have not yet asked AI, AR,GE, JU, TI and VD.

@baryluk The list of answers is getting longer. Should we create a file or page for the overview? This way I could add the link to my answers to the cantons, so that they can double check.
I suggest that we add this to "Cantonal / FL data ressources" in the readme. Sould I make a suggestion?

@zukunft that would be great :)

We're solving this by declaring differences in reporting in the sources repo:

https://github.com/openZH/covid19_sources_caseNumbers

The information coming from the sources-file is integrated into the dashboard as well:

https://www.web.statistik.zh.ch/covid19_dashboard/#/

I have seen the dashboard. Thanks for the great work.

Was this page helpful?
0 / 5 - 0 ratings

Related issues

jeanbrouille picture jeanbrouille  路  6Comments

zukunft picture zukunft  路  6Comments

rsalzer picture rsalzer  路  3Comments

netzwerg picture netzwerg  路  6Comments

dominikgehl picture dominikgehl  路  6Comments