unable to search by proprietary drug name e.g tazocin

[drcjar]

should return PIPERACILLIN WITH TAZOBACTAM which is what tazocin is

[davidmiller]

We currently don't have either a good mapping or an exhaustive list.
Extraction isn't really a good long-term option - the time costs of
scraping are high.

Crowdsource it?

On 14 May 2013 14:27, Carl Reynolds [email protected] wrote:

should return PIPERACILLIN WITH TAZOBACTAM which is what tazocin is

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18
.

Love regards etc

David Miller
http://www.deadpansincerity.com
07854 880 883

[simonh10]

I've already done this analysis as part of my attempt to do generic/branded budget comparison. I'll see if I've got it somewhere as a json dump. It was done by finding the lowest equivalent priced product and assigning that to the generic option.

[drcjar]

Using other sources e.g www.bnf.org it's quick to look up a drug that we don't match correctly on e.g Tazocin

I'd be happy to manually add if we need to know when we fail and we don't at present #14 and even if we did we'd miss incorrect matches #19 .....

[davidmiller]

Could be misremembering, but weren't there concerns raised about the er,
clinical accuracy of that mapping?

On 14 May 2013 14:33, Simon Hargreaves [email protected] wrote:

I've already done this analysis as part of my attempt to do
generic/branded budget comparison. I'll see if I've got it somewhere as a
json dump. It was done by finding the lowest equivalent priced product and
assigning that to the generic option.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18#issuecomment-17876042
.

Love regards etc

David Miller
http://www.deadpansincerity.com
07854 880 883

[simonh10]

It's as accurate as the BNF codes are. Everything that follows a pattern of e.g. 0409020C0__ACAC is equivalent in chemical formulation, dosage and delivery form.

[drcjar]

seems to handle drugs behind #19 and #13 well

[davidmiller]

Great.

Now all we need is an original scrape of the BNF data that takes BNF codes
as well.

Could prob write something to imply it from the breadcrumb data.

On 14 May 2013 14:53, Carl Reynolds [email protected] wrote:

seems to handle drugs behind #19https://github.com/nhshackday/mobileformulary/issues/19and
#13 #13 well

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18#issuecomment-17877229
.

Love regards etc

David Miller
http://www.deadpansincerity.com
07854 880 883

[drcjar]

could raise as an issue here https://github.com/tomtaylor/bnf-html if are absent?

[davidmiller]

Last time I looked the raw codes weren't on the page anywhere - not an
issue with the scrape but the content.

Could raise with the BMJ ? :)

Codes are semantic, which means we can write something to figure it out...
Should have enough in the database currently.

On 14 May 2013 15:38, Carl Reynolds [email protected] wrote:

could raise as an issue here https://github.com/tomtaylor/bnf-html if are
absent?

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18#issuecomment-17880067
.

Love regards etc

David Miller
http://www.deadpansincerity.com
07854 880 883