Breakdown of drug spending by use(s)/purpose(s)

[mattgawarecki]

Task

Analyze Medicare Part D spending to build an understanding of the most treated conditions, as well as what's the most expensive to treat.

This is your chance to break out the tables, graphs, and charts!

How this will help

If we can get an idea of what Part D is being used to treat, we can do further exploration to find ways of reducing cost and improving overall quality of care.

[dhuppenkothen]

I'd like to take a stab at this over the coming few days, unless that would duplicate someone else's work?

[mattgawarecki]

Go for it! I don't think you'll be duplicating anything. Even if you do, though, that's the kind of thing that sparks discussion and builds great ideas! Make sure to assign the issue to yourself, and don't forget to shout if you run into problems! 👍 On Feb 1, 2017 18:38, "Daniela Huppenkothen" <[email protected]> wrote: I'd like to take a stab at this over the coming few days, unless that would duplicate someone else's work? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#15 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAo5c8HSPOMxe3P7NAjDb1c97qzbSMb4ks5rYSVvgaJpZM4Lw7Dc> .

[jenniferthompson]

Agreed! I think this will be hard to finalize till we wrap up issue 6, but it can definitely get worked on in the meantime!

[dhuppenkothen]

Thanks! It doesn't look like I can actually assign the issue to myself. "Assignees" is greyed out for me. Could one of you assign me?

@jenniferthompson I realized that, too, last night. I'm currently hacking my way through a (perhaps temporary) solution. Not sure yet whether it'll work; let's see.

[mattgawarecki]

@dhuppenkothen I'll work on getting you assigned to the issue. I think we need to get you officially added to the D4D organization to make that happen :-) Will update when I know more, and/or when it's taken care of.

[mattgawarecki]

You should now have an invite to the org -- whenever you accept, let me know and I'll officially assign the issue to you. :-)

Remember to fork the repo into your own account first, then do your work there. When you're finished, submit a pull request ("PR") to this repo and we'll review/merge it. Thanks so much for pitching in!

[jenniferthompson]

All set @dhuppenkothen!

[dhuppenkothen]

Thanks, @jenniferthompson

[jenniferthompson]

@dhuppenkothen Do you have a vision in mind for what you want to do here? In the process of writing up what we're working on, I'm wondering if this and issue 7 are basically getting at the same thing (Issue 7 is pretty vague and was added awhile back, before we started to get rolling on organization). Since you and I are assigned to each one, I think we should either combine forces (which may mean I hand it over to you!) or differentiate what we're working on, whichever makes more sense. What do you think?

[dhuppenkothen]

@jenniferthompson Yes, we should! My hope for this issue was to explore the data and proto-type informative and useful visualizations (for some yet-to-be-determined definition of "informative" and "useful"). It may end up that these visualizations will have an interactive component, in which case they might tie in really well with the Dashboard?

At this stage, though, my goal is to figure out what's in the data, and how to best display it in 2D. Once we've got that, it probably shouldn't be that hard to either re-code the figures in R and stick them into the Dashboard. Does that seem like a reasonable approach?

I was also going to look at your conceptual drawings to guide some of the initial ideas I have for visualization. :)

[jenniferthompson]

@dhuppenkothen Definitely reasonable! Yeah, I think a combination of static and interactive visualizations could work really well for this.

How about I focus on getting things really well organized (repo structure, data dictionaries) for the next few days while you dive deep into this, and once you've got your plan in mind we can talk about specifics? I'm open to adding to the current dashboard or developing a new one (or other vis type), whatever gets the message across best!

[dhuppenkothen]

Yes, that sounds good.

[selahlynch]

Hey, I'd love to help with this task also. Any chance you could add a link some of the relevant datasets to be used. I'm having some difficulty figuring out what they might be.
PS... I pledge to do some data cleaning/grunt work as well

[jenniferthompson]

@selahlynch All our datasets are in our repo on data.world - so far most of the work has been with the CMS Medicare Part D data (spending-201x files). Not sure what @dhuppenkothen has going on so far!

[dhuppenkothen]

I haven't used any data sets from data.world, as far as I know. The read_data.py does all the data reading for what I've been doing so far, but maybe some of those data sets should go up on data.world?

[jenniferthompson]

@dhuppenkothen Looking through read_data.py - yes, it would be excellent for the prescription drug profiles and RxNorm data to get on data.world (and any other datasets I'm missing). Then non-Pythoners can use it too! That would be really helpful.

[dhuppenkothen]

Could you perhaps open an issue and attach it to me? Then I'll get on that after work tonight.

[jenniferthompson]

@dhuppenkothen yep, I was already on it :) done!

[franckjay]

Hi all, I can try and help with the classification of drugs by use. I had a thought that we could do a Wikipedia scrape for each drug, looking for keywords?

[darya-akimova]

Closing in order to restructure the issues and to change project scope.