EOB math

[John Caruso]

I am working with the values in the EOB and trying to understand the logic of 

1. how to add across item / categories, and 
   
2. how to add items together to get totals

From the prior release, the benefitBalance had more info and was slightly confusing, so the new total (i.e. the "payment" object) makes my second question more straightforward.

for example, resource / id = carrier-22273336765

patient = resource / patient / reference = Patient/20140000008325

This claim resource has 11 claim line 'items', with a total payment / amount = $150

Relative to question 1 above, can you point me to examples of claim lines with non-zero amounts in each category of an 'item' - illustrating the payments going from submitted charges and outlining all steps to provider payments and patient payments? If there is no such example, can you provide the math that you would expect to work within each 'item' to get from submitted to paid? If an "other" is needed on our end, what would you suggest we call it - assuming we will get questions on it from users :) ?

Relative to question 2 above, iterating through the 'adjudication' lines yields a total of $130 (not $150). Is this an artifact of synthetic data? If they don't always sum to the total, which is more representative of the amount that was paid - the 'payment / amount' or the sum of the items?

 

Thanks,

John

[karl....@cms.hhs.gov]

John,

Thanks for posting this question over here! I was just about to reply to your email and request that. Regardless, sorry for not getting back to you sooner.

So, you've got a lot of requests here and unfortunately I can't get to them all right now, but I'll try to hit some highlights:

1. Q: It doesn't seem like these values add up like I'd expect from lines to claim totals?
1. Answer the First: That is largely a function of how the synthetic data was produced. The production data should be more consistent in this regard.
2. Answer the Second: Keep in mind, though, that there's a LOT of Medicare claims data, and like all real-world data, it's going to be messy in "interesting" ways.
3. Answer the Third: Some claim types have "outlier payments" that will not appear in the revenue center/line items, but will appear in the claim totals. I've been asking around about this today and haven't had a chance yet to get full details, but this is the case for at least some Outpatient claims.
2. Q: Are there examples of each adjudication type?
1. A: Great question! Unfortunately, the answer is: maybe? Ensuring comprehensive examples like that wasn't a goal of our synthetic data generation process, so if it happened, it was by accident. That said: it's entirely possible that it did, since we have a lot of synthetic data.
   We're hoping to eventually release the synthetic data in a form that's easier for developers to query/analyze for questions like this, but it just hasn't bubbled up to the top of our priorities for a sprint yet.

On the off chance that you're motivated enough to take advantage of it, tomorrow I'll reply with a sqlite dump of the source CSV data that gets fed into our ETL process and eventually results in the FHIR resources in our sandbox. I can't really provide support for it (it's a hack I put together to do that type of exploration myself), but if it helps you explore the data too, that's great!:

• A sqlite dump of our source synthetic beneficiary and claims data: TODO (will send tomorrow)
• An out-of-date spreadsheet with some more information on that source schema: rif-layout-and-fhir-mapping.xlsx
• Note: All of the links there are broken, but you can look up the fields in the CCW Data Dictionaries: https://www.ccwdata.org/web/guest/data-dictionaries

[karl....@cms.hhs.gov]

John,

On the off chance that you're motivated enough to take advantage of it, here's a sqlite dump of the source CSV data that gets fed into our ETL process and eventually results in the FHIR resources in our sandbox. I can't really provide support for it (it's a hack I put together to do that type of exploration myself), but if it helps you explore the data too, that's great!:

• A sqlite dump of our source synthetic beneficiary and claims data: https://drive.google.com/open?id=1m5nhgyypCXYN1ynm37WbUYtfNDbtN6bj
  

• An out-of-date spreadsheet with some more information on that source schema: rif-layout-and-fhir-mapping.xlsx
• Note: All of the links there are broken, but you can look up the fields in the CCW Data Dictionaries: https://www.ccwdata.org/web/guest/data-dictionaries

Best regards,
Karl M. Davis

[Vahndi Minah]

Could I get the password for the Sqlite dump please?

[karl....@cms.hhs.gov]

Vahndi,

I can't check right now, but I'm pretty sure it doesn't have one. Are you getting an error trying to access/open it?

Best regards,

Karl M. Davis

Blue Button 2.0 API, Lead Engineer

[Vahndi Minah]

Hi Karl, I am getting the following message in DB Browser for SQLite. Not sure if I am doing something wrong:

[Joey Hawilo]

Hi Everyone, 

Has anyone figured out the password to open the file? Trying to access it now, and get the same error. 

[karl....@cms.hhs.gov]

Joey,

Sorry for any confusion, but there's no password on the file. It's a SQLite dump, though -- not just a SQLite database. To use it, create a new SQLite 3 database and run this to restore the dump:

sqlite> .read bluebutton_synthetic_sqlite_dump_20180329.sql

(Took quite a while to restore on the slow i5 I just tested it on. Enjoy a cup of coffee!)

Best regards,

Karl M. Davis