Extra plots derived from Encore Pro Database...

littlefeat · Post by **littlefeat** » Sun Mar 20, 2005 2:55 pm

These new plots in sheet 4 are awesome--the lower left plot (AHI against pressure) would be particularly helpful. How complicated is it to get the spreadsheet to talk to the database for a given user?

Mikesus · Post by **Mikesus** » Sun Mar 20, 2005 3:16 pm

Yes it does. It appears that if you tell Encore Pro to not delete data it will keep retaining compliance data, and a rolling 7-8 days of detail. So every time you download, you will get overlap.

Mikesus · Post by **Mikesus** » Sun Mar 20, 2005 3:17 pm

Also, from that last round of graphs, it would suggest that my "ideal" pressure is around 13...

littlefeat · Post by **littlefeat** » Sun Mar 20, 2005 3:25 pm

I think that you're talking about the data-hold capacity of the smartcard, not the database tables. The "erase data" routine only erases (or doesn't erase, if that's what you have in the registry) the smart card after each download. I assumed that doning a download each day, and erasing the smart card only weekly (which is about its storage capacity), did not create duplication in the database, or that the program seems to be able to report a day's data only once, even though it may have been downloaded from the smart card six times. For example, I erase my smart card every Tuesday AM, and then capture Tuesday night through the following Monday night in seven daily downloads. This gives me an every-morning look at the night before, then a seven-day report for my doc every Tuesday AM.

Mikesus · Post by **Mikesus** » Sun Mar 20, 2005 3:31 pm

Yes it does duplicate the data. It does not appear that way in the reports that you see, but it is very clear in the database that there are overlapping datapoints. For me this started when I did not erase the card. (so by not erasing the card, and downloading daily, it will have about 4 days on average that are downloaded in each daily download.)

derek · Post by **derek** » Sun Mar 20, 2005 3:41 pm

Mike -
I went through your data with a filter in Excel and deleted all the duplicate records (phew!) and redid the plots. You will see that the indicies have hardly changed at all , although the total hours is way smaller.

Click here for Mikesus's revised pressure dependence

littlefeat · Post by **littlefeat** » Sun Mar 20, 2005 3:42 pm

Roger that. I assumed that maybe multiple downloads of the same night's data might be handled in the db by successive overwrites with respect to the same night so that, in the end, the night would exist only once in the db. But I don't know how to look at the db tables themselves, so obviously was wrong in that assumption.

Does that mean that the trend stuff that you guys are working on would be skewed by the duplication, whereas the canned reports apparently are not skewed? Thanks again.