04-23-2014 10:07 AM
I am aware that this possibly is a question on a novice level, but I would appreciate your help.
I've used EG quite successfully to link databases and select specific cases for further analysis. I am now wondering what would be the best way to enter new data for these new cases. The source databases are updated regularly. After processing these databases, linking them, a query selects eligible cases (a couple of hundred) for the pending analysis. For this analysis, I have to perform additional measurements and enter these in a new table, with the unique case identifier. I'm not quite sure what would be the most efficient way to input and store this additional data.
04-23-2014 10:47 AM
You would probably want to edit this information in a separate table, stored where you have write access.
Then you can use this in your project, using Query Builder/user written code to merge7join onto your source data.
The design of this table depends on the source data and the information you wish to add. If you give some examples on what you wish to add, source data, and desired merged result, we could could give some more detailed help.
04-23-2014 12:05 PM
's suggestion is spot-on. One factor that would definitely be used to determine the best mode for this is where your new data comes from. Will it be produced by an automated system, or will someone in your group type it into some form of table?
In the case that it's the latter, what are you most comfortable with...Notepad, Excel, Word, Access...?
The good news is there is a TON of good ways to do this; it's just a matter of finding the best.
04-24-2014 10:15 AM
Thank you for your help.
The main output from the ultimate query that selects eligible cases is caseID (and some logistical information). The new data I want to collect have to be typed in (manual assessment) and consist of date/time variables and some numerical variables with labels. Some other points:
1) the number of eligible cases is increasing, the data sources are dynamic
2) it is possible a case changes status from eligible to not eligible
3) The new variables have to be available in the EG project
4) I want to prevent having to do measurements twice.
The most ideal picture would be: I periodically run the project, I get presented with a list of eligible patients without the new variables, I enter the new variables, these variables are saved and imported in the project, in the end I have a data file with all performed measurements of said variables by caseID.
04-28-2014 02:17 PM
It sounds to me like using Excel to capture your manually-entered data might be the easiest. How does this sound?
1. Using your automated system, select your set of eligible cases, and associated data, as a SAS dataset.
2. Using one of the many options available, copy these columns to an Excel spreadsheet on your PC.
3. Using Excel, add your new variables. I've had really good luck going from Excel to SAS with the 28APR2014 date format.
4. Again, using one of the many options for this, transfer your completed spreadsheet back into SAS for further processing.
P.S. Another option would be to use the Add-in for MS Office to handle the Excel conversion part.