Forum Discussion

ChristineKirk's avatar
ChristineKirk
Qrew Cadet
8 months ago

Extrapolating nuanced data

Hello again - one more for today!

I'm self-taught in Quickbase and data management altogether. I'm trying to suss out the best way to pull some data. We capture a dataset on consults in our Quickbase. Every call our program receives, gets one record entered into Quickbase. Sometimes callers call us back for more than one consult on the same case. To this end, we have a field capturing "New consult" or "Continuation of consult". Then we capture services performed and patient demographics of the consult.... many situations can transpire in which we got that data on the first call, or on a subsequent call, and it's quite varied (for instance, first call may be very brief, they give us limited info and say they'llc all back/we don't get that info right then, but do on the longer subsequent call; the inverse of that, they call, we spend meaningful time, capture the data the first time, and they call back for some follow-up questions).

Now, to extrapolate services (how many times did X happen on a case we consulted on) -- how can I pull that nuanced data out? I guess I can manually look through the data, which is how we do indeed do this to report quarterly on how many cases we served..... There must be a better way to look across the life of our program and pull the data.

  • Thanks Mike - I have tons and tons of charts - haven't yet explored the new dashboards.

    So this specific use example - we have "Call Records" with a field, "Type of Call". Every call = 1. New Consult, 2. Continued Consult, or 3. Not a Consult

    As mentioned above, with options 1 and 2 - it's not predictable to know which of those call types we'll receive basic patient demographics. We might receive it on the first call, or that one might be very brief and we capture it on the second "Continuation". Sometimes we capture it on both. Basically a nurse enters it if she knows it.

    My goal is to make sure that for every 1 case we served (New plus all its subsequent consults), we're counting it only once. My suspicion is that this is maybe a hair too complex to really do here..... Like I'd need maybe a way to tie the records from the same table together in order to filter.

    With that expanded info, any other ideas?

     

    Thank you!

    • MikeTamoush's avatar
      MikeTamoush
      Qrew Commander

      There are a number of ways to manipulate and look at data. If you have a specific use case, you can post it and we might be able to help more.

      However, in general to look at numbers you will want to create Reports. Can use table reports or charts, depending on the use case. To get specific totals, you can use the filters. In some cases, you will simply filter the report itself (in the report customization options), in other cases you may just want to use a dynamic filter on the report. Bar graphs can sometimes be quite useful to look at something over time (x axis can be a date access, orgnized by day,week,month or year).

      You can also utilize the new Dashboards, and then utilize the global filters on those new dashboards. ie: Put a pie chart or bar graph on the dashboard, and tie a filter to it (date filter for example). When you filter by date periods, the charts will update their numbers.

      • ChristineKirk's avatar
        ChristineKirk
        Qrew Cadet

        Thanks Mike - I have tons and tons of charts - haven't yet explored the new dashboards.

        So this specific use example - we have "Call Records" with a field, "Type of Call". Every call = 1. New Consult, 2. Continued Consult, or 3. Not a Consult

        As mentioned above, with options 1 and 2 - it's not predictable to know which of those call types we'll receive basic patient demographics. We might receive it on the first call, or that one might be very brief and we capture it on the second "Continuation". Sometimes we capture it on both. Basically a nurse enters it if she knows it.

        My goal is to make sure that for every 1 case we served (New plus all its subsequent consults), we're counting it only once. My suspicion is that this is maybe a hair too complex to really do here..... Like I'd need maybe a way to tie the records from the same table together in order to filter.

        With that expanded info, any other ideas?

         

        Thank you!

  • Hi!

    Reopening this. I'm just starting to think through how to build this out now. So. In my instance, I may create a "CASE" parent, with "CALLS" as the child.

    My issue comes down to - we have certain required fields which then should be at the parent field - but they're only required based upon certain conditions of the CALL. For instance, if we never entered the exam room, we don't require those data points.

    Is there a way to collect the CASE data points on any of the subsequent Calls, but have it cleanly roll up to the CASE? Or like, is there ANY way to put daisy-chained fields from the parent table into the child record? (It would be fine if those remained the same through all subsequent child records). The data points of a CASE don't (or maybe exceedingly rarely) change (it's things like demographics, services provided which could change if something additional were offered later on). The goals and tasks of separating them is 1) to ensure the data of a case is only counted once, and 2) to ensure that if we do take certain actions on a CALL, certain fields become mandatory to include about the CASE.

     

    Post-Script Idea: Maybe this is a little out there/too convoluted. I just thought to myself, maybe there's some way to write a formula field that sums up the existing child records. If any child record has "Entered exam room" = Checked, then require these Parent fields. 

    Though, I don't know that there are any such formulas to trigger something becoming mandatory, let alone a whole other field. 

    Thanks very much!

  • Having a Case Parent and then Child records for the calls is a good architecture.  You can do lookups from the Case down to the child and then use formulas to put up Rich Text (say red) warning messages on the Calls record if the Call involves a data point on the case which was not initially provided.  The user would be able to save the child but they would be trained to obtain the missing information from the Caller and update the Case before closing off the Call.  The warnings messages from all the calls could be rolled dup in a Combined Text Summary field up to the Call and then converted to regular Rich Text with a formula (I cvan hjelp with that if you are stuck). 

    So the contact would be that your agents would not be done with the data entry until they cleared all the red warning messages.