Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
SGiambrone
Frequent Visitor

"Missed Check-ins" & Compliance Rate Measures

I am working on developing a dashboard where patients will be entering 4 health measures on daily basis: weight, blood pressure, temperature, and pulse ox readings.

 

Assuming the data flows through like such (Note: Data below is dummy data and in no way associated with a real patient):

SGiambrone_0-1594233028167.png

 

I am trying to arrive at two tricky measures:

  1. How can I calculate the number of "missed check-ins"? I'm hoping to slice specific to a patient, day, "health action", and in aggregate across the whole data set. (Note: When there is no data entered by the patient, nothing flows through to the data extract (as opposed to populating with values missing)
  2. Compliance rate - ratio of how many data points are being "captured" out of the total that "should" be captured (again, assuming all 4 must be recorded daily per patient). If possible I'd like to see this measure react to a date filter in case the end user is looking at a specific day, or in aggregate over another timeframe (1 week, 1 month, etc.)

Thanks in advance!

5 REPLIES 5
V-lianl-msft
Community Support
Community Support

Hi @SGiambrone ,

 

I hope I understand your intention correctly.

1. Try measure like this:

Measure = COUNTROWS(FILTER('Table',ISBLANK('Table'[VALUE1])&&ISBLANK('Table'[VALUE2])))

2. Try measure like this:

Measure 2 =
VAR total_reading_status =
    CALCULATE ( COUNT ( 'Table'[Reding status] ), ALLSELECTED ( 'Table' ) )
VAR total_n =
    CALCULATE (
        COUNT ( 'Table'[Reding status] ),
        FILTER ( ALLSELECTED ( 'Table' ), 'Table'[Reding status] = "N" )
    )
RETURN
    DIVIDE ( total_n, total_reading_status )

Sample .pbix

 

Best Regards,
Liang
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

Hi Liang,

 

Thank you for sharing. Unfortunately, when a patient does not complete their "check-in", there is no record created in the data at all. Ideally, I would love it if the data would flow through as blank for those missed events.

 

Any other thoughts on how to accomplish this? 

HotChilli
Super User
Super User

I hope there is a date table in your model, and a patient dimension table.

 

For 1 (assuming a 'missed check-in' is a row in the date dimension that doesn't have an entry in the Fact table) , create a simple measure like 

NumVisits= COUNT(TableVisits[datevisit]) + 0

Pull date from the date table and id from the patient dimension on to a table visualisation.  That should show all dates and number of visits.  Once you've tested it and you're happy, you can use the filter pane to filter the measure = 0.

 

2. The ratio should be the same measure divided by 4.  Or leave off the +0 if you're not interested in the 0% ones

Hi @HotChilli ,

 

Thanks for sharing your idea. So, assuming I followed this correctly, here is the resulting output when I create the visual table. I chose to show it by patient, by date. It does appear to be accurately capturing the # of check-ins and compliance rate at the date level. (ie. 4 = 100%, 3 = 75%, and so on).

SGiambrone_0-1594318710797.png

How can I get the compliance to calculate correctly for each patient across the full date range? For example, overall compliance rate for Charlie Brown should be total check-ins divided by total "opportunity" (assuming 4 required each day) for whatever timeframe is selected.

 

Further, my ultimate end goal is to create two KPI card visuals (as opposed to just the table) that show the total # of missed check-ins, and overall compliance rate. These measures should be dynamic based on the filter selections (date range, patient, health action) - however high level or detailed the end user would like to see.

 

Thank you for your help! Any thoughts how to proceed further?

As you've written, create a measure that shows "total check-ins divided by total "opportunity" (assuming 4 required each day)"

To start (and for testing), create 2 measures, 1 to count visits, 1 to count days (x4).

Add them to the visual to see if they are correct.

When happy. create the final measure which divides visits/days * 4

 

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.