Calculating 30 day hospital readmission rates for unique patients

[jcoranx1]

MRN	AdmitDate	DischargeDate
2	2019-01-05	2019-01-07
2	2019-01-30	2019-02-03
3	2019-11-20	2019-11-25
4	2019-05-12	2019-05-15
4	2019-05-19	2019-05-26
4	2019-07-29	2019-08-04
5	2019-04-08	2019-04-14
6	2019-03-09	2019-03-13
6	2019-03-18	2019-03-23
6	2019-08-01	2019-08-09

I am trying to calculate 30 day readmission rates for unique patients in a large dataset. I am familiar with how to use the datetime difference formula to calculate length of stay or 30 day intervals between a discharge and admit date. Where I am stuck, is how to calculate the 30 day readmission rates with the condition of unique patient identifiers?

I included dummy data with similar structure to see if anyone on the forums can assist me. Is anyone able to walk me through the tool or code that can accomplish this?

[joshuaburkhow]

When you say "with the condition of unique patient identifiers" What does that mean exactly that you just want to calc the readmissions by individual patient? 

[jcoranx1]

Correct. I would like to flag (column with a 0 and 1) acute admissions encounters for each 30 day readmission event per unique patient. This will help me identify the total number of readmissions (for network and by patient) when calculated across the entire data set.

[afv2688]

Hello @jcoranx1,

This gives you the amount of times somebody has been readmitted in the last 30 days. Not perfect but kinda works? lol:

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

Regards

[jcoranx1]

This solution was close to what I am looking for, but not quite the exact result. I placed an example of the new column (Readmit30dayFlag) I am trying to create in the table below. It flags the readmit encounter as a 30 day readmission for the unique patient, but it requires the qualifying event to populate a "1" in the readmission flag column. 

Qualifying event is an acute hospital admission within 30 days of discharge for the same patient. 

MRN	AdmitDate	DischargeDate	Readmit30dayFlag
2	2019-01-05	2019-01-07	0
2	2019-01-30	2019-02-03	1
3	2019-11-20	2019-11-25	0
4	2019-05-12	2019-05-15	0
4	2019-05-19	2019-05-26	1
4	2019-07-29	2019-08-04	0
5	2019-04-08	2019-04-14	0
6	2019-03-09	2019-03-13	0
6	2019-03-18	2019-03-23	1
6	2019-08-01	2019-08-09	0

[afv2688]

Hello @jcoranx1 ,

Here you go!:

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

Regards

[OllieClarke · Accepted solution]

Hi @jcoranx1 I think this workflow does what you're after (flagging admittances that are <=30 days since a MRN's last discharge)

Hope that helps,

Ollie

[jcoranx1]

This workflow is what I was looking for and is simplified. Will mark as solution. I like the previous discharge date field and will be keeping that in the data model for reporting purposes. 

Thanks!

[artism]

Thank you! This is very help! How would you include the actual number of days in took for the patient to be read admitted? I would like for the column to have a number instead of a true or false. 

[DonnaB]

I'm a student and having a hard time following the pics. Can you provide the formula for the creating the column Readmit 30day Flag?