Lodi Memorial focused on readmission types that they could reasonably affect when setting goals.
We set our original goals for reducing readmissions based on our overall hospital readmission rates, but we didn’t meet those goals. One reason for that is that our rates were already relatively low, so shrinking them even more was difficult. Another reason was that we set the percentage goal without looking at which readmissions were actually preventable.
What We Tried
We drilled down into all of our readmissions to understand where they were coming from and why they were coming back. How many readmissions were coming from SNFs vs. home health vs. home? How many were readmitted because of medication noncompliance vs. lack of access to social services vs. substance abuse? This took a long time because we had to dig into each readmission, calling the patient, SNF, or PCP when needed.
We then looked at which of those factors we thought we could reasonably affect. Some readmission cases are just worsening disease and patients should come back. Other patients will get readmitted no matter how hard you try. We used the subset of readmissions that we thought we could change to set our goals for the following year.
Targeting our goal based on data really helped us focus our efforts. We were able to reduce our preventable readmissions significantly and are now extending our program to include the readmissions that we have less control over. It will be a bigger challenge, but we’re ready for that now.
As we were doing this analysis we realized how hard it is to measure readmissions. One unexpected factor we realized was that our admissions team wasn’t recording where patients were coming from correctly, so it was hard to do accurate analysis.