As health care continues to implement electronic health records and analytics with a focus on interoperability, the tried-and-true advice still rings true: Information is only as good as the data entered, and to be effective, automation must start with streamlining and enhancing best practices and data integrity.
One step in that direction will be to stop using reduced readmissions as an indicator of success without determining why the trend occurred. Acute care reporting is performed using large data sets often from claims data. If we define success of improved quality of care delivery as reduced readmissions, based on aggregate of patients with readmission of fewer than 30 days, decision makers may assume patients are not returning because the quality of care was improved. However, this may not be the case if any of the following are true:
- The review of results looked only at readmissions with the same DRGs as those assigned at discharge, with no initial effort made to account for coding errors. The DRG assigned to a patient at discharge may erroneously differ from the DRG assigned at admission, for example, or the patient may be discharged with the DRG that was assigned at admission but then readmitted under a new DRG with the original problem unresolved.
- The review did not account for total visits, and the correct order of coding was not verified, especially in cases where there were multiple DRGs. For example, a patient is coded with congestive heart failure (CHF) with a comorbidity of diabetes mellitus (DM) on his first visit, and then CHF is coded as a comorbidity of DM on his second visit.
- The review did not consider changes of inpatient or outpatient status during emergency department (ED) visits, and such visits are not compared with previous visits. When a patient returns to the hospital with the same condition or associated condition for which he or she was admitted previously (e.g., shortness of breath), the patient’s status would be listed as “outpatient,” and the encounters are thus not associated together.
- Tracking of the patient does not reflect a readmission to another facility or free-standing ED.
- The patient died or moved outside of the hospital’s area.
To provide a reliable indicator of the quality of care in readmission reporting, hospitals should validate their readmission statistics by performing a patient-centric process involving a 30- to 45-day follow up with patients, tracking the reasons they did not return, ensuring these points are considered in their review of results.
Rose M. Rohloff is a healthcare organization consultant and HC Consumer presenter, Phoenix.
