Predict readmissions with follow-up calls

As hospitals face penalties for excess readmissions starting in October, they can predict which patients are most likely to bounce back to their facility by simply picking up the phone, according to a recent CipherHealth study of congestive heart failure patients (CHF).

Follow-up calls for more than 600 CHF patients discharged from Charleston (W.V.) Area Medical Center identified patients that were 65 percent more likely to be readmitted than the cohort average.

The medical center's program, which gave discharged patients one call within 48 hours of discharge and a second call one week later, found that patients who responded positively to both calls had a 13 percent readmission rate, the study noted. Meanwhile, patients who displayed a negative reaction (or positive on the first call and neutral or negative on the second call) had almost three times the readmission rate at 38 percent.

"The results from this CAMC study show that, with a multi-call program, you can identify additional patients at risk of readmission who previous would have fallen through the cracks," Alex Hejnosz, a CipherHealth executive who helped design the study, said today in a statement.

And with rehospitalizations for chronic conditions such as CHF substantially higher than for acute conditions, predicting CHF patients most at-risk for readmission can help hospitals appropriately target interventions and avoid costly repeat visits.

The study also adds evidence that predictive modeling programs are identifying at-risk patients successfully and keeping them from returning to the hospital unnecessarily.

For example, New York City's Bellevue Hospital Center incorporates phone calls into its care coordination efforts to keep patients from coming back to the emergency department. It holds weekly conference calls among primary care doctors, visiting nurse services, methadone programs and substance abuse programs. As a result, overall ED visits dropped 10 percent, and hospitalizations fell about 40 percent, FierceHealthcare previously reported.

To learn more:
- here's the research announcement