Study Identifies Lower-Cost Way To Predict Health Crises Post-Discharge

New research has identified an approach to patient monitoring that can be put into place at a relatively low cost after providers discontinue more resource-intensive telehealth programs post-discharge.

The paper providing the results of the study, which appears in JMIR Medical Informatics, notes that most existing hospital telehealth programs monitor patients post-discharge for 30 to 60 days.

While such programs have been successful at reducing 30-day readmissions and post-discharge emergency department visits, they typically focus only on the costliest patients. In other words, while telehealth monitoring programs have been effective, they’re apparently still too expensive to use for any but the highest-risk segment of the population.

However, researchers found that by monitoring patients using inexpensive personal emergency response system (PERS) technologies, they were able to predict which elderly patients were at high risk for emergency hospital transport. This approach can be deployed without additional investments by hospitals, as it relies on equipment and services the patient or family is paying for.

To conduct the study, researchers used deidentified medical alert pattern data from 290,434 subscribers to build a PERS service to build a model predicting 30-day hospital transport. The predictive factors used in the model include subscriber demographics, self-reported medical conditions, and up to two years of retrospective PERS medical alert data.

The team tested the model’s performance by linking EHR and PERS records for 1,815 patients from a home health program, then comparing actual emergency encounters documented in the EHR with risk scores developed from PERS data.

As it turned out, PERS data did an excellent job of predicting which patients were at risk for emergency encounters. In fact, when researchers compared the predictions made by PERS data analysis with clinical outcomes documented in the EHR, there were 3.9 times more emergency encounters among those predicted to be at high risk than low-risk for the year following the prediction date.

The research team concluded that patient data collected remotely using PERS could reliably predict 30-day emergency hospital transport. Given that this method of monitoring is significantly less expensive to administer than telehealth programs, it sounds as though PERS could easily supplement or serve as a substitute for post-discharge telehealth monitoring.

As I’ve noted elsewhere, pundits say that population health management technologies will need to move beyond addressing only the highest-risk patients. At present, doing this seems a bit of a stretch, in part due to the need to develop better protective analytics models.

That being said, identifying lower-cost ways of tracking patients instantly step in the right direction. Perhaps academic researchers will identify other ways to make broader PHM models affordable for most providers.

About the author

Anne Zieger

Anne Zieger

Anne Zieger is a healthcare journalist who has written about the industry for 30 years. Her work has appeared in all of the leading healthcare industry publications, and she's served as editor in chief of several healthcare B2B sites.