IT Leaders Focus on Readmissions Reduction Programs
Beyond care coordination, all three discussion panel participants said their organizations were executing sophisticated data analytics using business intelligence software to identify high-risk patients for readmissions.
“That’s been a collaborative activity with a researcher at Parkland,” said Velasco. “He’s developed a very sophisticated risk prediction model looking not just at clinical risk factors, but socio-economic status and other data that are captured in the EHR, and using this prediction model to give a more fine-tuned assessment of patients. We’re working with him to better direct intensive risk-reduction efforts to those patients that are at highest risk.”
Velasco said that THR’s predictive models are not only focusing on morbidities and co-morbidities, but zip code, how many times the patient has moved, how many relatives the patient has for social support, and all things that are documented as a byproduct of care in the EHR to stratify the risk and to target the most at-risk patients. Through these efforts, THR has been able to reduce heart failure patient readmission rates by 25 percent.
Velasco admitted that readmission rates are just the tip of the iceberg and really a lagging indicator of bad care coordination; and that there’s much more, like the entire measure of access to care, that goes into reducing readmissions. He lamented the fact that most commercial EHRs lack the functionality to assess and predict readmissions and urged healthcare organizations to develop these competencies independently through various programs.



