From the many evidence-based health risk stratification tools available, Community Health Network has adapted a popular hospital readmissions indicator for use with medically complex patients at high risk of readmissions. Deborah Lyons, MSN, RN,NE-BC, network disease management executive director for Community Health Network, describes the adaptation process.
HIN: Where do home visits for heart failure patients enter the picture?
Deborah Lyons: We do a high-risk home assessment while we have patients in the hospital. Fully 100 percent of our patients that are admitted to inpatient status are automatically screened and ranked in terms of readmission risk. That’s where we use the LACE/ACE tool. We embedded that tool in our software so it can predictively tell us which patients to focus on.
HIN: How did you decide on the LACE tool? Is the ACE tool different than the LACE tool?
Deborah Lyons: The LACE itself is evidence-based. We work with the advisory board. And they had just done an analysis of all the predictive models out there in terms of readmission risk when we started this work. There were only two tools that were moderately predictive for risk. LACE was one of them. LACE looks at length of stay (L), acute admission (A), (meaning they came in through the emergency room), their Charleston Comorbidity score (C) and the number of ED visits (E) they’ve had in the past six months.
All this information was easily available to us at the time that we did this because we were on a different computer system. But the concern was that the L factor (length of stay), might lead us to place the patient at high risk when they were leaving the hospital. Maybe they started at low risk and then on the fourth day of stay, because they had been there four days, now they moved to high risk but they’re being discharged. You really can’t do anything at day of discharge. We first set a threshold for LACE, which we tested and validated and then ran a correlation and asked ourselves, “If this threshold is a LACE high risk, what would a correlating threshold be if we dropped the length of stay?” That’s how we moved to an ACE score.
Stratifying High-Risk, High-Cost Patients: Benchmarks, Predictive Algorithms and Data Analytics Reviews a range of risk stratification practices to determine candidates for health coaching, case management, home visits, remote monitoring and other initiatives designed to engage individuals with chronic illness, improve health outcomes and reduce healthcare spend.