If you’re struggling to find ways to benefit from the massive data being collected on patients and devices at your hospital, you’re not alone. Healthcare leaders have often felt unsure of which data points to focus on and feared that the benefits of data analysis may not outweigh the costs. But recent success stories showing how predictive analytics can cut costs and boost patient satisfaction have clearly caught their attention. Many are now predicting a big increase in investment in predictive data analytics.
According to a recent poll by the Society of Actuaries, only 47 percent of healthcare executives currently use predictive analytics, but 89 percent report they plan to create analytics initiatives or increase the ones they already have. More than half say that at least 15 percent of their total budgets will be dedicated to predictive analytics within the next five years.
Here are three stories that illustrate recent ways hospitals have reached specific goals with predictive analytics:
Fall Prevention Through Data
Almost one million patients fall in U.S. hospitals each year, according to the Agency for Healthcare Research and Quality. Those who are injured in a fall add an average 6.3 days and an additional cost of $14,000 to their hospital stay. Healthcare News reports that El Camino Hospital in California wanted to improve its fall prevention program, which had long depended on using colored slippers and wristbands to alert clinicians to patients at high risk of falling.
The hospital integrated a predictive analytics program that hones in on patients identified at admissions as being at high risk for falling. Through machine learning, the system combines information about which ones set off a bed alarm or use the call light most frequently, and looks for certain triggers that send alerts to nurses that a patient is at imminent risk of falling.
Nurses responding to this predictive information helped to reduce falls at El Camino by 39 percent in just six months. “When certain data elements line up, you take action in the moment so that your prediction does not come true,” says Cheryl Reinking, Chief Nursing Officer at El Camino. “You can change the outcome in the moment.”
A Better Readmissions Picture
Reducing preventable readmissions is a priority at many treatment centers, partly due to Medicare penalties for high readmission rates under the Patient Protection and Affordable Care Act. The experience of Conway Regional Medical Center in Arizona shows how even smaller medical centers can cut readmissions with predictive analytics.
Conway Regional, a 160-bed acute care center in north central Arkansas considered using the L.A.C.E. index to identify patients at the greatest risk for early readmission. But using L.A.C.E. is highly-labor intensive, so Conway turned to a predictive tool called RAPID, a business intelligence solution from Medisolv. RAPID pulls data from Conway’s EHR daily, scores every patient, and makes recommendations for intervention that are available to everyone in the organization. The process is vastly more efficient than having caregivers spent time manually analyzing data for each patient.
Patients identified as being at high risk for readmission are given extra education by nurses about their reasons for hospitalization, new medications, care plans and other factors. Primary care physicians of high risk patients are also alerted, so they can schedule follow-up appointments within three days of hospital discharge.
As a result of using the RAPID predictive readmission risk tool, Conway saw a 3.31 percent decrease in overall readmissions in just 10 months. The downtrend was steady, suggesting more improvement ahead.
Cutting Hospital Admissions For Heart Failure Patients
At Aurora Health Care of Milwaukee, WI, a predictive analytics tool for patients with heart failure (HF) and chronic obstructive pulmonary disease (COPD) was pivotal to its move from fee-for-service care to a value based approach.
HFMA.org reports that Aurora launched a pilot program using an analytics tool to identify patients with HF or COPD who were highly likely to be admitted to the hospital during the next six months. Through proactive use of this data, RN care coordinators formulated care plans and monitored all aspects of these patient’s health. The results: Hospital records showed a 60 percent reduction in HF-related admissions among 126 patients in the pilot program. At the same time, admissions for COPD patients in the program were reduced by 20 percent.
According to Sylvia Melter, MD, Chief Medical Officer-Population Health at Aurora Health Care, predictive analytics can be a great tool to identify opportunities. The true benefit comes when all members of the care team use the data in a care model that is consistent and replicable.