Software aids in employee retention

Turnover among healthcare and social assistance workers is on the rise, with the rate inching toward one-third annually.

Not only is turnover disruptive, it's also resource-intensive, with providers spending tens of thousands on job searches and new employee training.

Every time a nurse leaves and is replaced, for instance, a healthcare organization spends $37,700 to $58,400.

Part of the solution, according to data and analytics company Arena, comes upfront during the hiring process.

Arena wants to keep providers from bearing the costs of turnover by improving retention rates. To do so, the company uses prediction models to cull from among a pool of applicants those most likely to stick with the job at hand and to be engaged, along with other metrics. Thanks to machine learning, the prediction models continuously improve.

Arena founder and CEO Michael Rosenbaum compares the technology to online recommendations: “Similar to the way Amazon uses your purchase data to predict what products you might want to purchase, Arena predicts how each person will improve outcomes like employee retention, patient satisfaction” and others, he said. Overall, organizations that use Arena to screen candidates see a 38% reduction in turnover.

“It's not an end-all, be-all,” said Sarah Colley, chief human resources officer at Regional One Health, a Memphis, Tenn.-based health system that's used Arena for two years. “But I like it as a tool. It allows us to see what we might not see as an HR or hiring manager.” Even whether someone will take the assessment in the first place is a tipoff to his or her success, Colley said. “If they're willing to take the assessment, it tells you if someone is really serious.”

Arena's algorithms use information about an organization's existing employees and from applicants. Arena data scientists build customized models for each client and for each role to predict, based on job applicant assessments, how likely each applicant is to improve certain metrics—including retention.

For example, for a client with two facilities across the street from each other—one for long-term care and the other for acute care—the Arena assessment, which takes about 15 to 30 minutes, asked each applicant if he or she was a leader in a community organization. Those who said yes were more likely to stay in their jobs at the acute-care facility and more likely to leave the long-term-care facility.

But Michael Finn, Arena's vice president of marketing, cautions against reading too much into individual questions, which may be more correlation than causation. Instead, Arena's predictions come from data models. “The results of the predictions are better than any single correlation,” he said.

Since MultiCare Health System in the Pacific Northwest began using Arena in 2015, its emergency room nurses who were recommended by Arena have turned over 33% less than others after 90 days. Now, the organization is looking into using the tool to predict employee engagement.

“When we have a more engaged workforce, that translates to higher patient satisfaction and the ability to achieve our quality goals,” said Kevin Dull, MultiCare's chief human potential officer. “The Arena tool is helping us put the right person in the right unit.”

For MultiCare and other systems using Arena, that translates to cost savings, Finn said. When asked why Arena is so important to hospitals, he said bluntly, “A lot of it is money.” Occasionally, organizations are reticent to use Arena's assessments, Finn said, saying they don't want to let computers dictate their decisions.

“In order to convince a hospital CFO to spend money on our solution, we focus on the hard financial benefits,” he said. He pointed to less money spent creating job listings, training and finding fill-ins when someone leaves. “Over time,” he said, “if a hospital is able to meaningfully reduce their turnover, they can save a lot.”