Skip to main content
Skip to main navigation
Skip to footer



Creating a Value Loop

Predictive analytics plays a pivotal role in nursing staffing and scheduling and results in better patient care.

Bob Dent, Senior Vice President, Chief Operating Officer, Chief Nursing Officer, Midland Memorial Hospital


For health care executives seeking to deliver value back to their organizations, value can mean different things and take on different forms. Workforce satisfaction may not always be the first area where leaders seek to derive value, but research shows that an engaged and effective workforce can have a positive cascade effect into other areas such as outcomes and financial performance. As health care organizations are under increasing pressure to deliver value in patient outcomes and budgetary measures, it is worth making sure that front-line employees are working in the most effective way possible. Data mining and predictive analytics tools can now provide health care organizations with advanced insight into their workforce and the ability to act on that understanding to enhance employee performance and satisfaction.

In 2009, Midland Memorial Hospital set out to use predictive analytics to manage its nursing workforce. They wanted to respond to staff feedback about scheduling and find a better way to meet the daily changes in patient needs. They also recognized the increasing pressure that nurses face due to overwork. Nationwide, nurses increasingly report feeling stressed, overwhelmed and overburdened, in part due to increasing patient loads and the complexity of care required. Predictive analytics can help alleviate this by scheduling nurses based on the anticipated level of care needed for a patient, determined by analyzing historical data on care and outcomes for similar patients. This approach, known as acuity-based staffing, has been recognized by the American Nurses Association and the American Organization of Nurse Executives as an effective way to improve nurses' job satisfaction and patient outcomes. We sat down with Bob Dent, senior vice president, chief operating officer and chief nursing officer at Midland Memorial Hospital, to talk about his organization's experience with using predictive analytics to aid in the planning and staffing challenges of the nursing department and how that led to increased value realization for patients and for the budget.

Describe how Midland Memorial uses predictive analytics to manage its nursing workforce.

Using predictive analytics pulled directly from the electronic health record, we can identify the number of nursing staff we need on any given day based on anticipated patient care, and therefore staff by acuity. We're able to track acuity levels throughout a patient's stay, and that historical data gives us a better data set for predictive analytics. We use nursing documentation and the outcomes we want for our patients to understand the number of staff and the skill mix we need to address the care of the patient. Now we can recognize staffing peaks and troughs to bring in staffing at the right time of the day. We've developed staggering and alternating shifts to identify when appropriate staff will be needed throughout the day to better meet patient needs.

What value has Midland Memorial Hospital experienced since implementing predictive analytics?

We've found that it improves nurse satisfaction, patient experience and quality outcomes.

In 2008, we formed our nurse staffing committee and standardized allowable consecutive shifts among nurses and hours worked per pay period to minimize fatigue. The policy mandates nurses work no more than 12 hours a day, three 12-hour shifts in a row or 60 hours in a seven-day period. We incorporated this policy into our workplace staffing software, so if a nurse says they are feeling fatigued, we can pull reports to see how often they have been working outside of the fatigue management guidelines.

Since implementing this system, we've seen a 32 percent reduction in overall nursing turnover and a 43 percent reduction in turnover among new, first- and second year nurses. By not working our associates longer than we need, we're also able to deliver a better, safer experience for the patient. Between 2014 and 2016, we saw a reduction in centralline- associated bloodstream infections by 22 percent, catheter-associated urinary tract infections by 64 percent and ventilator-related events by 38 percent. Satisfaction among patients in the ED also increased to the 90th percentile.

Describe how predictive analytics impacts the patient experience.

The nursing workforce wants to show up every day and give their best, deliver positive outcomes for their patients and create a great patient experience. This means providing compassionate care and giving nurses time with their patients. Using predictive analytics, we can make sure we have the right people assigned every day and that we are staffing by acuity to make sure nurses aren't overloaded. We have a clinical operations team to watch and understand what is happening at the hospital any hour of the day and to deploy resources to assist where it's needed. When doing rounds, we can know ahead of time what kind of day staff might be having. By creating synergy between the nursing staff and the patient's needs, we can respond to the patient quickly and skillfully, and provide a higher quality of care.

As other health care leaders embark on a similar journey, or are trying to make the business case for predictive analytics, what advice would you offer?

Hone in on what you want to accomplish and start with a small test. Have a clinical operations expert run your data on a regular basis and communicate regularly with that person so you can continuously adjust.

You'll also want to get buyin and trust from the finance department and the health information systems (HIS) department. It's important that these departments understand the value of the investment; this ensures it will be supported from a financial and an IT perspective. For a period, we were behind on updates for our workforce management software, and that was a hindrance. Building a good relationship with HIS allowed us to better manage updates and be more successful with predictive analytics. Without the functionality in the optimized version of our workforce management software, we wouldn't have been able to implement the tools necessary for our improved staffing and scheduling system.

Predictive analytics is a financial investment and it was important that the finance department had visibility to the outcomes. We use data out of our workforce staffing system to set our yearto- year budget, which has led to a more efficient financial planning season. Because we're better able to plan for the number of nurses needed and the skill level needed, we set a more realistic budget and can stick to it. Our chief financial officer would tell you that nursing has it right.

How do you see predictive analytics progressing in the health care industry over the next five to 10 years?

There is tremendous opportunity. As we do more research about planning demand around the needs of the patient and making our organization more patient-centric, we can look at predictive analytics from a care team perspective. We must continue to think differently about how we staff and look at interprofessional staffing to care for the patient. We need to ask ourselves how we balance nursing hours with that of other care professionals to get balanced, holistic care for patients.

For example, we have a patient who comes in with pneumonia. They might need more time with a respiratory therapist, as opposed to a nurse. Or if you have a patient with a total hip replacement, they might need more from a physical therapist versus a nurse. Predictive analytics can help balance the care provided based on what the patient needs. We can deploy the right resources at the right number of hours, which helps more easily manage the financial, retention and recruitment aspects of our workforce. All of that creates value for the patient and for the organization. 

Where do you see predictive analytics and artificial intelligence providing support to hospital operations in a value-based reimbursement model?

We must make sure the things we are doing today - treatments, medication administration, having the right skill mix in our workforce - are going to get us rewarded in a value-based model. What matters is that you are getting the outcomes you need with the team that you have. Having data that creates transparency on the front end validates we are doing the right things to manage better outcomes, increasing our success in a value-based model. Research shows that an increase of even one hour of nursing care helps improve outcomes, so making sure that nurse has the right skills for the patient’s needs is critical. Predictive analytics lets us have the right number of staff with the most appropriate skills for the patient's needs, providing the best care possible.