The Art and Science of Nurse Staffing

Aug 01, 2019

Add to Collection

Added to Collection

Staffing is going to be much more scientific and analytically based — we will be able to crack the acuity code.

Therese Fitzpatrick

Therese Fitzpatrick is both a nurse and a self-described “data and technology person.” It stands to reason that she seeks to study and advance the intersection between nursing and technology to address one of healthcare’s most enduring and vexing challenges: appropriate nurse staffing.

Fitzpatrick is a senior vice president with the financial advising firm Kaufman Hall in Chicago and an assistant professor in the Master of Health Administration program at the University of Illinois at Chicago. In her article “Using Science to Improve the Art of Staffing,” Fitzpatrick views nurse staffing as a complex system and outlines new ways to merge an artful yet data-driven approach to this complicated issue. She notes that the application of big data analysis by companies such as Google, Walmart, UPS and United Airlines could inform processes for healthcare staffing.

How do you view the challenge of appropriate staffing?

Research demonstrates the correlation between poor staffing and nurse burnout, job dissatisfaction and intent to leave a position. Nurse executives are trying to balance staffing with clinical outcomes while also ensuring a healthy work environment. We often presume that the number of nurses is the most important element of staffing. It’s more complicated than that. My paper discusses the blend of contemporary science with the art of staffing.

In your article, you discuss “challenging the flat Earth approach.” Please describe it.

I’m a data person, and the technology I’m most excited about is what we’re seeing in big data analysis and artificial intelligence. It has the opportunity to revolutionize what we’re doing with both patient care and understanding analytics.

So, for example, a report just released says Google Artificial Intelligence (AI) does a better job in predicting lung cancer in patients. AI has superseded what radiologists have been able to do. When you think about that, the power of AI is going to be pretty phenomenal.

In another example, Walmart used data analysis to identify 800 radiologic diagnostic centers in order to send employees to those that do the best job of diagnosing certain conditions. When you think about that, it’s pretty powerful.

How do we “embrace a new world view” of staffing?

Challenging the status quo as it relates to staffing instructs us to stop looking at averages in our data.

That big data analysis allows us to look at demand patterns and utilization patterns at the hourly level — or in some cases at the 15-minute level — to determine how to best right-size the staffing for a unit and help right-size flexible staffing or a float pool to adequately support staff on the unit so that they can leave the unit for education, for time off and things like that, and to be able to do that in a much more scientific way.

We have tended to think about staffing as something that we can schedule our way out of. But I would suggest that you have to do this high level of analysis and pre-strategy work in order to have the right configuration of staff. Then you can leverage the great scheduling technology that’s available to us.

Tell us more about this contemporary approach to staffing.

We’re still using antiquated techniques to staff our critical care units. What I mean is that we’re still using an average daily census, typically marked at midnight, to determine demand. And that’s a very critical number in how we budget for staffing.

Those midnight averages don’t really reflect what’s actually happening on a busy critical care unit at 10 a.m. or at two-o’clock in the afternoon when patients are beginning to come back from surgery. So, it requires a revolutionary way to look at this — to look at demand at the hourly level.

That’s where big data comes in?

Yes. I have determined that looking at three years of data taken right out of the electronic medical record — which is a very large data set to manage — allows us to predict with about 95% confidence what the demand patterns will look like over the course of the next year.

The other big data implication here is that we typically have not looked at the implications of various work rules on how we budget and deploy staff. For example, in an organization where the staff nurses work every other versus every third weekend, you need a very different configuration of staff.

In organizations that limit the number of 12-hour shifts and limit the number of 12-hour consecutive shifts that a nurse can work, that has tremendous implications on making sure we have budgeted the right resources for a particular unit. Without this big data approach our tendency is to use averages — a notion that I look at as the “flaw of averages.” Additionally, we have not really dissected what the term staffing means.

How do you define staffing?

There are four very distinct, critical processes that are part of staffing: correctly budgeting, correctly scheduling, correctly deploying staff in the moment and how we assign staff based on patients’ needs.

I use the same sort of science that United Airlines uses to put planes up in the air and that UPS uses to deliver packages. I’ve taken the science from big logistics industries and applied it to staffing and scheduling within nursing, which is a very complex logistics problem.

What does appropriate staffing look like with this approach?

Staffing in the future is going to be much more scientific and analytically based, and I think that with our deployment and focus on electronic medical records we will be able to crack the acuity code. We have had a hard time measuring and monitoring patient acuity over the course of a patient’s stay. Now, with AI and big data analysis, we will do a much better job in terms of quantifying and monitoring patient acuity. That needs to be the first driver of how we begin to think about appropriately budgeting. As critical care nurses, this opens up a host of opportunities for us to become very creative about how we document the value we bring to patients’ outcomes.

I think we’ll see radical changes within the next five years. Our challenge is how critical care nursing will intersect with these new technologies. But, it is also an amazing opportunity for us in terms of being able to lead the change. The implications for us are really quite astounding. It means that we must focus on certification, skills and education in order to serve as leaders in this change.

What are your recommendations for CNOs, other nursing leaders and front-line nurses to begin to innovate solutions for appropriate staffing and balanced schedules?

Staffing and scheduling are a team sport. We have developed this notion called the “staffing ecosystem,” coming to the table to support both the staff and the chief nurse. They need to be partners in decision support and finance, in IT if they have an electronic scheduling system, in clinical education and, in some organizations, bringing physician partners to the table in support of the CNO.

It’s also important to partner with our data science colleagues, because it’s precisely this type of collaboration that I have had with statisticians and operations research scientists that has allowed me to combine their mathematical skill with my knowledge of what happens related to our clinical staffing that has really been the powerful underpinning of this solution.