July 22, 2020

Data mining yields tool to predict OR staffing needs

By: Judith M. Mathias, MA, RN

Staffing to a block duration can be inefficient. Uncontrolled variables such as increased case complexity and unexpected staff absences can cause chaos. Leaders from Penn Medicine—Pennsylvania Hospital in Philadelphia used data analytics to measure the efficiency of matching staffing against an optimized OR schedule, which allowed them to predict staffing hours to case hours and determine a productivity score. By matching staff to case demand, they realized a 6.4% labor savings and improved productivity.

Pennsylvania Hospital is part of the multihospital Penn Medicine Health System. At Pennsylvania Hospital, 24,000 cases a year are performed across 28 in-hospital ORs and 8 ambulatory surgery center (ASC) ORs.



On her arrival at Pennsylvania Hospital in January 2017, Frances J. Strauss, MSN, MBA, MHA, RN, CNOR, NE-BC, clinical director, perioperative services, says she heard from many different people about the unmet needs of the OR—staffing was erratic, morale was poor, and staff burnout was high. On-call teams were being used to do elective cases and finish the schedule. Surgeons said the OR did not have enough skilled nurses to complete cases that ran into the evening.

In April, Strauss says she began looking at data from the previous year to understand which services had higher demand compared to others and what hours and days of the week were running the latest with electively scheduled cases.

By June, she and her team were ready to present their recommendations to the service chairs and executive leadership team. They included different shifts and an educational timeline that would prepare more skilled and competent staff.

In July, they took their recommendations to the OR advisory council, a panel of staff that facilitates bidirectional communication between perioperative leadership and perioperative staff.

In August, Strauss says they also began looking at their leadership structure and found differences in how leaders were managing day-to-day operations (ie, differences between charge nurse and clinical coordinator roles). They needed to understand more about the boundaries of decisionmaking and the responsibilities within the two roles.

The new staffing model was implemented in September, and changes were made to the hours of operation with a focus on supporting those services that ran late with skilled and competent staff.

Between March and October 2018, surgical volume dropped significantly because two high-volume orthopedic surgeons left the hospital. This necessitated a look at how to manage staff to volume, which led to the implementation of a new productivity model in October. Transformation to the new model was done in five phases:

• Phase I: Burning platform

• Phase II: Changing shifts

• Phase III: Leadership structure

• Phase IV: Financial implications

• Phase V: Sustainability and flexibility.

“We wanted to focus on what we heard versus where we heard it,” says Strauss. “What we heard was, we needed to align staff expertise with patient needs, and we needed to align those needs with quality, safety, and the patient experience.”

If there wasn’t a highly competent, skilled workforce, cases could take longer and patients could be under anesthesia longer. More work also was needed around leadership—developing leaders, understanding their needs, and better role delineation.

“We had a tremendous amount of support from our executive leadership team,” says Strauss. “With the perioperative leadership structure, we were able to put the plan together and support the initiatives we knew we needed to improve the process.”


Changing shifts

The objective was to redesign the OR staffing model to support surgeon blocks, improve efficiency, and meet the needs of the staff by providing a more predictable schedule.

It was not uncommon for someone working at 11 pm to be asked to come in the next morning at 7 am, says Strauss. “We had to be better able to predict our schedule. We were using our call teams to finish the elective schedule, which compromised our ability to support a ‘now’ case.”

Their current-state analysis included existing staffing gaps, case mix, historical trends of surgeons’ block utilization, case duration, overtime utilization, on-call utilization, and bonus pay.

The outcome of their analysis was to start RN and surgical technologist (ST) shifts earlier and extend their shifts later into the evening.

The OR advisory council, which is made up of RNs and STs, served as consultants to the perioperative service leadership team. The council also used the Pennsylvania Hospital strategic plan, the Magnet Model, and the hospital’s Nursing Model for Relationship-based Care as the foundations for the work they did and the recommendations they made.

The Model for Relationship-based Care, a framework for team feedback, was used to create a caring and healing environment. “A lot of what we needed to do was rebuild relationships, re-engage staff, and develop a level of trust with staff so that when their shift was over, they knew they would be able to leave on time,” Strauss says.

The leadership team presented three options to the advisory council for OR staffing and scheduling:

• Seniority: The pros of the seniority option were that it was an equalizer, and it was easily defendable. “When all else fails, you can fall back on seniority,” says Strauss. The cons were that it was a much longer process, it might not meet the goals and expertise of the staff, and it might not achieve the equity they were looking at for all staff.

• Staff survey: The pros of this option were that it was a faster process, and it asked staff their preferences while ensuring broader options. “They could have 8-, 10-, and 12-hour shifts, and it factored in the experience level of staff for all shifts and reflected Magnet standards for shared governance,” she says. There weren’t any cons for this option.

• Random selection: This option had only one pro—it would be a faster process than options one or two. The cons included staff not getting a choice and not having the expertise they were looking for throughout the day and into the evening, and it didn’t reflect a culture that values the voice of the staff.


Leadership structure

Strauss says staff told her they rarely saw their charge nurses or clinical coordinators, and they wanted them to be more available in the OR.

The new model provided an opportunity to create a patient-centered leadership structure that could support what staff needed.

This included:

• fostering a culture of professional development in leaders

• ensuring leaders were able to have the autonomy to make decisions

• inspiring gratitude and recognition in leaders that would transcend to the staff.

Strauss says they set an objective, which was to clarify performance expectations of service line coordinators throughout perioperative services to shift their focus toward patient progression and minimize focus on tasks. This allowed them to work more closely with their staff, troubleshoot, and ensure that staff had what they needed throughout the day.

They also standardized the role of the clinical service line coordinators by eliminating the charge nurse role and aligning their title and structure with existing nursing leadership structures (eg, time card approval, performance appraisals).

Finally, they discussed with the Penn Medicine Academy (corporate education support and development) how to enhance the role of the coordinators to be leaders for the staff. “We wanted to build and strengthen them, develop a robust plan for change, establish a change agent competency model, and provide change management tools and resources for them,” says Strauss.


Financial implications

The two orthopedic surgeons who left took about 10% of the volume from the main hospital OR. “Knowing that our volume was decreasing and our labor costs were rising, we knew we had to do something,” says Megan E. Sanders, BS, business manager, perioperative services, Penn Medicine—Pennsylvania Hospital. She says they came up with several strategies to attack the labor spend year-over-year increases.

One strategy was to look at the RN to ST ratio, knowing that in the systemwide rate variation, RNs compared at a 2:1 ratio against STs. Though AORN suggests a 70:30 ratio, it was important to take patient safety and case complexity into account.

Staffing ratios were adjusted as follows:

• FY 2018: 77% RNs and 23% STs

• FY 2019: 74% RNs and 26% STs

• FY 2020: 70% RNs and 30% STs.

“Compared to when we started this plan in FY 2018, we are on track to save more than half a million dollars,” says Sanders. The next goal is to have 68% RNs and 32% STs because the case mix justifies it, she says.

The second strategy was to develop a productivity tool that compares in percent an ideal OR day with the reality of an actual day. The tool is loosely built on the nursing hours-per-patient-day (HPPD) model.

To build the model, a broad range of subject matter experts were brought together, including Sanders, Strauss, nurse managers, clinical coordinators, medical secretaries, and the advisory council.

The two main constants in building the model were the number of hours the OR needed to staff and the number of staff needed per room, per hour, and per day. AORN’s generic staffing number is 2.5 people per room per hour per day. However, Sanders says they wanted to see if they could get a more accurate number and increase buy-in by getting to a Pennsylvania Hospital-specific complexity number.

“We wanted to get down to what our number was every day of the week, and we wanted to make sure our homegrown metric was simple to understand and use,” says Sanders. “We used 100% as our perfect OR day, which means we had the exact staff hours to cover the case durations we had that day.”

The complexity factors are filled in for each day (sidebar below).

Sanders says they start with the number of rooms running, add the RNs and STs needed for each room, add double scrub personnel needed in the head and neck surgery rooms (Pennsylvania Hospital does a large volume of head and neck complex flap cases), and then add the extra ST who runs between three robotics rooms.

Once the sum of RNs and STs are figured for each room, each day, Pennsylvania Hospital specifics are added, including the call out rate (eg, protected time based on ordinances by the City of Philadelphia, which is around 11%), and the nonproductive rate (eg, vacations, sickness, which is 14.2%). This nonproductive rate is higher than the hospital budget rate, which is around 13%, but it accounts for the experienced nursing staff who have accrued more vacations and extended time off, she says.

This total number is divided by the number of rooms running to get the complexity factor score, which varies throughout the week. For example, on Monday, the total number of staff is 75, and this is divided by the number of rooms, which is 27, to get the complexity factor score of 2.77.

The complexity factor score is held constant while, each day, the “chaos” in the OR is recorded, says Sanders. The staff schedule, OR time, and staff time are tracked each day with Epic to see how they compare to the ideal OR day of 100%.


Daily use of tool

“We look at this tool every day, prospectively and retrospectively,” Sanders says. “Prospectively, we are looking a few days ahead and prepping for the number of OR minutes we’re going to need, and then we translate that to OR time and compare that against our perfect schedule.”

In an example for July 9, under the number 3 (in red), the medical secretary pulled in the OR minutes from Epic. The minutes included wheels in to wheels out plus turnover time (sidebar below).

These minutes are converted to OR hours, which are then entered under the number 4 (in red). Then the clinical coordinator pulls in the number of paid productive hours scheduled for that day under number 5 (in red).

The ratio of OR hours to the number of staff hours gives the productivity score (in yellow). On this day, the tool predicted a productivity score of 91%. “Anything under 100% means we are overstaffed. Over 100% means we are understaffed and wouldn’t have enough staff hours to cover the cases,” Sanders says.

“Every day we use four numbers to give us two percentage scores,” she adds. “We look prospectively at OR minutes and OR staff hours, and we look retrospectively at OR minutes and OR staff hours.”

It takes the clinical coordinator and medical secretary about 5 minutes each day to complete this tool. Daily and monthly summaries are sent to the whole team.

In the past, if a day was lighter than expected, staff were asked to go home midway through their shifts. While the productivity tool was being developed, there was a “leave early” list with a process approved by the perioperative advisory council. Now, extra staff are no longer scheduled in the first place, and it rarely happens that a large number of staff are sent home because of inappropriate scheduling.

“We have less grumbling over long days and not getting out on time, staff are given notice of chances to take an extra day off, and there is more consistent use of per diem staff and thus better competencies in the OR,” Sanders notes.

Looking at year over year, productivity is outpacing increased costs of inflation and intensity and is responsible for a decrease in overall labor cost per weighted case. “We are bending the cost curve and are on track to spend less per case in FY 2020 than FY 2019, and OR utilization has jumped 5%, resulting in a savings of $1.9 million overall,” she says.

Since rolling out this process change, a few subsequent changes were made, such as adding new service lines, expanding cases into evening hours, and pulling back slightly on morning shifts. “We have found that we need more staff at 7 am than 6 am,” says Sanders. The model has now been rolled out to the ASC and endoscopy suites.

Staff engagement also has improved. “We asked staff whether they think we are implementing new methods and technologies to support them and whether they are participating in improvement work,” she says. “Our engagement with the OR advisory council allowed us to make significant improvements in those two categories.”

In addition, leadership changes have improved the perception that managers are open and actively engaged with staff, a staff survey shows.

“Seeing the category of ‘I have a manageable workload’ go up every quarter really shows our team believes its productivity score is working, and they notice a difference,” she says (sidebar below).

And finally, Sanders says, “we really think it’s important to pay attention to staffing patterns. My recommendation is to get into the numbers and really dig through them. Understand that it will take time, and that a lot of work goes into getting the numbers and getting the data, but we think it is worth it. It is also really important to listen to your team and create a culture of collaboration, so that all develop a sense of ownership of productivity.” ✥



AORN Position Statement on Perioperative Safe Staffing and On-Call Practices. 2014. https://www.aorn.org/-/media/aorn/guidelines/position-statements/posstat-personnel-safe-staffing-on-call-practices.pdf.


Sanders M, Strauss F J. How nimble is your staffing model? OR Business Management Conference. 2020.

Get OR Manager Updates

Sign Up Now
Latest Issue of OR Manager
February 2022
Video Spotlight
Live chat by BoldChat