Our participation in the Transforming Care at the Bedside (TCAB) initiative at the University of Pittsburgh Medical Center (UPMC) Shadyside began in 2003. At the time I was unit director of 3 Main, a 38-bed cardiothoracic and vascular surgery unit. On the day shift, a nurse cares for four or five patients; on the night shift, it can be as many as six or seven. Patient acuity is high on this unit, and patient care assignments are determined by acuity and care requirements.
The unit is laid out on two 200-foot parallel hallways, with rooms totaling 20 beds along one hallway and 18 beds along the other. Patient rooms are only on the outer sides of the hallways. In the middle, between the two hallways, there is a central nursing station, a utility room, and a patient kitchen. Three pathways connect the long hallways.
Nurses often had patients on both hallways and sometimes in all four corners of the unit. This caused them to waste steps and their energy. Dissatisfaction with this situation was high among both the staff and the patients, who had to wait longer than they liked for nurses. We decided to address the problem through TCAB because it affected all four TCAB focus areas (patient-centered care, teamwork and vitality, value-added processes, and safe and reliable care).
Our history as an established TCAB unit that encouraged openness enabled our nurses to bring up their concerns and issues at staff meetings and elsewhere, and several staff members told us that the assignment method we were using was challenging and unproductive. The primary issue was that the distance between their patients' rooms required them to do a great deal of walking. This issue was important not just in terms of its effect on overall productivity, but also because of the extra physical demands it placed on the nurses. On our unit, 29% of the RN staff were near or beyond the average age of nurses in the United States, 46.8 years.1 The potential this presented for increased injury and reduced ability to meet some of the physical demands of their work2 prompted us to reassess the way the RNs performed their work.
In my quest for evidence and ideas about how to restructure our assignments, I found minimal literature on care assignments among nurses. The most comprehensive article dated back to 1973 and suggested that it was important to keep assigned patients in close proximity.3
Along with the hospital's improvement specialists, the nurse on the unit who was on the hospital's TCAB committee and I decided to divide the unit into four "pods" with two nurses assigned to each pod during each shift. We thought this would decrease the number of unnecessary steps and improve nurses' ability to monitor and be accessible to their patients. Within each pod, patients would be equally divided between the nurses by their acuity, which would be reassessed at every shift change. However, when we presented the idea, the staff was resistant and raised concerns about the potential unfairness of assignments.
We then went back to TCAB basics and collected data to support the need for making the change and to address the staff's concerns. We planned to develop tests of change based on the information we gathered. First, we created a spaghetti diagram to show the path a nurse traveled during a four-hour period. Spaghetti diagrams are used in manufacturing to expose inefficient layouts.4 We created our baseline spaghetti diagram by directly observing a nurse who was chosen because she was organized, thorough, and able to complete her work in a timely fashion. The spaghetti diagram of her steps before we implemented the pod design revealed an erratic pattern. According to a pedometer, the nurse walked 1,075 steps in four hours and 3,928 steps in eight hours.
We then assessed the patient complaints and scores collected on our Press Ganey patient satisfaction surveys. We paid particular attention to patients' assessments of nurses' promptness in responding to their calls, attention to their personal needs, and overall care. We felt that the metrics supported performing a test of change of the pod design.
We created four pods that covered 100 feet each, two for each hallway. Each pod had eight rooms—with 10 beds each in pods A and B and nine beds each in pods C and D—whose occupants were divided between two RNs. We were careful to distribute acutely ill patients, confused patients, those in isolation, and empty rooms as equally as possible among the pods. The charge nurse quickly reassessed patient care assignments at the end of each shift to ensure fairness in the assignments for the following shift's nurses. Midshift changes were made only in the event of an emergency.
We began using the pod design in April 2007. The entire unit participated in the test of change for 30 days, which we thought would give staff enough time to adjust to and be able to articulate pros and cons about the new arrangement.
We saw a consistent and sustained improvement in patient satisfaction scores in the months immediately after we implemented the pod design (see Figure 1). Patient complaints, which are reported, processed, and tabulated by the hospital's patient relations department, dwindled to a single complaint in June and in July, less than the usual two or three complaints and one or two grievances per month. (A complaint is a verbally communicated patient or family concern or issue that can be resolved quickly by the staff, whereas a grievance is a written or verbal complaint that was not resolved at the time it occurred.) The spaghetti diagram, created with the same nurse and technique as before, showed an improved, less erratic work flow. The number of steps the nurse had to take decreased significantly, to 877 steps in four hours and 2,291 steps in eight hours.
Using personal digital assistants (PDAs) and a work sampling study, we documented a consistent increase in the amount of time spent in direct patient care (see Figure 2). Time devoted to value-added care—patient-centered actions that directly benefit the patient—also increased. Work sampling involves applying statistical sampling techniques to work activities and is typically used to estimate the proportion of a worker's time that is devoted to different tasks.5 The PDAs were set to vibrate randomly approximately 22 times over a 12-hour period. The nurse then selected her or his location and the most accurate description of the activity being performed.
We have been using the pod design for patient care assignments for more than two years. At times, we have adapted the staffing patterns somewhat by slightly overlapping a nurse's patients in adjacent pods. We also sometimes change the pod division, depending on the number of RNs working in a shift.
The pod design for patient care assignments has improved patient satisfaction by increasing the visibility and accessibility of nurses and has enhanced nurses' ability to provide safe and reliable care. This care assignment design has also improved staff vitality by reducing the number of unnecessary steps nurses take during a shift.
The pod design has been spread to and adapted by other nursing units at UPMC Shadyside that have the same physical layout. We supplied our design plan and outcome metrics to the spread units and presented them at our hospital's weekly TCAB meeting and the UPMC health system's TCAB forum. Unit directors who presented the idea to their staffs reported that they also initially encountered resistance to the idea, but the TCAB philosophy of adapting changes enabled these units to make the design more suitable for their use.
This TCAB experience proved to me that the cautions I had read in the literature about resistance to change are not exaggerated. Although the unit's staff had identified that there was a problem with the old way of making assignments, they didn't want to try something new. We found that nothing can break down resistance to change like good metrics. The measurements we gathered opened the eyes of many doubters on our staff, who then supported changing to the pod design. Tests of change are the TCAB way of life, and the staff has largely become accustomed to them. Although they still create some discomfort simply because they involve change—though it may be as small as relocating a printer—the negativity that accompanied our first test of change has abated.
I also learned that the sustainability of a change often hinges on the ability to adapt as necessary. For example, our nurses work a mixture of eight- and 12-hour shifts, which makes adjusting patient assignments necessary. In the end, the adaptations made the pod concept stronger.