Visuality is a fundamental tool that organisations can deploy to increase problem identification and to support their lean transformation. Alice Lee, Vice President of Business Transformation at Beth Israel Deaconess Medical Center in Boston, guides LMJ through a recent project to optimise patient flow in a hospital department, presenting a few examples of how visual management has supported it.

A colonoscopy is no fun.

It starts with a large dose of laxatives the day before the procedure. Then, after a restless night with multiple trips to the bathroom, no coffee or breakfast, the patient heads to the Digestive Disease Center, one of the busiest outpatient services at BIDMC. She’s not relaxed; she may even be thinking about everything her family went through when her father was diagnosed with colon cancer after his colonoscopy.

Many people are involved in her care. She is checked in by front desk staff; then a volunteer gets her from the waiting room. Nurses in admitting review her medical history, allergies and meds, and then update the online medical record.

She is then wheeled by a technician to the procedure room. The physician explains the procedure to her and obtains consent to proceed, and the procedure nurse administers medication to sedate her. After the procedure, she is handed off to a recovery nurse who cares for her until she is discharged.

Throughout these steps, our staff does its best to provide care and comfort to the patient, as well as perform numerous tasks for compliance with regulations, billing and equipment maintenance. They repeat this for 70-100 patients each 8.5-hour day.

The nurses, doctors and support staff would tell you they are there to care for our patients. Yet, the patient’s unflattering comment that things work like “in a factory” surprised us.

There had been many previous attempts to improve operations, but the complexity of systems had been paralysing. Providing high quality with efficiency is paramount, but so is patient centricity. What could we do differently?

A project was launched, led by Digestive Disease Center staff, with support from Business Transformation – a department of management engineers – to tackle the daunting task of trying to better synchronise the equipment (scope), procedure room, procedure nurse, doctor and patient for each procedure. The goal was to build capacity to support and grow volume utilising existing resources, equipment and staffing.


The Digestive Disease Center (DDC) offers comprehensive and innovative endoscopic, surgical and medical treatment for digestive disorders and is among the top five programmes in the nation. The DDC performs over 23,500 visits annually in a facility that opened in March 2008.  At that time, the architects estimated the facilities could accommodate 27,000+ cases per year. However, when the volume reached 23,000 cases per year, the DDC began to experience bottlenecks. A project team was formed to study the issue on the unit within the DDC with the highest volume.


By going to the gemba to observe practices and workflow as they existed, team members saw problems firsthand. There was a lot of waiting. Waiting staff, waiting patients, waiting physicians. Even scopes were waiting to be cleaned and brought back to the procedure rooms.

The team saw idle procedure rooms while patients waited for their procedures, as well as the utilisation of these rooms for tasks that could be done elsewhere. It became apparent that flow of patients from one step of the process to the next required an added level of staff-to-staff communication and effort.

An important component to this process is the scope, which is the primary equipment used for procedures. You need a clean scope of the particular type that the procedure requires. We saw physicians bringing dirty scopes from the procedure rooms to be reprocessed (Figure 3). In fact, we saw four physicians bring dirty scopes within five minutes. This was interesting to us because it means four procedures are ending at the same time. Are we causing that? Is it just a coincidence? Can we study the schedule to understand this better?

What we discovered was that each of our physicians had his/her own schedule: at 7:30am, we could have five patients scheduled in five rooms. Bottlenecks resulted, as there were five scopes to reprocess, five rooms to turn over, five patients to transport out, and five new patients to transport in, all occurring simultaneously. This cycle repeated itself throughout the day as the schedule was built this way and had a peak and valley effect on the system as our patient batches moved through the schedule (Figure 2 Peaks & Valleys).

So the question becomes, how do we staff for this? Do we staff to the peak? Or to the average? This is a real problem the staff experiences when they have too much work at points in the day and none at other points. How do we start to level this out to create a steady workflow?


The main value-added step in the process is the procedure – our pacemaker. Therefore, we designed the system to synchronise around the procedure. We started from the end of the process to create pull.


Despite the finding that the process itself is fairly reliable, the wait time is the main source of variation.

We graphed the arrival time of scopes throughout a day with 82 scheduled procedures. Figure 4 depicts the WIP (work in progress) which builds up throughout the day so the techs can never keep up.


At the end of a procedure, the nurse and physician document their notes, the room is cleaned, the patient is transported to Recovery, and the room is set up for the next procedure while the next patient is transported in from Admitting. These activities are shared amongst five different roles.

Are we using the rooms to the best of our ability? Are we utilising them appropriately? We also asked who was doing the turnover activities – was everyone working to the top of their license? A central goal is to assign the right person to the right job at the right time and at the right place.

With these questions in mind, we categorised all room turnover activities as “in room” or “out of room”. We then moved as many “in room” activities out of the room as we could to improve the productive time in the procedure room (Figure 5 top).


The nurse had 24 minutes of work to turnover a room (Figure 5 bottom). This means each case couldn’t start sooner than 24 minutes after the last one. No wonder we could not stay on schedule – we allocated 30 minutes per procedure without any time in between to turn over the room.

Activities were appropriately reassigned by role and work continues to further balance the turnover activities to takt.

To further improve the procedure room productivity, a visual signal is used to trigger turnover activities sooner. As the physician leaves the room, she flips the “Start Turnover” flag to alert the turnover tech to clean the room.


The team ran experiments to create a visual line of sight between admitting and the procedure area to better coordinate which patient is needed next. They settled on a switch of the admitting and recovery areas, which greatly reduced the number of phone calls (82% decrease in average calls from eleven to two per day). The recovery area is quieter, which is a patient and staff satisfier. The visibility between the two areas allows for tighter communication and is foundational for tackling our next challenge – create visual cues to improve communication and ultimately patient flow.


To capitalise on the line of sight created, the team aimed to create a system where everyone knows what needs to be done, when it needs to happen and who needs to do it. A rule of thumb we learned, “5 feet in 5 seconds” means anyone should be able to gather the information needed to complete their work from any visual board from five feet away in less than five seconds.

The team set out to design a better system, first with pen and paper, then by mocking up their ideas, building in colour coding and picking strategically which information truly needed to be shared. After seven iterations, the team settled on a transparent board that can be viewed from both sides (Figure 7).

The system takes judgement and error out of who to admit next and decreases the time spent at the board by 71% (less than two minutes out of every hour for nurses). It also reduces waiting for our patients, staff and physicians.

With pull throughout the process and flow initiated through the unit, we now have a stable foundation for improvement and a team with a strong will to continuously improve their systems for flow. Next, we are turning our focus on how to best schedule procedures to create even better level flow by smoothing arrivals. Stay tuned!