How Lean Thinking and Practice Helped Put Shots in Arms—Part 1: Building the Process

This essay is the first of two focusing on the design and management of a single, mass vaccine delivery site at the Ohio State University Wexner Medical Center (OSUWMC) and examining the contributions of lean thinking and practice to the cycles of design, fulfillment, and use in this one site. Look for the second, which will be published on August 16.

For sixteen months, we have mainly seen two kinds of stories about the Covid-19 pandemic. Stories of heroism have abounded—caregivers of all stripes treating desperately ill patients selflessly despite significant personal risk. The other type of story was equally prominent—stories of mismanagement. We have heard about failures of strategy, planning, process, and communication, coupled with glitches in supply chain and organizational failure and finger-pointing at all levels of government. Both stories drive lean thinkers to tears.

However, one (mostly) good story surrounds vaccines. They were developed more quickly and are more effective than suggested by even the most optimistic forecasts. In the U.S., manufacturing ramped up quickly (good management), and the complicated logistics of distribution have been more effective than many expected (more good management). As a result, large numbers of people have gotten shots in their arms more quickly than anticipated. While the process to administer vaccines varies within and between states and among the medical centers and pharmacies that provide them, it appears that many sites have taken on the new work of mass vaccination with success, as indicated by the large numbers of shots provided.

A question that surely comes to the lean thinker’s mind is about the extent to which lean thinking and practice can or should be applied to mass vaccination in particular and, more generally, to responses to large-scale exigent population health challenges. We begin to address the question by exploring the contribution of lean thinking and practice to the Covid-19 vaccination process in one large medical center.

Identifying the problem to be solved

This note focuses on designing and managing a single, mass vaccine delivery site at the Ohio State University Wexner Medical Center (OSUWMC). With similar efforts going on across the world, our purpose is not to illustrate best practice but to examine the contributions of lean thinking and practice to the cycles of design, fulfillment, and use in this one site.

The Ohio State University Wexner Medical Center (OSUWMC), located in Columbus, OH, a metro area of more than 2 million, had a critical new job: safely vaccinate as many individuals as vaccine supply and space would allow. Although providing vaccinations was far from new to OSUWMC, both the massive scale and the immediacy of the community need were different.

For OSUWMC, the potential customer includes everyone in the area, a demographically diverse population, including many people with vaccine hesitancy. The quality of the experience itself is paramount, requiring that scheduling be convenient and the injection process is fast, pleasant, and safe. Process design began in November 2020 for a rapid ramp-up that would start in late December. (Two of the authors of this note, Ryan Haley and Crystal Tubbs, are primary architects of the process.)

Planning the Approach

A cross-functional team was established to design and implement a system to deliver the vaccine to the public. Initially, this group’s priority was to design a process to provide vaccinations to frontline health care workers within OSUWMC, transitioning to the general public shortly thereafter. The team included individuals from IT, Marketing/Communications, Revenue Cycle (Registration, Scheduling, Billing), Physicians, Nursing, Pharmacy, Legal Services, Facilities Management, and others to design the workflow. Although mass vaccinations at this scale are uncommon, the OSUWMC team did have experience, including seasonal flu vaccinations, mass Covid screening, and early experience with vaccinating OSUWMC staff and physicians, which all helped inform the design of the process.

The process needed to have several components that would mesh seamlessly to provide a good patient experience: an inventory management system to track, receive, and safely store the vaccines from multiple suppliers; a scheduling system that could handle a high volume of online and phone appointment requests; a day-of-visit process that could accommodate fast check-in, screening, administering the vaccine, post-vaccination monitoring, and scheduling a second visit, if required; and post-visit processes for billing; patient support, and safety reporting.

The fast ramp-up required that the basic system be built quickly using existing knowledge and components and then be tested and rapidly improved through a series of experiments. For example, the existing online patient portal (MyChart) system was modified to provide vaccination invitations, scheduling, and reminders  (MyChart is a part of the current electronic medical records system). As expected, as volumes increased, the capacity of this system was challenged, but improvements were made through several Plan-Do-Check-Adjust (PDCA) cycles so that the high volume could be managed.

Similarly, the day-of-visit process was designed starting with a value-stream map with cycle times and a number of stations based on expected demand and previous experience. Estimating times was complicated by the fact that many of the staff were “volunteers,” borrowed day-by-day from other parts of the medical center. There could be anywhere between 60 to 120 people working at a single point in time, and well over 1,000 different workers might set foot into the facility in a given week. Variation in capacity could occur because of the different experience levels of people working in the system. All vaccinators had substantial clinical experience—the issue was adjusting to the mass vaccination pace. Standardized work was developed and used to minimize variation in time, but individual learning curves remained.

New vaccinators were required to train a day or two prior, so there could be confidence in their readiness on the day of the session. Runners and other roles that required fewer skills were trained before their first session in a few minutes. All were positioned near an experienced team leader, who could provide help and coaching as needed.

Work sessions were scheduled in four-hour blocks (3 per day). After accounting for change-over and a pre-session huddle, A four-hour session yielded about 3 hours and 30 minutes of patient contact. A value stream map of the Day-of-Visit process follows.

Creating the Daily Management System

The designers understood the importance of including a daily management system to deliver expectations and communicate with key stakeholders, including frontline people. The nature of the frontline staff, who were volunteers or otherwise borrowed from other medical center duties, made clear and regular communication even more important than it usually is. Daily management was also required to capture issues close to their occurrence in time and place and escalate as needed. In addition, daily management would provide a forum for quick feedback on the various process improvement experiments that were underway.

The essential components of the daily management system touchpoints, from bottom to top, are as follows.

• Session Huddles:  A team huddle occurred before every four-hour session. Important information about team member assignments, safety updates, volume predictions, and technical education were all provided during this brief huddle. These would last about 5 to 7 minutes and occur roughly 15 to 20 minutes before patients were to be seated. Once released from the huddle, team members would procure supplies, head to their posts, and start working.
• Weekly Operation Meeting: Core members of the management team (Clinic Management, Nursing, Pharmacy, IT, Scheduling, and other operations leaders) participated in this 60-minute meeting. Feedback and ideas from the session huddles were brought forward and acted upon. This meeting was also the source of improvement ideas generated by the core team and their organizations. An outcome of each weekly operations meeting was to identify and assign vaccine slots based on projected supply provided by the state, assuring that they used the entire seven-day supply. 
• Daily/Weekly Leadership Call: The cadence of this leadership meeting was daily early in the mass vaccination program. As the process stabilized, meetings were reduced, eventually to a weekly occurrence. At this meeting, the Vaccine Operations Co-Leads would brief over 100 leaders across OSUWMC on volume outputs, operational challenges, and resource needs. They would also seek approval for significant changes to operational plans. Approved items would advance to a weekly Senior Leader call for final approval.
• Weekly Senior Leader Call: This short call served as an opportunity for the CEO of the Hospital Division and the Chief Medical Officer to provide progress updates to the Chancellor regarding vaccine distribution and its progress. It also provided a mechanism for senior leaders to communicate, express gratitude, and respond to recommendations by the frontline workers.

The tiered daily management system played a critical role in the success of the program. The quick ramp-up to high volume necessitated frequent two-way communication both up and down the hierarchy. It also served to transmit process changes and changes to standardized work across the three daily sessions. In addition, visual management was used to keep and communicate the “score” and, more importantly, to clearly indicate flow to largely uninitiated frontline workers. Finally, daily management also provided a mechanism for requests to be made and decisions rendered quickly. In short, the daily management system was a necessary element in delivering mass vaccination to the community.

Understanding the Role of Experimentation and PDCA

Experimentation was a big part of management work. For example, vaccinators were assigned to sections with a certain number of seats. The question of the right number of seats that a vaccinator could efficiently handle in a section needed to be determined experimentally. Similarly, visual management cues were vital for patient flow and to indicate the status of each patient in the section (i.e., ready for prep, ready for vaccination, vaccinated, and waiting under observation before discharge). So, the placement and design of the visual cues evolved significantly based on daily experimentation. Likewise, the placement of equipment and supplies to reduce wasteful steps also led to several changes based on almost daily experimentation.

PDCA provided a framework for experimentation. As an illustration, consider the problem of understanding the system’s capacity when there was variation in worker speed, number of workers available, vaccine available, etc. Data, observation, and prior experience were insufficient to estimate how many shots could be delivered in a session. Through many iterations of the PDCA cycle, the assigned team concluded that 5.5 chair turns per three-hour, 30-minute session was the effective rate of the process. At that rate, by filling every seat in the Schottenstein Center, roughly 1,250 to 1,300 patients could be vaccinated during a session. This information allowed the management team to safely predict the total available slots that could be handled in a day if all three sessions were filled. This understanding was critical in determining staffing needs, patient appointments, and whether the process was on track to meet its session and daily goals.

Understanding Lean Management’s Contribution

At the beginning of this note, we suggested that lean thinking and practice contributed to designing, executing, and improving a process to vaccinate thousands quickly. Most would agree that the practices described are mostly good, but why are they lean practices?

The answer begins with the focus on the customer -- what the customer needs right from the start — which is how to make getting vaccinated quick, easy, and safe. The process itself is characterized by extensive use of standardized work and visual management, arguably two of the most salient aspects of lean practice. Additionally, the process was shaped and re-shaped by continuous experimentation framed by PDCA, also a characteristic aspect of lean practice. Finally, it was all managed daily with a system that facilitated surfacing problems, escalating them when required, and making quick adjustments.

Lean management contributed to a mass-vaccination system that worked under extreme pressures. Consequently, lives were saved, and we all inched closer to understanding the new, post-Covid normal.

This article was written by Dr. Peter Ward, retired Fisher College of Business Professor, Ryan Haley, Senior Director of Ambulatory Services at The Ohio State University Wexner Medical Center, and Crystal R. Tubbs, Associate Director of Pharmacy at The Ohio State University Wexner Medical Center.