How to Vaccinate 100 Million People in 100 Days

This week the incoming U.S. President Biden announced his intentions use FEMA and the National Guard to establish COVID vaccination sites across the country. The goal is to vaccinate 100 million people his first 100 days in office. At less than a third of the U.S. population, that is far from herd immunity. It’s a good start, perhaps a proof of concept. In general, I’m not a fan of apparently ambitious, round-numbered targets. They may be good for setting a vision. But as marching orders they tend to cause people to do unreasonable things. In order to meet “100 by 100” type targets, people jump to solutions that don’t always address causes of failure. If the thing is important enough to do, we should arrive at our target conditions and achieve-by dates in a more scientific manner.

There is a scientific way to design and execute even such large-scale rollouts. Organizations that practice lean management have been doing it for decades. It’s a fairly simple matter of following a set of principles, starting smaller, testing them out rapidly, learning and adapting as you scale up. Here are some of the lean process design principles and how they might apply to vaccinating 100 million people in 100 days.

Learning from 3P – the Production Preparation Process

The lean management approach to designing processes and equipment is known as 3P: the production preparation process. In essence, 3P designs equipment and processes inside-out. We start from the value-added transformation and wrap the “box” including fixtures, power source, and peripherals around it. This approach allows us to design small, scalable, low-cost, high-quality equipment that is just good enough to perform the value added step.

This is in contrast to the so-called catalog engineering approach. This starts outside-in. We select machines based on their advertised features from a catalog. These are boxes full of bells and whistles. We do this because we didn’t take time to deeply understand our process requirements. Solutions vendors love catalog engineers.

Many people in position to spend government money are more interested in a quick, promised solution than in understanding the process at the local level and how to design the minimum viable and scalable value-added step. Fortunately, in this case we’re talking about a very simple process. The value-added transformation is to add mRNA into a human arm. We have a proven way to do this involving needles. We don’t need to reinvent the core process, only how to make it flow.

Pacing the Work to Takt Time

One of the first lean principles to consider when evaluating process design options is whether it can keep pace with demand. Can we deliver value one customer at a time, when they want it? This rate of customer demand is known as takt time. We calculate takt time by dividing the net available time per period by the customer demand per period.

In terms of the goal of vaccinating 100 million people in 100 days, basic math tells us that we’ll need to poke 1 million people in the arm every for one hundred days. But this is not takt time. One million pokes per day is only our daily demand. We need to know the other half of the equation, which is net available time.

In terms of gross available time per day, we have 24 hours per day, 60 minutes per hour, and 60 seconds per minute. That’s 86,400 seconds per day available per day. Now we can calculate takt time.

(86,400 seconds / day) ÷ (1,000,000 vaccinations / day) = 0.0864 seconds per vaccination

Yikes! That’s really fast! Another way of looking at this is that we will need to vaccinate 11.57 people every second of every hour of every day for one hundred days to hit 100 million.

Sounds like we’ll need to build a highly automated delivery systems using rows of robot arms loaded with syringes with people cruising by on conveyor belts. How else could we poke so many people at such speeds? A process like this would be expensive, take a long time to build, and probably break down on the first day. Luckily, this is not how lean process design works. But this is not too far from how many big-budget, large-scale projects are engineered.

Using Additive Capacity

The lean approach is not to design processes based only on maximizing their individual speed. When maximizing output of one process, equipment often becomes complex, changeover times crippling, quality suspect, and maintenance costs high. One process that is much faster than others can create a traffic jam. This slows down the overall performance of the delivery system.

Other lean thinking principles help balance out the need for customer-driven speed with stability. The fable of the tortoise and the hare is worth remembering. Rather than having one speedy but unreliable rabbit to deliver vaccines, we need an army of steady tortoises. This is known as using additive equipment to scale capacity up or down based on customer demand.

Indeed, some of the States that are currently enjoying success in delivering vaccines rejected the top-down approach proposed of the federal government. Instead, they collaborated with chains of pharmacies or other small local centers. Larger bureaucracies can bring powerful, large-scale change, but can also be slower and less adaptive to local conditions.

The added benefit of small, scalable, additive processes is that any individual failure is small and contained. Part of the plan in discussion is to fill giant arenas with vaccine stations and move people through. There are apparent economies of scale. It also creates a large single point of failure. When a single large operation breaks down or struggles to ramp up, it sets the whole plan back. When capacity is dispersed across many small, local centers, the risk is also dispersed.

Letting the Customer Pull

Another core lean principle is to let the customer define and pull value. At the moment, not everyone who wants the vaccine is considered most deserving, and not everyone who should get it is willing to get it. Managing this demand-supply imbalance introduces a lot of complexity. We just don’t need this for process design purposes, if we take the goal of 100 million in 100 days at face value. Just follow the lean principle and let the customer pull value.

Currently, we are putting energy towards trying to figure out who gets the vaccine first, prioritizing, scheduling and validating that the most at-risk or most deserving are showing up at the vaccination centers. The simplest way to vaccinated 100 million people is to give it to whomever wants it. The people who want it will get it quickly. Will some people who are “less deserving” get it before some at-risk people? No doubt. But sooner more people who are vaccinated in absolute numbers, the sooner we will all be safer.

Of course, this is based on the giant caveat that we have enough vaccine to supply based on demand-pull rather than scheduled-push. But availability is an issue that needs solving regardless.

Fortunately, about sixty percent of Americans surveyed in November 2020 said they would get the vaccine. This percentage is also growing. The goal is 100 million people. The U.S. population is north of 300 million. If we can design a process to deliver it, at 60% we’ll get more than enough takers for the 100-day goal.

Designing Delivery Systems to Flow Value to the Customer

Lean processes are designed under the principle that they flow continuously until they deliver value to the customer. This means connecting processes and moving the delivery as point close to the customer as possible. This not only minimizes transportation losses it also enables customer pull. When we design this system not around the customer but around the efficiency of the vaccination centers, we end up with big arenas, lots of people movement, traffic, logistics and other non value-added effort. Instead, we need to be scientific and pragmatic how may vaccination centers we stand up and where we put them.

What’s the easiest way to reach 100 million people? First off, we’ll need to access to about 167 million people, assuming a 60% of the population will show up. The top 40 metropolitan areas in the U.S. are home to about that many people. For ease of calculation, and so that we don’t completely ignore low population areas, let’s add ten more to make it 50 hubs.

Image source: https://www.statista.com/statistics/183600/population-of-metropolitan-areas-in-the-us/

Now we can return to our takt time. Instead of one giant multi-armed robot operating at 0.0864 seconds per poke, we have 50 hubs each with a target of one vaccination every 4.32 seconds (0.0864 sec/poke/hub x 50 hubs) . That’s still too fast to be safe or sustainable. But if we go with additive capacity, each “hub” or metropolitan area will have multiple centers, probably dozens in each metro area. Let’s say 50 locations on average per metro hub.

We multiplying our 4.32 second takt again by fifty centers per hub, now we are at 216 seconds per vaccination. That’s just over three and a half minutes for a person to sit down, swab their arm, poke, hold gauze and tape on, and stand up and walk out. I’m not a nurse, but I’ve been vaccinated before. If ask people to get really good at doing just that, I think we can poke a person and move them along safely every three minutes. If we can’t, we add more centers, based on the math.

Keeping the Doors Open

The takt time math above, and the resulting number of 250 centers across 50 metro hubs, is based on a non-stop operation, 24 hours per day seven days per week. This assumption may or may not be reasonable across all metro hubs. If so, we’ll need to redo our calculations based on adjusted available times per center.

For example, if all vaccination centers were only staffed 8 hours every day of the week, we would need triple the number, or 750. If the plan was five-day weeks, with only 40 hours of available time per week instead of 144, the number of centers would need to be 30% higher.

News reports suggest that the Biden administration is planning for “a few thousand locations”. It would be interesting to see their number, the math, logic and assumptions behind it.

These are calculations for demonstration purposes only. We shouldn’t ignore rural areas, low population states or the smallest cities if we want to contain the virus. Once we have a viable, scalable model, the feds should fund local efforts to stand up vaccination centers bottom-up. But this is the sort of math-based thought process that lean thinkers use to design future states of value streams.

Mapping the Entire Value Stream

Relatively speaking, figuring out where and how many centers to stand up in order to vaccinate 100 million people in 100 days is the easy part. The value-added process of vaccination could barely be more basic. Walking people past a needle and poking them with it every few minutes is something we can figure out how to do at scale. We know how to calculate the takt time, measure cycle time per person vaccinated, and estimate the number of units of additive capacity in each metro hub.

It’s the coordination of all of the ancillary steps that will trip us up, unless we keep it very simple. We’ll need to be careful not to go so fast that we outrun our supply lines. We need to look across the entire value stream and map its steps. How will the 100 million people arrive across the centers? How will the materials and the federal workers all arrive at the point where the value-added work happens? Again, the smaller, simpler, additive solution makes managing these flows easier.

Regardless of the delivery system, we’ll need to figure out how to move large quantities of vaccines safely, on-time and under conditions that maintain their quality to many hundreds of centers. How much stock will we need at each point to keep the supply lines moving steadily? How much buffer will we need at various handoff points to account for delays due to weather, spoilage or other unforeseen losses? The fewer steps there are between production of the vaccine at the factory and its injection into a person’s arm, the better it will be. Ideally, we would manufacture the vaccines in all 50 hubs, within miles from their consumption point.

Embracing Uncertainty and Experimenting

There is a lot of uncertainty when trying to tackle a big, ambitious projects like this. Leaders are not comfortable with uncertainty. When the stakes are high, people fear failure and demand certainty on how we’ll get something like this done. Unfortunately, it’s not always possible to find these answers. The right question for one regional hub might be the wrong one for another. There is too much unknown for a top-down solution to work.

At the same time, there are aspects of this that a fully bottom-up, leave-it-to-the-Sates approach will be unable to address. The federal government pays for and supplies the vaccine. Only they have the agencies such as FEMA, CDC and manpower and other services to coordinate this across the whole country. The key is to avoid blame, passing of the buck, or hiding the truth. We can succeed by agreeing to experiment with processes, failing in good faith, communicating, learning and quickly bouncing back, together.

Calling All Lean Thinkers in the U.S. Government

There are many people across various U.S. federal government agencies and services with the knowledge and expertise to do what’s described above. Lean management, while not yet widespread in government, has made significant inroads. Are there people at the highest decision-making level with the knowledge that such process design practices exist? Are there leaders with the curiosity to Google, “how to vaccinate 100 million people in 100 days”? Hopefully, some of them are in positions of influence where their voices can be heard. Time will tell.

10 Comments

  1. Anand Nicodemus

    January 19, 2021 - 6:41 pm
    Reply

    Jon,
    Firstly, wish you the very best in 2021.
    Really appreciated this well written article loaded with the science of takt and capacity.
    Let us know if we – a not for profit organisation Lean Practitioners Canada – a subscriber to the Gemba Academy material can make a reference to this and advise our members to read it.
    Thanks Jon.
    Anand

    • Jon Miller

      January 19, 2021 - 9:10 pm
      Reply

      Hello Anand
      Nice to hear from you again and best to you in 2021 as well.
      Yes, please feel free to share this article with others.

  2. Mark Graban

    January 21, 2021 - 4:19 pm
    Reply

    Great post, very detailed Jon.

    The goal announced by the new Biden administration is “100,000,000 shots” or 50,000,000 people in 100 days.

    https://www.cnn.com/2021/01/21/politics/biden-national-coronavirus-plan/index.html

    As Dr. Deming said, “By what method?” There’s a lot of work to be done… but having a goal and support behind this helps. The details aren’t trivial, as you’ve written about so well here, but it’s solvable.

    • Jon Miller

      January 21, 2021 - 4:48 pm
      Reply

      Thanks Mark. With the Defense Procurement Act being put in motion today, I’m optimistic that we are finally starting to do some of the “obvious” things that need to be done.

  3. Chris MItchell

    January 21, 2021 - 5:12 pm
    Reply

    Jon,
    Glad to see the politics side of the equation is missing, politics just makes it all fuzzy math anyway. Great article and I did steal some clips for my Facebook. Twitter put a disclaimer that the 2021 election results are in… Man, let it go and enjoy the science.

  4. Mark Graban

    January 22, 2021 - 8:46 am
    Reply

    Here is an article that says we are actually almost at “1 million people a day”:

    https://www.nytimes.com/2021/01/21/health/biden-covid-vaccine-supply.html

    “President Biden’s goal of one million shots a day for the next 100 days, they say, is too low and will arguably leave tens of millions of doses unused. Data collected by the Centers for Disease Control and Prevention suggests that the nation has already reached that milestone pace. About 1.1 million people received shots last Friday, after an average of 911,000 people a day received them on the previous two days”

    Currently, vaccine production and availability is said to be the bottleneck…

    “…federal health officials and corporate executives agree that it will be impossible to increase the immediate supply of vaccines before April because of lack of manufacturing capacity. The administration should first focus, experts say, on fixing the hodgepodge of state and local vaccination centers that has proved incapable of managing even the current flow of vaccines.”

    BTW, here is a video from Salem Health that shows their process at a decent level of detail:

    https://valuecapturellc.com/salem-health-presentation-on-their-covid-vaccine-process-and-lean/

    • Jon Miller

      January 22, 2021 - 11:41 am
      Reply

      Mark,

      Thanks for sharing the case study from Salem Health. It gives me hope. Keep up the good work at Value Capture!

  5. Mark Graban

    January 26, 2021 - 8:24 am
    Reply

    Updates — the goal has been raised to 1.5 million injections per day.

    And, it’s said that vaccine production will be higher:

    “Pfizer and Moderna, makers of the two vaccines that have been authorized for emergency use, have promised 200 million doses, enough for 100 million people, by the end of March.”

    I’d like to know the rate of how many doses per day, or doses per week, that it represents… not just the cumulative number.

    https://www.nytimes.com/2021/01/25/opinion/covid-vaccine-plan-biden.html?referringSource=articleShare

  6. Shelby

    February 9, 2021 - 2:54 pm
    Reply

    Hey Jon!
    This was a great read! I am currently earning my yellow belt for lean six sigma and found this very insightful to relate back to our hectic world. I completely agree with your point that to succeed we must learn from our failures, “failing in good faith” and bouncing right back to improve the process. Due to process improvements, do you think the original rollout vaccination plan will be on the slower side initially?

    – Shelby

    • Jon Miller

      February 9, 2021 - 3:00 pm
      Reply

      Hi Shelby
      I’m cautiously optimistic that the vaccination will pick up pace and be close to hitting the 100 million target. Availability of vaccine is the constraint in many places. This is giving people time to prepare for the delivery process. The key is to put a good process in place that we can ramp up, adjust and duplicate, not hit a target at an artificial date.

Leave a Reply

Your email address will not be published. Required fields are marked *