How to speed mass vaccination

How to speed mass vaccination

Daniel Penn Success Stories

Process improvements at Hartford’s Dunkin’ Donuts Park clinic increase COVID-19 vaccination rate by 207 percent in five weeks

By Tony Rodriguez, President, and Mike Beauregard, Senior Consultant, Daniel Penn Associates (DPA)

Emergency use authorization of the Pfizer BioNTech and Moderna vaccines ignited hope for stemming COVID-19 infections and deaths in 2021. It also pushed city health departments’ plans into overdrive to vaccinate their most vulnerable and marginalized residents.

Three months before vaccines became available in January, the City of Hartford, Connecticut began working with the region’s hospital systems, community clinics, and volunteers to coordinate its mass-vaccination infrastructure, according to Hartford Health and Human Services (HHS) Director Liany Arroyo.

To speed shots in shoulders, cities’ mass vaccination sites need their processes to flow like warm syrup – smoothly and quickly. To complement the work of its mobile vaccination clinics and expand access, Hartford needed an indoor vaccination site close to all city neighborhoods, bus lines, and with free parking. Leaders of the four-years-new Dunkin’ Donuts Park (home of the Hartford Yard Goats minor league baseball team) offered the stadium’s event space for a clinic.

“Working with the park’s management team, multiple city departments, our community organization partners, regional medical reserve and youth services corps, it took just two weeks to get our first Saturday clinic up and running at Dunkin’ Donuts Park in early February,” Arroyo said. “We then added volunteers from local colleges and universities who had been supporting our mobile vaccination effort.”

Lightning-speed setup of the clinic was challenging. “Our biggest dilemma was how to keep everything appropriately spaced. We tried locating different elements on two floors, then three floors. There was a lot of angst and nervousness among staff and volunteers on how to make it work,” said Arroyo.

To help Hartford’s Department of Health & Human Services offer more vaccinations faster at the Dunkin’ Donuts Park Saturday clinic, the Daniel Penn Associates (DPA) team volunteered our operational expertise to Hartford Mayor Luke Bronin and Vas Srivastava, his chief of staff. They supported the idea and introduced us to Ms. Arroyo.

Applying the lean principles, we use to improve our clients’ operations, we identified ways the clinic could make simple, non-cost-intensive changes to shave time from each step of the process, smooth flow and speed throughput. The DPA team helped them apply more visual controls, balance their processes, and eliminate wastes, especially the Waste of Waiting. We helped improve throughput by 207% in five weeks. These improvements set the stage for the Saturday vaccination clinics to nearly quadruple their output. Here’s the story.

The Team Effort

We started working on the process on February 27 at the Dunkin’ Donuts Park clinic.

During our observations and data collection, we met on site with Ms. Arroyo and her staff, Dr. Bruce Gould, Hartford Yard Goats President Tim Restall, volunteers from local universities, and some of the volunteer nurses and doctors administering the vaccines. The city’s health and human services staff welcomed us to their team. “Tony and Mike worked respectfully and unobtrusively, observing the entire process and making notes,” said Faith Palmer, Operations Manager for Hartford’s Health and Human services department.

Unique Challenges

From the very beginning, the city’s health and human services team worked hard to create an efficient vaccination clinic at Dunkin’ Donuts Park. But to improve the operation, they faced challenges. The Saturday prior to our first visit, the Hartford team scheduled and vaccinated 274 patients. The initial goal was 500 vaccinations at each 10 am to 3 pm Saturday clinic. Achieving this goal would require overcoming some constraints:

  1. Elderly population – Connecticut was still in the 65 and older age brackets for vaccinations.
  2. Mobility issues – many of the patients had limited mobility and needed assistance.
  3. Language barriers – some patients spoke no English.
  4. Space constraints – Dunkin’ Donuts Park is a great place to watch a baseball game. We love it. But it is not optimal as a vaccination clinic site in the late winter/early spring.
  5. Volunteer-supported processes – while many from Hartford’s professional staff work the vaccination clinic, the bulk of the personnel are volunteers with many changes in faces from week-to-week.
  6. Budget constraints – vaccination clinics are important, but city resources are limited.

Initial Effort

We observed the process and broke it down into these steps:Covid-19 Initial Process

Evaluation 1 (E1) = general evaluation time period of 15 minutes for most people after their vaccination. Evaluation 2 (E2) = 30 minutes post-vaccination evaluation period for patients with some medical conditions.

We then collected data on each of the process steps.

Covid-19 Initial Process Steps

Check-InTemp CheckWaitRegistrationVaccinationEvaluation
Cycle Time1.3 m0.3 m0 to 16 m6.7 m4.8 mE1 – 15 m
E2 – 30 m
Staffing2 stations2 stations31 chairs9 stations28 stationsE1 - 6x4 =
24 chairs
E2 - 12 chairs

This gave an average total cycle time of 36.1 minutes for patients in the 15-minute evaluation grouping and 51.1 minutes for patients requiring a 30-minute evaluation.


With a Takt Time of 0.6 min/patient, even with 9 stations, registration was the bottleneck. Under the clinic’s initial operating practices, this would limit vaccinations to 403 patients per Saturday session. On the opposite end of the spectrum, the actual vaccination step had far too much capacity. This was supported by other data we captured where random samplings revealed 14.6 vaccination stations idle at any given point during the day.

Registration as the Bottleneck

We focused on registration. We found that the average wait time was being driven up by four causes.

  1. Language barrier – there was often a need to find and then use a translator in the registration.
  2. Health screening – the registration staff was filling out the pre-vaccination questionnaire for each patient.
  3. Patients were not ready for the registration process.
  4. Patient information was incorrect or missing from the Vaccination Administration Management System (VAMS).

To improve the registration process, we recommended:

  • Take advantage of the clinic’s bilingual registrars by focusing their efforts on serving non-English speakers.
    • Post colorful “Hablo Español” signs on the plexiglass in front of registrars who speak Spanish.
    • Guide Spanish speakers to those registrars.
  • Move health pre-vaccination screening from registration to the waiting area to balance flow in the process.
    • Provide medical professionals in the Waiting Area to assist in filling out the questionnaire. [Having the patient fill the questionnaire out on their own was deemed to be impractical because of the challenges discussed earlier.]
  • Have the patients ready for the registrars.
    • Put up large, colorful signs in the Waiting Area and along the path to Registration to remind patients to have their IDs, health insurance cards, health assessment forms, and vaccination card (if second vaccination) out and ready for the registrars.
    • As the volunteers release patients to Registration, remind them to have their documents out and ready for the registrar.

Other Improvement Ideas

  • Place laminated instruction sheets at every Registration station.
  • Designate a volunteer responsible for escorting people from registration to the vaccination area.

Other Process Steps

Other opportunities for process improvements included:

  • Only allow patients scheduled for the active time slot to enter the Gate and Check-In.
    • Having Check-In start to process patients and then have to refuse them because their time slot was not active yet bogged down the line there.
    • In inclement weather, allow all in, but direct those with later time slots to a bullpen area.
  • Move the temperature check from just prior to the Waiting Area to the arrival and Check-In area at the main gate.
    • This was also a safety measure to prevent infected patients from moving through the waiting area.
  • Increase the use of visual controls in all steps in the process.
  • Have the patients ready to receive the shot as they enter the Vaccination Area.
    • It was taking some patients up to 2.5 minutes to peel off layers to free up an arm for the injection once the medical provider was ready.
    • If they could not be ready before arriving at a Vaccination Station, the escort should help them get ready as the Medical Provider started the Health Assessment review.
  • Develop one-page laminated instruction sheets for all volunteer functions and use them in the pre-clinic training sessions.
    • This will consistently train the ever-changing volunteer staff and give them a reference to use throughout the day.
  • Cut down the overall number of volunteers and the number of vaccination stations.
    • The need to manage extra people was adding unneeded confusion to the processes.

Revised Process

Based on the modifications we recommended to each part of the vaccine process, the clinic’s steps would become:

Covid-19 Revised ProcessWe projected the cycle times if our recommendations were implemented as:

Revised Covid-19 Vaccination Process

Estimated Cycle Time2.4 m1.5 m with 6.5 wait4.2 m2.8 mE1 - 15 m
E2 - 30 m

This would yield a total estimated cycle time of 32.4 minutes for patients in the 15-minute evaluation bracket and 47.4 minutes for those in the 30-minute evaluation bracket.

The DHHS team decided to make these improvements the following Saturday.

Dramatic Improvement After Two Weeks

The Dunkin’ Donuts Park clinic’s team worked hard to implement DPA’s recommendations. We returned to the clinic on Saturday, March 13 to determine the results of the improvements, look for further improvements, and find opportunities for refinements.

As a result of more preparation, training, tools, and staffing in the pre-registration waiting room, patients received support in filling out their health screening forms there. Once their forms were completed, they waited less than a minute to move to the staging area in front of the registration.

Despite occasional bottlenecks in the registration areas, the team had reduced patient registration times from an average of 6.7 minutes to an average of 2.6 minutes.

In the vaccination area, we recommended two people to fill syringes and a decluttered layout for the syringe-filling tables, changes to the placement of supplies and syringes, and ideas for stamping lot numbers on patients’ vaccination record cards.

That Saturday, it took patients an average of 25 minutes to move completely through the vaccination process, a 31% improvement compared with 36 minutes on February 27. The number of patients vaccinated on March 13 jumped to 430 from 274 two weeks prior, a 57% improvement. More patients could have been vaccinated if more Harford residents had signed up.

Outstanding Improvement After Four Weeks

By March 27, the clinic team’s continuous improvement efforts were really paying off.

The registration process had been bifurcated to serve patients who required language assistance. The bulk of the patients were now visible in the VAMS system. The time in Registration for the downloaded patients ranged from 1.1 to 2.2 minutes. Patients not in VAMS ranged from 2.5 to 5.6 minutes. Overall, the patients sampled averaged 2.2 minutes in Registration.

The nine tables in registration were balanced with twelve stations in Vaccination. The DHHS staff elected to staff more stations that theoretically were needed on average to account for the large wave at the beginning of each time slot. With this approach, no more than three patients were waiting at Vaccination at any given time. This occurred at the 11 am peak demand for the day. Those three individuals waited for less than one minute. For the most part, people walked over from Registration and were immediately taken to a Vaccination Station.

Most patients were ready for their vaccination as they sat down or got ready while the health assessment was being reviewed.

We observed no more than two idle vaccination stations until the scheduled lunch period – a remarkable change from the first week we visited the clinic. The HHS workers were quick lean learners and applied lean techniques themselves too – turning the vaccination tables 90 degrees to allow movement through the area more quickly.

In the samples taken, Vaccination averaged 3.0 minutes per patient with more experienced vaccinators being statistically faster than student vaccinators.

The DHHS staff asked us to review the syringe Drawing operation again. We found there was plenty of capacity with three medical professionals assigned to that operation; however, it was being affected by interruptions. Not all of these interruptions could be controlled, so we elected to try to increase the capacity when they were loading syringes. We suggested an assembly line approach with a spider: one person assembling syringes and one person drawing the syringes with the spider bringing needles and the vaccine to the drawing station and running filled syringes to the vaccination stations on regular ‘milk runs’. Vaccination stations would be filled to a visual kanban of 5 units until after 2 pm when that would be adjusted down as demand dropped at the end of the day.

The following Saturday, the DHHS team with their volunteers hit 840 vaccinations (including 90 extra patients beyond reservations). With their new processes and the continued application of lean techniques, they can easily achieve 1000 vaccinations, 100% more than their initial goal, which is critical as vaccine registration is now open to all age groups in Connecticut.

Feedback from the Field

“The DPA team suggested small process improvements that none of us saw in the aggregate that in fact affected the entire operation. Having their outside-eyes view how all the elements of our vaccination process work together helped us get a larger picture of our flow.” ~ HHS Director Arroyo

“The DPA team made observations and provided detailed reports. They met with us, told us why they were making their recommendations, and asked for our feedback. When our daily numbers were low, they gave us the confidence we had the foundation to increase our vaccination rates.” ~ HHS Operations Manager Palmer

According to Arroyo and Palmer, Hartford has applied a few of the practices DPA suggested at the Dunkin’ Donuts Park clinic in the city’s other vaccination efforts. “We’re helping people fill out their pre-vaccination surveys right as they line up at our mobile clinics and at our health department site before they register and wait for their shot. Our staff have the clipboards they need. We’ve improved our directional and instructional signage at all locations,” ~ Arroyo explained.

Hartford’s vaccination team leaders are also empowered to make suggestions and quickly improve their processes without having to take their ideas up and down the approval chain. “The leaders and backup staff and volunteers in each section we have put in decision and leadership roles are now more purposeful and powerful in seeing what needs to change and make those changes,” ~ Arroyo added.

The City of Hartford’s Dunkin’ Donuts Park DHHS vaccination team can be extremely proud of the improvements they put in place during five weeks of Saturday clinics. The entire vaccine process is running smoother. The number of people being vaccinated in one day has increased from 274 to a record daily high of 840, a 207 percent improvement. As process flow became more efficient, people we queried as they exited the facility after getting their vaccinations unanimously thought the process ran very smoothly.

Tony Rodriguez is president and Mike Beauregard is a senior consultant at Daniel Penn Associates, LLC.


Ready to discuss your operational goals and challenges?

We’re ready to listen. Book a discovery call or meeting with us.

Share our content