I’m not a scientist or medical researcher, just a regular working person who has been in the workplace a long time. Today we face challenges related to a pandemic. The situation has been described as unprecedented. On the disease management front that may or may not be true; I’m not qualified to comment. However, some of the business challenges created are very familiar to me. They’re not unprecedented – or even new – they’re just common problems faced by anyone operating a manufacturing operation. This article will review some Lean manufacturing tools that St. John’s Health Care Corporation has applied to manage the health of employees and residents in a skilled nursing facility.

Opportunity 1: Entrance Screening and the Seven Wastes of Lean

In March 2020 it was apparent that we had to change our business processes in order to continue operations in a pandemic. The skilled nursing segment was determined to be an essential function, and having employees work remote wasn’t an option for most of our workforce. We began to wear masks and cancel anything we could that grouped people together. We established a health screening function for employees as they reported for work.

I don’t believe anyone on the St. John’s team, even those with deep clinical backgrounds, had ever given a lot of thought to managing thousands of entrance health screenings every week. We’ve all been to airports so were familiar with the concept but never imagined running these kinds of operations ourselves.

We are very proud of the team at St. John’s ability to, in rapid order, set up an effective entrance screening operation. It took us only days to get things going. After only a month we had a model in place that could effectively handle 600 screening activities each day without disrupting work times. (One downside of a pandemic is that it kind of destroys your interest in and ability to celebrate, but the work our team did was worthy of celebration.)

Our initial problem that we had to solve was managing entrance screening. How were we going to pay for entrance screening? We decided (or defaulted) to put the screening cost in the human resources budget. Our annualized HR budget went up by 84 percent and the number of employees in the group more than doubled.

We looked at our entrance screening as if it were a screening factory, and applied some simple Lean tools to processes. We used the 5 Whys of the Toyota Production System. We challenged every assumption about what we were doing in the process to find waste, and we first were able to attack the waste of over-processing.

Screening is a good tool. It encourages people to inform us about issues concerning their health. It’s great to have a culture of keeping infection out of the workplace, but at the same time we need to recognize the limitations of entrance screening. We’re not the TSA. We can’t check travel records or search pockets for lozenges and used tissues to determine if you might be ill. We rely fully on the honesty of people being screened.

In time we concluded that having paid screeners asking everyone about their health and recording temperature readings was, in fact, an over-processing waste. We could display the screening questions in print, and ask people to let us know if they had symptoms or travel that might concern us. We also mounted touchless thermometers and let people check their own temperature. Unfortunately, our facility was shut down during the pandemic and we had no visitors. However, the lack of visitors provided the opportunity to do self-screening as everyone being screened completed the same operation every day.

You may wonder about self-administered temperature screening. Can’t people just falsify their answers? Yes, they can. You have to go back to the 5 Whys and ask what value the screening really brings. When we looked at the screening data, we had done over 50,000 temperature screens and had no employees disqualified for work because of fever. That’s because entrance screening for temperature has some unique limitations in the Finger Lakes region of New York. Except for a welcome time in the summer, which is a three-week period from mid-July to early August, the rest of the year is quite cold. If you look at our body temperature data from screening when people first walk in the door you would conclude, based on commonly accepted science, that most of our workforce was in the early stages of hypothermia. The plot of the temperature recordings we collect are almost exclusively below 97 degrees. So, temperature screening is an idea, and you should not be at work if you have a fever, but we have to rely on you to tell us if you’re feeling unwell; we won’t “catch” you.

The move to self-screening helped us recover some of the cost increases we incurred, but we were still spending too much on the process. We started looking for other wastes and focused on what Lean processes describe as the waste of waiting. Remember that we were shut down to visitors? We started to examine this waste buried in a number of other jobs. We had people who, prior to 2020, spent part of their time managing visitors to our site. But now we had no visitors. Although we couldn’t quite see it, we knew that there had to be spare capacity in some jobs with a visitor focus.

We had a number of cross-function meetings trying to determine how to tap this capacity hidden in a number of different jobs. We worked to try to reinvent our work for pandemic times. (If I wanted to try to sound like a Lean guru I’d tell you this was a Kaizen event, but we just called it a bunch of folks kicking ideas around.)

In one instance we took a concierge function and relocated it to the screening room location so people – highly skilled with managing people coming through the doors – could handle vendors and medical providers coming to our campus, as well as oversee screening functions. In another situation we relocated a screening station to an area where security had an overnight presence, and our very capable security officers could also manage screening for overnight employees.

The result of our intense focus on waste reduction, using tools like the 5 Whys and Kaizen, was that we reduced 70 percent of the additional costs focused on screening. Now this is not as good as finding a way to conduct screening with no additional costs, but it got us a lot closer a cost structure we may be able to live with.

We believe our screening process has been successful as an infection management tool. We conduct extensive contact tracing on each positive employee case we encounter, and with nearly 200 of those, we have a good sample size. We have no data suggesting any of our COVID-19 cases among employees (or residents) was related to a failure of entrance screening. We believe that screening has prevented a number of cases as several dozen employees have informed us of exposures outside of work, and we have been able to test and quarantine those employees with full salary while we determine the risk they bring to our workplace.

Opportunity 2: Using Statistics to Understand Vaccine Behaviors

If you visit any modern manufacturing facility you will notice that everything is measured and plotted. After a while, it all starts to seem quite normal. I’ve spent most of my life in this world and have come to appreciate the value of statistical applications in problem solving.

St. John’s was one of the fortunate healthcare facilities offered the COVID-19 vaccine starting in December 2020. We spent time trying to decide how we would prioritize limited vaccine availability among our workforce, but then something happened that surprised us. Despite 20 percent of our employees testing positive for COVID in 2020, many who were quite sick, only 40 percent of our employees signed up for the vaccine. In some way this was great because it saved us from the horrible dilemma of rationing supplies, but we had hoped for more interest.

We began to run a number of excellent communication and education events to attempt to increase vaccine interest. We wondered why people didn’t want the vaccine, and did surveys, but didn’t come up with much of interest. Then we noticed some trends. Older people signed up in greater numbers for the vaccine which was not surprising, but people of color in our workforce, representing 60 percent of our employees, seemed to be much less interested. We turned to statistics for help.

The statistics we used weren’t of the process control variety, but grow out of the Lean discipline of making decisions based on data. We are required to collect and report all kinds of data on our employees about the vaccine, and the data reminded me of something from a long time ago, my college master’s thesis.

I had this idea in college that I could predict job attachment (turnover) based on the design of benefit incentive employees were offered by their employers. I predicted that employees who had unvested pension plans would be less likely to leave their job than employees with vested plans. I used multiple regression analysis to look at relationships among variables.  (Unfortunately, I had this idea before Excel was invented, so the analysis was somewhat time-consuming.)

I was unable to demonstrate the validity of my hypothesis, thus beginning a long career of me guessing wrong. It turns out that even though I had thousands of data elements to work with, the data was from an earlier generation when hardly anyone actually left their employer. So I couldn’t demonstrate my hypothesis because not enough people in a rather large sample quit their jobs.

In 2021 we employed the same statistical tool to look at relationships among variables concerning vaccine interest. I copied our regression results below, and the greatest factor we found influencing our employee’s vaccine decision was not age, but race and ethnicity.

We were surprised that age wasn’t the most predictive factor of vaccine decisions. I think this is what is happening. If I walk into a group of 100 randomly selected people and have to wager concerning who is most likely to want a COVID-19 vaccine, I’d be smart to bet on the oldest person in the room.  I think what our data showed us is when you work with a large number of ages ranging from 16 to 79, age doesn’t predict as well because there’s no difference in vaccine interest among people aged 31 or 43. The interest may only significantly peak after age 60.

Our experience, given some time to reflect, doesn’t seem at all surprising. A January study published by Hannah Recht and Lauren Weber at the Kaiser Health Foundation was one of the first to report this trend. As we were one of the first places in the U.S. to administer the vaccination process in December 2020, we just weren’t aware of the issue.

Our learning was valuable, however, in that it helped us focus future communications on groups who may need more information to get comfortable with a vaccine decision. We have better recent experience with vaccine participation for people of color. We haven’t found that communication campaigns alone make a difference, but people who have received the vaccine sharing their personal experiences, especially if they are people of color themselves, has permitted us to more than double our interest in underserved populations in our current round of vaccine administration. This is not data, but my impression is that sharing vaccine experiences of working people has had a greater impact than sharing studies or having prominent national figures discuss the issue.

Opportunity 3: Safety Compliance

As St. John’s is in the senior life business, we are fortunate to have every known tool available to us to fight the spread of COVID-19. The federal government and the state of New York have provided us with incredible resources to help our residents and employees stay safe. We’re going to discuss personal protective equipment (PPE) but let’s first take a quick diversion and look at COVID-19 testing.

In addition to entrance screening and vaccines, another tool at our disposal is COVID-19 testing. All our employees are tested one or two times each week. There are two types of COVID testing we use; both have some limitations. Genetic-based polymerase chain reaction (PCR) tests are highly accurate, but it takes us an average of five days to get results. So depending on infection and test dates, one of our employees can have an asymptomatic COVID infection and work a week before we know about it. The other type of test is a rapid point-of-care antigen-based test. The results from these tests are immediate, but the tests don’t always detect asymptotic infections accurately. Neither tests type keeps people with an asymptomatic infection out of work, which is a problem when we believe that asymptomatic spread is the greatest threat in our workplace.

Entrance screening, testing and vaccines are all good and necessary, but they all have limitations in keeping infection out of the workplace. We have one other tool at our disposal, and that is the use of PPE. PPE is not new or high tech and had been extensively used as a job tool far before the rise of the pandemic. We have entered a period where our employees need to wear masks, face shields, gowns and gloves to protect themselves and our residents. Wearing all this gear isn’t comfortable, and so we have chronic worries of getting people to comply. The challenge is how to get compliance and, again, statistical process control tools can help us out.

Staffing was critically short in 2020 and continuing to 2021 for a number of reasons. One is that people are afraid to work in healthcare, and another is that there is financial support available to people out of work. Most kids are also not in school and childcare issues prevent many from working. When your staffing is at crisis levels, some tools for PPE compliance aren’t really available. Discipline-based tools might work, but they also might upset employees and encourage them to leave, which is a horrible outcome. So, what to do?

PPE compliance is an old problem in both manufacturing and construction. PPE historically wasn’t face masks, but rather safety glasses and lock-out tag-out systems, fork truck speed management, machine guarding, etc. A statistical-based tool commonly used in these types of businesses is something known as behavioral-based safety. It relies on observation of behaviors (such as PPE compliance) tracking data trends, problem solving based on data and coaching.

We’re using behavior-based safety; observing and collecting data addressing problems and trends with coaching. It isn’t perfect and we still have challenges, but at least we have a tool to work with and develop. Generally, you are more likely to get results if you try something than if you just curse the darkness.

Our observation data is showing that PPE compliance is sustained at greater than 90 percent over multiple months, and employees often coach their peers to maintain compliance because nobody wants to get sick. In some ways, this situation reminds me of safety compliance in chemical plants, where peers encourage compliance because noncompliance by one person can endanger many.

Our compliance still isn’t at 100 percent; we’d like to do better. One creative approach some our clinical areas are taking is creating safe zones where employees can safely distance to get a break from wearing full PPE gear during the day. It’s too soon to know if this will increase compliance further, but it’s a clever approach we hope will help us.

Conclusion

In healthcare these days, we’ve got 99 problems and the ones we’ve reviewed here are just some of them. Among the situations we’ve addressed, I wouldn’t consider them solved, but do consider them being attacked. Lean tools developed in the manufacturing world may offer solutions to new problems in very different industries. In some instances, the St. John’s team has achieved remarkable cost and infection reductions, and in other areas we may not have results yet but do have tools helping us approach our new challenges. Being an essential worker is stressful for everyone. One of the most effective techniques to reduce stress is to take action to gain some control over your circumstances. What we’ve reviewed here are just that; they are actions we’re taking to get control in an incredibly uncertain environment.