The Nurse in the Stair Closet

Prefer to listen in your podcast app? Apple Podcasts · Spotify

In 1897, New York City hired one hundred fifty doctors to inspect schoolchildren for about an hour a day.

They were checking for contagious disease: trachoma, ringworm, scabies, impetigo, lice. The Board of Health wanted classrooms free of anything that could spread. So the doctors examined, diagnosed, and wrote exclusion notices. Sick children went home. The doctors moved to the next room.

That was the whole job. Find the disease, remove it from the building. No one was tasked with confirming the note reached a parent, or that a parent could read it, or that a family could afford a doctor. No one tracked whether a child came back treated, came back still sick, or never came back at all.

Inspection happened. Treatment did not.

In September 1902 alone, the city excluded 10,567 children from school.

The city hadn't set out to just count sick kids. It had asked a reasonable question: how do we keep contagious disease from spreading through a classroom? The system it built answered that question well. One hundred fifty physicians, one hour a day, a steady stream of children sent home—in fact more than 10,000 children excluded from school in a single month.. It just answered a narrower question than the one the city actually needed answered. Protecting the classroom and healing the child turned out to be two different jobs, and New York had only staffed one of them.

Somewhere along the way, the proxy quietly became the mission. The city became exceptionally good at measuring the beginning of care while almost completely overlooking whether care actually happened. We still do this. Organizations count what is easiest to count and assume it represents what they truly value. Sometimes it does. Often it doesn't. A completed inspection is measurable. A child walking home untreated with trachoma, never returning to school, is much harder to capture. The numbers looked like success because they described activity. They said almost nothing about outcomes. There is a name for this pattern, and it's one worth knowing. Economists call it Goodhart's Law: when a measure becomes the target, it stops being a good measure. This is one of my favorite discussions, but we'll come back to it in a future Field Note. For now, it's enough to notice what happened here. The city counted what it did, mistook the count for care, and thousands of children disappeared into the gap between finding a problem and solving it.

Enter the Board of Health, and a borrowed nurse.

Lina Rogers, a Canadian-trained pediatric nurse living at the Henry Street Settlement, started on October 1, 1902, covering four schools with about ten thousand children between them. In one of them, Old School No. 12, the only space anyone could find for her was an unused stair closet. She couldn't stand up straight in it. A radiator did duty as her dressing table, and a highchair rescued from an ash heap served as the chair for her patients. Her first patient in that closet had a rat bite.

It's easy to romanticize that image. One nurse. One closet. One impossible job. But it was work that took a singular focus on the now, an ability to hold aspirational outcomes, and the grit to close the gap between the two.

In her first month, Rogers treated 829 cases, made 137 home visits, and returned ninety-three already excluded children to their classrooms.

That's the version of this story we usually celebrate. One nurse, one month, one thousand kids seen.

It isn't the important part.

The important part is what happened next.

The city read her numbers and understood something the hero story misses. Rogers wasn't the fix. Her results were the diagnosis, proof the system was missing nurses, not proof it was missing a hero.

On November 7, 1902, New York formally appointed Rogers, making it the first city in the world to take financial responsibility for school nursing. Within months the city hired assistant nurses, created the role of Superintendent of School Nurses, funded the program, established treatment hours in every school, created records for every child, and required weekly reports.

By September 1903, exclusions had fallen to 1,101, a drop of nearly ninety percent in one year. By 1914, New York employed 374 school nurses.

Lina Rogers did something remarkable. Then the city did something even more remarkable.

It refused to leave the work dependent on Lina Rogers.

That distinction matters more than it gets credit for.

Organizations love to celebrate the person who saves the day. The employee answering emails at midnight or going in on weekends. The administrator who somehow keeps the unit running despite three broken processes stacked on top of each other. The one person carrying fifteen undocumented systems around in their head.

We call them indispensable. What we usually mean is that our infrastructure has failed, and one person is covering for it.

In every system I've worked in, there was always someone everyone depended on. The person who somehow remembered every unwritten rule. The scheduler who knew every family's story. The clinician who caught problems no checklist ever mentioned. The administrator who magically knew exactly which notebook to reference. We admired them because they made the impossible look routine.

What I have realized is that these people are not just exceptionally capable. They are carrying invisible infrastructure. Every day they compensate for systems that haven't yet been built. They aren't extraordinary because they are superheroes. They are extraordinary because they temporarily stand in for the organization itself.

This pattern is not unique to healthcare. In the 1890s, Japanese textile mills ran on the same kind of vigilance. A power loom kept weaving even after a thread snapped, so a worker had to watch it constantly to catch the break before it ruined the cloth. One loom, one set of eyes, one job that never let up.

Sakichi Toyoda built a device that stopped the loom itself the instant a thread broke. No one had to watch it anymore. A single operator could run several looms at once, and the cloth coming off each one was more consistent than any watchful worker had managed. The idea became known as jidoka, and it is still one of the two pillars of the Toyota Production System. The lesson traveled from cotton mills to car factories: build the stop into the machine, and you no longer need someone heroic enough to catch it in time.

Heroics are easy to notice because they're visible. Infrastructure is easy to overlook because, when it's working, nothing dramatic happens. No one posts about the checklist that closed off an entire category of mistakes. No one gives a standing ovation for correct documentation.

Yet those are usually the moments that matter most.

The real compliment isn't that someone rescued the organization. It's that the organization stopped to measure what counts, and build what actually solves.

None of this means heroics don't matter. Sometimes a person has to hold the line while the system catches up, and that's real work worth recognizing. But if the rescue keeps happening, around the same problem, the organization isn't looking at a string of heroes. It's looking at a system it hasn't built yet and a field of questions it has not asked.

Rogers didn't stay the hero. She wrote the manual instead. In it, she warned against the kind of paperwork that would 'make complex what should be simple,' a caution worth keeping next to the case for infrastructure, not against it. It would be another 30 years before Einstein came along and shared his own version of this idea for theoretical physics during a lecture delivered at Oxford University. Which I share as another illustration of the brilliant mind of this nurse.

Lina Rogers didn't change New York only because she worked hard. She changed it because the city treated her results as a diagnosis, not a tribute.

Every organization has a stair closet.

A place where someone compensates for infrastructure that doesn't exist yet.

The question isn't whether your organization has one.

It's whether you've mistaken it for evidence of excellence.

Seen this pattern where you work? Hit reply and tell me. That's the harvest.

Want a bit more? Listen on Spotify or Apple for One More Signal about this topic.