Why culture is the most important clinical variable in care

We say it on our website and in my letter: culture is the most important clinical variable in care. Most people hear that as a slogan. I mean it the way a nurse means blood pressure, a number you measure and act on before it hurts somebody.

How sure I am. Sure enough to run the company on it. The studies below tie the way a nurse’s workplace runs to how patients do, in hospitals and in home health. Most of them are observational, so they show strong links, not proof, and there are fewer of them in home care than in hospitals. Nobody has ranked every clinical variable against the others. “Most important” is my call as an operator, and this note shows how I get there.

Nobody from the office is in that room

Think about how a home health visit actually happens. Our nurse parks out front, walks in with a bag and, for that visit, is the whole hospital. There’s no charge nurse down the hall and no doctor on the floor. The nurse decides what to check, what to teach the family, what has to go to the doctor today and what can wait. The aide who comes the next morning decides how much time the bath gets.

Nobody from the office is in that room. Every one of those calls gets made by a person carrying whatever we gave them that week: a caseload they can handle or one they can’t, a manager who picked up the phone or one who let it ring.

That’s what I mean by culture. It’s how we treat the people who make those calls: their caseload and schedule, their pay, their manager, and whether they trust us to back them when they call a doctor at six at night about something small. All of it walks into the house with them.

The best argument against me

The strongest pushback goes like this: culture is a soft word for hard things. Staffing, training and pay are the real clinical variables, and culture is the feeling they leave behind. Fix the hard things and the culture takes care of itself.

Part of that is right. Those hard things are inside what I mean by culture. A nurse with too many patients can’t give any of them a good visit, and a thank-you lunch won’t change that.

Where the research pushes back is on which comes first. A team at the University of Pennsylvania’s nursing school studied 665 hospitals in four states. Giving each nurse fewer patients lowered deaths mostly in hospitals where nurses rated their work environment well. In hospitals with poor work environments, cutting the patient load brought no benefit at all, for deaths or for failure to rescue, which is when a patient dies after a complication that wasn’t caught in time.¹

Same investment, more nurses. It paid off in one kind of workplace and not in the other. The environment decided whether the staffing worked, and that’s why culture goes first for me. It’s the variable that decides whether the others pay off.

Nurses figured this out in 1983

This isn’t a new idea. In 1983, the year I was born and the year Medicare started paying for hospice, the American Academy of Nursing went looking for hospitals that could attract nurses and keep them. Of the 163 it studied, 41 stood out, and they became known as magnet hospitals.² Hospitals now apply for Magnet status, and the first one was named in 1994.²

Thirty years after that study, Penn researchers compared more than 600,000 surgery patients in 56 Magnet hospitals and 508 other hospitals in four states. After accounting for how sick the patients were, death rates were 20% lower in the Magnet hospitals. Better work environments, better staffing and more nurses with bachelor’s degrees explained much of the gap.³

The places nurses wanted to stay turned out to be the places patients were more likely to survive.

What the researchers actually measure

Work environment sounds vague until you look at the survey behind it. A survey used in much of this research is the Practice Environment Scale, 31 questions built at Penn.⁴ It asks nurses about five things: whether they have a say in how the place runs, whether the basics of good nursing care are in place, whether their manager is able and backs them up, whether there are enough people and resources, and whether nurses and doctors work well together.

Every one of those is a decision somebody in management makes, usually in a meeting no patient will ever hear about.

What it looks like in home health

Hospitals are easier to study, so most of the research comes from them. Home health has less, and what it has points the same way.

Researchers from Penn and Rutgers matched nurse surveys from 118 home health agencies with Medicare’s own data on those agencies’ patients. Agencies with good work environments had fewer patients end up in the hospital, and more patients finish care still living at home or in the community, than agencies with poor ones.⁵

Then there’s continuity, whether the same nurse keeps coming. Researchers at NYU’s nursing school followed 23,886 people with dementia who got home health care after a hospital stay, at a large nonprofit agency in New York. About one in four went back to the hospital. Patients whose visits were spread across more nurses had about 30% higher odds of going back than patients who mostly saw the same nurse, after accounting for how sick they were.⁶

Now put that next to how often people leave. In the latest national survey of hospices, about one in four registered nurses left their hospice within the year.⁷ At home care agencies, about one in three aides and nursing assistants did.⁸ Every one of those departures is a family meeting someone new and explaining Mom all over again.

The people who stay can burn out, and that shows up too. A 2024 review pooled 85 studies from 32 countries. Nurse burnout was linked to worse patient safety, including more infections, falls and medication errors, and to lower patient satisfaction and quality of care.⁹

Line those up and you can see the machine. How we treat our people decides whether they stay and whether they burn out. A nurse who stays knows your mom. One who isn’t burned out catches the small change before it turns into an ambulance ride.

At the end of life, it’s most of the care

Hospice makes it even plainer. After a hospice patient dies, Medicare surveys the family member or friend who cared for them and publishes the results.¹⁰ Look at what it asks about: communication with the family, getting help in time, treating the patient with respect, emotional and spiritual support, help for pain and symptoms, and training the family to care for the patient, then the family’s rating of the hospice and whether they’d recommend it.

Most of that is how a team behaves once it’s in the room. I wrote in my letter that at the end, what’s left is people, time and whether there was kindness in the room. Medicare’s survey asks families about the same things.

How we run HealthView because of it

We were told culture and taking care of people were too expensive, and I get why. Medicare pays home health by the 30-day period and hospice by the day, so every hour a nurse spends on anything but a visit looks like cost on a spreadsheet. We spend it anyway.

We measure it. Every year Great Place To Work runs an anonymous survey of everyone here, and we publish what comes back. In 2026, 98% of our people said HealthView is a great place to work. At a typical U.S. company it’s 57%.¹¹ Fortune ranked us No. 16 on its Best Small Workplaces list for 2026.¹²

We put nurses in charge. Monica Cutia, a nurse, runs operations with me, and most of our management team are nurses who still see patients. The people deciding caseloads have carried one.

We wrote down how our best managers already worked and made it the way every team runs: a few goals for each team every quarter, a brief before every one-on-one, and promises tracked until they’re kept. It runs on CultureAI, which we built inside HealthView for exactly this.

And we hire carefully. This year has been a hard one for home care in Los Angeles County. As other agencies closed, more clinicians applied to us than ever. We made our interviews longer, added ride-alongs so candidates saw real visits before anyone decided anything, and sent the people who weren’t the right fit for us to neighbors who were hiring.

It hasn’t cost us growth. HealthView is No. 97 on the 2026 Inc. 5000.¹³

Where I could be wrong

Most of these studies show links, not proof. A place that treats its nurses well might also have more money or healthier patients, and researchers can only adjust for what they can measure. Most of the evidence comes from hospitals. And our own survey measures how our people feel about working here. That’s the input. It doesn’t tell you how your mom’s care went, and I won’t pretend it does.

You should know my incentive too. I run a company that does better when you believe this, which is why every number in this note links to where it came from.

What would change my mind is research showing an agency can burn through its people and still give families the same care. I haven’t seen it. If it turns up, I’ll write about it here.

None of this is a promise about any one patient. What we control is how we treat the people who walk into your home, and that’s where I start.