What Holds Healthcare Together

I've been trying to understand Indian healthcare from the ground up. I don't come from a medical background. Most of my work has been in growth and marketing. I came in with beliefs about healthcare I'd held for a long time. Most of them turned out to be incomplete. This is a record of what I learned so far.

If I had to say it in one sentence: healthcare built everything except the part that holds it together, because the patient's own need did that job.

Here's how I got there.

1. I thought every system has to win its people

In growth, one of the earliest and most expensive problems is getting people through the door. We measure it as CAC, the cost of acquiring each customer. Most companies pour a large share of everything they have into it, because without it nothing else matters.

So I assumed healthcare had a similar problem. It doesn't, at least not in the same way.

Hospitals do advertise. But look at what the ads compete for. They compete for which hospital you choose. They don't have to convince you that you need one. When someone has chest pain, nobody has to persuade them to get care. The only question is where.

Economists have measured how strong this pull is. In the RAND Health Insurance Experiment, run in the US in the 1970s and early 80s, families were randomly given insurance plans in which they paid different shares of their own bills. The finding, published in 1987, was that a 10% rise in what people paid themselves cut their spending on care by only about 2%. For many things we buy, price matters far more than that.

In India, you can see the same pull in a stronger way. In the government's national health survey of 2017–18, about 13% of hospital stays in rural India were paid for mainly by borrowing.

Years ago I read Anselm's argument for God, which he described as "a single argument that needed nothing but itself alone for proof." I think the need for health is close to that. It needs nothing outside itself to keep it going. People delay, negotiate, wait for the weekend. But across a whole population, the drive to get checked comes back on its own.

That sounds like good news. It took me a while to see the other side of it.

In product, every step after acquisition matters, because people leave. If they don't experience value, they don't stay, or don't tell anyone, the funnel leaks and the money spent bringing them in is lost. That pressure is what forces companies to fix everything after the front door.

Healthcare doesn't feel that pressure in the same way. When people drop out after a visit, new patients still arrive the next morning. So I think the gaps we see now never cost the system enough to make anyone build what comes after the first visit: the memory, the routing, the follow-up.

But someone still has to do that work. In healthcare, that someone is mostly the patient.

2. I thought the system was held together from the inside

In my first piece, I described care as a loop: learn → decide → act → learn (again). Each part hands something (a prescription, a scan, a report) to the next. And for anything to move from one hand to another, someone needs a reason to give it, carry it, and to receive it. That reason may take any form like money, trust, duty, or policy.

When I looked for who carries things between the parts I described, I kept finding the same answer.

Start with learning. A doctor's best tool is often not a machine. It's the story of what has already happened. In a 1975 study in Britain, doctors noted their thinking at each step for 80 new patients. For 66 of them, the history alone, what the patient said and what the referring doctor had written, already pointed to the final diagnosis. The examination and the tests added far less. I don't think the basics of diagnosis change much from one country to another.

That history has to travel from one visit to the next. Who carries it? The patient: a folder of reports, photos on a phone, sometimes their own memory. You can see the system leaning on this. Under India's national programme for long-term illness, patients often carry a notebook to every visit. Given how strong the patient's need is, building around the patient made sense.

Then comes deciding. Here the patient carries something different: trust.

In 1963, the economist Kenneth Arrow pointed out what makes medical care unlike almost anything else we pay for: "the customer cannot test the product before consuming it, and there is an element of trust in the relation." You can return a shirt. You can't return an operation, and often you can't even know whether a different choice would have gone better.

So patients do the only thing they can: they trust. Since they can't judge the decision, they judge the person. Who recommended this doctor? Do they listen? Does my family trust them? How did the last visit go? And when that trust breaks, they leave and find someone else. The patient becomes the quality check on every decision made about them.

Put together, the list looks like this:

• Who's responsible for remembering what happened last time? The patient.
• Who decides whom to trust, and where to go next? The patient.
• Who decides when to come back? The patient. (In some cases, the doctor.)
• Who notices, over the long run, whether the treatment worked? The patient.

That is not a bad design. It's shaped around the one person who will always show up. Each part of healthcare is built from the inside out. The whole system is held together from the outside, by the patient.

3. I thought care always happens at the clinic

I used to picture healthcare as something that happens in a hospital or a doctor's room. That's where the decision is made. But a decision changes nothing in the body by itself. Someone has to act on it.

Most of that acting happens at home. A consultation lasts a few minutes. Taking a tablet every day for months, changing what gets cooked, going back for the next test: all of that happens in the patient's own life, long after they leave the room. In long illness, the patient and their family do most of the work of care. It's hard work, even when everything else is right.

In a 2003 report, the WHO found that in developed countries, only about half of people with long-term illness take their treatment as prescribed. In poorer countries, it's even lower.

Then there are the brakes. I assumed that if people knew about a condition, they'd act on it. That's only partly true.

Some illness gives you nothing to know. High blood pressure causes no pain for years. In India's national family health survey from 2019 to 2021, only about one in three adults aged 15–49 with high blood pressure knew they had it. They weren't ignoring their health. There was nothing to feel, so the drive never started.

And knowing isn't enough. In the same survey, about six in ten of the people who knew weren't on treatment. Between knowing and acting sit the brakes, like:

• Urgency: something else always feels more pressing.
• Money.
• Family: whether they can support an elder through it.
• Fear: hesitation, or plain fear of doctors.
• Tablets: the discomfort of taking several every day.
• Doubt: not being sure the body can handle any of it.
• Acceptance: sometimes a quiet decision, "let me live like this for the time I've got."

That's why prevention is so hard. It isn't only about value; it's about human behaviour. The benefit is distant and unfelt. The cost and effort are here, today.

4. I thought follow-up meant reminding people

I assumed follow-up was a scheduling problem. Send a reminder, get the patient back.

Then I thought about what the first doctor actually sees for a common condition. If the patient's condition gets worse, they come back. If nothing helps, they go somewhere else. If they get better, they usually say nothing. A small study from AIIMS Bhopal, published in 2019, found that 25 of 60 people with diabetes had left their first doctor, and 9 of them had switched again. Their reasons included getting no relief, the cost, the distance, and how they were treated.

To the first doctor, silence looks the same whether the treatment worked or the patient left. The patient takes the result with them. In most of healthcare, nothing standard is built to bring the result back. Usually, the result/feedback only comes back if the patient does.

It reminds me of a computer. Everything on a computer sits in one place, so whoever is using it can see what's happening, try something, and see what changed. Healthcare has no such place. Nobody, not even the government, can see the whole picture from one seat.

But where results do come back, a system can learn. One initiative I found, the India Hypertension Control Initiative, tracks each patient's visits in a real-time system. Because the results come back, it can see what's working and change course.

5. I thought technology would close the gaps

My first instinct was that the missing piece was software. Digitise the records, give patients access, and the gaps close.

Then I looked at the US. Nearly every American hospital runs on digital records. Yet in 2024, 59% of Americans had their records spread across more than one patient portal, and only 7% used an app to bring them together. Far more software, and the patient is still the one joining the pieces.

India has gone further on the foundation than most countries. By July 2026, more than 93 crore health accounts (ABHA) had been created, with more than 104 crore records linked to them. Records stay with whoever created them and move only when the patient agrees, and a national incentive has paid hospitals more than ₹107 crore to create and link digital records. It's a careful, sensible foundation.

Take UPI, where India did an amazing job. In August 2026 alone, UPI handled about 2,450 crore transactions. Similar to UPI, as health has the foundation, I assumed everything would work. But UPI has something a health record doesn't. Every payment has a sender who wants the money to go and a receiver waiting for it, at the same moment. Both ends have a reason.

A health record is different. It's created today, by someone thinking about today. It's needed months later, by someone who may not know it exists. There's a reason to create it. There's rarely anyone other than the patient with a reason to carry it to the next visit.

Technology makes a hand-off easier. It doesn't create the reason for it to happen. India built the foundation first, which was the right order. What comes next is the reason to build on top of it.

What it adds up to

By nature, the need for health drives itself. Because of that, healthcare never had to build the part that holds care together across time. The patient did that work: remembering, choosing whom to trust, carrying out the treatment at home, coming back, noticing changes in the body. For illness you can feel and fix quickly, that has worked surprisingly well.

But illness in India has been changing. Recent estimates suggest that about six in ten deaths in India now come from long-term conditions like heart disease, diabetes and chronic lung disease, many of which stay silent for years. That's exactly where the patient's drive is weakest, and where the loop I described in this essay has to keep running the longest.

Think about it this way: If one person had to run all of India's healthcare and could see everything at once, I think the first thing they'd notice is that, the only person who ever sees a whole loop firsthand, from the first symptom to whether the treatment worked, is the patient.

None of this is permanent. In a system like this, anything can change the moment the value and the forces change. The foundation is built. What's missing is a reason for someone other than the patient to care enough to build something and deliver value.