This World Head and Neck Cancer Day in July, a set of hospitals ran a campaign asking Indians to spend two minutes a month in front of a mirror, checking their own mouths for a patch or an ulcer that won't heal.

It's good advice.

I believed it too, until I read what happens to the people who follow it.

At Lok Nayak Hospital in Delhi, researchers tracked 110 people with mouth cancer from the day they first noticed something wrong in their mouth to the day their treatment started.

Half of them waited more than 202 days, i.e. nearly seven months. 74 of the 110 were diagnosed at Stage IV, the last and most serious stage.

My first guess was that the patients were to blame. I assumed they ignored the sore for months and went to a doctor too late.

The study says otherwise, because the researchers split the wait into three parts. Half the patients went to someone for help within two weeks of noticing the symptom.

For the typical patient, getting from that first visit to a confirmed diagnosis took about three months. Starting treatment took almost two months more.

So the longest wait started after they had already gone to someone for help. I wanted to know what happened on that first visit.

For 60 of the 110, the first person they saw called it minor and told them to come back later.

52 said the person they saw didn't suspect cancer.

As a doctor, this is hard to read.

We tell people to get checked early. Half of these patients did that, and it still took months to get an answer.

In medical college, they teach you one line: when you hear hoofbeats, think horses, not zebras. Look for the common thing first.

It's good advice, because most mouth sores really are harmless. But the few that aren't are the ones that get missed.

Why would someone trained to spot illness not suspect cancer?

One experiment shows how it can happen. When you're looking for one thing, you can miss something else, even something big.

In 2013, researchers asked 24 radiologists, the doctors who read scans, to look through lung scans for small white spots that can be early cancer. In the last scan, they had added something else.

Look at the top right. That's a man in a gorilla suit, shaking his fist. He is 48 times bigger than the spots the doctors were looking for.

20 of the 24 never saw him.

Eye trackers showed that most of the ones who missed him had looked straight at him. They were looking for white spots, and a gorilla wasn't what they expected to see.

A mouth isn't a scan, and cancer isn't a gorilla someone slipped in. But the habit is similar.

If you expect a harmless sore, it's easy to see a harmless sore.

I can't tell you that's what happened in Delhi. The paper didn't test why the first person missed it. Its authors point to gaps in the system and say clinicians need more training.

But the patients describe the same pattern in themselves.

The patients made that call about their own symptoms. At the first visit, the person they saw often made the same one.

India has more mouth cancer than any other country.

In 2022, there were 143,759 new cases and 79,979 deaths, and three in four were men. One recent analysis puts tobacco and areca nut behind roughly 70 to 80% of them.

One answer is to make looking for it part of someone's job.

A trial in Kerala did that, and The Lancet published the results in 2005.

In 7 areas with 96,517 adults aged 35 and above, trained health workers who weren't doctors visited homes and looked inside people's mouths, up to three times, three years apart.

Six other areas, covering 95,356 adults, got no screening.

Anyone who screened positive was sent to a doctor, and 63% of them went.

And 41% of the cancers in the screened areas were caught at stage I or II, against 23% in the other areas.

Across everyone, though, the drop in deaths was too small to rule out chance.

A 2013 review of the evidence found this was the only randomised trial of its kind, and flagged weaknesses in how it was run.

So it isn't proof. But it points to the same thing Delhi does: finding a sore early only helps if the next steps happen quickly.

The Health Ministry's oral health guidelines ask people to examine their own mouths and watch for an ulcer, or a red or white patch, that hasn't gone away after two weeks.

A clear check doesn't rule out cancer. It helps you notice a change early enough to ask someone.

If a sore or patch is still there after two weeks, get it examined.

And don't leave with only an ointment and a "come back later."

In Delhi, patients whose first visit ended with treatment for the symptom and no referral waited far longer for a diagnosis than those who were referred.

Ask directly: could this be cancer, and does it need a biopsy?

Hit reply and tell me: what were you once told was "nothing" that turned out to be something? How long did it take to find out?

I read every email.

Until next week,
Ritesh