The Third Opinion: Why Patients Don't Trust Doctors Anymore

The phone comes out about four minutes into the visit with their loved one in the ICU.

Sometimes the patient hands it over. More often it is a daughter, or a son who flew in overnight and has not slept, and they turn the screen toward me with a thumb still resting on the scroll bar, holding the place.

Half of them apologize first. That is the part I have never gotten used to. A woman whose mother is dying will say I'm sorry, this is probably stupid before she asks the most important question she has. The ones who open with I did my own research are easier, honestly. At least the terms are clear.

They are not being difficult. Somebody they love is sick, and at two in the morning they did what any of us would do, which is go looking for something that would answer.

By the time I walk into that room I am not the first opinion. I am the third.

That is what has actually changed in twenty years. Patients always had questions, and the sharp ones make me better. What is different is that people now arrive holding a conclusion, and my job has quietly shifted from explaining what I found to explaining why I disagree with what they already believe.

Last week made that job harder.

Two events, four days apart

On July 29, Dr. Anthony Fauci sat before the Senate Homeland Security and Governmental Affairs Committee, chaired by Senator Rand Paul, and declined to answer more than a hundred questions, invoking the Fifth Amendment on advice of counsel. He accused Paul of an obsession with putting him in prison. Paul said the committee would vote on contempt. At one point Capitol Police removed Fauci's attorney from the room.

Update, August 6, 2026. The committee voted along party lines to hold Fauci in contempt, and Senator Paul hand-delivered a criminal referral to the Justice Department the same day, bypassing a full Senate vote. Whether the pardon voided his Fifth Amendment protection is a real legal question and I am not the person to settle it. Nothing in the argument below changes either way.

Update, August 19, 2026. Senator Ron Johnson, who chairs the Permanent Subcommittee on Investigations, asked Fauci to sit for a separate voluntary transcribed interview. On August 14 his attorney declined, arguing that a second appearance would harass him for political purposes. Johnson says his staff will keep collecting records and interviewing former colleagues, and that the subcommittee is weighing its legal options. The Justice Department has not said whether it will act on the contempt referral.

Congressional oversight of a federal official is legitimate. It is supposed to happen. A Fifth Amendment invocation is not a finding of fact, which is precisely why the protection exists.

But almost nobody watched three hours of that hearing. They watched a clip, and in a clip an unanswered question becomes its own answer.

Four days later, on August 2, HHS Secretary Robert F. Kennedy Jr. sat down with Dana Bash on CNN's State of the Union. For twenty minutes it was less an interview than a contest. When Bash pressed him on the measles outbreak, he pointed to Mexico, Canada, and England and asked why nobody was talking about their numbers. When she raised autism, he told her she would not understand because she is not a scientist.

She answered that neither is he. At one point he told her she was talking nonsense.

Then, pressed to look into the camera, he said parents should vaccinate their children against measles, and that the vaccine stops it about 97% of the time.

He is right about that. It is a good vaccine.

What troubles me is that the endorsement arrived at the end of twenty minutes spent attacking the institutions and the evidence-weighing process that produced the number he had just quoted. You cannot spend years teaching people the referees are corrupt and then expect them to accept the call.

The number everybody cites, and the part they leave out

You have seen this statistic. In JAMA Network Open, Perlis and colleagues surveyed 443,455 American adults across 24 waves. The share reporting a lot of trust in physicians and hospitals fell from 71.5% in April 2020 to 40.1% in January 2024. Lower trust tracked with lower vaccination.

I have used that number in talks. I want to be more careful with it than I have been.

April 2020 was a strange peak. Health care workers were briefly national heroes, and trust sat at a rally-round-the-flag high that was never going to hold. Measuring the decline from there overstates the fall. The survey also asked about a lot of trust, which is a high bar, and it ran on a nonprobability internet panel, and it ended two and a half years ago.

Pew released data in January of this year, fielded the previous October, showing 77% of Americans still have at least a fair amount of confidence that scientists act in the public interest. Essentially flat over two years. Scientists ranked as the most trusted profession Pew asked about. Lower than 2019, yes. Collapsing, no.

A piece arguing that medicine should show its work has an obligation to show its own.

The honest version is that trust in the abstractions took a beating while trust in the person in the room held. KFF finds this repeatedly: around 83% of adults trust their own doctor on vaccines, and the number holds across party lines, Democrats at 93% and Republicans at 78%. Trust in the CDC as a source sits near half.

The relationship survived. The institution did not.

Three different things wearing the same word

We say mistrust as though it were one thing. It is at least three, and they need different answers.

Earned mistrust

This comes from what actually happened to people. When Perlis's team read the open-text responses, the reasons that came back were financial motives, quality of care, outside agendas, and perceived discrimination. A month's wait for a new appointment, and that figure comes from the fifteen best-supplied metro areas in the country. A day off work. A bill nobody can parse. Women who spent years being told the pain was anxiety, and turned out to have endometriosis. Black and Indigenous patients carrying a documented history that did not end with Tuskegee. Just trust your doctor lands as an insult in that room.

Uncertainty-driven mistrust

This comes from watching guidance change with nobody explaining the reasoning. People heard one thing in March and something else in June, and nobody walked them through the middle.

Engineered mistrust

This is cultivated on purpose, and it teaches a specific lesson: revision proves dishonesty, disagreement proves conspiracy, expertise is itself the tell. It is profitable. It scales.

Confusing them is where we go wrong. Earned grievance does not yield to more data, and manufactured suspicion does not yield to defensiveness.

What patients are actually asking

When someone challenges me with a video, they are rarely asking me to prove the video wrong. The question underneath is usually some version of: are you telling me everything, or are you managing me?

That is fair. We have earned it.

The answer cannot be my credentials. It has to be the work. Here is what I found, here is how confident I am, here is what I do not know, here is what would change my mind.

I do not need to be believed. I need to be checkable.

---

Next in this series: the parallel health system your patients carry in their pocket, and the one number that should worry every clinician reading this.