When Good Doctors Give Bad Advice: How to Trust but Verify

Almost anyone who watched television in the 1970s remembers Marcus Welby, M.D. Robert Young played the kind, wise family doctor who seemed to have unlimited time for his patients and almost always knew what was best for them. At one point it was the No. 1 show on TV. Young had previously starred in Father Knows Best, and the Television Academy later observed that Marcus Welby, M.D. might more appropriately have been called “Dr. Knows Best.” That was a time when Americans were told to trust their doctor. Today, we think patients need a somewhat different guiding principle: trust but verify.

We were reminded of this recently by a reader who asked us about chickenpox and the shingles vaccine.

When a Doctor Doesn’t Always Know Best

Q. When I was 38 years old, I had chickenpox for the first time. It was a mild case with red spots but no blisters or itching. I am sure this was my first case because a blood test was negative the previous year. I got tested because my son had come down with chickenpox.

I am now 65 years old. My primary doctor says I don’t need the shingles vaccine because I would still have immunity from almost 30 years ago. Do you agree that I don’t need the shingles vaccine?

A. We are perplexed by your doctor’s position. Any chickenpox infection puts a person at risk for shingles later in life. That’s because the virus that causes chickenpox (varicella-zoster) persists in the nervous system and can emerge years later to cause shingles.

The CDC recommends the Shingrix vaccine for older adults to protect against shingles and its complications. This vaccination is about 95 percent effective in preventing outbreaks in healthy adults (Reviews in Medical Virology, Sep. 2026). The vaccine may also modestly reduce the risk for developing dementia in old age (Vaccines, Aug. 10, 2026).

You may wish to seek a second opinion or ask your doctor to check the CDC’s recommendations. The CDC is quite clear on this point. It recommends two doses of Shingrix for immunocompetent adults 50 years and older. Previous shingles or previous vaccination with the older Zostavax vaccine does not eliminate that recommendation.

The reader’s question raises a much larger issue, though. What should you do when a healthcare professional gives advice that doesn’t sound quite right?

Do Doctors Always Know What’s Best?

For much of the 20th century, medicine operated according to an unwritten principle that could be summarized in three words: doctor knows best. Medical ethicists now call this model paternalism. The physician possessed specialized knowledge and decided what was best for the patient. The patient’s role was largely to cooperate and not question. Many people were expected to swallow their pills without knowing what they were taking or what the side effects might be.

One review of clinical ethics actually describes the word paternalism using the old adage “the doctor knows best.” Beginning in the 1970s, however, patients increasingly demanded a greater voice in decisions about their own bodies and treatment.

The book Our Bodies, Ourselves: A Book by and for Women was published by Simon & Schuster in 1973. My book, The People’s Pharmacy: A Guide to Prescription Drugs, Home Remedies and Over-The-Counter Medications was published by St. Martin’s Press in 1976. I received a lot of criticism from physicians and pharmacists. One healthcare professional even suggested that the book was dangerous and should be available “by prescription only.”

I would like to think that those two publications helped change the trajectory of the patient-doctor relationship and eventually led to today’s emphasis on patient autonomy and shared decision-making. That change took a long time, however.

We have enormous respect for physicians. We have interviewed thousands of health professionals during half a century of writing and broadcasting about health. Most are conscientious people who want to do the very best for their patients.

But even an excellent doctor cannot remember every drug interaction, adverse reaction, dosing instruction and constantly changing treatment guideline. Relying exclusively on artificial intelligence or some online algorithm scares us. We also recognize that medicine has become too complicated for anyone to know everything.

That’s where the principle, trust but verify comes in.

Trust but Verify Is Not the Same as Distrust

President Ronald Reagan made the expression “trust, but verify” famous during arms-control negotiations with Soviet leader Mikhail Gorbachev. Reagan had learned the Russian proverb doveryai, no proveryai and liked it so much that Gorbachev reportedly teased him for repeating “trust but verify” at their meetings.

We are not suggesting that every interaction between a patient and a physician should turn into a negotiation or cross-examination. Nor are we urging patients to challenge every diagnosis, second-guess every prescription or arrive at an appointment armed with 43 pages of Internet printouts. That can undermine the very relationship that good medical care depends upon.

But respectful skepticism is different from distrust. If a recommendation surprises you, contradicts something you were previously told or doesn’t make sense, asking a few questions is entirely appropriate.

Here are some questions we developed for our book Top Screwups:

Questions to Ask to Reduce Diagnostic Disasters:

1. What are my primary concerns and symptoms? (Ask the doctor to summarize your primary concerns and symptoms. This helps confirm that the doctor has been listening and understands what matters most to you.)
2. How confident are you about this diagnosis?
3. What further tests might be helpful to improve your confidence?
4. Are there any findings or symptoms that don’t fit your diagnosis?
5. What else could it be?

Question to Ask When You Get a Prescription:

1. Is there another way to treat my condition besides this drug?
2. What is the evidence this drug will produce a meaningful outcome?
3. What is the number needed to treat (NNT) or percentage of people who get benefit?
4. What are the most common side effects?
5. What are the most serious side effects?
6. What symptoms require immediate medical attention?
7. How do I take this drug–with food or without, morning or evening?
8. Are there any special instructions for stopping this medicine?

Those are not hostile questions. They are signs of an engaged patient.

Doctors Can Misjudge Benefits and Risks

There is another reason patients should participate actively in medical decisions. Doctors themselves may not always have an accurate picture of how much a treatment is likely to help or hurt.

Researchers Tammy Hoffmann and Chris Del Mar conducted a systematic review of 48 studies involving 13,011 clinicians (JAMA Internal Medicine, March, 2017). They examined doctors’ “expectations of the benefits and harms of treatments, tests, and screening tests.”

The results were sobering. When there was a correct estimate against which physicians’ answers could be compared, most clinicians got the answer right for only 3 of 28 benefit outcomes and 9 of 69 harm outcomes.

Errors went in both directions. But when clinicians were wrong, they were more likely to overestimate benefits (32%) and underestimate harms (34%).

Patients Can Misjudge Benefits and Risks

Before anyone becomes smug about doctors getting things wrong, patients have the same problem. The same researchers reviewed 35 studies involving 27,323 patients (JAMA Internal Medicine, Feb. 2015). Where investigators could assess overestimation, most patients overestimated benefits for 65 percent of the relevant outcomes. Where underestimation of harms could be assessed, most patients underestimated harm for 67 percent of the outcomes.

In other words, both doctors and patients may have a natural tendency to think medical interventions work better than they really do while paying too little attention to potential downsides.

That makes genuine shared decision-making especially important.

Trust but Verify: How Much Will This Treatment Actually Help?

Numbers can be slippery. Suppose a treatment is advertised as reducing the risk of something terrible by 50 percent. That sounds spectacular. Drug companies love to share such relative risk reductions because they are so convincing. And uninformed reporters love headlines that state:

Drug X Lowers the Risk of Condition Y by 50%

But imagine that without treatment, 2 people out of 1,000 would experience that problem. With treatment, 1 person out of 1,000 would. That is indeed a 50 percent relative risk reduction.

The absolute risk reduction, however, is 1 person out of 1,000.

We are not suggesting that such a benefit is necessarily unimportant. Depending upon the severity of the condition, even a small absolute benefit could be worthwhile. But patients deserve both numbers.

Whenever possible, we think people should ask:

“Out of 100 people like me, how many would be expected to benefit?”

And then:

“Out of 100 people like me, how many might experience an important side effect?”

Those questions can turn an abstract percentage into something much easier to understand. We wish people would ask their healthcare professionals those questions when it comes to medications like statins to lower cholesterol. There is no doubt that atorvastatin, rosuvastatin and simvastatin can lower LDL-cholesterol very effectively. The more important question is:

“Out of 100 people like me, how many would avoid a heart attack or a stroke if they took this statin for 5 years compared to those who took a placebo?”

How to Trust but Verify a New Prescription

Suppose your doctor writes you a new prescription. What should you check with both the prescriber and the pharmacist?

Q. How Should I Take This Medicine?

That sounds ridiculously basic. It isn’t.

• Should the drug be taken with food or on an empty stomach?
• Morning or evening?
• Can the tablet be split?
• Can the capsule be opened?
• Should you avoid grapefruit?
• Can calcium, magnesium, iron or antacids interfere with absorption?
• What happens if you miss a dose?

Some medicines are remarkably forgiving. Others have very particular instructions.

The FDA recommends its approved medication information as a source for directions, precautions, interactions and adverse effects. The National Library of Medicine also provides DailyMed, a searchable database containing current drug labeling submitted to the FDA.

Do not overlook one of the most accessible medication experts in the healthcare system: your pharmacist.

Q. Could This Drug Interact With Something Else I Take?

People frequently see several different specialists. A cardiologist may prescribe one medicine, a urologist another and an orthopedist a third. The patient may also be taking aspirin, an antihistamine, a pain reliever, vitamins, herbs and dietary supplements. No prescriber or dispenser can recognize an interaction with a medicine he or she doesn’t know about.

That’s why every patient should maintain an up-to-date list of:

• prescription medicines
• over-the-counter drugs
• vitamins
• minerals
• herbs
• dietary supplements

Bring the list to medical appointments and show it to the pharmacist. Yes, I know your online medical record is supposed to have a current list of all that information available at all times. It has been our experience that the list is often out of date.

Q. What Side Effects Should I Watch For?

This question has been at the heart of The People’s Pharmacy from day one in 1976. Over the decades, we have heard from countless readers and listeners who developed fatigue, dizziness, coughing, muscle pain, memory problems, insomnia, sexual difficulties or some other symptom and never imagined that a medication might be responsible.

Sometimes their doctors didn’t make the connection either. That doesn’t mean every symptom is caused by a drug. Far from it. But whenever a new symptom appears after a medicine has been started, stopped or the dose changed, asking whether there could be a connection is not just reasonable, it is crucial.

We would never suggest stopping a prescription medicine suddenly because you read something alarming online. Some drugs can cause serious complications or withdrawal problems if discontinued abruptly. Instead, contact the prescriber or pharmacist and ask how to proceed safely.

Don’t Let an AI Summary Be the Last Word

Searching for health information has changed dramatically. Once upon a time, patients went to a library and searched through medical textbooks or journals. Then came Google. Now artificial intelligence can produce an apparently authoritative answer in seconds.

We use these tools ourselves. They can be extraordinarily helpful. But there is a big difference between using an AI response to begin an investigation and using it to end one. Artificial intelligence can misunderstand a question, miss an important qualification, rely upon outdated information or occasionally produce a confident answer that is simply wrong.

The same warning applies to snippets at the top of search results. When the stakes are high, keep going until you reach the underlying source.

• For medicines, look for FDA-approved labeling or DailyMed.
• For vaccines, check CDC and the Advisory Committee on Immunization Practices.
• For research, look for the actual study through PubMed.

When something remains confusing, ask your physician or pharmacist. Don’t hesitate to ask your healthcare professional to repeat back the key points you have described. That helps verify that what they heard is what you actually said. This is what pilots and air traffic controllers do for every communication between the cockpit and the tower.

The goal isn’t to become your own doctor. The goal is to become a better-informed patient.

Trust but Verify Without Driving Your Doctor Crazy

There is a fine line between becoming an activated patient and assuming that three hours online are equivalent to medical school and years of clinical experience. They aren’t.

Doctors see patterns that patients may not recognize. They know when a symptom that appears trivial could signal something dangerous. They also understand that laboratory results and imaging findings need to be interpreted in context.

But patients possess something doctors do not: intimate knowledge of their own bodies, priorities and experiences.

The best medical relationships combine both kinds of expertise.

Instead of announcing:

“The Internet says you’re wrong,”

try:

“I found something that seems different from what I understood. Could you help me make sense of it?”

Or:

“I saw that the CDC recommends this for people my age. Is there a reason you think that recommendation doesn’t apply to me?”

Those questions invite a conversation rather than confrontation.

When Should You Get a Second Opinion?

Not every disagreement requires another doctor. But we think a second opinion deserves serious consideration when:

• the diagnosis is uncertain
• the proposed treatment carries substantial risk
• surgery or another irreversible procedure is recommended
• there are several reasonable treatment choices
• the treatment isn’t working
• serious symptoms remain unexplained
• the recommendation conflicts with recent research
• your questions are repeatedly dismissed rather than answered

A good doctor should not be threatened when a patient facing an important decision asks for another perspective. Sometimes the second doctor will confirm the first doctor’s recommendation. That reassurance alone may be valuable. Sometimes the second opinion changes everything.

From “Doctor Knows Best” to Partnership

Medicine has come a long way from the days when “doctor knows best” ended the conversation. Perhaps the opposite extreme is equally undesirable. A patient who assumes doctors cannot be trusted and believes every provocative claim on TikTok, Facebook, YouTube or an Internet discussion board can get into serious trouble.

We favor a middle path. Doctors bring training, judgment and clinical experience. Patients bring their values, observations and questions. Pharmacists bring specialized knowledge about medicines. Reliable medical resources can help everyone check the evidence.

We are big fans of shared decision-making. And it brings us back to our reader who was told that because she had chickenpox at age 38, she didn’t need a shingles vaccination at 65. She didn’t accuse her physician of incompetence. She didn’t ignore the advice and decide on her own what to do. She asked another question.

That may be the simplest definition of an activated patient we can think of. Good doctors can occasionally give bad advice. Good patients can misunderstand medical information too. That’s why health care works best when everyone is willing to trust but verify.

Final Words

Did you find this article helpful? Please share your own experiences with healthcare professionals in the comment section below. If you would like to learn more about how to avoid medical mistakes, why not check out our book, Top Screwups at this link. We admit that the title was not very kind, but the book is loaded with practical information. Of the more than 20 books we have written, this is the one we think is most important. Thank you for supporting our work.