A matter of trust

 

One of the most essential elements in the doctor-patient relationship is trust. Because you are unlikely to know as much as your doctor does about your body, your symptoms and what they mean, and how you should be treated, you place trust in your doctor to diagnose you properly and prescribe an appropriate course of treatment that will offer the maximum chance of benefit and minimum risk of harm.

In the light of this, a systematic review published in the March 6 2017 edition of JAMA Internal Medicine makes for rather worrying reading. As it turns out, clinicians are not terribly good at assessing either the risks or the benefits of medical treatment.

The review, which was conducted by researchers at Bond University, examined 48 studies, involving a total of over 13 000 clinicians. Of these 48 studies, 20 focused on treatment, 20 on medical imaging (including x-rays, MRIs and CT scans), and 8 on screening.

When it came to assessing expectations of benefit from treatments, doctors provided a correct estimation for only 3 out of 28 outcomes. They overestimated the likelihood of patients benefiting from treatment in 32% of cases, and underestimated potential benefit 9% of the time.

Their strike rate for expectations of harm wasn’t any better: in only 9 out of 69 outcomes did doctors correctly estimate the risk of a treatment harming a patient. They were far more likely to underestimate harm (34% of outcomes) than overestimate it (5% of outcomes).

What does all this mean? The average doctor recommends investigations and prescribes treatments that he or she thinks are going to be more beneficial and less harmful to you than they actually are.

When the potential harms are minor, such as a mild skin rash, brief episode of diarrhoea or minor headache, you might be prepared to put up with them if you’re likely to benefit greatly from the treatment. But when the potential benefits are small, any risk of harm is a big deal.

Take a look at the table below, from an article called Are preventive drugs preventive enough? A study of patients’ expectation of benefit from preventive drugs, and pay particular attention to the numbers in the far right hand column:

What this all means is that if you haven’t yet had a heart attack or other coronary event, but you have very high cholesterol (over 7 mmol/L), taking a statin drug – in this case, Pravachol – will reduce your risk of having a coronary event by 2.3%, and your risk of dying by 0.9%. That’s right – there’s less than a 1% chance that the drug will keep you alive any longer than if you just kept on going about your business with a ridiculously high cholesterol level.

What if you’ve already had a heart attack? Taking the statin will reducing your risk of dying by 3.1%, and your risk of having another heart attack by 2.9%. How enthusiastic do you feel about taking this drug now? It doesn’t fire me up, especially when compared to the results gained by Dr Caldwell Esselstyn, whose dietary program reduced the risk of a major cardiac event by over 60%.

As the authors of the article point out,

“Even high risk patients have less than 5% chance of benefiting from a cardioprotective drug taken for 5 years; 95% of patients will take the drug for 5 years without benefit.

These statistics are seldom shared with patients.”

Based on the Bond University study, doctors aren’t not sharing these statistics because they’re evil people who want to harm their patients. They’re not sharing them because they don’t know them themselves.

Doctors are often too busy to read medical journals, and lack the training to read studies critically and understand statistics presented in them even if they did so. Most doctors rely heavily on pharmaceutical company representatives for information about new drugs, and this information influences their prescribing behaviour even though only 9% agree that it is “very accurate”. When they do their own research, roughly 50% of doctors use Wikipedia as their primary source of health information; meanwhile, even primary school children are not permitted to cite Wikipedia as an information source for assignments. If they rely on clinical practice guidelines for their prescribing decisions, they’re probably not aware that the people serving on the committees that write these guidelines are riddled with conflicts of interest such as working for pharmaceutical companies and receiving industry funding for their research, as the article Conflict of Interest in Seminal Hepatitis C Virus and Cholesterol Management Guidelines, in the same edition of JAMA Internal Medicine, revealed.

As the researchers concluded,

“Clinicians rarely had accurate expectations of benefits or harms, with inaccuracies in both directions. However, clinicians more often underestimated rather than overestimated harms and overestimated rather than underestimated benefits. Inaccurate perceptions about the benefits and harms of interventions are likely to result in suboptimal clinical management choices.”

Where does this leave you, the patient?

The first priority is learning how to take care of yourself properly to reduce your need to engage with the medical system; healthy people don’t need doctors, barring accidents.

After that, you need to learn how to assess your options for investigation and treatment, in case you ever do get sick. My EmpowerEd program is all about becoming an informed health-care consumer by learning how to critically analyse health and nutrition information; ask for the right tests and know how to interpret the results; avoid useless screening tests and make the most of those that are useful; understand the risks and benefits of medical treatments; and choose the right medical and health professionals for their care team.

The EmpowerEd program includes monthly in-depth webinars covering topics like cancer screening, bone mineral density assessment, and how to read your own blood tests, as well as an open question-and-answer session every month.

Check out the benefits of EmpowerEd membership here – your first month of membership is free!

Leave your comments below:

Leave A Response

* Denotes Required Field