What makes an exceptionally good doctor?

PhD's are boring with long titles that only make sense to a specialist in the field, right?

Not this one - the title is "What Makes an Exceptionally Good Doctor?" and the findings are perfectly understandable.

What are the findings? Simple: Doctors make an at times big difference to patients' physical health, exceptionally good doctors exist and leave a large impression on those who worked with them and on their patients. They are just as likely to get attacked as praised and those who also listen to their patients stand out even among that group.

So, why did I get a PhD for stating the obvious or easily guessed? The reason is that nobody actually guessed this before and was able to write - in this case 7 - scientific papers on the matter.

Imagine, you have an operation coming up or you have been to five doctors and none of them could help you diagnose an issue. Would you look and ask around to find a good doctor or the best doctor possible?

Obviously, yes. At least most of us would do that.

There are billions of medical interactions each year consisting of a doctor, a patient, and an intervention and there are millions of published papers about interventions but only a tiny amount of papers that ask the question 'what makes a good doctor'. In fact, the British Medical Journal had a special edition asking what makes a good doctor and can we learn from them? They concluded that it was not possible to know what makes a good doctor, as strange as that sounds.

Since then, there have been a number of surveys with inconsistent outcomes establishing for example that doctors wearing shirts and ties are considered by patients to be better doctors.

In other words, as absurd as it sounds, before my PhD it was not scientifically established what makes a good doctor. My supervisors, Dr Mark Jones and Dr Jeremy Howick, and I spent a year investigating until the Eureka moment hit: If you want to know what makes a good doctor you first need to establish that doctors make a difference. There is lots of research showing for example that more experienced surgeons and other doctors have on the whole better results and there are websites that compare doctors but these websites have major flaws such as not adjusting for patient risk - an emergency surgeon will have a higher death rates than a surgeon who only operates on relatively healthy people regardless of any difference in ability.

I therefore started to look for scientific papers that showed whether doctors make a difference to patients' physical health even after accounting for all known factors such as experience or patient risk. It turns out - great discovery but not exactly a surprise - that doctors make anything from a negligible to a very, very large difference to patients' physical health, depending on the intervention and outcome measured. For example anaesthetists have little influence on patient survival (though they may have a large influence on post-operative pain but nobody has investigated that), but, for example, shoulder surgeons have a huge influence on patients' SANE score (a measure of shoulder health). However, the results presented in the 79 papers we found by looking through 10,063 papers were a dog's breakfast with no consistency in presentation of any kind.

Systematic reviews: Doctor, Surgeon, Methodological Review

This meant that checking for doctors' performance was an afterthought for most investigators, hence despite co-author Justin Clark's amazing search string we had to look all over the place to even find 79 papers.

What we also found was that exceptionally good doctors with consistently superior patients' physical health outcomes appeared regularly in those papers but nobody in any of the papers commented on these outliers and whether these exceptionally good doctors should be followed up. One problem was that the original data is usually anonymised so the names of these exceptionally good doctors were simply not known but nobody was looking for them anyway. To resolve that, Mark and I wrote a methodological paper to make it easier to get much more detail when and where a doctor makes a difference. That should help with future research. It also allows many existing dataset to be re-investigated specifically for a doctors' effect.

Sometimes it is really surprising what science hasn't done. Here is something (a doctor) that is applied billions of times a year worldwide but there is no systematic research whether some doctors are inherently better than others, yet anybody I know would make thorough investigations if they consider they need a really good doctor.

Anyway, we now knew that there are exceptionally good doctors. Who would be most likely to know such doctors? Other doctors!

So I interviewed 13 doctors from all over the world on whether they ever met a doctor they considered to be exceptionally good and what was that experience like? It turns out that they all have met at least one such doctor, a doctor who is either an extremely skilled surgeon or a doctor of any specialty, who is both very knowledgeable but also very good at communicating. Such doctors tend to be humble despite their superior abilities and, a big surprise, were as likely to be attacked by colleagues and regulators, as praised and lauded, precisely because they are exceptionally good. In other words, medicine eats its own, best people and a cultural shift may be in order. Not only does it badly affect the best doctors but what signal does it set for other doctors?

We published two papers about these interviews.

By this stage we had quite a good list of characteristics of exceptionally good doctors and we decided to ask patients - 500 of them to be exact. The large majority had encountered at least one doctor they considered to be exceptionally good. They were also clear how they defined such a doctor - a doctor who is great at diagnoses or great at interventions (treatments) or a great communicator was considered to be an exceptionally good doctor.

That makes sense - if we have been to many doctors and we finally find one who provides the correct diagnosis, then that is enough to an exceptional doctor.

About one third of the 500 respondents ticked a box that said "the doctor willingly listens to the end". That one third of patients gave their doctor substantially better grades than the other two third of respondents and not only considered the exceptionally good doctor they nominated to be better communicators but also more knowledgeable and more humble! Therefore, if a doctor wants to stand out even among exceptionally good doctors then the key is to listen. Such a seemingly simple thing. Here are the two papers.

However, there are two flies in the ointment again: The patients graded exceptionally good doctor better than average doctors in almost everything of 34 items we asked, except that they didn't consider the doctor they nominated to be more popular. Also, patients whose doctor listened to them were substantially more critical of average doctors, giving motivation to the colleagues of exceptionally good doctors to attack such doctors.

In other words, medicine is different. High performers are not publicly known or identified and if they are identified, their colleagues and regulators may praise or attack them.

All of this is very important for evidence-based medicine. Originally Sackett in 1996 said that evidence-based medicine is a doctor combining their clinical expertise with the best available evidence. Today this has morphed to a doctor being required to follow guidelines for diagnoses and interventions and the doctor risks being sued if they modify those recommendations by following their clinical judgment. As doctors themselves have an influence on patients' physical health it may be useful to go back to Sackett's original impulse.

Also, once we know who exceptionally good doctors are we can also learn from them and follow their wisdom.