How many lives does a doctor save? (Part 3)

This is Part 3 of an updated version of a classic three-part series of 80,000 Hours blog posts. You can also read updated versions of Part 1 and Part 2. You can still read the original version of the series published in 2012.

It’s fair to say working as a doctor does not look that great so far. In general, the day-to-day work of medicine has had a relatively minor role in why people are living longer and healthier now than they did historically. When we try and quantify the benefit of someone becoming a doctor, the figure gets lower the better the method of estimation and already is low enough such that a 40-year medical career somewhere like the UK would be on a rough par with giving $20,000 dollars to a GiveWell top charity in terms of saving lives.

Yet there is more to say. The tools we have used to arrive at estimates are general, so they are estimating something like the impact of the modal, median, or typical medical career. There are doctors who have plainly done much more good than my estimates of the impact of a typical doctor.

So, what could a doctor do to really save a lot of lives?

Doing doctoring better

What about just being really, really good? Even if the typical doctor’s work makes a worthwhile — but modest and fairly replaceable — contribution, maybe there are brilliant doctors that stand head-and-shoulders above the rest. The Gregory Houses (no relation) and Doogie Howsers of the world, who save lives with feats of medical brilliance their average peers can’t achieve, loom large in the public imagination.

But those examples are fictional for a reason — indeed, several reasons:

• Some fields have very heavy tails for individual practitioner performance — meaning some workers are far more effective than others. But it is hard to imagine clinical practice, where you are treating one patient at a time, is like this. It is hard to imagine a doctor who can effectively treat (e.g.) 10 times more patients than average. This would look something like doing a heart transplant in half an hour or completing a primary care consultation in 90 seconds.
• In terms of quality, good medicine looks much more like ‘follow the usual protocol’ than ‘be a genius to figure out the diagnosis.’ When doctors err, it’s usually because they stray too far from guidelines too readily rather than following them too strictly, and cases where deviation from the standard of care is the right call are rare.
• Practitioner assessment is murky, but you do see some significant differences — e.g. some surgeons have much lower rates of complications than others even when controlling for procedure, how sick their patients are, etc.¹ Yet large relative differences typically translate into smaller absolute gains. Suppose you are a 10 times ‘better’ surgeon, so your heart transplants have a mortality risk of 2% versus the batting average of 20%. You being 10 times ‘better’ nets out to saving 18% more lives than your peers.
• For this to be a worthwhile strategy, you would need to know in advance you have the makings of a great — rather than good — doctor. Yet predicting performance is murkier than the already challenging task of assessing performance, even setting aside the fact that folks often overestimate themselves. So you may have all the statistical makings of an exceptional doctor, and still, your chances of turning out to actually be a truly exceptional clinician — despite being better than almost everyone else — remain low.

Go where doctors are needed most

The estimates I gave before were for doctors working somewhere like the UK. What about working somewhere unlike the UK? In places with much greater health needs and much fewer doctors (e.g. lower-income countries, conflict zones), we would expect the next additional doctor to make a significantly bigger impact on the margin. So what about working in a lower-income country, or for Medicins Sans Frontieres (also known as MSF or Doctors Without Borders)?

It is hard to assess how much bigger of an impact you could have there. There are a couple of reasons to be cautious about the likely benefit.

• Taking MSF as an example. They are pretty selective (often demanding extensive post-graduate experience or further degrees) and aren’t always recruiting across all relevant medical specialties. Given they’re always keen for donations, this suggests willing medical personnel may not be the key bottleneck — so your contribution may be fairly replaceable.
• Consider again the argument that the social determinants of health have driven the global increase in life expectancy rather than day-to-day medical practice. This suggests that the things that countries with worse health outcomes need most isn’t more doctors, but better nutrition, sanitation, and other social factors that prevent disease. These considerations should lower our expectations of the impact of an additional doctor practicing in a low-income country.
• I’d also speculate the ideal mix of skills for providing healthcare in resource-constrained environments will have a lower proportion of very highly trained staff than medical practice in the UK. Many basic health needs unmet in lower-income settings, do not require postgraduate medical training to be met. In these cases it may be more ‘marginal community health workers’ rather than ‘marginal doctors’ that are called for.

To give it a quantitative stab, let’s plug in the number of doctors in different countries into the model in the last post and see what the marginal return is like when you increase it by one. These estimates are even more uncertain than our analysis was for the UK, but here are some indicative figures:

So indeed, compared to working in a country like the UK, your doctoring is likely to go further in a lower-middle income country like India and perhaps 1.5–2 orders of magnitude further in Burundi, the lowest-income country on earth. Even so, the results are less compelling than they might first appear.

On this model, an additional Burundian doctor has an impact equivalent to around $1,000,000 given to a GiveWell top charity. This is a lot of money, but donating this much over a career (especially a highly paid medical career) would be remarkable, not impossible. It amounts to ~$25,000 each year over 40 years; the average salary for a US doctor is ~$200,000.

So the chequebook could still beat the stethoscope — even one wielded by an MSF doctor in the world’s poorest country.

Doctoring to give

Thinking like this pointed me in the ‘earning to give’ direction, as mentioned in the last post: if my chequebook can indeed do more good, perhaps it is worth making it as hefty as possible. Medicine is highly paid, and one could aim towards the highest-paid medical work.

Yet if the great bulk of one’s impact comes from earnings as a doctor rather than practice as a doctor, it might be worth (especially if early career) looking outside medicine for even higher-earning careers. Pharma and consulting are potentially lucrative exit options, and one could consider retraining for something further afield.

Further, as medical practice is not the only thing you can do to make money, it is also not the only thing you can do in terms of making an impact directly. Earning to give might beat medical practice, but some other form of direct work could be more impactful than either.

Medical research and other medicine-related work

One example closely related to clinical practice is medical research — which 80,000 Hours covered in a separate career review.

Typically, if you look at lists of doctors who’ve saved a lot of lives, you’ll see physician-scientists responsible for major health breakthroughs like oral rehydration therapy or vaccines.

The underlying reason this looks more promising than direct patient care is that scientific contributions scale much better than clinical ones. The researcher making an advance can improve the practice of all clinicians; Jonas Salk and Albert Sabin, who developed the first poliovirus vaccines, did not have to vaccinate every child themselves to drive the disease to near-eradication. Rather than healing the sick one person at a time, you are doing it one population (or one disease) at a time.

This is an oversimplification, and the complications defy easy calculation of ‘expected impact of a marginal medical researcher.’ It is much higher variance (maybe your ideas don’t work, even if they’re good ones). And there’s an issue of replaceability: even if Salk or Sabin never worked on polio, someone would have developed a polio vaccine, so their impact comes from potentially accelerating when polio gets mostly eliminated, rather than whether it ever does.

But the back-of-the-envelope calculation for developing the vaccine still looks promising. Polio had hundreds of thousands of paralytic cases annually worldwide prior to vaccination, so if their research efforts could have sped up the timeline by even just a week in expectation, that amounts to thousands of children not becoming paralysed.

There are other medicine-adjacent roles that could have a high impact. Such work which is commonly (but not exclusively) performed by doctors, and your contributions can reverberate much further than the patient immediately in front of you. Some other examples:

• Running highly effective global health charities
• Medical leadership (e.g. running a hospital or directing the World Health Organisation)
• Public health (e.g. discovering cholera is waterborne or pushing for policy to reduce rates of pesticide self-poisoning)
• What I work on — reducing global catastrophic biological risks
  • You can read more about career paths for reducing biorisk.

Doctoring, what is it good for?

Perhaps clinical practice (or clinical experience) can be a valuable complement to these (or other) paths for impact. Perhaps two days a week in the clinic and three days a week in the lab means one does better research than five days a week in a laboratory, regardless of the (likely modest) impact from one’s clinical activity. Perhaps practical experience is invaluable to a later career directing things from afar. Or perhaps, even if the clinical experience itself is mostly irrelevant, it is a crucial credential or piece of career capital for certain roles.

Perhaps. But perhaps only sometimes, and (I think) it’s true less often than many — especially medical students or doctors rethinking their careers — suppose. Although being a doctor is often useful for impactful roles, it’s not usually necessary.

A medical background is occasionally handy for my work on biological risk, and it was somewhat more handy when I worked in public health. But many without a medical background enter and excel in these fields. Although Salk and Sabin trained as physicians, Maurice Hillman — who developed 40 vaccines, perhaps saving even more lives — did not.

And becoming and practicing as a doctor involves significant time costs. The hypothetical researcher who spends two days a week in the clinic is sacrificing 40% of their research activity. It’s just far from clear that jumping through the usual hoops to become a doctor is necessary when you think your impact will come from doing other work. If you already are a doctor though, and you want to pivot into higher-impact roles, you certainly can make use of the career capital you gained by getting an MD.

Non-medical careers as a doctor

If you look at some of GiveWell’s top charities, they are typically concentrated on a small subset of possible global health interventions. One reason for this is that some interventions are much more promising than others. Although charity performance does not only rely on the promise of the intervention, this does explain a lot of the variation in cost-effectiveness (very efficiently deploying a cost-ineffective intervention is still ineffective). Restricting yourself to only give to, or work in, a particular subset of interventions (e.g. education, AIDS, charities in a particular country) may greatly limit your impact if better options lie further afield.

In the same way, if it turns out medical practice does not do that much good, but you’re interested in doing the most good (e.g. you write a pretty cringe personal statement asserting being a doctor is the most worthwhile thing versus anything else you could do instead), perhaps your best option is not just ‘not clinical practice’ — but nothing related to medicine at all.

Biorisk is only tangentially related to the day-to-day practice of medicine; working on transformative AI is not really related at all. These could be more important than more medicine-adjacent areas.

Conclusion

What to make of all this? Although it is not quite ‘medical school: just say no,’ it is perhaps ‘medicine: not a no-brainer.’ The typical medical career I had in mind when I applied, rather than ‘best out of all the other things I could do,’ is not nearly as promising an opportunity to do good, in my view, as more atypical paths or non-medical career options would be. This includes the path I ended up pursuing. Realising this sooner would have been valuable for my career.

This lesson generalises beyond medical careers to approximately everything about doing good. The really good global health interventions are many times more effective than the merely worthwhile ones, and one needs a lot more than rough heuristics to find them. Which problems are more important, or more pressing, than others may be better found by careful reflection than by snap judgement. And spending some of your 80,000 hours figuring out how your career might do the most good is time well spent.

Learn more:

• Read our in-depth career guide
• Problem profile: Preventing catastrophic pandemics
• Career review: Biorisk research, strategy, and policy
• Podcast episode: Dr Pardis Sabeti on the Sentinel system for detecting and stopping pandemics
• Podcast episode: Elie Hassenfeld on two big picture critiques of GiveWell’s approach, and six lessons from their recent work
• Podcast episode: Andy Weber on rendering bioweapons obsolete and ending the new nuclear arms race