Medical doctors in clinical practice prevent deaths, relieve suffering, improve quality of life, and support patients and families through difficult experiences. This one-on-one clinical care frequently has profound impact on the individual patients being treated and their families - just ask anyone who has had to attend a hospital in the midst of a serious medical emergency. Many doctors also increase their impact through teaching, advocacy, developing new services or identifying ways to improve the quality of care.

But how much benefit can be attributed to any one doctor? In a three-part article for 80,000 Hours, Dr Greg Lewis attempted to estimate the number of lives saved or improved by a “typical” clinical doctor over their career. Lewis’s calculations produced a surprisingly low estimate of impact, although the available evidence doesn’t support a precise answer and comparisons with other career paths must be undertaken with caution and while applying the same standards.

His analysis points towards three ways doctors may be able to have substantially more impact:

1. Providing clinical care where there is a significant shortage of doctors

2. Using their training to help solve important, solvable, and neglected problems by by developing, improving and/or implementing programmes, policies, technologies, or health systems 

3. Donating part of their salary to highly effective charities, enabling significant additional impact alongside that achieved through their work.

Medicine is only one determinant of health

Much of the improvement in life expectancy achieved over the past two centuries resulted from better nutrition, sanitation, housing, education, and working conditions. Medical advances (e.g. vaccines, antibiotics) have also prevented illness and premature deaths, and provision of good clinical care remains essential for population health. See our blog post for more discussion about this.

Medical doctors helped produce some of the knowledge behind improvements conventionally classified as social or public-health advances. John Snow’s work on cholera and water supply is one familiar example. Doctors can have substantial impact by helping to change the conditions that determine whether populations become ill.

A first estimate: around 70 lives over a career

Lewis’s starting point comes from a 1994 analysis by John Bunker, Howard Frazier, and Frederick Mosteller. The researchers examined preventive and therapeutic interventions for major causes of illness and death, drew estimates of their effectiveness from clinical studies, and considered how widely those interventions were used in the US population. They concluded that medical care accounted for approximately five years of the roughly 30-year increase in US life expectancy during the twentieth century. They estimated that it also relieved around five years that would otherwise have been lived in poor health. 

Lewis converted these gains into disability-adjusted life years, or DALYs. Five additional years of life equal approximately five DALYs averted. He assigned the five years relieved from poor health an illustrative disability weight of 0.25, producing another 1.25 DALYs. This gives a total estimated 6.25 DALYs averted by medical care for each person in the population.

Lewis then attributed all this benefit to doctors and, using doctor density data from the UK (approximately 3 doctors per 1,000 people), calculated about 2,100 DALYs averted per doctor over a career. Assuming 30 DALYs are equivalent to one life saved, he estimated approximately 70 lives saved per doctor’s career, or just under two lives per year of clinical practice.

The additional value of one doctor

This “70 lives saved” estimate is reached by dividing the overall benefit of medical care equally among doctors. It does not tell us what changes because one particular person chooses to become or remain a doctor.

Lewis therefore considered a doctor’s counterfactual impact: the difference between what happens with that person in the role and what would otherwise have happened. In a well-staffed health system found in most high income countries, if one doctor was removed then much of that doctor’s work would be covered by another doctor or redistributed among colleagues. Thus that individual doctor’s counterfactual impact is much lower than we might initially have thought. The additional doctor may still reduce waiting times, improve the quality of care, or allow the service to treat more patients, but their impact is smaller than the total benefit received by everyone that doctor personally treats.

Impact of an additional doctor where doctors are scarce

Lewis used international data to compare physician density with disease burden across countries. He saw that disease burden fell sharply as physician density increased from almost zero to around 50 doctors per 100,000 people, but declined much less above that level. World Bank data show that many African countries have fewer than 50 doctors per 100,000 people, and some have fewer than 10 (compared with 240–450 in Canada, the UK and Germany). 

The pattern suggests that an additional doctor may have much more impact where physicians are scarce. But caution is required, because these data do not prove that adding doctors causes a difference in health burden, or establish a precise threshold. Countries with fewer doctors also tend to have greater poverty and shortages of medicines, facilities and other health workers.

Lewis uses these doctor density/health burden data to estimate the marginal benefit of adding one doctor to a health system. At a density of around 300 doctors per 100,000, his calculation suggests that one additional doctor would avert approximately 5 DALYs annually, or about 7 lives saved in the doctor’s career (This is lower than his original estimate of 70 lives because the two calculations answer different questions. The first assigns each doctor an equal share of medicine’s total historical benefit; the second estimates the additional benefit of one more doctor in an already well-supplied system).

This illustrative example is not a reliable causal estimate, but nevertheless demonstrates how long-term clinical work in severely underserved health systems or humanitarian settings may have considerably greater direct impact. True impact is difficult to predict, however, because the doctor shortage may not always be the main constraint: medicines, facilities, nurses, community health workers, referral systems, or appropriate local knowledge also matter and may be the primary bottleneck.

Why even the revised estimate may be inaccurate

Lewis’s calculations initially assign all the benefits of healthcare to doctors. In reality, healthcare is provided by nurses, pharmacists, community health workers, therapists, laboratory staff, administrators, and others. Dividing the benefits more realistically among these contributors would further reduce the amount attributed to doctors.

Combining this consideration into his calculations, Lewis proposed a final estimate of approximately one DALY averted for each year worked in the UK, perhaps one or two lives saved over a career. This is not a reliable empirical result because of various speculative adjustments in the calculation. Its value lies in challenging the assumption that the benefits received by a doctor’s patients are all wholly attributable to that doctor’s career choice.

The calculation also omits aspects of medicine that DALYs capture poorly, including reassurance, dignity, support for families, advocacy, and the value of having a dependable health system. A typical clinical career remains highly impactful for the patients and families involved, even if its additional population-level effect cannot be measured accurately.

Extending impact beyond clinical care

Doctors may have a much greater impact when their work focuses on solving health problems that affect populations, especially those that are important, solvable and neglected. Possible pathways include: developing and evaluating treatments, vaccines, and diagnostics; improving how healthcare is delivered; creating services for underserved populations; leading public-health programmes or health organisations; influencing government policy; founding an organisation addressing an important unmet need; reducing pandemic and other biological risks; and teaching, advocacy, and changes to clinical practice. 

Doctors have made important contributions throughout history to improvements in nutrition, water safety, housing standards, occupational health, and disease prevention precisely because some used their knowledge beyond individual patient care. Medical training and experience is not essential for every role, but can provide uniquely useful knowledge, skills and insights, as well as credibility for this work.

Moving from treating one patient at a time to improving a policy, programme, technology, or health system can sometimes allow one person’s work to benefit thousands or millions of people. This large potential impact is, however, also uncertain: many potentially high impact projects will fail, while a few achieve exceptional results.

Counterfactual impact applies to other careers too

When comparing a medical career with other options, Lewis’s reasoning should also apply to the evaluation of other forms of high-impact work, such as in non-profit or for-profit organizations, research, government, and entrepreneurship. For example, most job openings at highly effective global health or biosecurity organizations are highly competitive, often with hundreds and sometimes thousands of applications. If the chosen strongest applicant had not applied, the next best applicant would presumably also be highly capable. Thus the counterfactual impact of the successful applicant is limited.

The work of impactful organisations is also undertaken by teams, as it is in medicine too. The resulting impact must be attributed across all the team members, not just the leader or other chosen individual.

Effective giving can add substantially to a doctor’s impact

Doctors in many countries earn relatively high salaries. Donating a proportion of that income to highly effective charities can therefore add considerably to their career impact.

Lewis illustrates this using his first year as a junior doctor. He donated approximately £10,000, including about £5,000 to highly effective global-health charities. Using the cost-effectiveness estimates available at the time, he concluded that these donations probably produced substantially more measurable health benefit than his additional clinical contribution that year.

If the additional impact of a typical doctor in a well-supplied health system is relatively modest, and highly effective charities can save or improve lives at low cost, regular giving will add greatly to the overall impact of a medical career. A doctor’s work and their effective giving are complementary rather than competing choices.

What should doctors take from this?

Lewis’s estimate of one or two lives saved over a medical career is far too uncertain to be treated as a reliable fact. Nor does it capture all the ways doctors help patients, families, colleagues, and communities.

Its more useful message is to highlight that doctors may be able to accomplish far greater positive impact by:

• Practising where physician shortages leave important needs unmet

• Focusing their work on solving important, solvable, and neglected problems by developing, improving and/or implementing programmes, policies, technologies, or health systems

• Supporting highly effective charities by donating a proportion of their salaries through their career

A conventional clinical career does a substantial amount of good. Medical training can also be a starting point for forms of work and giving that improve health far beyond the patients one doctor can treat personally.