Intended for healthcare professionals

Observations Ethics Man

Doctors in danger

BMJ 2016; 352 doi: (Published 23 February 2016) Cite this as: BMJ 2016;352:i1075
  1. Daniel Sokol, barrister and medical ethicist
  1. 12 King’s Bench Walk, London, UK
  1. Sokol{at}

There’s no obligation to stay, no matter the risk

In March 2011 the earth shook off the coast of Japan. The ensuing tsunami breached the walls of the Fukuyama Daiichi nuclear power plant, releasing vast quantities of radioactive material into the atmosphere. Hundreds of thousands of people evacuated the area. In a psychiatric hospital 35 km from Fukushima three of the six full-time doctors resigned and moved away. One explained, “I’m concerned about radiation because I’ve got a small child.”1

A feature of Ebola epidemics since 1976 has been the high death toll among clinicians and the tendency of local doctors and nurses to flee. In the recent outbreak in west Africa many foreign doctors returned home for safety reasons, leaving behind patients and local colleagues. So far, there have been 881 cases of Ebola virus disease among medical staff in Guinea, Sierra Leone, and Liberia and 512 deaths.2

Last October several clinicians working in a hospital supported by Médecins Sans Frontières in Kunduz, Afghanistan, were killed when their hospital was bombed in an aerial assault.3 On 10 January this year an MSF clinic in Yemen was also destroyed, and another two hospitals were bombed in northern Syria last week. Increasingly, humanitarian workers in conflict zones are targeted by armed groups with no concern for law or ethics. It is well known, for example, that improvised explosive devices are often positioned in pairs. After one explodes, the other is set off once helpers arrive to help the injured.

Such occupational risks are not limited to doctors. On 25 January 2016 a senior judge and his family were killed in their home in Yemen by airstrikes. Lawyers in many parts of the world are assaulted, imprisoned, or killed for defending the “wrong” person or, as in China, “subverting state power’.4

Where there is a real risk to personal safety, do doctors have a duty to remain? Should the care of their patients be their first concern?

There is, of course, a utilitarian argument for allowing doctors to walk away from danger. A dead doctor is useless, and an injured one a burden. But, utility aside, no doctor should have to make unreasonable sacrifices. If the danger—be it a risk of infection, radiation, or physical harm—is disproportionate to the good that will result, a doctor is not morally required to help. If a doctor’s competing obligations (to spouse, child, or others) outweighs the duty to patients, then a doctor can derogate from the usual rule of “making the care of the patient your first concern.” The decision to stay or leave is the product of a balancing exercise specific to each doctor and his or her unique circumstances.

In practice, the way to keep clinicians at work in perilous conditions is to reduce the risk by providing training, equipment, and resources; to minimise inconvenience by arranging transport, accommodation, and facilities; and to anticipate the challenges in good time. This is easier to achieve in places like the United Kingdom and United States than in Yemen and Afghanistan.

Will this non-absolutist position, which allows doctors to walk away in times of danger, dilute the nobility of being a doctor? I doubt it. Do we think less of the doctor who left the Fukushima hospital for fear of radiation to his young child? The exception to the duty of care will apply so rarely that most doctors will never invoke it.

William Osler wrote that the “human heart by which we live” must control our professional relations.5 A doctor’s heart is receptive to the needs of patients and colleagues, but it is not closed to others. To pretend otherwise is myopic and, in times of crisis, dangerous.


  • Competing interests: None declared.

  • Provenance and peer review: Commissioned; not externally peer reviewed.

  • Ethics Man Doing the right thing (doi:10.1136/bmj.h5288)


View Abstract