Intended for healthcare professionals

Observations Ethics Man

Is doctor-patient confidentiality dying a slow death?

BMJ 2018; 363 doi: https://doi.org/10.1136/bmj.k4313 (Published 15 October 2018) Cite this as: BMJ 2018;363:k4313
  1. Daniel Sokol, medical ethicist and barrister
  1. 12 King’s Bench Walk, London
  1. daniel.sokol{at}talk21.com
    Follow Daniel on Twitter @DanielSokol9

The threshold for divulging information to other professionals should remain very high

Mary, an acquaintance, called me early one morning. She said her husband had grabbed her throat and slapped her a few days previously. Their children were asleep at the time, oblivious to the incident. That same evening Mary read online that victims of domestic abuse should visit their GP.

Mary shared her story with her GP. Despite the GP’s encouragement to disclose the incident to the authorities, Mary insisted that nothing be said to anyone. She wanted the marriage to work. Her husband had never hit her before; the injuries were limited to minor bruising; and the children were not in any danger.

A few days later some police officers knocked on Mary’s door. When Mary let them in, the police officers arrested her husband in the presence of her terrified children. “Where are you taking my daddy? Leave him alone,” one of their children shouted in tears. As he was leaving, one of the policemen told Mary, “The doctor wants you to know that she’s not the one who called the police.”

Although the GP had reassured Mary in the consultation that she would not breach her confidentiality, she had changed her mind and called social services, who then called the police.

The family is now in turmoil. Mary felt betrayed by the GP and has lost all faith in the medical profession.

There is no doubt that the GP had good intentions, but did she act ethically?

One moral justification for keeping patients’ secrets is that it tends to lead to better health. The promise of confidentiality helps patients overcome their natural reluctance to divulge sensitive information, such as bodily functions and personal feelings, which may assist the doctor in treating the patient. Remove confidentiality, and patients may withhold medically important information or not visit at all. This could harm the patient and, in the case of certain transmissible or psychiatric disease, others.

Another moral reason for respecting a patient’s confidence is the implicit promise that the patient can entrust secrets to the doctor without fear of disclosure. Armed with this trust, the patient can run her life safely in the belief that the doctor will not break her promise and infringe her autonomy. This is why Mary felt so betrayed. The doctor broke a promise.

Those moral justifications for confidentiality do not trump all considerations. Doctors should breach the confidence of a suicidal commercial pilot or a driver who is unfit to drive, for example, even if the patient refuses to consent to the disclosure. Although it is not an absolute moral obligation, the threshold for breaching confidentiality should nonetheless be very high.

Mary’s case did not reach that high threshold. There was no credible risk of significant harm to others. To find otherwise would set a worrying precedent: many victims of domestic abuse would refuse to see their GP or would lie to them about the true cause of their injuries. This would be undesirable on a personal and societal level. It would further damage the reputation of doctors, earned over centuries, as professionals who can be trusted with secrets.

The GP, undoubtedly with good intentions, tried to salvage the relationship by inviting Mary to a follow-up appointment. This served only to infuriate Mary, who didn’t want to see the GP again. An apology would have been more helpful.

When I put the phone down I wondered if doctor-patient confidentiality, once so robust, was dying a slow death. If so, it needs urgent treatment.

Footnotes

  • Competing interests: None declared.

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

  • Patient consent obtained.

View Abstract