Do No Harm … Whatever That Means

      The phrase “first do no harm” (in Latin, primum non nocere) is fundamental to the ethics of medical treatment in the Western world and has endured for approximately 2,500 years. It is attributed to Hippocrates, who wrote that “The physician must ... have two special objects in view with regard to disease, namely, to do good or to do no harm.” These are the principles of beneficence (“to do good”) and nonmaleficence (“to do no harm”). The Hippocratic oath contains similar language. Many practitioners today still pledge that oath or at least receive a copy of it early in their training. The phrase is well known to almost any practitioner, yet despite its ubiquity, and perhaps even because of it, it is worth asking this question: What does “first do no harm” actually mean today? Has its meaning or importance changed over time? I will discuss the historical meaning of the phrase, its relationship to other ethical principles, the changing nature of harm, and the changing ethics of health care, as health care delivery and health care culture continue to evolve.

      Why “Do No Harm”?

      The primary ethical principle that forms the foundation for medical care (and the foundation of every society throughout history) is the concept of beneficence — the obligation to do good for others. The ability to do good is also the obligation to do good. The word hospital shares the same Latin root (hospes) as the word hospitality. It implies a duty to shelter and take care of a stranger or guest.
      The desire to do good often compels health care practitioners to perform some action in virtually any clinical situation, regardless of its effectiveness or even lack thereof. It is a reflection of our training, if not our DNA. Even so, medical and surgical treatments are inherently dangerous. Moreover, countless studies around the world have demonstrated that physicians and the public perceive the benefits of almost every treatment to be far greater and the risks substantially lower than they really are. There is a biased belief in the good that will be done. This belief is based on hope, fear, an exaggerated sense of the power of medicine, surgery, and the expectation that doctors will do good, as well as a cultural and historical awareness of “miraculous” advances in medicine and surgery over the past 150 years.
      The development of antibiotics, insulin therapy, and antihypertensive treatment in the 20th century, for example, resulted in the cure of previously dangerous infections and transformed conditions such as diabetes and hypertension from fatal diseases into chronic illnesses. As the science advanced, the goals of treatment and the ethos of health care changed; death became the enemy, rather than an inevitable consequence of life. Fighting death, therefore, became an important, if not primary, goal of American health care. Moreover, patients were expected to fight death as well, no matter how awful the treatment or what harm it caused.
      Preventing illness and screening for disease (or potential disease) in asymptomatic people is now considered an imperative. The benefit of various preventive measures, and the perceived absence of risk of testing and treatment (e.g., from screening tests such as prostate-specific antigen and mammography, or from statin therapy) are likewise far from reality.
      More can be done now to prevent or treat illness than at any other time in history. That more can be done compels further action despite the risks. More is done, and more harm occurs as a direct result.
      Many treatments such as chemotherapy, surgery, and even hospitalization for older patients virtually guarantee harm. Almost every patient receiving them will get worse in some way to even have a chance of getting better (though the risk of problems arising from forced bedrest such as deconditioning, muscle loss, pressure ulcers, and falls are not unavoidable). The admonition to first do no harm is a cautionary tale: the risk may outweigh the benefit. This, of course, demands that practitioners must know the risks. “Do no harm” requires knowledge, discipline, and maturity. It requires training and professionalism. The word “first” is also highly significant — a deed once done may be irreversible. As the proverb says, we should “measure twice, cut once.” But first, also, because doing no harm is not enough; the goal, the duty, is to do good — and to do good safely.
      The risk of adverse drug effects increases almost exponentially as the number of prescribed medications increases. The risk of harm from medical treatment is higher among older patients, in those with multiple comorbid conditions, and in patients of smaller size (i.e., women and children). One size never fits all. “Do no harm” therefore requires that drug prescribing take into account aging physiology, body weight, renal function, and other concomitant medications or diagnoses, rather than rely on standard dosages for all adults or using other aspects of cookbook medicine.


      In recent decades, the principle of autonomy — the right of patients to make their own decisions — has been ascendant. The clinician must still do good and avoid harm, but the patient ultimately has the right to decide. Patients may choose to accept a greater risk of harm, either from treatment or from the decision to forgo treatment.
      Rather than absolving clinicians of responsibility, however, autonomy raises the bar even farther. We are responsible for communication. It is incumbent upon clinicians not just to know the risks, benefits, and alternatives, but also to teach them to our patients. This forms the basis of the principle of informed consent. We must not only respect the choices that others make, but we must respect others regardless of the choices they make. This is especially important when their choices differ from those we would make for ourselves.
      Not giving patients and families accurate information and not respecting them or their choices is harmful. Abandoning them as a result of their choices is unconscionable.

      New Kinds of Harm

      Harm can take many forms. Some are intrinsic to a specific treatment or medication, but others are caused by care delivery more broadly. Some are age-old, and some change with the times. In the mid-1800s, physician Ignaz Semmelweis discovered that the high incidence of maternal death from sepsis after childbirth in Europe was caused by poor hygiene among physicians. Nearly 200 years later, health care–acquired infections are a major cause of harm and death, along with antibiotic-resistant microorganisms.
      Increasingly fragmented care across settings, and fragmented responsibility among an increasing number of care providers for a single patient and a single episode of illness, systematically introduce error, risk, and harm. Discharging sick patients from hospitals causes harm, as evidenced by the rising rates of hospital readmissions with decreasing lengths of hospital stays. Lower staffing levels in health care settings also increase the risk of harm. The proliferation of “information” in the form of medical records generated for the purpose of maximizing reimbursement minimizes the effectiveness of communication among providers, reduces productivity and thus access to care, and causes harm. Likewise, defensive medicine practices, which expose patients to unnecessary testing and treatment, are offensive and harmful. Defensive medicine is especially reprehensible insofar as its purpose is not to benefit the patient.

      Economic Harm

      The cost of health care in this country is so high, and so out of step with the rest of the world, that even paying for insurance causes severe financial hardship for many. In the extreme, the high cost of care may force some patients to choose between financial ruin caused by treatment or death without it. A cruel and immoral irony is that the patients without insurance — those who can least afford the cost of care — are charged more for it.

      Trained To Harm

      As doctors and nurses, we are trained as apprentices to do things a certain way, with the expectation that we must always do things that way. We develop habits that are intended to persist decades beyond our formal training. We assume without question that we are taught to do things the right way, so we are highly resistant to change. Unfortunately, all of us have unwittingly been trained to do some things that are ineffective, harmful, and out of date.
      These learned practices persist even when overwhelming evidence accrues over the years to attest to their harm. For example, numerous drugs are still prescribed in the United States even though the Food and Drug Administration has withdrawn any and all clinical indications for their use. Other interventions such as vertebroplasty for osteoporotic vertebral fractures have been shown to have absolutely no benefit compared with placebo and confer significant harm to many, yet they are still widely performed.
      Every medical and surgical specialty is now trying to reckon with this problem through a concerted national effort to engage and empower the public. The national Choosing Wisely campaign encourages patients to question the common medical and surgical practices performed by their esteemed board-certified diplomates ( This strategy may prove effective in reducing harm, but it is unlikely to increase the public’s trust in health care.

      Harm: A Sin of Commission or Omission?

      Historically, “do no harm” has been taken to be an admonition against action, or at least an admonition against haste. “Do no harm” may be given as an excuse for not acting. But harm results from the lack of action as well. In the extreme, a lack of action equals neglect. It is a mistake to think that doing no harm is a passive duty — it is an affirmative obligation that requires constant diligence.

      Do No Harm Today

      In 2016, doing no harm means lifelong learning and a willingness to embrace necessary change.
      More explicitly, to do no harm requires routinely reviewing, reducing, and discontinuing medications. It means avoiding antipsychotic drugs and other medications altogether when their sole purpose is to make people with dementia or delirium behave. It means avoiding physical restraints in all settings. It means an obligation during care transitions to participate in an active handoff of care between practitioners, with scrupulous medication reconciliation to avoid medication errors. It means knowing your patient well and accepting responsibility as your brother’s or sister’s keeper. To do no harm means to advocate for high-quality, affordable, universal health care for all. It means active listening, cultural competency, being present and engaged, teaching patients and families, and guiding them through the fragmented, byzantine, and potentially dangerous American health care system.
      It means striving to be the best that we can be, the best that anyone can be for their mother or father. We have an obligation to do good. We also have an obligation to be good — to be good people, and to be good at what we do. We have an obligation to care. These are the ethics for us and our time.
      Dr. Evans is a full-time long-term care physician in Charlottesville, VA, and medical director of two skilled nursing facilities. A past Society president, he serves on the Caring for the Ages Editorial Advisory Board. Read this and other columns at under “Columns.”