Stop saying “first, do no harm”. Seriously, just stop. A moment’s reflection will reveal how inane and unhelpful the phrase is. Might doctors forget and start harming their patients if we don’t regularly remind them not to? However, more than inane, “first, do no harm” paints an unrealistic picture of medicine that actually causes immense harm.
Doctors as idiots
Don’t harm your patients. That seems simple enough. We obviously don’t want doctors purposefully hurting patients, but do we really think that they require the reminder? (And if anyone is psychopathic enough to be hurting their patients on purpose, do we really think that hearing someone say “first, do no harm” is going to stop them?)
Taken at face value, I find the statement insulting. It seems to imply that we are dumb or absent minded enough to require constant reminders not to hurt our patients.
The harms of “First, do no harm”
However, the insult really isn’t what bothers me. The real reason that the phrase needs to be banned from the medical corpus is that it perpetuates a misguided understanding of medicine that, somewhat ironically, results in significant harm.
Medicine cannot be practiced without causing harm. Those supposedly sacred words are a lie. Any treatment that has an effect will also have side effects. Any hope of benefit will also be accompanied by a risk of harm. That is a fundamental, but often overlooked, rule of medicine.
Harm cannot be avoided. Do you withhold aspirin from a STEMI patient because it will occasionally cause a GI bleed? Do you withhold antibiotics in meningitis because of the risk of Clostridium difficile? Do you place chest tubes? Do you intubate? It might be catchy, but “first, do no harm” is consistently ignored in medicine, because every medical intervention will cause some harm.
I would like to assume that no one means for “first, do no harm” to be interpreted in this senseless way. However, at almost every conference I attend, someone seems to pull out the inane “first, do no harm” straw man. “Didn’t you know that succinylcholine can cause hyperkalemia. First do no harm!!” “Thrombolytics can cause head bleeds. First do no harm!” “Thoracotomies cause open chests. First do no harm!!!” Mic drop.
I don’t believe that I have ever heard “first, do no harm” invoked as part of a balanced medical discussion. The phrase is almost exclusively used as an inane rhetorical device; as a way to obfuscate intelligent conversation. (If we were willing to engage in intelligent conversation, we wouldn’t need to parrot thoughtless slogans.)
When harms clearly outweigh benefits (or when benefits clearly outweigh harms), the appropriate action is obvious, and “first, do no harm” is, at best, redundant. When harms and benefits are closely balanced, “first, do no harm” could act as a guiding principle, but it seems to fail both theoretically and practically in modern medicine. Strictly following this advice seems to lead to an incredibly paternalistic practice of medicine, in which physicians judge these difficult decisions independently, ignoring patients’ values, and excluding patients from the decision making process. Perhaps more importantly, I think the evidence is fairly overwhelming that doctors ignore this advice.
Doctors aren’t listening
Doctors consistently overvalue benefits and discount harms. Despite the endless repetition of “first, do no harm”, the evidence suggests that we tend to do the exact opposite. We err on the side of intervention, regardless of the harms.
Consider the use of heparin for NSTEMI. When I read the evidence, I see clear harm without any benefit. However, in the very best light, heparin might cause a small decrease in non-fatal MI, but an equal increase in major bleeding. (Andrade-Castellanos 2014) If “first, do no harm” was truly a guiding principle in medicine, we would give more weight to the harm in this balance, and advise against heparin. Instead, guidelines enthusiastically recommend using heparin, completely ignoring the harm. (Amsterdam 2014) When it really matters, we ignore the words of Hippocrates.
There are endless examples. When harms and benefits are equally balanced, we consistently discount the harms and tout the benefits. PPIs are still widely prescribed for acute GI bleeds, despite no evidence of benefit and some clear indications of harm. We prescribe antibiotics for otitis media. We cast buckle fractures and Salter-Harris 1 injuries. We debride burn blisters. We acutely manage asymptomatic blood pressure. We use adenosine for SVT. We consistently ignore harms and fixate on benefits.
The culture of “do no harm”
Despite being rather inane, “first, do no harm” is repeated consistently enough to influence medical culture. Those 4 silly words establish an expectation of perfection. Harm becomes taboo. When a treatment I prescribe harms a patient, I have violated my oath. (“First, do no harm” is not actually part of the Hippocratic oath, but the feeling is the same.) This culture dissuades us from openly discussing the harms we routinely cause. We spend much less time teaching our medical students about the harms of our treatments than we do the benefits. Our patients often assume that anything we suggest must be perfectly safe, because a doctor would never do harm.
The taboo of harm is ingrained in the culture of medicine. We talk about “risks and benefits”, but this is an inherently unbalanced equation. The benefit is stated as a given, whereas harm is only mentioned as a possibility. We won’t even say the word “harm”.
Studies are powered for benefit and often don’t even look for harms. Systematic reviews always report benefits, but often fail to mention harms. The same is true of guidelines. We have made harm a dirty little secret in medicine.
The constant repetition of this phrase creates a false impression. It leaves medical students, patients, and maybe even experienced doctors with the impression that there are risk free treatments. There are not.
Ironically, the idea that doctors would never do harm underlies the proliferation of harmful practices. “What’s the harm?” is a common refrain in medicine. We offer unproven therapies with little regard for their potential harms. (The COVID pandemic is a perfect example, with medications like hydroxychloroquine being widely prescribed despite no indication of benefit, and plenty of harms.) We see patients undergo invasive angiography, despite a lack of benefit. We see futile rounds of chemotherapy. We see unnecessary surgeries. We see patients demanding antibiotics for their viral illnesses. We routinely order full body CTs in trauma. “First, do no harm” fosters a medical culture in which harm is shameful, thus ignored, and allowed to proliferate.
I will admit, “always ensure that benefits outweigh harms, and if there is uncertainty use your judgement and involve the patient in decision making” is not nearly as catchy. However, “first, do no harm” is either inane, insulting, disingenuous, or misleading. It is either a self-evident cliche, or it is a pernicious distortion of true medical practice. Every medical intervention causes harm. It is impossible to “do no harm”, and by continuing to pretend we can, we are hurting our patients.
Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC guideline for the management of patients with non-ST-elevation acute coronary syndromes: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014; 130(25):2354-94. [pubmed]
Andrade-Castellanos CA, Colunga-Lozano LE, Delgado-Figueroa N, Magee K. Heparin versus placebo for non-ST elevation acute coronary syndromes. Cochrane Database Syst Rev. 2014;6:CD003462. PMID: 24972265
Justin Morgenstern. Stop saying “First, do no harm”, First10EM, 2020. Available at: