Although medical science is frequently celebrated for its extraordinary achievements, there is a less comfortable reverse side to its history: the legacy of concepts once regarded as absolute truths that science itself subsequently debunked.

For centuries, competent, well-intentioned, and respected physicians defended theories and treatments that today seem strange, useless, or even dangerous. Some were based on incomplete knowledge; others, on misread observations; some were perpetuated by tradition; others, by the authority of their proponents. There were also situations in which a novel practice, later proven correct, was initially met with suspicion or outright ridicule.

This does not mean medicine is a fraud or that “science knows nothing.” It signifies precisely the opposite: science is valuable because it admits the possibility of being wrong.

 

When Handwashing Looked Like an Eccentricity

Few stories illustrate this characteristic of medicine better than that of Ignaz Semmelweis.

In the 1840s, puerperal fever was a leading cause of maternal mortality after childbirth. Semmelweis observed that mortality was substantially higher in one of the obstetric clinics at the Vienna General Hospital, where women were examined by physicians and students who also worked in autopsy rooms.

He noted something unsettling: when doctors washed their hands with a chlorinated lime solution before examining women in labor, mortality plummeted. Between 1847 and 1849, the disparity between the clinics virtually disappeared (Semmelweis, 1847–1849; Best; Neuhauser, 2004).

Today, this seems obvious. At the time, it was not. The germ theory of disease was not yet established. The premise that a physician’s hands could transmit an invisible agent from a cadaver to a healthy woman was difficult to accept. The proposal encountered resistance, and the routine adoption of obstetric antisepsis took decades. The subsequent work of Joseph Lister and the broader emergence of germ theory contributed decisively to shifting surgical practice (Tröhler, 2015).

Lister published his classic paper on the antiseptic principle in surgery in 1867. His proposal also faced debate and skepticism, particularly in Great Britain. Acceptance did not occur overnight.

The narrative is often retold as if Semmelweis had uncovered a self-evident truth rejected by obstinate doctors. The reality is more nuanced: he held an exceptionally vital clinical observation, yet lacked an adequate microbiological explanation.

This distinction is essential. One can be right about an empirical phenomenon while mistaken about its underlying mechanism. Conversely, a seemingly elegant theory can persist for years without yielding genuine clinical benefits.

 

Bleeding the Patient to Cure Them

For centuries, bloodletting stood as one of the primary therapeutic pillars of Western medicine. Phlebotomy was applied across numerous diseases, sustained by physiological models that appeared coherent within the medical framework of each era.

Today, we know that for the vast majority of conditions in which it was prescribed, bloodletting was either ineffective or actively harmful.

The historical trajectory is instructive because its obsolescence did not simply follow an instantaneous revelation of the truth. It was a gradual retreat driven by a fundamental shift in how the human organism and disease itself were conceptualized (Drakman, 2025).

Contemporary medicine still uses therapeutic phlebotomy in specific contexts, such as hemochromatosis and polycythemia vera. Hence, concluding that “bloodletting is bad” remains overly simplistic. The exact same intervention can be useless in one setting and essential in another. Context dictates validity.

 

The Cigarette That Calmed Nerves

It may be hard to fathom today, but tobacco was once far more than tolerated: for centuries it was regarded as a medicinal plant and prescribed by physicians for diverse therapeutic aims (Greene, 2020).

By the 20th century, tobacco industry advertising leveraged the prestige of physicians to project safety. During the 1930s and 1940s, cigarette advertisements made direct appeals to medical authority, associating specific brands with lower throat irritation or greater health margins (Gardner; Brandt, 2006).

A subjective perception supported this notion: smoking appeared to relieve anxiety.

Today, the phenomenon is clearer. Nicotine produces acute mood shifts and temporarily relieves unpleasant symptoms; however, in dependent smokers, much of this perceived “relaxation” simply reflects the reversal of acute withdrawal symptoms induced by the dependence itself (Parrott, 2003).

What presented as an anxiolytic was actively contributing to the very problem it purported to soothe. This serves as a vital reminder for modern clinical practice: immediate symptomatic relief does not establish long-term clinical benefit.

 

When Food Refinement Was Heralded as Progress

Nutritional history contains its own paradoxes. With industrial milling and polishing, grains such as rice became increasingly refined. White rice was visually appealing, had a longer shelf life, and was perceived as a premium commodity.

However, the polishing process stripped away the outer grain layers containing essential micronutrients.

The research of Christiaan Eijkman—later recognized with the Nobel Prize—revealed the link between polished rice consumption and a condition mimicking beriberi in animal models. Subsequent investigations demonstrated that the syndrome stemmed from the deficiency of a vital nutrient now known as thiamine, or vitamin B1 (Eijkman, 1897; Carpenter, 1995).

The paradox is striking: technological progress yielded a product that was cleaner, more refined, and more marketable, while simultaneously stripping out crucial nutritional value.

This does not imply that all refined food is inherently toxic. It means that visual appeal, palatability, shelf life, technological complexity, and nutritional value are not synonymous—a distinction that remains intensely relevant.

 

Alcohol and the Myth of the “Healthy Dose”

Perhaps no substance has generated more ambiguity in medical communication than alcohol.

For decades, the notion spread that moderate alcohol intake—particularly red wine—could exert cardioprotective effects. Observational cohorts consistently identified associations between moderate consumption and lower cardiovascular endpoints, giving rise to the famous J-shaped curve.

The critical issue is that statistical association does not establish causality. The comprehensive Global Burden of Disease study published in The Lancet in 2018 analyzed data from 195 countries and concluded that overall health risk escalates with any alcohol intake, finding that the level of consumption minimizing health loss is zero (GBD 2016 Alcohol Collaborators, 2018).

This does not imply that earlier clinicians recommending a daily glass of wine acted in bad faith. It demonstrates that the interpretation of empirical evidence evolves. When more sophisticated analytical methods control for confounding factors and selection biases, long-standing medical assumptions often require revision.

 

Promising Pharmaceuticals That Went Wrong

Pharmacological history offers stark examples. Thalidomide, introduced in the late 1950s as a sedative and antiemetic for morning sickness, was prescribed internationally and subsequently linked to severe congenital malformations. In the United States, FDA reviewer Frances Kelsey blocked domestic approval due to insufficient safety data. This watershed event catalyzed sweeping reforms in regulatory frameworks, establishing strict demands for substantial evidence of efficacy and rigorous pre-market toxicity evaluations (FDA, 2023).

Another prominent case is diethylstilbestrol (DES). Prescribed for decades to pregnant women to prevent miscarriage and premature birth, it was later shown to lack efficacy for these indications while significantly elevating the incidence of clear cell adenocarcinoma of the vagina and cervix in female offspring exposed in utero (NCI, 2026).

These were not fringe nostrums; they were fully integrated into standard medical care. This highlights an enduring truth regarding “conventional” medicine: conventional describes what is accepted in a given era, not what is immutable or final.

 

Surgeries That Worked—at Too High a Cost

Surgical history mirrors these trajectories. In bariatric care, the jejunoileal bypass—developed in the 1950s and 1960s—induced drastic weight reduction by radically shortening functional intestinal surface area.

However, this mechanical efficacy carried devastating penalties: intractable diarrhea, profound electrolyte imbalances, severe malabsorption, hepatic failure, and oxalosis-related nephropathy. The procedure was ultimately abandoned (Kremen et al., 1954; Griffen, 1981).

Here emerges a core medical tenet: an intervention cannot be judged solely by its primary target outcome. A procedure might dramatically improve a surrogate marker, alleviate an acute symptom, or drop total body mass, yet remain unacceptable if the risk-benefit balance is fundamentally flawed.

 

Lobotomy and the Hazard of Vested Authority

Few interventions remain as cautionary as the prefrontal lobotomy. Developed by António Egas Moniz, this psychosurgical intervention gained rapid prestige. Moniz was awarded the Nobel Prize in Physiology or Medicine in 1949 for his work on prefrontal leukotomy, and the surgery was widely adopted across multiple continents.

Over time, catastrophic consequences became undeniable: irreversible personality alterations, profound blunting of affect, severe executive dysfunction, and high neurological morbidity. The introduction of chlorpromazine and early antipsychotics in the 1950s accelerated the method’s abandonment (Mundalil Vasu; Bunch, 2014).

This chapter is disquieting because it proves that even an idea celebrated with the highest scientific accolade can prove destructive. An award does not turn a clinical hypothesis into permanent truth.

 

When Oxygen Proved Toxic

Neonatology provides another pertinent example. During parts of the 20th century, supplemental oxygen was administered liberally to premature infants. Decades of clinical observations ultimately linked unmonitored hyperoxia to retinopathy of prematurity, a primary cause of childhood blindness.

The conclusion is nuanced: an intervention that rescues lives in acute hypoxemia induces severe microvascular harm when delivered in uncalibrated excess. There is no simple binary between “good” or “bad” molecules. Dose, timing, target population, and clinical context dictate the net clinical outcome.

 

Hormone Replacement Therapy

This dynamic extends into modern medicine. In 2002, the Women’s Health Initiative (WHI) halted its trial arm of combined estrogen plus progestin prematurely when interim analyses revealed elevated risks of coronary heart disease, stroke, pulmonary embolism, and invasive breast cancer. In 2004, the estrogen-alone trial was similarly stopped following increased stroke risks and a failure to demonstrate the expected cardioprotective benefit (WHI, 2026).

This does not mean hormone replacement therapy is universally inappropriate. It indicates that the clinical question can never be simply: “Does this therapy work?” Rather, the relevant questions must be far more granular:

• For whom?
• Under which clinical indication?
• At what age or chronologic interval from menopause?
• At what dose and route of delivery?
• For what duration?
• Compared to what control or baseline?
• What are the absolute risks and absolute benefits?
• What are the clinical endpoints after five, ten, or twenty years?

 

Critiquing Science Is Not an Attack on Science

This iterative critique is precisely how medical science matures. The problem is rarely ignorance itself; it is the failure to recognize that current premises may be flawed. Ignorance acknowledges an absence of data. Scientific error occurs when a community presumes certainty, only for superior evidence to prove the consensus mistaken.

The latter scenario is far more dangerous because it is accompanied by misplaced clinical confidence. History illustrates several recurring mechanisms that generate false certainty:

• Authority: accepting conclusions solely because an eminent figure endorsed them;
• Tradition: assuming an intervention is valid merely because it is historical;
• Anecdotal experience: treating isolated clinical successes as systematic proof;
• Biological plausibility: assuming that because a biochemical mechanism sounds elegant, it must translate into clinical efficacy;
• Statistical association: conflating correlation with underlying causality;
• Surrogate endpoints: presuming that normalizing a laboratory parameter equates to preventing morbidity or mortality;
• Commercial interests: market promotion outrunning clinical evidence;
• Technological bias: presuming a modern modality is inherently superior to prior alternatives.

Contemporary evidence-based medicine developed robust methodologies to minimize these biases: randomized controlled trials, blinding, intention-to-treat frameworks, systematic reviews, meta-analyses, post-marketing pharmacovigilance, long-term registries, and causal epidemiology.

Yet, none of these safeguards guarantees perfection. Science is not a repository of immutable truths; stating that medical science can err is not an assault on the enterprise, but an accurate appraisal of its epistemology.

A scientific model does not need to be timeless to be useful. It must simply represent the best operational hypothesis given the available data, while remaining receptive to revision as superior evidence surfaces. This capacity for self-correction is precisely what distinguishes science from dogma. Medical knowledge is inherently provisional—though this does not imply that all claims carry equal weight. There remains a profound gap between an unsubstantiated conjecture and a clinical consensus rooted in dozens of blinded, replicated trials. Scientific humility is not scientific relativism.

Recognizing the frontiers and flaws of medical evidence does not mean equating rigorous clinical trials with unsubstantiated internet claims; it means demanding continuous scrutiny even for established practices.

 

Prudence Must Not Breed Nihilism

A natural temptation when reviewing historical errors is complete cynicism toward innovation. That would be an equally damaging mistake.

Had historical clinicians rejected every novelty under the pretext that established paradigms were sufficient, antisepsis, vaccination, antimicrobial therapy, modern anesthesiology, cardiovascular surgery, and critical care medicine would never have evolved.

The mature stance avoids both extremes: neither credulous acceptance of every new trend nor reflex rejection of novel paradigms. What is required is disciplined curiosity coupled with methodological caution. A new hypothesis warrants attention. A novel therapy justifies rigorous testing. An unexpected observation demands formal investigation.

However, a biological hypothesis must not be prematurely deployed as standard care, nor should a therapy be turned into an indiscriminate, universal guideline before outcome data mature.

 

Progress Resides in the Willingness to Change One’s Mind

One of medicine’s most noble attributes is the capacity to state: “We were wrong.” There is no shame in abandoning an obsolete practice when superior evidence emerges. The true professional failure lies in perpetuating an ineffective or hazardous intervention solely out of institutional habit, commercial interest, or deference to authority.

• Semmelweis was correct regarding hand hygiene, even if his mechanistic explanation was incomplete.
• Bloodletting fell out of use when its theoretical foundation dissolved under scientific scrutiny.
• Tobacco ceased to be viewed as a healing herb once its systemic toxicity was documented.
• Thalidomide and DES underscored the absolute necessity of rigorous reproductive toxicology and regulatory independence.
• The jejunoileal bypass demonstrated that anatomical effectiveness without systemic safety is substandard medicine.
• Hormone therapy trials clarified how large prospective cohorts can reshape therapeutic paradigms.
• The evolution of alcohol research illustrates that dietary recommendations must account for residual confounding and selection biases in observational epidemiology.

 

The Medicine of Tomorrow Will Correct the Medicine of Today

Inevitably, some current certainties will be overturned. Dietary recommendations considered robust today may be qualified. Pharmacological interventions may prove excessive in specific subpopulations. Biomarkers routinely tracked today may lose clinical relevance. Surgical techniques will be superseded. Phenotypes currently grouped under a single diagnostic label will likely be disaggregated into distinct pathophysiological entities.

This should not inspire cynicism; it represents the organic progression of knowledge.

The clinician must balance two qualities: receptivity and skepticism. Receptivity to novel paradigms, and skepticism against adopting them before validation; humility to acknowledge diagnostic and therapeutic limitations, and courage to retire ingrained habits when evidence invalidates them.

 

A Medicine Humble Enough to Continue Learning

After more than a century of scientific progress, medicine’s most critical lesson remains clear: revising a consensus in light of better evidence is a sign of rigor, not defeat.

Science is not an immutable monument. It is an ongoing, empirical process. Precisely because errors occur, clinicians must design better trials, evaluate clinically meaningful endpoints, monitor patients over longer horizons, and interrogate both ingrained traditions and therapeutic fads.

Medicine must protect the critical disposition that abandoned phlebotomy, retired hazardous operations, recognized the lethality of tobacco, retracted harmful pharmaceuticals, and adopted aseptic techniques.

The exemplary scientist is not one who claims infallibility, but one who builds checks and balances into clinical practice to discover errors as swiftly as possible. That remains the most rigorous way to honor science—and to protect the patient.

 

Disclaimer

This article is intended solely for educational and historical reflection. The history of medicine is presented here in summary form and must not be used to initiate, modify, or discontinue medical therapy. The fact that an intervention was historically modified or abandoned does not imply that all modern adaptations or related indications are invalid. Clinical management requires individualized assessment considering formal indications, updated evidence, individual risk-benefit balances, and professional oversight.

 

References

BEST, M.; NEUHAUSER, D. Ignaz Semmelweis and the birth of infection control. Quality and Safety in Health Care, v. 13, n. 3, p. 233–234, 2004.

CARPENTER, K. J. Eijkman’s contribution to the discovery of vitamins. The Journal of Nutrition, v. 125, n. 2, p. 155–163, 1995. DOI: 10.1093/jn/125.2.155.

DRAKMAN, A. The open body closed: a rationale for the abandonment of bloodletting, based on nineteenth-century Swedish medicine. Social History of Medicine, v. 38, n. 2, p. 270–290, 2025. DOI: 10.1093/shm/hkae069.

GBD 2016 ALCOHOL COLLABORATORS. Alcohol use and burden for 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. The Lancet, v. 392, n. 10152, p. 1015–1035, 2018. DOI: 10.1016/S0140-6736(18)31310-2.

GARDNER, M. N.; BRANDT, A. M. “The doctors’ choice is America’s choice”: the physician in US cigarette advertisements, 1930–1953. American Journal of Public Health, v. 96, n. 2, p. 222–232, 2006.

GREENE, J. A. The rise and fall of tobacco as a botanical medicine. Journal of Herbal Medicine, 2020.

KREMEN, A. J. et al. An experimental evaluation of the nutritional effects of extensive intestinal bypass. 1954.

LISTER, J. On the antiseptic principle in the practice of surgery. British Medical Journal, v. 2, p. 246, 1867. DOI: 10.1136/bmj.2.351.246.

MUNDALIL VASU, N.; BUNCH, S. António Egas Moniz (1874–1955): lobotomy pioneer and Nobel laureate. Journal of Neurology, Neurosurgery & Psychiatry, 2014.

NATIONAL CANCER INSTITUTE. Diethylstilbestrol (DES) exposure and cancer. Bethesda: National Cancer Institute, 2026.

PARROTT, A. C. Cigarette-derived nicotine is not a medicine. World Journal of Biological Psychiatry, v. 4, n. 2, p. 49–55, 2003. DOI: 10.3109/15622970309167951.

SEMMELWEIS, I. Die Ätiologie, der Begriff und die Prophylaxis des Kindbettfiebers. Pest: C. A. Hartleben, 1861.

TRÖHLER, U. Statistics and the British controversy about the effects of Joseph Lister’s system of antisepsis for surgery, 1867–1890. Journal of the Royal Society of Medicine, v. 108, n. 7, p. 280–287, 2015. DOI: 10.1177/0141076815593720.

U.S. FOOD AND DRUG ADMINISTRATION. A brief history of the Center for Drug Evaluation and Research. Silver Spring: FDA, 2023.

WOMEN’S HEALTH INITIATIVE. About the Women’s Health Initiative: hormone therapy trials. 2026.