When Treatments Were Considered Effective Before Their Risks Were Proven
Medicine does not exist outside of time. What was considered safe and progressive in one period may appear unacceptable decades later. The reason is the limitation of knowledge, tools, and methods of risk assessment.
The safety of a drug is not an absolute characteristic. It is an assessment based on the data available at a given moment. If the data are limited, the drug is new, or the observation period is short, an impression of safety is created that is not always confirmed in the long term.
It has long been known that when re-examined, a significant portion of medical methods are revised. Analyses of the clinical literature show that among medical practices re-evaluated in large randomized controlled trials, approximately 40% are later found to be ineffective, less beneficial than previously believed, inferior to alternative treatments, or even associated with an increased risk of harm to patients.
Historically, many interventions were introduced on the basis of a visible effect. If a drug reduced symptoms, suppressed an infection, or influenced behavior, this was considered sufficient for its use. Delayed complications, cumulative toxicity, or rare severe reactions became noticeable only after widespread use.
In addition, many drugs were perceived as acceptable simply because no alternatives existed at that time. In the absence of antibiotics, modern analgesics, and psychopharmacology, even toxic methods could be regarded as a step forward.
This article examines how the criteria for the safety of drugs were formed and why the history of pharmacology is regularly reassessed. Drugs that have recently entered the market and do not yet have sufficient long-term data may initially be considered safe-only until accumulated statistics confirm this assessment or force it to be reconsidered.
In this context, the proportion of medical practices that are later revised receives a logical explanation. This is not a sign of the failure of the system, but a reflection of how science works when analyzing large datasets and long-term observations.
Heavy metals as therapy and hygiene
In the nineteenth and early twentieth centuries, heavy metal compounds were part of routine medical practice. They were used not as extreme measures, but as standard therapeutic agents.
Mercury was used in laxative preparations, for example as a component of calomel. It was prescribed to adults and children as a “purifier” of the body. A pronounced physiological effect-diarrhea and increased salivation-was perceived as evidence of therapeutic action.
Mercury ointments were widely used in dermatology and ophthalmology. They were used to treat skin infections, blepharitis, and styes. They indeed had pronounced antimicrobial activity, which made their use appear justified. Mercury was also used in the treatment of syphilis-before the introduction of penicillin this was considered standard therapy.
It was also present in hygiene products. Mercury compounds were added to some tooth powders and antiseptic preparations as a bactericidal component. At that time there was no understanding of the cumulative toxic effect during chronic exposure.
Arsenic was also actively used in medicine. In the nineteenth century Fowler’s solution (potassium arsenite) was used for skin diseases, anemia, bronchial asthma, and chronic conditions. Arsenic was included in “tonics” and preparations “for the blood.” In small doses it produced a subjective stimulating effect, which strengthened confidence in the therapy.
At the beginning of the twentieth century, organic arsenic compounds appeared, for example arsphenamine (Salvarsan), which became one of the first targeted chemotherapeutic agents against syphilis. Before the antibiotic era this was indeed perceived as a scientific breakthrough.
Modern toxicology evaluates these practices differently. Both mercury and arsenic have neurotoxic, nephrotoxic, and cumulative effects. With chronic use they are capable of causing systemic damage that in the past was not directly linked to treatment.
At the same time, heavy metals have not completely disappeared from medicine. Arsenic trioxide is now used in acute promyelocytic leukemia in strictly regulated doses and under control of complications. Here the risk-benefit principle applies: in a life-threatening disease the use of a potentially toxic agent is acceptable if the expected effect significantly outweighs the risk.
Mercury has also not completely disappeared from medical use, although its role has fundamentally changed. Mercury compounds were historically used as preservatives in multidose vaccines and in solutions for contact lenses. Debates about the possibility of the safe use of even low doses of mercury continue to this day.
Dental amalgam still contains metallic mercury in a bound state and for decades was used as a durable filling material. In many countries dental amalgam is being phased out, including for environmental reasons-due to contamination of the environment during production and disposal.
Radioactivity as a Means of “Strengthening”
At the beginning of the twentieth century, radioactivity was perceived as a scientific breakthrough. The discovery of radium and X-rays generated not only scientific interest but also therapeutic enthusiasm. Radioactive radiation was associated with energy, renewal, and the “stimulation of life.”
Radium tonics and so-called “healing waters” containing traces of radioactive elements appeared. They were advertised as remedies for increasing vitality and for treating fatigue, anemia, and chronic diseases. The mere presence of radioactivity was perceived as proof of modernity and effectiveness.
X-ray therapy was used not only for oncological diseases. It was also applied for enlarged thymus in children, acne, ringworm, psoriasis, and other skin conditions. The assumption was that suppression of cellular proliferation could correct the pathological process.
At that time there was no full understanding of cumulative radiation exposure and dose-dependent effects. Long-term observational registries were not maintained. Delayed oncological consequences began to be identified years later, when clinical experience had accumulated.
Over time it was established that even relatively low doses of radiation, when repeatedly applied, increase the risk of malignant neoplasms, especially in children and adolescents. Radioactive “tonic” remedies were withdrawn from use, and the application of ionizing radiation became strictly regulated.
In modern medicine, ionizing radiation is used primarily in oncology, as well as in diagnostics. The dose is calculated according to the principle of minimal necessary exposure. Its use is accompanied by strict radiation safety standards and monitoring of cumulative exposure.
Teratogenic Catastrophes
The middle of the twentieth century became a turning point in understanding drug risks during pregnancy. Before that, it was assumed that the placenta reliably protected the fetus from most exposures, and therefore many drugs were considered relatively safe.
Thalidomide was introduced to the market as a mild sedative and sleeping aid, as well as a remedy for nausea. It was widely prescribed to pregnant women. It did not produce pronounced acute toxicity in the mother, which was perceived as confirmation of safety.
Several years later it became clear that thalidomide has pronounced teratogenic effects. Thousands of children were born with severe limb malformations (phocomelia), as well as abnormalities of internal organs. This case led to stricter requirements for preclinical studies, including mandatory evaluation of embryotoxicity and fetotoxicity in animal models, as well as reforms in regulatory procedures for drug approval.
Diethylstilbestrol (DES), a synthetic estrogen, was prescribed for the “maintenance of pregnancy” and the prevention of miscarriage. It was used for decades based on the logic of hormonal support. Later it was established that daughters of women who took DES during pregnancy had an increased risk of rare vaginal and cervical tumors, as well as reproductive disorders. This is an example of a delayed effect manifesting in the next generation.
Both cases demonstrated a fundamental problem: the absence of immediate complications in a pregnant woman does not guarantee the absence of long-term risk for the fetus. After these events, requirements for drug testing changed substantially.
In modern practice, hormonal support during pregnancy, when indicated, is based on the use of bioidentical progesterone. Estrogens are not used for the “maintenance of pregnancy.”
In addition, drugs that do not have sufficient data on safety during pregnancy are considered potentially risky until convincing evidence to the contrary is obtained. This reflects the modern precautionary principle with respect to embryofetal safety.
Opiates and Cocaine in Everyday Medicine
At the end of the nineteenth and the beginning of the twentieth century, the boundaries between a “medicine” and a “dangerous substance” were far more blurred than they are today. Morphine, heroin, and cocaine were not perceived as a social problem-they were considered achievements of pharmacology.
Opiate-containing syrups were prescribed to children for cough, colic, and general restlessness. Dosages were often indicated imprecisely, and sometimes the preparations were sold without strict control. The main criterion was the clinical effect: the child calmed down-therefore the remedy worked.
Diamorphine (heroin) was introduced into medical practice as a “milder” alternative to morphine. It was used as a cough suppressant and an analgesic. Initially it was believed to produce less dependence.
Cocaine was widely used as a local anesthetic. It indeed had a pronounced analgesic effect and became an important stage in the development of surgery. However, its psychoactive properties and potential for dependence were underestimated.
Over time, the accumulation of clinical observations revealed the problem of tolerance, dependence, and toxicity with repeated use. This led to stricter control over the circulation of narcotic substances and the formation of modern regulatory regimes.
In modern medicine, narcotic drugs still retain their importance, but they are used under clearly defined indications and under control. Opioids are used for the treatment of severe pain, including cancer pain, as well as in the postoperative period. Their prescription is accompanied by restrictions on dosage, duration of use, and monitoring of the risks of dependence.
Sedative Drugs of the Past
Each generation of medicine had its own “safe” class of sedative agents. At the beginning of the twentieth century this role was played by barbiturates. They were widely prescribed for insomnia, anxiety, and emotional instability.
Barbiturates indeed produced a pronounced sedative effect. The problem lay in a narrow therapeutic range: the difference between the therapeutic dose and the toxic dose was small. Overdoses, drug interactions, and dependence became a serious clinical problem.
Later, barbiturates were partially replaced by benzodiazepines, which were initially presented as safer. They indeed had a wider therapeutic index and lower acute toxicity.
However, long-term use of benzodiazepines revealed another problem: the development of dependence, reduction of cognitive function, and withdrawal syndrome. Many patients received the drugs for years because the risk of chronic use was initially underestimated.
Today, a significant body of data has accumulated on these classes of drugs.
Barbiturates are recognized as drugs with a high risk of overdose and pronounced drug dependence. Their use in psychiatry and in the treatment of insomnia has been largely discontinued.
Benzodiazepines were initially considered a safer alternative to barbiturates. However, as more data accumulated, it became clear that repeated use leads to the development of tolerance and dependence. Clinical guidelines emphasize the need for the shortest possible course. In a number of clinical guidelines it is indicated that continuous use should not exceed several consecutive days (a Canadian recommendation-up to four days), after which therapy should be reconsidered and the need for further prescription evaluated. Associations have been established with memory impairment, decreased concentration, and an increased risk of falls in elderly patients.
Cardiovascular Risks Identified After the Introduction of Drugs
For a long time, cardiology was based on the logic of suppressing symptoms and correcting electrophysiological parameters. If a drug eliminated an arrhythmia or normalized a measurement, this was perceived as an improvement in prognosis.
Class IC antiarrhythmic drugs were widely used to suppress extrasystoles and other rhythm disturbances. They effectively reduced the number of arrhythmias on ECG, which was considered a therapeutic success.
Later, large clinical studies showed that in patients after myocardial infarction the use of these drugs was associated with increased overall mortality. Suppression of arrhythmia did not lead to a reduction in the risk of sudden death, and in some groups the risk increased.
Selective nonsteroidal anti-inflammatory drugs of the new generation, such as rofecoxib, were initially positioned as safer for the gastrointestinal tract. As data accumulated, increased risks of myocardial infarction and stroke during long-term use were identified, which led to the withdrawal of the drug from the market.
Fenfluramine, used for weight reduction, also demonstrated delayed cardiovascular complications. Heart valve damage and cases of pulmonary hypertension were identified, after which its use was discontinued.
At present, the evaluation of cardiological drugs and agents with potential cardiovascular effects includes mandatory analysis of clinical outcomes-mortality, myocardial infarction, and stroke-rather than only changes in symptoms or electrocardiographic parameters.
Psychiatry and Invasive Methods
Twentieth-century psychiatry actively sought biological solutions for severe mental disorders. With limited understanding of the mechanisms of schizophrenia and severe depression, methods were introduced quickly, often before the appearance of a strict evidence base.
Lobotomy (prefrontal leukotomy) was a surgical destruction of connections in the frontal lobes of the brain-an area associated with behavior, emotions, and decision-making. It was assumed that the intervention would reduce the severity of psychotic symptoms and severe behavioral disturbances. The method was widely used in the mid-twentieth century, then was recognized as excessively traumatic and discontinued.
Insulin comas were used in the treatment of psychoses. Patients were deliberately placed into a state of severe hypoglycemia with loss of consciousness. The method was associated with risks of seizures, brain damage, and fatal complications and was discontinued after the appearance of pharmacological alternatives.
Electroconvulsive therapy (ECT) was used for severe depression and psychoses. Early procedures were performed without anesthesia or muscle relaxation, which was associated with injuries. In modern practice ECT is performed under anesthesia and used under strict indications.
With the introduction of antipsychotics, a transition to pharmacological treatment began. First-generation antipsychotics (typical) effectively reduced the severity of psychotic symptoms, but caused extrapyramidal disorders-movement disorders associated with effects on the centers controlling movement. These included muscle rigidity, tremor, spasms, and pronounced motor restlessness. With long-term use, some patients developed tardive dyskinesia-involuntary repetitive movements of the face and tongue that could persist even after the drug was discontinued.
At present, second-generation (atypical) antipsychotics are predominantly used in clinical practice. They less frequently cause pronounced extrapyramidal symptoms, but are associated with a different profile of side effects, including weight gain and disturbances of carbohydrate and lipid metabolism.
Thus, invasive methods have left practice, while pharmacological treatment has evolved with the changing understanding of the risks of different generations of drugs.
Pediatrics and Overestimated Safety
For a long time, pediatric medicine developed by analogy with adult medicine. Many drugs were prescribed to children according to a simplified principle: a “reduced adult dose adjusted for body weight.” Age-related features of metabolism, the activity of liver enzymes, the rate of elimination, and the sensitivity of the central nervous system were not fully taken into account.
Combination cold remedies were widely used for viral infections in children. They contained antihistamines, vasoconstrictors, and cough suppressants. Later it became clear that in uncomplicated acute respiratory viral infections their clinical benefit is limited, while the risk of side effects-respiratory depression, tachycardia, rhythm disturbances, and overdoses when several drugs are combined-is real.
Opiate-containing preparations were historically used in children much more widely than today. At the same time, in early childhood the respiratory center and the liver systems responsible for drug detoxification are not yet fully developed, which increases the risk of respiratory depression, especially with imprecise dosing.
Codeine-containing syrups remained in practice for much longer. Codeine is a substance that is converted in the body into morphine-this is what produces its analgesic and antitussive effect. However, the rate of this conversion varies among individuals. In some children morphine can be produced faster and in larger amounts than expected. This can lead to dangerously high concentrations of the active substance in the blood and respiratory depression. The identification of such cases became the basis for reconsidering the use of codeine in pediatrics. At present, its use in children under 12 years of age is not permitted in most recommendations.
Antibiotics also required a differentiated approach. It became known that certain classes have age restrictions because of specific risks: tetracyclines can disrupt the formation of teeth and bone tissue in children, fluoroquinolones can potentially affect cartilage tissue, and chloramphenicol in newborns can cause a severe toxic reaction due to the immaturity of enzyme systems.
Modern pediatric practice is based on age-specific standards, minimum acceptable thresholds for use, and strict dosage calculations taking into account body weight and age. The pediatric population is considered a separate category with distinct requirements for evidence of drug safety and effectiveness.
Hormones and Expanded Indications
Hormonal therapy was often perceived as a tool for correcting “deficiency” or age-related changes. In certain periods its use extended beyond the original indications.
Early oral contraceptives contained higher doses of estrogens compared with modern preparations. Over time, increased risks of thrombosis and cardiovascular complications were identified, especially in women above a certain age, smokers, or those with additional risk factors.
Hormone replacement therapy (HRT) during menopause was for a long time positioned not only as a means of controlling symptoms but also as a method of preventing cardiovascular diseases and maintaining overall health. The expansion of indications occurred faster than the accumulation of long-term data.
After the publication of large studies in the early 2000s, increased risks of thrombosis, stroke, and certain forms of cancer were identified with certain treatment regimens and durations of therapy. This led to a sharp decrease in the frequency of HRT prescriptions and a revision of recommendations.
In subsequent years, the analysis of data became more differentiated. It was shown that the risk significantly depends on the age at which therapy begins, the time since the onset of menopause, the presence of risk factors, and the form of the preparation. When prescribed in the early postmenopausal period and in the absence of contraindications, the safety profile differs from that observed in older women or with long-term use.
In modern practice, hormonal therapy is considered primarily as a treatment for significant menopausal symptoms rather than as universal prevention. Prescription is carried out taking into account individual risk, the type of preparation, dosage, and duration of therapy.
Use of Antibiotics
With the introduction of antibiotics, medicine underwent a revolution. The rapid cure of bacterial infections created the impression of the universality of this class of drugs.
Antibiotics began to be prescribed for conditions where their effectiveness is absent, primarily for viral infections. This practice developed under the influence of clinical habit, patient expectations, and the desire to reduce the risk of complications “just in case.”
Over time it became clear that excessive and unjustified use contributes to the formation of resistant strains of bacteria. Antibiotic resistance has become a systemic problem in healthcare.
In modern practice, drugs with a narrow spectrum and a broad spectrum of action are distinguished. Broad-spectrum antibiotics affect a large number of different bacteria and are used mainly in severe conditions or before the pathogen is clarified. Their use is considered justified only when there are indications, since broad action increases the selection of resistant bacterial forms.
The approach to therapy has become more strict: selection of the minimally necessary spectrum of action, adjustment of treatment after receiving test results, limitation of the duration of the course, and refusal to prescribe in viral or fungal infections.
In addition, the influence of antibiotics on the microbiome is taken into account, especially with repeated courses of treatment.
Dentistry and Local Toxicity
In dentistry, materials and methods were used for a long time that are now evaluated differently from the perspective of safety.
Formaldehyde-containing devitalizing pastes were used for the chemical destruction of pulp tissue during root canal treatment. Resorcinol-formalin mixtures and paraformaldehyde preparations provided a pronounced antiseptic effect and technical simplicity of the procedure. Over time it became clear that such substances have high local cytotoxicity, can damage surrounding tissues, and complicate repeated endodontic treatment.
Arsenic-containing pastes were used to “kill” the nerve before canal treatment. They indeed caused necrosis of the pulp, but with inaccurate application they could damage surrounding tissues and cause complications. Their use was gradually discontinued.
Mercury compounds were also present in dental practice. In addition to amalgam fillings, antiseptic preparations and pastes with added mercury-containing components were used. In addition, in earlier periods some tooth powders and toothpastes contained mercury compounds as bactericidal substances. Understanding of the cumulative toxic effect was limited at that time.
Amalgam, containing metallic mercury in a bound state, was considered a standard filling material for decades due to its strength and durability. The question of chronic low-dose exposure is still discussed. In a number of countries its use has been discontinued or reduced for environmental reasons and because of the availability of more modern alternative materials.
In modern dentistry, priority is given to mechanical treatment, controlled antisepsis, and biocompatible filling materials.
Why It Is Important to Know This
As you can see, history shows that medical approaches are regularly revised as data accumulate. The reason for these changes is most often a larger body of knowledge, the emergence of new tools for risk assessment, and the availability of longer observations by the time a method is introduced.
In many of the examples described, the drug or procedure did indeed produce a clear effect: symptoms were reduced, infections were suppressed, and laboratory indicators changed. A rapid clinical result created the impression of sufficient safety.
The absence of a significant portion of side effects at an early stage is often associated with a short history of use. Rare complications, cumulative toxicity, and long-term consequences become apparent only with prolonged observation and widespread use.
Even now, before a drug enters the market, clinical trials are conducted on a limited group of patients and under controlled conditions. This allows the assessment of effectiveness and the identification of common adverse reactions, but does not always reveal rare complications, interactions during long-term use, or consequences in special groups of patients.
After widespread use begins, the volume of observations increases many times over, which makes it possible to identify effects that could not be detected during limited clinical trials. There is also the issue of conflict of interest - studies are conducted by the same companies that plan to profit from the introduction of the drug.
Accelerated Drug Approval
In most cases, new drugs undergo a lengthy evaluation process that includes laboratory studies, animal testing, multiple phases of clinical trials, and post-marketing surveillance. While this process cannot identify every possible risk, it provides substantially more information about a drug’s safety and effectiveness before widespread use begins.
However, there are situations in which regulators choose to accelerate this process. During epidemics, pandemics, public health emergencies, periods of high mortality, or when no effective treatments are available, drugs may be authorized on the basis of a more limited body of evidence than would normally be required for standard approval.
The rationale is straightforward: to provide patients with potentially beneficial treatments as quickly as possible. At the same time, however, the amount of information available at the time of approval is reduced. Follow-up periods are shorter, fewer patients are studied, and rare adverse effects, delayed complications, drug interactions, or risks affecting specific populations may be more difficult to detect.
For this reason, the later discovery of additional safety concerns in drugs approved through accelerated pathways should not be surprising. If a significant portion of the standard evaluation process has been shortened or deferred until after authorization, the likelihood that initial conclusions will later be revised increases accordingly.
It is important to understand that the absence of identified risks at an early stage does not necessarily mean that such risks do not exist. Sometimes it simply means that there is not yet enough data to detect them. The smaller the evidence base and the shorter the observation period at the time of approval, the greater the likelihood that a drug’s safety profile will continue to evolve as experience accumulates across millions of patients.
This does not necessarily mean that accelerated authorization was the wrong decision. In the face of a serious threat, the potential benefits may outweigh the uncertainty. However, accelerated approval and a high degree of confidence in long-term safety are not the same thing. The less information available when a drug enters the market, the greater the chance that important safety findings will emerge only after widespread use has begun.
Not Only Medications
The revision of views has affected not only medications but also everyday consumer products. In these cases, marketing and commercial rhetoric played a significant role in shaping public perceptions of benefit before long-term data appeared.
Cigarettes were actively advertised in the mid-twentieth century. They were associated with relaxation, social success, and sometimes even with “benefits for breathing.” Advertising messages were built on an image of modernity and safety. Later, accumulated epidemiological data established a link between smoking and lung cancer, cardiovascular diseases, and chronic lung diseases, which led to strict advertising restrictions and warning labels on packaging.
Alcohol for a long time was considered an acceptable element of the diet, and interpretations of certain studies contributed to the spread of the idea of moderate “benefit.” As data were clarified, the emphasis shifted to dose-dependent risks and the association with oncological and metabolic disorders.
Margarine was promoted as a healthier alternative to animal fats. Only later was the role of industrial trans fats in increasing cardiovascular risk recognized, which led to changes in production technologies and regulatory restrictions.
Sugar for many years was not considered a factor in chronic metabolic diseases. Its influence on the development of insulin resistance, obesity, and liver disease became the subject of active discussion as clinical and population data accumulated.
In all of these cases, the initial information environment was shaped by commercial interests and a limited amount of long-term observation. As evidence accumulated, advertising claims were revised, and some of them were prohibited because of their misleading nature.
Conclusion
The history of medicine shows that our understanding of safety and effectiveness evolves as more data become available. Many treatments, drugs, and medical practices once considered safe and beneficial have later been re-evaluated.
It is important to understand that the absence of evidence of harm is not the same as evidence of safety. Sometimes it simply means that insufficient data are available. Even among medical practices that have undergone the standard research process, a substantial proportion are later found to be less effective than originally believed, ineffective, or associated with previously unrecognized risks.
For this reason, the smaller the evidence base, the shorter the observation period, and the faster a drug or intervention reaches widespread use, the greater the likelihood that important information will emerge only after large numbers of people have been exposed to it.
The revision of previous conclusions should not be viewed as a failure of science. Rather, it reflects science’s ability to refine its understanding as new evidence accumulates and longer-term observations become available.