Time for evidence-based medicine

Evidence-based medicine (EBM) was formed in the eighties of the twentieth century. The words “Evidence-Based Medicine” were translated into Russian 10 years ago from the English “Evidence based medicine” (EBM; there are other versions of the translation: evidence-based medicine; evidence-based medicine).

EBM is a type of medical practice characterized by applying only interventions that have been demonstrated to be beneficial in good scientific research to treat an individual patient.

Features suitable

There are two components to the scientific approach to medicine: deductive and empirical. They are both scientific because in different ways they provide generalizable knowledge, that is, knowledge that can be transferred to other situations, to other patients, to other ages and in other countries.

Traditionally, the main thing in medicine was considered to be the deductive component of knowledge - understanding the causes of diseases, the mechanisms of their development, the mechanisms of action of drugs (drugs) and surgical operations. Unfortunately, medicine and science in general have insufficient knowledge about many diseases; some diseases remain mysterious - neither the cause nor the mechanisms of development are known. Sometimes it only seems that the cause of the disease is known. So, before the discovery of the microorganism that causes peptic ulcer, it was treated with soda, sleep, gastric resection and many other methods. Each method of treatment had its own explanations and justifications, its own theory of peptic ulcer disease: nervous, self-digestion of the gastric mucosa, etc. For this reason alone, an intellectually attractive deductive approach cannot be effective, that is, with a certain probability bringing the expected result: if the idea is not true, the conclusions from this idea are usually false. History, of course, knows many cases when correct practical conclusions were drawn on the basis of false theoretical premises, but these are still rare exceptions.

In medicine, at the same time, there has always been another approach, empirical (pragmatic): the method of treatment is better, which helps patients better. It does not matter whether we know how this method works or not [1] . For example, it is not known exactly why a single injection of the hormonal drug dexamethasone to a woman going into preterm labor significantly reduces the likelihood of death of the immature newborn and complications in the mother. But it is known that this intervention works, which means it must be used.

Of course, it is interesting to know why and how effective intervention helps prevent an undesirable outcome. This knowledge is especially useful for scientists developing new drugs based on knowledge of the mechanisms of disease development and knowledge of the action of the drug. This knowledge is also useful for doctors. Sometimes deduction - the conclusion about the correct course of action based on existing knowledge - allows you to successfully treat patients in situations where there is no empirical evidence of the effectiveness of treatment. For example, the effectiveness of many drugs is not tested on children. Accordingly, pediatricians have to rely on ideas about the mechanism of action of the drug and apply it based on knowledge of the effect in adults or not apply it. On the other hand, sometimes it is possible to use a proven, effective drug more effectively if its mechanisms of action are known; we can assume possible side effects of treatment, interactions with other drugs that the patient is taking.

In an ideal world, a professional's knowledge is complete, and everything he does has a certain effectiveness. Unfortunately, in the real world, knowledge about many diseases is incomplete, and available treatments may be untested. Moreover, this situation can persist for a long time. For example, the supposed healing effect of thick chicken broth for “colds” [2] has not been tested experimentally, despite the fact that this remedy has been used since biblical times. Likewise, drinking heavily when you have a “cold,” which has been recommended by doctors for at least 100 years, may not provide relief and may even make you feel worse.

Knowledge of the mechanism of action of a drug suggests that some intervention is useful, but does not prove the usefulness of the drug for the treatment of a specific disease. For example, a significant proportion of patients with damage to the heart muscle from a lack of blood flow (myocardial infarction) die from heart rhythm disturbances. There are drugs that successfully help against arrhythmia, but do not reduce, and even increase, mortality from heart attack. Another example. In older people, bone density often decreases and fractures become more frequent. There are drugs that successfully increase bone density, but do not prevent fractures. There are plenty of examples of how a medicine produces the desired changes in the body, but does not help in treating the disease.

Therefore, to justify the treatment of a particular patient, it is better to have not so much scientific considerations about how the disease develops, but scientific evidence that in such patients such an intervention helps to achieve such results (desired, beneficial for the patient).

The Age of "Enlightened Skepticism"

The distinction between “pre-scientific” and “scientific” theories is not as simple as it seems. It is clear that from the observation made in Ancient Egypt of the sacred ibis bird allegedly inserting its beak into its anus and releasing water into it, it does not follow that rinsing the colon with water is beneficial to health. However, more bizarre constructions continue to live in the modern world. An example is homeopathy. This relatively recently appeared treatment system is based on the assumption that with a decrease in the dose of a substance, its effect on the body does not disappear, but is reversed: a substance that in normal doses causes phenomena similar to illness (swelling of the mucous membranes), in small doses will eliminate such swelling . This assumption contradicts the body of knowledge of scientific medicine, and the creator of homeopathy, Samuel Hahnemann, directly stated that his system was opposed to medicine (the latter he called allopathy). However, homeopathy is thriving in some modern medical settings.

The empirical component is what we call evidence of effectiveness. Regardless of whether there are theories or how correct they are, experience provides knowledge about whether an intervention works or not. Experience should not be idealized. It is necessary to distinguish between experience as a set of repeated actions in known situations and experience as a technologically more or less complex practice of studying the consequences of certain actions (research, synonymous with experiment). If you don’t make a distinction, don’t feel the difference, then bizarre phenomena like the “experimentalism” of the times of Trofim Lysenko [3] arise . Then thousands of methodologically unprepared Soviet people, mostly peasants, were involved in haphazard manipulation of plants with noble goals, but with no chance of obtaining a result that would be interesting to anyone. Since Lysenko, Lepeshinskaya, Boshyan and many other promoters of Stalin’s times did not have sufficient education, they did not understand the criticism of their results from “normal” scientists.

What is needed is scientific evidence, i.e. the results of benign clinical studies. Clinical research in this sense is opposed to laboratory research: they are focused primarily not on finding explanations for the disease, but on testing means of treatment or prevention, the accuracy of diagnostic methods and the accuracy of predicting the course of the disease.

The first mention of proposals to rely on data on disease outcomes in assessing the work of a doctor or the effectiveness of drugs dates back to the 18th century. However, these attempts were a rare exception, and until the end of the 19th century, a method of treatment was evaluated on the basis of reference to authoritative opinion and based on agreement with generally accepted theory. Ignace Semmelweis, who successfully prevented puerperal fever, was not recognized by his contemporaries and died in the yellow house. The reason is simple - he assumed the incredible: the cause of fever (infection) is in the hands of obstetricians, and hands must be washed.

In order to move into the current “evidence-based” phase by the end of the twentieth century, medicine had to go through periods of changing enthusiasm for different theories of disease. The most important stage was the refusal of scientific medicine to search for a universal theory for all diseases. Perhaps the last surge of this ancient aspiration was the so-called. the theory of nervism, which in the USSR the Communist Party imposed under the name of “Pavlovian theory”. This theory claimed to explain all or almost all painful phenomena by processes in the nervous system [4] .

In the 19th century, medicine experienced an important period of disillusionment and “therapeutic nihilism,” as physicians learned of the limited effectiveness and even futility of the most common treatments (bloodletting, enemas, and laxatives). Medicine emerged from this state thanks to basic disease research that offered truly effective interventions, such as sulfonamides to treat infections instead of bloodletting and cupping, and vaccines to prevent infections. Actually, the understanding that in the arsenal of medicine, along with highly effective methods of treatment (antibiotics for pneumonia), ancient remedies with unknown effectiveness (blood-sucking cups on the back for pneumonia) are preserved, led to the emergence of DM.

The “founding fathers” of evidence-based medicine saw its essence in “enlightened skepticism.” Indeed, an “evidence-based” physician should view proposals to use any remedy with skepticism and accept them only when evidence of its effectiveness is presented. An “evidence-based” doctor must also consider existing practice critically. It was thanks to the emergence of benign data proving the uselessness of blood-sucking cups and mustard plasters that they were no longer used for pneumonia [5] .

Medicine has always been a science, but ideas about what “science” is, scientific proof, have changed over time. Relatively late, only in the second half of the twentieth century, an idea was formed about what methods of “scientific experience” (experiment) should be used in order to evaluate the clinical effectiveness of diagnostic and treatment methods. Medicine becomes “evidence-based” only in the sense of basing the doctor’s actions on direct evidence of the correctness of the actions.

Benign evidence

Knowledge about the effect of drugs on the concentration of sodium in the blood or the nature of cell damage in a given disease can be obtained in the laboratory, on animals, or on cell culture. Knowledge about the effect of drugs on the course of the disease can only be obtained in the clinic. Therefore, for a doctor, for medical practice in general, fundamental research is not as important as clinical research. Doctors intuitively know this and read clinical journals, not theoretical (fundamental) journals. True, authoritative doctors, especially those who teach students, like to talk about the mechanisms of action of drugs and the mechanisms of diseases. This makes an impression on others and may look very convincing, but is not proof of the correctness of the actions resulting from the reasoning.

Clinical trials are different. The main difference between studies is not the intervention or disease being studied, but the structure of the study (its architecture, design). Different diseases and different interventions can be studied in studies of the same design.

The simplest study is a description of a case or several cases (case series). Such research usually indicates an interesting rare variant of a disease or combination of diseases, suggests to the doctor possible outcomes, but says almost nothing about how such patients should be managed.

More difficult research is the “description of current practice” type. The doctor not only describes how he treated his patients, but also indicates the usual outcomes, the timing of discharge from the hospital (calculates the average severity of symptoms, the average length of hospitalization, etc.), and the frequency of disability. Such research is also not conclusive. If no one died from acute bronchitis, this does not at all prove that the treatment described is the best. People do not die from bronchitis without treatment. Only in rare cases, when we are talking about an incurable disease, the description of some practice, as a result of which a significant proportion, for example 1/4 of patients, recover, is remarkable. Such a study is enough to consider that patients should be treated this way.

If a doctor describes not just hospital practice but the results of some treatment, such as the results of a new version of incontinence surgery, the study is called a clinical trial. A doctor has created a new method, applied it to a group of patients (tested it), and reports the results. Such a clinical trial is of great interest to the readers of the journal. This is not just work experience. This is data for assessing the effectiveness of the invention.

The trouble is that the invention of highly effective treatments is rare. More often than not, a new, more effective treatment method is only 10-15% better than the old one. What does it mean? This means that the new method reduces the length of hospitalization from 15 days to 12, and the frequency of exacerbations from annually to four in five years.

Obviously, the patient cannot notice such a difference in the frequency of exacerbations, and the doctor cannot notice a decrease in the number of bed days by such an amount, since in practice, hospitalization periods do not change due to holidays or slow laboratory work. That is why special research structures have been created to reliably identify relatively small advantages or disadvantages of new and old treatments.

For treatments, the most robust research design is the controlled trial (CT). Its main difference is that the result of applying the method being studied is compared with the result in the control group. If the comparison groups differ only in how they were treated, then there is good reason to believe that the difference in outcome is due to different treatments. This is reliable evidence of the advantages or disadvantages of a new treatment. DM operates with such evidence.

Of particular importance is a type of clinical trial called a “randomized, controlled, blinded clinical trial.” This option was first implemented only in 1946. Its difference is that to create identical groups of patients, the randomization method is used (from the English random - random). In this study design, patients are randomly assigned to a new treatment group and a comparison group (receiving standard treatment or nothing at all). “Random” does not mean randomly, as one pleases, but in an unpredictable order, for example, the way a tossed coin lands on heads or tails. As a result, the groups, of course, are not exactly the same, but it is important that in all respects they differ only slightly, randomly.

Randomization also facilitates “blinding” (sometimes called “masking” for obvious reasons). “Blinding” is the conduct of a clinical trial in such a way that patients in the compared groups do not know whether they are receiving new or old treatment. This is necessary so that patients take the prescribed treatment with equal diligence and are equally interested in the use of additional remedies, for example, “folk” ones. To ensure that medical personnel treat patients in the comparison groups equally, in clinical trials, which are called “double-blind,” the trial is conducted in such a way that medical workers do not know who they are treating and whose condition they are assessing—patients receiving a new treatment or an old one. This is necessary so that a doctor assessing, for example, changes in an x-ray, is equally critical of it, regardless of how the patient was treated.

This complex design of testing the effectiveness of medical interventions makes trials expensive and complex, but this complexity is the price of obtaining reliable evidence of the superiority of one treatment method over another. Therefore, sometimes with an unacceptable degree of simplification they say that EBM is about treating patients based on the results of double-blind randomized clinical trials. This, of course, is not so, since in the absence of the results of double-blind randomized CTs, the “evidence-based” doctor uses existing data, the results of simpler studies, and even in the presence of the results of randomized CTs, does not ignore the results of simpler studies, but only gives them an appropriate, smaller weight.

Limits of application

The situation with the use of evidence of the effectiveness of interventions is complicated by the fact that medicine is not limited to “science”. It has two terms that reflect its properties, determined by the interaction of people - doctors and patients.

1. Empathic component: features of relationships and actions arising from the understanding of the patient as a friendly human being.

2. Ethical component: restrictions and responsibilities of a doctor arising from ethical standards.

The role of EBM in solving ethical problems and building relationships with the patient is still small. The fact is that the impulses given by a person’s “inner ethical feeling” may be alien to other people, and aspirations are not necessarily rational. Therefore, in modern medical practice, EBM norms are combined with ethical rules, the effectiveness of which has not been experimentally proven. The irrational behavior of patients and doctors is clearly demonstrated, for example, in the case of suffering in which medicine cannot effectively help the patient. Many people, including doctors, feel the need to do something. Following this desire is dangerous: interventions that do not bring benefit are sure to cause harm. Because all interventions are harmful. The only rational course of action in this case is to use only interventions that are useful, and in the absence of such, non-intervention. Naturally, this approach does not exclude the use of symptomatic remedies, palliative operations, and pain therapy. These interventions are effective, although not in treatment, but in reducing the patient’s suffering, at least temporarily. It is usually difficult for people to accept that non-intervention is the best course of action, but this is often the case.

It follows that evidence-based practice is the greatest good a doctor can do for a patient. Therefore, we can say that EBM lies at the foundation of a doctor’s moral choice. As for psychological and social problems, the same scientific principles apply in this area as in medicine.

DM doesn't cancel anything. It only draws the doctor's attention to direct evidence of the usefulness or uselessness of interventions. Already because there is no reliable evidence on the effectiveness of many interventions, for many situations, the doctor will inevitably in such cases have to make decisions based on his knowledge of the mechanisms of disease development and the mechanisms of action of drugs.

Of course, all the components of a medical decision are united by the doctor’s experience, including his knowledge and skills (see diagram). Scientific evidence is only relevant if it is in an area that is acceptable to the patient and appropriate to the circumstances of the particular case. The doctor is required not only to know scientific evidence, but also to master the techniques of examining the patient to identify the main features of a particular case of illness and be able to inform the patient and identify his preferences.

DM urges the doctor to abandon the questions: What is this? How is this explained? Why is this happening? The answers to these questions may be interesting, but they have little relevance to the management of a particular patient.

DM points to the right question - what should be done? More specifically, is intervention A better for state B to achieve outcome C? This method of questioning applies not only to therapeutic interventions (therapeutic, surgical), but also to diagnosis and prevention. Only when searching for an answer to such a question can one rely on the results of scientific research, which prove the correctness of the decision to use or not use an intervention in a particular patient.

New information environment

EBM reflects the new state of medical science. This progress is based on the development of several branches of medicine. This is, first of all, the development of epidemiology as a science not about the spread of infectious diseases, but a science about the patterns of development and spread of all diseases, a science about methods for studying medical reality [6] . It is epidemiology that provides answers to questions about whether a new disease exists and how much, for example, exposure to toxic gas in the workplace can influence the development of the disease. In the clinic, epidemiology provides the research physician with methods for conducting research.

Secondly, this is the development of medical information. Until the end of the twentieth century, a doctor, just like any other specialist, had to study and read a lot in order to keep abreast of new discoveries, technologies, and knowledge. A system of postgraduate (postgraduate) training has been created for doctors and other specialists in all developed countries. In the USSR and now in Russia, it exists in the form of periodic “improvement”. In the world, the main direction of its development is the transition to continuous postgraduate education. It is believed, not without reason, that listening to lectures once every five years is not enough to ensure that a doctor is constantly ready to correctly use more and more new methods of diagnosis, treatment and prevention. In order to be a modern doctor, to possess today’s and not yesterday’s knowledge, the doctor must systematically monitor new scientific achievements. This means keeping an eye on the medical literature. There are a great many medical journals and books published in the world. They are simply impossible to read and comprehend.

Information technologies of the late twentieth century and the concept of EBM solved this problem. Electronic media and the Internet have given physicians the ability to quickly access the latest issues of journals. Moreover, electronic means have made it possible to quickly search for necessary information in databases. For example, thanks to funding from the US government, the MEDLINE database is available to the whole world. It contains titles and abstracts of articles from the best medical journals in the world. Each entry is supplied with a set of keywords. Therefore, if a doctor needs to find publications on methods of endoscopic repair for inguinal hernia, then using these keywords he can easily find everything published in the world, more precisely, in the best journals in the world.

EBM provided additional search tools: from all publications on the issue, it is possible to select only those that are directly related to clinical practice and are evidence-based. For example, select only those articles in which, during a clinical trial, outcomes were compared between patients operated on conventionally and through an endoscope. An analysis of such articles can quickly answer the question: are there any advantages to endoscopic surgery for a given patient?

Another important problem was solved in the twentieth century. This is the problem of multiple studies. As you know, scientific problems are studied by many scientists at the same time, and the more important the problem, the more research is carried out. Accordingly, a situation often arises when a doctor finds not one, but several articles with test results for a clinical question, and these results may not coincide. What to do in such a situation? Previously, the problem was solved by “voting”: three studies “for”, one “against”. But it has always been obvious that studies can be of varying methodological quality. This means that we cannot approach them according to the principle of “one study, one vote.”

Modern methodology for summarizing scientific evidence consists of systematic review and meta-analysis. A systematic review is a review of the literature that: 1) includes all the world's literature on a problem and 2) discards low-quality (non-evidence-based) studies. It became possible to include all the world literature only after the advent of electronic search tools. Before this, doctors could not conduct such a search simply because the collections of journals and books in the libraries available to them were poor, and the doctor did not have enough time to spend several days in library research to solve the problem of an individual patient. In old dissertations and journal articles, expressions like “we did not find ...” in the literature available to us are not uncommon. Today it is no longer possible to write such a thing.

Discarding non-evidence studies has become possible as the quality of research has improved. This has been proven by analyzes of the content of medical literature at the end of the twentieth century. The researcher who prepares the systematic review sets the criteria for inclusion of studies in the review. Those studies that do not meet the criteria are discarded. Conclusions are drawn from benign studies. For example, in the countries of the former USSR, the use of low-energy laser radiation to treat various diseases is widespread. Does it help? There are thousands of published studies of different structures in different diseases. However, if we put aside case reports and descriptions of current practice, as well as trials that were not blinded or randomized, there are very few studies of acceptable quality. For many diseases for which lasers are used, there is no benign evidence at all.

If several benign studies are found, the systematic review writer has the additional task of summarizing the results of these studies to provide the clinician with one final assessment. This assessment cannot be simply qualitative: drugs help. The assessment must be quantitative, that is, indicate the size of the effect. For example, with influenza type A, the antiviral drug rimantadine helps. But how much? If treatment is started on the first day of manifestations of the disease, regular use of the drug leads to a reduction in the disease by one day. In order to obtain such a quantitative assessment by summarizing the results of several trials, statistical (mathematical) procedures called “meta-analysis” are used, that is, analysis beyond the usual analysis of a single study.

Of course, it would be nice for a doctor not to read thousands of articles, but to have access to summaries of the results of individual studies, i.e., systematic reviews. Just 20 years ago, systematic reviews were very rare. 10 years ago, a large group of doctors from different countries created the Cochrane Collaboration ( www.cochrane.org ), an organization named after the English epidemiologist Archie Cochrane, who first called for assessing the effectiveness of all treatment interventions by summarizing (systematic review) the results of all clinical trials. The organization's mission is to create systematic reviews, maintain their relevance (freshness) and provide the world's doctors with access to the results of these reviews. On the organization's website, anyone can find the most reliable information about the effectiveness of individual interventions. All reviews follow the same principle mentioned above - answering the question of whether intervention A for condition B is better than another intervention or no intervention to achieve outcome C.

The idea of ​​generalization turned out to be fruitful. ) was created in the Cochrane model Recently, the Campbell Collaboration ( http://www.campbellcollaboration.org . This organization prepares systematic reviews of the effects of interventions in the social sphere (pedagogy, criminology, etc.).

In other words, EBM as a modern phenomenon is based on the use of intellectual achievements of the second half of the twentieth century. The evidence-based clinician today has excellent reliable sources of information, most notably Cochrane reviews. A new type of reference books has also appeared, in which recommendations are based on an analysis close to the analysis in Cochrane reviews (the reference book “Evidence-Based Medicine”, http://www.mediasphera.ru/books.htm ).

Evidence-based medicine could not have emerged before the 1980s, simply because the means to find better information were not available and doctors did not truly know what practice-relevant information was. Now, with some specialized knowledge of EBM, every doctor can quickly and effectively search for, critically evaluate information and apply it in their work.

Enthusiasts believe, not without reason, that the development of evidence-based medicine will contribute to the progress of healthcare to a much greater extent than large scientific projects proper, such as, for example, the Human Genome Project. Practice so far lives up to these hopes.

---

[1] We use the term “effectiveness” - the equivalent of the English. "efficacy" to indicate the effectiveness of an intervention, such as a treatment, in an experimental setting. This way of measuring and expressing efficiency is otherwise called “ideal efficiency” or “effectiveness under controlled conditions.” In real medical practice, the effectiveness is usually lower for various reasons. This is sometimes called "real efficiency".

[2] The popular name “cold”, based on the assumption that the disease arises from cold, corresponds to various diseases. A hundred years ago, doctors called this set of diseases “catarrh of the upper respiratory tract”, in the mid-twentieth century - “acute respiratory diseases”, then “acute respiratory viral infections”, and now “influenza and influenza-like diseases”. Such uncertainty exists because these diseases, caused by different microorganisms, are similar in their course, and their treatment is mainly reduced to symptomatic, i.e., to alleviating the manifestations of the disease.

[3] Soifer V.N. Power and science: the history of the defeat of genetics in the USSR. Tenafly, NJ: Hermitage, 1989.

[4] Yaroshevsky M. G. How Ivan Pavlov was betrayed. Repressed science. St. Petersburg: Nauka, 1994. pp. 76-82.

[5] Vlasov V.V. Introduction to evidence-based medicine. M: Mediasfera, 2001.

[6] Vlasov V.V. Epidemiology in modern Russia // International. magazine honey. practices. 2001. No. 2 P. 27-31 [ http://www.mediasphera.ru/mjmp/2001/2/r2-01-7.htm ].