Prevention, diagnosis, treatment: What is tested in evidence-based medicine?
Doctors are not only responsible for diagnosis and treatments – they also give advice about how to prevent a disease. The following applies to treatments, preventive measures and diagnostic procedures: To find out whether they actually have a benefit, they must be tested in correctly designed studies.
An example of a typical patient visiting his doctor can demonstrate what this means: Mr. Miller is in his late fifties and is slightly overweight. His family doctor has been encouraging him to get more exercise and eat fewer calories for years now. Losing some weight should lower his risk of developing diabetes. Now the doctor thinks that Mr. Miller has developed type 2 diabetes despite all of his efforts to lead a healthier life, so she does several tests to find out. For example, she checks the glucose (sugar) levels in his blood to find out if they are too high. If Mr. Miller really does have diabetes, the next step is to decide which treatment is most suitable for him.
How can you test preventive measures for a medical condition?
In principle, preventive measures should be tested in the same way as treatments are tested – in randomized controlled trials. But this is often difficult. For example, to find out whether getting more exercise and eating low-calorie and low-fat foods can help to prevent diabetes, the following would have to be done:
- A large number of healthy people would have to be willing to be randomly assigned to a group that changes their diet and participates in a sports program for years, or to a control group that is not instructed to do so.
- These people would have to be monitored for decades to notice any health-related differences later on.
Randomized trials on the prevention of medical conditions in healthy people involve a lot of time and effort. Because of this, these kinds of studies often look at smaller groups of people who are already at a greater risk of developing the condition instead. In the case of type 2 diabetes, that means people over the age of 45 who are overweight or already have slightly increased blood sugar levels. Randomized trials for this group of people have already been completed, looking into how lifestyle can affect their diabetes. The results suggest that type 2 diabetes can't be prevented completely by making adjustments to your lifestyle, but it could be delayed at least. So the doctor's advice that Mr. Miller get more exercise and change his diet is based on scientific evidence.
How is the reliability of diagnostic methods assessed?
In order to make good, informed treatment decisions, the medical condition has to be diagnosed correctly. The diagnosis will determine whether treatment is even needed and if so, what the options are. One of the biggest advantages of a diagnosis is its potential to help the patient get the right treatment. So studies have to be done, on the one hand, to see whether the method used for the diagnosis delivers reliable results. And on the other, to find out whether the diagnostic test actually helps people to find the best treatment, thereby improving their health.
Ideally, diagnostic tests should be easy to use, inexpensive and highly reliable. They should detect the medical condition being tested for in everyone who has it (this is also called test sensitivity). And they should correctly determine who doesn't have the condition (test specificity). In other words: Reliable tests shouldn't produce "false positive" results (in our example, saying that Mr. Miller has diabetes, although he doesn’t). There also shouldn't be any "false negative" results either (saying that Mr. Miller doesn't have diabetes, although he does).
But hardly any diagnostic tests are so reliable that they are always right. Most tests produce wrong results every once in a while. So it is good to know how often the test produces false positive results and false negative results, and what the best way to check the test result is.
Fortunately it's easy to find these things out using studies. There is often a standard diagnostic test for detecting a specific medical condition – known as the gold standard. Each time a new test is developed, its results can be compared with those of the gold standard test in a study. That might involve doing both tests in people who have the condition at various stages. The study is "blinded," meaning that the people carrying out the tests don't know what the results of the other test were. Then the outcomes are compared.
How are the benefits of a diagnostic method assessed?
Studies that not only test the accuracy of a diagnostic test, but also check whether it affects the success of treatment, are even more useful. This is the critical point for patients because diagnostic tests are only medically advisable if they lead to better treatment. This is another question that can be answered in randomized controlled trials.
In cancer treatment, for example, a special test can be used to help tell whether the patient would benefit from chemotherapy. This diagnostic test looks for specific changes in the genetic material of the cancerous tissue. Some of these changes can accelerate the growth of cancerous cells. Drugs are now being developed to specifically slow down the growth of cancerous cells with these changes. This type of chemotherapy is called "targeted therapy."
The benefits of such combinations of diagnostic tests and treatments also need to be tested in studies. They would need to compare targeted therapy with conventional chemotherapy – both in patients who had cancer cells with the specific changes and in patients who had cancer cells without the changes.
If it turns out that the targeted therapy was more successful in patients with the changes than in patients without them, it would show that the test to detect these specific cell changes offers advantages. In other words, it is then a good idea to have the test done.
Is this also true for screening tests?
Screening tests are used to find out whether people who do not show any symptoms have a previously undetected medical condition. Here too, the tests are only as good as their reliability – in other words, how often they deliver correct results. More importantly, though, screening tests should improve your chances of getting better or provide other health-related advantages. Studies are needed in order to look into the possible pros and cons of screening tests.
How are treatments studied?
Treatment – be it medication, surgery or another approach – should ideally be tested in a randomized controlled trial. Here participants are randomly assigned to one of two or more groups.
- One group then receives the treatment that is being tested,
- while the other groups get a comparator treatment, no treatment or a dummy treatment (placebo).
It is only on the basis of the results of these kinds of comparisons that we can find out whether the treatment being tested really has any benefits or advantages over the other treatment.
But here, too, the devil is in the details: It is important to check a number of different criteria to tell whether a randomized controlled trial is high-quality. Systematic reviews analyze and summarize all the available reliable studies on a particular medical issue.
It can be more difficult to assess a treatment if it is made up of multiple elements – as could be the case for Mr. Miller. People who have type 2 diabetes are often advised to make a number of changes: In addition to losing weight and doing more exercise, medication or insulin therapy are often recommended too. Many people with type 2 diabetes also have other health problems, such as high blood pressure. So Mr. Miller and his doctor need to consider the recommended standard treatment and discuss whether they think it is suitable for Mr. Miller. To help them decide, they could use evidence-based guidelines or recommendations such as those offered by structured treatment programs (disease management programs).
Bucher H.C. Kritische Bewertung von Studien zu diagnostischen Tests. In: Kunz R, Ollenschläger G, Raspe H, Jonitz G, Donner-Banzhoff N (eds.): Lehrbuch evidenzbasierte Medizin in Klinik und Praxis. Cologne: Deutscher Ärzte-Verlag; 2007.
Greenhalgh T. Einführung in die Evidence-based Medicine: kritische Beurteilung klinischer Studien als Basis einer rationalen Medizin. Bern: Huber; 2003.
Kuhn J, Lampert T, Ziese T. Komplexe Interventionen – komplexe Evaluationen? In: Robert Koch-Institut (RKI). Evaluation komplexer Interventionsprogramme in der Prävention: Evaluation komplexer Interventionsprogramme in der Prävention: Lernende Systeme, lehrreiche Systeme? Berlin: RKI; 2012.
Zeeb H., Donath S.. Epidemiologische Methoden in Evidence-based Public Health. In: Gerhardus A, Breckenkamp J., Razum O., Schmacke N, Wenzel H. Evidence-based Public Health. Bern: Huber; 2010
IQWiG health information is written with the aim of helping
people understand the advantages and disadvantages of the main treatment options and health
Because IQWiG is a German institute, some of the information provided here is specific to the German health care system. The suitability of any of the described options in an individual case can be determined by talking to a doctor. informedhealth.org can provide support for talks with doctors and other medical professionals, but cannot replace them. We do not offer individual consultations.
Our information is based on the results of good-quality studies. It is written by a team of health care professionals, scientists and editors, and reviewed by external experts. You can find a detailed description of how our health information is produced and updated in our methods.