Can we trust studies on the risks of medications?
When researching or discussing the health benefits or risks of medications based on medical studies, one almost always comes across results and statistics that need to be questioned.
More information about the key concepts of risk will help to develop judgment skills that help to critically understand health benefits and harms.
An existing risk means only the Possibilitythat something happens
When estimating disease risks, the probability of a person or group of people becoming ill - or not - in a certain period of time is examined. This also involves research into which characteristics or behaviors are associated with the increased or reduced risk. In order to specify the risk more precisely, large groups of people are examined. This is to ensure that coincidences do not distort the results.
AN EXAMPLE:
In a 10-year study with 10.000 elderly patients with high blood pressure, half of the participants took a blood pressure medication and the other half received a placebo. In the placebo group, 500 people - 10% of the study group - died of a Heart attack. In the group that took the blood pressure medication, 300 people suffered a fatal heart attack - 6% of the study group. In addition, two people in the placebo group - 0,04% of the participants - died of liver disease during the study. In the group that received the blood pressure medication, four people - 0,08% of the participants - died.
This example illustrates risks as follows:
• Relative risk –
The relative risk associated with taking the blood pressure medication is expressed as a percentage. It expresses how much lower a 6% risk is compared to a 1% risk.
Compared with taking a placebo, the blood pressure medication reduced the risk of a fatal heart attack by 40%.
In terms of the risk of fatal liver disease, the percentage expresses the relative risk, how much higher a 0,08% risk is compared to a 0,04% risk.
For participants who took the blood pressure medication, the risk of dying from liver disease was 100% higher than for participants in the placebo group.
• Absolute risk –
Here, the risk of taking the blood pressure medication is given as an absolute number. It indicates how much lower a 6% risk is compared to a 1% risk.
For participants who took the blood pressure medication, the risk of a fatal heart attack fell by 4 percentage points. The phrase "a difference of 4 percentage points" is sometimes used to describe this.
In terms of the risk of fatal liver disease, the absolute risk means how much higher a 0,08% risk is compared to a 0,04% risk. That is, compared to the placebo group, the absolute risk of dying from liver disease was increased by 0,04% in the group that received the antihypertensive drug.
Risk means asking questions
It is therefore not always easy, especially for laypeople, to correctly assess the results of studies. In any case, the following questions should be asked - based on the example:
• Is the examined Disease frequent or rare? –
If the risk of a rare disease changes significantly, only a few people are affected. However, if the risk of a common and fatal disease, such as heart disease, changes only slightly, many people are affected.
A calculation example:
If one in 10.000 people gets a rare disease each year and this risk increases by 100%, only two of the 10.000 people are affected. On the other hand, if 1.000 out of 10.000 heart patients suffer a heart attack each year, an increase in risk of just 1% means that 100 additional patients will have a heart attack each year.
• Who is the study aimed at? –
Often only certain groups of people are included in a study and examined. Criteria can be: gender or age, a certain illness or a certain severity of an illness. In order to gain knowledge, the study participants do not have to fit a certain profile in all respects.
• How many study participants were there? –
In order to be able to estimate a risk, you should know approximately how many participants took part in the study. In general, the results of studies with few participants are more likely to have come about by chance.
• How long did the study last?
The duration of a study can also distort risk analyses just as much as other factors. For example, risks that only appear after 20 years may not be as important for an 80-year-old as they are for a 40-year-old.
Or: For example, a study result relates to an experimental drug for Cancer in the advanced stage, which is supposed to double the survival time, "doubling" no longer sounds so impressive when the predicted survival rate is only one or two months.