This document was published more than 2 years ago. The nature of the evidence may have changed.
Well-founded decisions concerning healthcare and social service interventions require a fair, comprehensive assessment of the benefits and potential harms. But the perceptions of professionals, patients and clients concerning various effects may be based on wishful thinking rather than solid evidence.
Risk assessment is not always a rational process, and factors other than actual probabilities may come into play.
Human nature makes it difficult to leave wishful thinking, creed, ideology and other subjective attitudes aside.
According to a recent systematic review  published in JAMA Internal Medicine, doctors rarely estimate benefits and risks correctly. The review was based on 48 studies that included 13,011 doctors and examined treatments, radiology results, screening and other diagnostic methods. A mere 11% of benefits and 13% of harmful effects were estimated accurately. Overestimating benefits and underestimating risks was more common than the other way around.
The same research team published a similar review  of 35 studies and 27,323 patients in 2015. Once again, incorrect estimates, and a tendency to overestimate benefits and underestimate harms, were frequent.
Undoubtedly, the included studies have their weaknesses. Many of them are small, and subjects' estimates of risk were compared to “true” magnitudes which the authors did not validate. However, the risk estimates of both doctors and patients clearly differed from those of scientists, who were generally less optimistic.
Researcher Lennat Sjöberg discusses ways that people perceive and assess risk in a chapter of a monograph  published by the Research Institute of the Stockholm School of Economics.
One of his points is that people have a proclivity for underestimating the risks to which they personally are exposed. They tend to be more objective about what other people face.
Smoking, drinking, diet and other lifestyle issues are typical in this respect. Smokers are clearer about the hazards when they look at other people. People see themselves as less vulnerable than their peers who find themselves in the same situation.
The sense of being in control is a contributing factor. If you believe that you can manage a risk, you may perceive it as smaller. Voluntary exposures may seem less dangerous than hazards that you think have been imposed on you.
Gender, previous experience and certain personality traits also affect risk perceptions to one extent or another. If one of your relatives died of a particular disease, you may overestimate your own chances of meeting the same fate. A doctor with a patient who experiences a serious adverse effect is likely to be more observant of similar medications going forward.
The scope for personal interpretation is greater in the case of uncommon and poorly researched risks. If a situation is complex and ambiguous such that the facts are elusive, the frequency of events which are easy to recall tends to be overestimated. Readily accessible memories assume exaggerated proportions at the expense of statistical probability.
A systematic review in the Cochrane Library  suggests that well-documented decision aids can help patients develop a more realistic view of the benefits and risks associated with various interventions.
The review is based on 105 randomised studies and 31,000 subjects. Decision aids included brochures, videos and online tools, as opposed to the general or minimal information provided by customary care services. All the approaches described the patient’s options and helped them evaluate the conceivable impact on their life. Among the various choices that patients faced were surgical procedures, medical treatment, screening and genetic testing.
The authors of the review concluded that decision support educates patients about the various options they face. They feel as though they are better informed and have a clearer sense of their own personal priorities. The authors also found moderate scientific evidence that patients become more realistic about the benefits and risks associated with various options.
Studies that deal with the way potential risks and probabilities are communicated indicate that various modes of presenting facts may influence how patients respond. While numerical figures are generally considered to be accurately understood, adding a verbal description can facilitate comprehension.
For example, when telling patients that an adverse effect arises 20% of the time, you might add that such a frequency is officially classified by public agencies as “very common”. Saying that “20 patients out of 100” experience the problem may be easier to understand than “20% of patients” experience it.
A doctor who says that the risk of headache is 10% generally means that one out of every ten people experience it. But the patient might think that anyone will have a headache every ten times they take the medication or throughout 10% of the treatment period.
Similarly, a doctor may be more inclined to recommend treatment that is said to reduce risk by 50% than from 2% to 1%.
The opportunity of obtaining benefit is often viewed more favourably than avoiding a risk. Surveys have shown that both doctors and patients look at a 68% survival rate as more desirable than a 32% fatality rate. [RL]
Medical and Social Science & Practice The SBU newsletter presents and disseminates the results of the SBU reports, describes ongoing projects at the agency, informs about assessment projects at sister organisations, and promotes interest in scientific assessments and critical reviews of methods in health care and social services.