How to talk to patients about treatment-related risks

August 1, 2011

Most clinicians consider communicating information to patients about risk among their most important duties; however, few have time to effectively convey all relevant information.

Key Points

Most clinicians consider communicating information to patients about risk among their most important duties; however, few have the time to effectively convey all relevant information.1 Compounding the problem are the ways in which patients misunderstand or dismiss what they are told by their physicians. Only by understanding the difficulties you and your patients face in communicating and comprehending risk can steps be taken to address this vital component of patient care.

Any effort to improve risk communication should include an understanding of the difference between absolute and relative risks. For example, suppose a new contraceptive has just been released. Among women taking it for a year, the number who become pregnant (compared with another contraceptive) decreases from 20 of every 1,000 women to 10 per 1,000 women. In other words, the risk of pregnancy drops from 2% to 1%, only an absolute improvement of 1%. The assessment is made by incorporating the 1,000-women population into the calculation; this is an expression of absolute risk. A relative-risk expression of this data leaves out the 1,000-women population and compares only the 2 risk figures: taking the new contraceptive decreases the risk of pregnancy by 50% (a decrease from 2% to 1% is in fact a 50% drop). This figure suggests that the new contraceptive is twice as good as the old one. In fact, the new contraceptive reduces the absolute annual rate of pregnancy by 1% compared with the old formulation, so the argument could be made that it is 1% better: not very impressive at all.

Unfortunately, physicians, the media, and other sources frequently communicate risk in relative, not absolute, terms. A study of risk communication in ethnic and mainstream newspapers found that they rarely expressed cancer risks numerically. When they did, only 26.2% of mainstream and 29.5% of ethnic newspaper articles presented risk information in both relative and absolute terms.2 In a review of 3 studies involving relative and absolute risk, researchers found that in all cases, relative risk or benefit statements influenced patients significantly more than did absolute statements or numbers-needed-to-treat data.3