Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Understanding risk


Understanding risk is hard enough even for those of us who are numerate, literate, and deal with these issues on a daily basis. Rafts of issues complicate informing patients about the risk of side effects of their medicines, predominantly that of how to present information about benefit and harm. Making a sensible, informed, choice is not easy. A study from Leeds shows how presentation of adverse event information can have huge effects on perceived risk by patients [1].


The setting was 120 adult patients attending a cardiac rehabilitation clinic after bypass surgery or myocardial infarction, and taking simvastatin or atorvastatin. The only exclusions were inability to read or where first language was not English.

There were two variables. Two different adverse events of statins were chosen, constipation and pancreatitis, and the adverse event information was presented using words or numbers (Presentation Box below), using EU guidelines (Table 1). Half of the participants received information about constipation, and the other half information about pancreatitis (atorvastatin users only). Within each group, half received the information using words, and half using numbers. Allocation was random.



.........statin is associated with some side effects. It can cause constipation. This is a common side effect of the medicine.


.........statin is associated with some side effects. It can cause pancreatitis. This is a rare side effect of the medicine.


.........statin is associated with some side effects. It can cause constipation. This side effect occurs in 2.5% (that is, 4 in 100) people who take the medicine.


.........statin is associated with some side effects. It can cause pancreatitis. This side effect occurs in 0.04% (that is, 4 in 10,000) people who take the medicine.

Table 1: EU verbal descriptors of harm

EU verbal descriptors of side effect probability
Very common
over 10%
more than 1 in 10
1 in 100 to 1 in 10
1 in 1,000 to 1 in 100
1 in 10,000 to 1 in 1,000
Very rare
less than 0.01%
less than 1 in 10,000

After reading the information, participants were asked to give the percentage probability likelihood of having the adverse event, and were able to refer to the information sheet they had been given. They also completed a sheet with six questions about attitudes and how the adverse event might affect them.


The age range of participants was 35 to 74 years (median 63), and they had been taking a statin for one to 70 months (median six). The majority (56%) had no formal educational qualifications.

For both constipation and pancreatitis, participants overestimated the percentage chance of the adverse event affecting them (Figure 1). The degree of overestimation was far higher using words than using numbers. Using words rather than numbers also increased patient estimates of the event happening to them using a six point scale, with small differences in some other questions, and satisfaction with information presented was somewhat higher with numbers than with words.

Figure 1: What patients thought versus actual risk


Clearly we are a long way from being able to convey information about risk in a standardised way that people will understand correctly. Words look to be less useful than numbers, but numbers themselves are hardly conveying the right message, especially for rare events.

The other variable is the level of education in the population. In this sample only 30% had qualifications gained in the UK at age 16 or so (O level), and only 13% had qualifications gained at 18 years or older (A level or better). We do know, though, that women overestimated the risks and underestimate the effectiveness of hormonal contraceptives, irrespective of educational attainment [2].


  1. P Knapp et al. Comparison of two methods of presenting risk information to patients about the side effects of medicines. Qual Saf Health Care 2004 13: 176-180.
  2. JE Edwards et al. Womens' knowledge and attitudes to contraceptive effectiveness and adverse health effects. British Journal of Family Planning 2000 26: 73-80.

previous or next story