Risk is in the eye of the beholder when it comes to communication

Perspective / 18 March 2024

Are you and your patient talking about the same thing when you discuss risk? Most likely not, and understanding this is key to effective risk communication.


A recent European survey mapping the population’s health literacy indicates that in some countries, as much as 40% of the population struggles with evaluating the advantages and disadvantages of different available treatments. Such is the case even though health information is more readily available now than it ever was. Studies also show that potential adverse events presented in patient information leaflets provoke fear and anxiety in a considerable percentage of the population (everything from 9% to 30%, depending on the study). Clearly, patients are struggling to understand and grapple with risk when faced with treatment options.

So, what makes risk communication so challenging? One key point is that risk is communicated with facts and numbers, but is perceived through a lens of individual factors, such as priorities, expectations, values, and bias. Take for instance the statement: “One in ten patients experience high fever for about one week after beginning treatment”. Whether this is considered a statement of high or low risk depends on many contextual factors, not least the expected benefit of the treatment. Because of such variability, when we communicate facts about risk and probabilities, we can never be sure about the way they are understood. Guidelines suggest communicating probabilities in the form of numbers and then putting them into context. An example of good practice would be: “Every year, about two in 7000 women who use oral contraception experience thrombosis. For reference, this happens to four in 7000 women who become pregnant”. All in all, we can say that risk perception is not all about facts, yet the way we communicate facts and numbers plays a vital role in the way the message is perceived by the patient.

However, can healthcare practitioners be sure that what they are communicating is being understood in the way they wish by the patient? Research indicates that health practitioners and patients sometimes understand common words like ‘cause’, ‘effect’, and ‘probability’ differently, which can create further confusion and misunderstanding. What about when it comes to risk? Might health practitioners communicate something that is understood differently by their patients?

 

 

To answer this question, we need to take a step back and ask the fundamental question: does ‘risk’ mean only one thing? Or are there different possible interpretations of the concept of risk? The short answer is yes. Since risk is a measure of probability for an undesired effect, there are as many interpretations of ‘risk’ as there are of ‘probability’. Let us take this step by step and first look at what professionals talk about when they refer to ‘risk’ of a certain event as usually provided by existing scientific evidence for the effect.

When statisticians and epidemiologists talk about risk, they refer to a precise mathematical concept. The risk of an individual’s outcome is predicted by looking at the past frequency of this outcome at the population level. This numerical concept of risk originated in the late 1800s with the rise of the life insurance industry, when companies had to find a systematic approach to predict an individual’s life expectancy. This approach to risk and probability is also called frequentism: the practice of inferring probability for one case by observing how often a certain event happened in previous, similar cases. The patient who is informed about risk is therefore given the counts of how often an event has happened in the study population.

However, there are other approaches to interpreting probability. Let us now switch to the user’s point of view. Patients need to make sense of this information in light of their situation. When told, for instance, “one in 10 who used the medicine experienced fever”, they need to evaluate whether they are more likely to be part of the group who was hurt by the medicine, or the group who was unhurt by it. Patients, in other words, need to evaluate their individual propensities or tendency to be hurt (or not) by the treatment. The propensity interpretation of probability is less known than the frequentist interpretation, nevertheless, it has been discussed by philosophers and scientists in many different contexts. Simply put, propensities can be seen as the explanatory understanding of risk and probability. They take into consideration the context, the treatment, and the patient who is predisposed to a certain event.

One of the aims of the CauseHealth Pharmacovigilance project is to highlight how different interpretations of terms such as 'risk', 'probability', and 'evidence' often remain unspoken, but can still influence the way we understand, communicate, and make decisions about evidence. The CauseHealth handbook for healthcare professionals (free access online resource) describes the propensity interpretation of probability with the example of a dose of morphine. The probability that a patient falls asleep after a certain dose of morphine may be explained by many properties of the drug. First, by the sedative power of morphine, but also by how often the patient may have used morphine or other sedative medicines in the past. The context in which the morphine is taken will also affect this. Combined, these properties give the overall propensity of that particular patient to fall asleep after that specific dose of morphine.

How can this help with risk communication? In our view, health professionals should always be aware that they are communicating frequencies, while patients need to hear propensities. Health professionals should also be aware that frequencies can be a good indicator of individual propensities, but they not always are. Often, we cannot infer propensities directly from frequencies. In those cases, it is important to communicate this source of uncertainty, for example: “One in 10 patients who used medicine X experienced high fever. Unfortunately, we do not know why, so we cannot say whether it will happen to you or not”. Talking openly about uncertainty further aids the health professional’s responsibility for transparency when communicating with their patient.

When the pathophysiology behind the adverse event is known, on the other hand, this should be communicated to help the patient make sense of the numbers. Here is an example: “One in 10 patients using medicine X experience fever. The majority of these patients are those who are susceptible to allergic reactions, as the medicine prompts the immune system, thus causing a fever response in these patients”. This way, the patient is provided not only a number but also a pathophysiological explanation for this number. This helps the patient evaluate their individual propensity to the adverse event.

To my knowledge, there is not yet enough evidence as to the benefits of providing more comprehensive explanations behind the numbers when it comes to communicating risk. Admittedly, this strategy will not fit every situation. But we suspect that in the reality of clinical practice, this is often done to some extent, especially when the stakes are high. If this is true, then clinicians would benefit from more research on how to best communicate evidence of individual propensities. This can be a complex endeavour since to communicate complex evidence, it is necessary to have a clear overview of the facts. As a part of the CauseHealth project, we have created a system for systematically evaluating evidence of individual propensities of a patient, a medicine, and a contextual situation to a certain adverse event. For the moment, this is intended as a guide for evaluating evidence of causal propensities, and not yet as a tool for communication.

However, we think that there is great potential in this system for building confidence in health professionals to better understand and communicate probability and risk. This being said, we are interested in hearing the thoughts and experiences of health professionals and patients in applying the propensity theory of probability to risk communication in their own practice. Perhaps by learning from one another, we may come to a better understanding of risk and become better risk communicators as a result.

Read more:

OR Herber, et al. “Patient information leaflets: informing or frightening? A focus group study exploring patients’ emotional reactions and subsequent behavior towards package leaflets of commonly prescribed medications in family practices”, BMC Family Pra

MA Jeraisy, et al. “Utility of patient information leaflet and perceived impact of its use on medication adherence”, BMC Public Health, 2023.

A Horwitz, L Reuther, SE Andersen, ”Patienters vurdering af medicinpakningernes indlægssedler”, Ugeskrift for Læger, 2009.

Belfolkningens helsekompetanse del 1 – English summary on pgs 12-15.

The CauseHealth handbook

The Causehealth website

Elena Rocca @Cause_Health
Associate Professor, Faculty of Health Sciences, Oslo Metropolitan University, Norway

You may also like


How can the ‘Black Swan Theory’ help us Prevent Adverse Drug Reactions?

How does an economics theory apply to drug safety, and how can we use it to better manage rare, unexpected, and serious adverse drug reactions?

Perspective / 23 September 2024

Risk Communications in Healthcare – Bridging the Gap

Effective risk communication is a mainstay of successful pharmacovigilance, yet examples of poor risk mitigation abound. Where are we failing, and what can be done about it?

Perspective / 20 November 2023

What’s trending in pharmacovigilance? New podcast episode

In this festive edition of Drug Safety Matters, we wrap up the year that was and look ahead to the year that will be – with a little help from one of our friends.

Perspective / 21 December 2023