Is it worth the risk? A systematic review of instruments that measure risk propensity for use in the health setting

Abstract

In this era of shared doctor–patient decision-making, eliciting and incorporating patients’ treatment choices is essential to ensure all patients receive the treatment that is right for them. Clinicians and researchers should fully understand the many factors that influence and guide patients in their preferences for treatment. One of these influences is an individual's general risk propensity or willingness to take risks, yet there is little in the literature about methods for measuring risk propensity.

A systematic review was undertaken to identify instruments that measure risk propensity and to appraise their validity and relevance for a clinical setting. Of 3546 articles, 139 were potentially relevant. From these, 14 instruments were identified. Eight measured risk propensity, whereas six measured personality traits associated with risk propensity. Most instruments demonstrated good internal reliability but their appropriateness for patients, particularly older adults, remains unclear. While no instrument was specific to or tested in a clinical setting, instruments that directly measured risk propensity were considered to be the most useful for clinical populations. The further adaptation and validation of these instruments among older adults are important avenues for future research.

Introduction

Eliciting and considering patients’ treatment preferences is a critical component of shared patient–doctor clinical decision-making (Elwyn & Charles, 2001). A growing body of evidence suggests that obtaining and making provision for patient treatment preferences has the potential to enhance patient adherence to such treatment and improve health outcomes (Bowling & Ebrahim, 2001). It is therefore essential that health-care providers fully understand the multi-factorial influences guiding patients in their choice of treatments within a clinical setting. The challenge for both the patient and doctor is to consider these factors in any decision-making process.

Patients with cancer, for example, may face difficult treatment decisions about the role of adjuvant chemotherapy, radiotherapy and different surgical options. The decision-making process involves a plethora of factors. Ultimately patients, with their clinician, have to weigh up various risks, including the benefit and harm of each treatment option. For patients with cancer, this may involve weighing up the chances of improved survival against the risk of treatment adversely affecting their quality of life. How the risk of treatment is communicated to patients and how such risks are interpreted have been the subject of many studies (Lobb, Butow, Girgis, Kenny, & Tattersall, 1999; Fuller, Dudley, & Blacktop, 2001; Edwards, Unigwe, Elwyn, & Hood, 2004), however, the influence of an individual's general risk propensity (willingness to take risks) on their treatment decisions has received little attention.

The small amount of research in this area could, in part, be attributed to the problems associated with defining, and therefore measuring risk propensity as a construct. One view is that it is an unstable trait across situations with decision risk domains including finance, recreational activities, health/safety, social, ethical, career and prestige (Jackson, Hourany, & Vidmar, 1972; MacCrimmon & Wehrung, 1986; Horvath & Zuckerman, 1993; Weber, Blais, & Betz, 2002). In addition to these domain effects on risk taking, Kahneman and Tversky (1979) state that risk taking (therefore utility function) depends on the actual framing of the ‘risky’ situation, whereby individuals will be risk averse when they perceive themselves in a domain of gain, and risk seeking when in a domain of loss.

An alternative conception of risk has been formulated in the risk-value framework. Here risk taking is a function of the decision makers’ risk perception and their willingness to take on these perceived risks for a return (Weber, 2001). Mellers, Schwartz, and Weber (1997) however, state that variation in risk taking is actually attributable to an individual's perceived risk attitude (PRA). PRA is based on the notion that choices between risky situations depend on both an individual's perception of risk and his/her preference for it. Finally, facets of personality, particularly sensation seeking and impulsiveness have also been demonstrated as being driving forces of risk taking (Zuckerman, 1970; Eysenck & Eysenck, 1977; Eysenck, 1978; Zuckerman, Eysenck, & Eysenck, 1978; Zuckerman, 1994).

A number of instruments have been developed incorporating these theories. Still, there is little information available for the health researcher, however, concerning the relative merits or deficiencies of these measures. This review sets out to systematically identify tools that measure risk propensity and to appraise their validity and relevance in a clinical setting.