Implications of Conflicting Definitions of Probability to Health Risk Communication: A Case Study of Familial Cancer and Genetic Counselling

Article excerpt


The question of what probability actually is has long been debated in philosophy and statistics. Although the concept of probability is fundamental to many applications in the health sciences, these debates are generally not well known to health professionals. This paper begins with an outline of some of the different interpretations of probability. Examples are provided of how each interpretation manifests in clinical practice. The discipline of genetic counselling (familial cancer) is used to ground the discussion. In the second part of the paper, some of the implications that different interpretations of probability may have in practice are examined. The main purpose of the paper is to draw attention to the fact that there is much contention as to the nature of the concept of probability. In practice, this creates the potential for ambiguity and confusion. This paper constitutes a call for deeper engagement with the ways in which probability and risk are understood in health research and practice.

Aust Health Rev 2007: 31(1): 24-3 3

THE MOVE FROM deterministic to probabilistic reasoning that seemed so radical some decades ago is now fairly well entrenched in health and medicine. It has even been said that this trend has gone too far and that uncertainty has become a valid analytical concept in and of itself.1 In this casual acceptance of uncertainty in both routine and extra-ordinary events, there is often an almost flippant treatment of probability. This paper takes a close look at the concept of probability and the way in which it is used in health science. The domain of genetic counselling for possible cases of familial cancer is used to ground theoretical debate in practical health concerns. However, these arguments have implications beyond this setting and are relevant to all forms of health risk communication and management.

The paper begins with a cursory outline of some of the dominant interpretations of probability. Examples are provided of how each interpretation can be seen to manifest in either the management or communication of risk in familial cancer. In the second part of the paper, some of the practical implications of debates on the philosophy of probability are explored. It is argued that a deeper understanding of the concept of probability is likely to lead to improved estimates of risk, enhanced clarity in risk communication, and better decision making in risk monitoring and management.

Interpretations of probability: theory and examples from familial cancer

Questions as to the nature of probability are not new. The beginning of the formal (mathematical) theory of probability is generally argued to stem from studies of games of chance in the seventeenth century.2 Since then, the philosophical interpretations underlying probability have been an ongoing topic for debate. Accounts of the historical development of probabilistic thinking are extensive3"" and the possible conceptualisations of probability are almost as varied as the number of probability theorists. An important conclusion from historical analyses of probability is that during the formative periods of probabilistic reasoning particular interpretations underlying probability were closely tied to the application in which probabilistic calculations were employed. More recently, however, interpretations have become utilised generally across applications and it has been pointed out that almost every interpretation "ever attempted by a probabilist is alive and well today in some form, however dubious its reputation in the intervening years."7 (p. 271).

In the academic literature there is much contention and argument as to the "correct" use and interpretation of probability. While this paper makes reference to this ongoing debate, it is not the main concern Io argue for or against any particular interpretation of probability. The fact is that, in practice, people draw on multiple constructs of probability. …