Introduction

Research on attitudes of health professionals has been a relatively marginal activity until recently. This has been because of what Marteau and Johnston (1990) describe as the implicit model of health professional attitudes and beliefs, i.e. that they are knowledge-based and invariant. Health professionals have been seen as having an empirically derived set of shared beliefs. However, a developing literature demonstrates wide variability in health professional attitudes. The importance of this variability is illustrated here in a range of studies which depict the presentation of treatment options for healthcare users, the professional choices made about access to services and the overall outcome of healthcare for patients and professionals. Many of the studies to date have been atheoretical, focusing instead on a description of attitudes themselves or on their associates. Attitudes of health professionals are often inferred rather than being directly assessed. For instance, previous experience is often assessed in relation to current behaviour with attitudes then inferred, as in the finding that patterns of greater hospital referral for childhood gastroenteritis by general practitioners was associated with prior specialist training in infectious diseases settings (McGee & Fitzgerald, 1991). Divergent current behaviour, e.g. higher levels of referral to cardiac surgery for male than female patients (King et al., 1992) are also documented and attitudinal influences inferred (in this case a range of attitudes including those relating to the severity of symptoms as presented by men and women).