of a range of prompts were most strongly associated with consultation following
symptom identification. Results showed that respondents who had greater numbers
of long-term health problems, were older, had a regular doctor or had greater social
support were more likely to consult (rs =.02–.19). However, consulting was most
strongly predicted by advice to consult from a member of their social network (r =.35),
the degree to which the symptom generated disability (r =.31), the perceived
seriousness of the symptom (r =.56) and, especially, the perceived efficacy of care (that
is, believing that medical intervention could alleviate or eradicate the symptoms) (r
=.69). This study was based on the health belief model (see Chapter 7) but the results
also support the theory of planned behaviour emphasizing the role of social norms and
the perceived benefits of acting (that is, positive attitudes towards consulting). The
results also emphasize the role of social support in consulting behaviour and the
importance of perceptions of medical effectiveness.
Anticipated effectiveness of consultation was also found to be an important trigger
in a population study of people with serious breathing difficulties, which compared
people who had and had not consulted. Controlling for smoking status and perceived
relative severity of symptoms, attribution of wheezing to smoking and lower self-
efficacy in relation to explaining breathing difficulties to a doctor differentiated
between those who did and did not consult (Abraham et al., 1999). The importance
of perceived causation was also highlighted by King (1982) who found that perceived
causes of elevated blood pressure predicted whether or not people attended for
screening. Thus causal understanding of symptoms and anticipation of positive and
effective interaction with health care professionals are also key determinants of health
service use.
Symptoms are not always clear so people may struggle to understand what they
mean. For example, Kendrick et al.(1993) found that, for 60 per cent of asthmatic
patients, there was no significant correlation between ratings of severity and
simultaneous peak flow measurements. This 60 per cent were not characterized by
less severe symptoms (as measured objectively by peak flow) or by age or gender. The
researchers concluded that a large proportion of asthmatic patients cannot reliably detect
changes in their lung function. Similarly, Cantillon et al.(1997) found that, for 86 per
cent of patients who believed they could predict changes in their blood pressure, there
was no significant association between patients’ assessments and clinical assessments.
Patients’ confidence in their ability to predict their blood pressure was, however,
associated with higher anxiety. Thus, when symptoms are unclear, emotional responses
to their detection are likely to be crucial to the effect that symptom perception has
on health behaviour and health service usage.
Interpretation of symptoms
Leventhal and colleagues (e.g. Leventhal et al., 1997; Leventhal, Nerenz and Steele,
1984) have identified five broad dimensions within which beliefs about symptoms and
illnesses can be categorized. First, identity, the way a symptom label is related to our
perception of cause and has profound implications for how we respond. For example
‘fatigue’ or ‘stress’ have very different connotations to ‘cancer’. Second, cause, refers
to our understanding of the processes generating symptoms. For example, ‘indigestion’
has very different implications to ‘heart attack’ and believing that symptoms are due
RELATING TO PATIENTS 231