The use of the TPB provided theoretical underpinning to the empirical work, identified factors that predicted behaviour (especially intention) and led see more to the collection of respondents’ beliefs underlying the direct constructs (TPB variables) reported. The overall combined response rate of 32% was less than expected and was likely to have been affected by the relatively complex nature of

the questionnaire and its length. However, even with this response rate the data derived from the 927 respondents had sufficient statistical power for all the regression analyses. Hence, the study had sufficient statistical power to achieve its objectives, that is, to examine the theoretical Vorinostat predictors of self-reported behaviour, together with respondent characteristics. The study was conducted in Scotland and few respondents were from ethnic minorities. Furthermore, because of the sampling strategy used, more female than males responded and respondents were also more likely to be older and to be married or living with someone. As such, the results might not be generalisable to individuals who are younger, living alone or from ethnic minorities. The sample was derived from the electoral register but

excluded individuals registered with the Mail Preference Service. While this is likely to have introduced bias into the sample, it was the most inclusive method available for this survey. The additional belief items were included in only half the sample to minimise the impact on the overall response rate. In general, respondents had positive perceptions regarding giving information to MCAs during consultations for NPMs. Previous research has shown that the extent of communication in terms of information exchange between patients and MCAs during consultations is often minimal, and that MCAs perceive that the public fail to realise

their role (the MCA) differs from that of general shop assistants[22] and that patients are reluctant to provide information.[4] Nevertheless, Interleukin-2 receptor family doctors and the NHS rather than MCAs, were identified as likely to have more influence on people’s behaviour, as indicated by the significant difference in these beliefs between those who did and did not give information. Patients’ desire to receive information during counselling for NPMs has been demonstrated previously,[23] as has awareness of the need to provide specific information during consultations.[8] Based on the results of this study, it seems that patients may be more positive about providing information during consultations than MCAs realise and the behaviour of MCAs may actually inhibit rather than facilitate information exchange. Patient demographics, such as age and gender have previously been shown to influence health professional communication behaviour.