Predicting and explaining uptake of cervical screening : the role of social cognitions
The aim of the prospective studies in this thesis was to identify factors relating to intentions to have a smear test as part of the national screening programme in the UK and actual uptake of this test. A second aim was to assess the sufficiency and efficacy of social cognition models in predicting these intentions and behaviour. Nearly 2,000 women died from cervical cancer in 1992 in the UK and a further 23,000 were affected by the disease, indicating that cervical cancer is a significant public health problem. It has been shown that screening can prevent death from this cancer and also it has been observed that the majority of women dying from cervical cancer have never been screened. Exploring how women behave in relation to cervical screening is therefore essential for identifying factors which may be used to ultimately increase uptake. Study 1 was based largely on the Health Belief Model (Becker, 1974). One hundred and eighty-three women (28 per cent response rate) completed a questionnaire assessing demographic characteristics; knowledge of, and attitudes towards, cervical cancer and cervical screening; and intention to attend for screening in the future. One hundred and fifty of these women subsequently attended for a smear test and 33 did not. It was found that intentions were predicted by perceptions of the costs and benefits of screening and whether women had ever had a smear test. The best predictors of uptake of screening were an intention to attend for screening and the woman having a current sexual partner. The second study attempted to extend and replicate the results of study 1. This study repeated the measures used to assess the Health Belief Model and also included measures based on the Theory of Planned Behaviour (Ajzen, 1991). One hundred and forty-two women participated in the study (59 per cent response rate), completing questionnaires before being invited to attend for screening. Seventy-two women subsequently had a smear test. Consistent with study 1, intentions were predicted by ever having had a smear test and having a positive overall attitude to attendance for screening. Years of education and frequency of contact with a GP also significantly added to the prediction of intentions to be screened. No individual variable contributed to the variance explained in behaviour. Beliefs derived from the Theory of Planned Behaviour accounted for considerably more variance in intentions than those derived from the Health Belief Model. The results of the studies indicate that the models used do not provide a complete explanation of influences on women's behaviour, accounting for relatively little of the variance in uptake of screening. This calls into question their efficacy in this behavioural domain. Suggestions for improvements to the models are given along with practical implications for increasing uptake of cervical screening.