A criticism of measures such as the IBIM is that they rely on self-report and do not record objective, multiple measures of behaviour [19]. Moreover, the prediction of actual behaviour from

the TPB is typically lower than the prediction of intention [33]. Thus, whilst previous research has found a strong association between antenatal ratings of the likelihood of immunising and the actual decision [34], access to children’s immunisation records would be needed to meet the behavioural criterion of the TPB. A related point is that the study was cross-sectional. A prospective GDC-0068 price longitudinal study could include test-retest reliability and would, ideally, measure clinic attendance. E7080 It is likely that parents interested in immunisation were more likely to respond to the invitation to complete our questionnaire. This interest could be due either to strong concerns about injections or to a strong belief that all children should be immunised, or for other reasons. Whilst it is therefore impossible to rule out selection bias, representatives of both extremes were included in our sample and many held more neutral beliefs. Although 27.6% of the questionnaires were removed prior to the main analysis (because some items were missed), excluded parents were similar

to participating parents in terms of sociodemographic characteristics. This indicates that, once parents had made the decision to take part, the completeness of their response was not influenced by issues such as educational level or ethnicity

(see Section 3.2). In addition, it was primarily the views of mothers that were measured, even though parents were encouraged by childcare staff to take Thymidine kinase a copy of the IBIM for their partner. It is possible, therefore, that it is mothers who take a greater interest in immunisation. However, this gender bias may also have resulted from recruitment through child groups as it was, in most cases, the mother who attended with their child or who collected their child at the end of the day when questionnaire packs were handed out. To improve its predictability, the IBIM could be tested with a broader sample of the population including fathers and those who do not use childcare facilities. Indeed, the finding that there was an unmediated effect of number of children on parents’ intentions to immunise with dTaP/IPV provides further evidence for the role of sociodemographic factors. It would also be interesting to see whether the measure could be applied to other vaccinations in the childhood immunisation programme. Since the IBIM was based on the qualitative interviews with parents of preschoolers [4] and parents of young infants [3], it may be possible to apply a revised version to the prediction of parents’ intentions to attend for primary doses and to compare the results with those described here.