It will therefore be critically important to highlight the need for screening, particularly for unvaccinated women, in materials sent with future screening invitations to these cohorts. Of course, this study measured screening intention almost 10 years before girls were due to be invited, and it is unclear to what extent this will reflect their future behaviour. The findings relating to ethnicity are also concerning, particularly as fewer women from non-white ethnic backgrounds tend to be screened for cervical cancer selleck inhibitor in the UK and elsewhere [6] and [44]. Rates of cervical cancer in women from black and Asian backgrounds have

been found to be higher than for white women in the 65+ age-group [45]. Incidence in women under 65 is currently lower among Asian women but is similar among black and white women, so lower vaccine uptake in black girls is of particular concern. Uptake may be low in non-white ethnic groups due to cultural barriers and parental concerns that vaccination may encourage sexual activity [46]. Studies have suggested the role of social sources of information and discussion (e.g. hearing about the HPV vaccine and discussing it with family or friends) are important for increasing perceived vaccine effectiveness [47] and increasing requests for the

vaccine [48]. This supports previous research showing cues to action (e.g. a recommendation from friends, family or a doctor) are the strongest predictors of vaccine uptake [49]. These factors should be taken into consideration when developing PD-0332991 in vitro health promotion campaigns

(e.g. narrative leaflets) aimed at reducing ethnic inequalities in vaccine uptake. As increasing numbers of countries, the including the UK, move to a two-dose HPV vaccine schedule [50], ethnic inequalities might be reduced. Research in the US has shown that ethnic disparities occur mainly between initiators and completers, with those from non-white ethnic backgrounds being equally likely to initiate but less likely to complete the three dose course [51]. As we had a single response category for ‘1–2’ doses, we were unfortunately unable to explore predictors of receipt of two or more doses in our sample. This study benefited from a large sample size, including girls from a variety of ethnic and socioeconomic backgrounds. Response rates in both waves of data collection were very high at over 98% but we acknowledge that there could be systematic differences between the schools that readily agreed to take part in the study and those that refused or failed to respond to our initial contact. In addition, a significant number of girls were absent at the point of data collection or did not know their vaccine status, which may reduce the generalisability of the findings. Because recruitment was limited to London, and to schools with levels of vaccine coverage within 10% of the national average, the results may not be generalisable to England more widely or to schools where uptake is much higher or lower.