Attitudinal and demographic predictors of measles, mumps and rubella (MMR) vaccine acceptance: development and validation of an evidence-based measurement instrument.
Brown KF., Shanley R., Cowley NAL., van Wijgerden J., Toff P., Falconer M., Ramsay M., Hudson MJ., Green J., Vincent CA., Kroll JS., Fraser G., Sevdalis N.
BACKGROUND AND OBJECTIVE: Parents' attitudes toward MMR vaccine and measles, mumps and rubella infections relate to their child's MMR status, therefore improving these attitudes is central to improving current suboptimal MMR uptake. However, no study has yet combined evidence-based, comprehensive and psychometrically validated assessment of these attitudes with reliable objective MMR status data, in order to identify through multivariate analyses the strongest attitudinal predictors of MMR uptake for interventions to target. The present study fills this lacuna by developing and testing a robust evidence-based MMR attitudes measurement instrument. DESIGN: Cross-sectional self-administered postal/telephone questionnaire with objective behavioural outcome. SETTING AND PARTICIPANTS: 535 parents of children aged 5-18 in London and north-west England, UK (response rate 18.1%). Recruitment via Primary Care Trust records, age-stratified purposive sample with suboptimally immunised cases oversampled. MAIN OUTCOME MEASURES: Parents' responses to evidence-based measurement instrument comprising 20 attitude/previous behaviour items (collapsing to 5 scales) and 7 demographic items, and their children's PCT-recorded 5th birthday status for MMR dose 1 (on-time, late or none) and MMR dose 2 (on-time or none). RESULTS: The attitudes measurement instrument was psychometrically robust: content valid, and demonstrating good or acceptable internal consistency (Cronbach's alpha=0.55-0.75 for all scales), test-retest reliability (Pearson's correlation >0.60-0.80, p<0.01 to <0.001 for all scales and 11 individual items), concurrent/construct validity (t-tests for difference between MMR status groups p<0.05 for four scales and thirteen individual items), and predictive/criterion validity (OR=0.66, 95% confidence interval=0.48-0.92 to OR=1.97, 95% CI=1.18-3.31 for three scales and five individual items). Black and minority ethnicity (OR=1.94, 95% CI=1.15-3.30 to OR=4.15, 95% CI=2.40-7.19), positive MMR attitudes (OR=1.63, 95% CI=1.00-2.66 to OR=1.97, 95% CI=1.18-1.31), and positive social attitudes (OR=1.64, 95% CI=1.23-2.40 to OR=1.72, 95% CI=1.13-2.38) independently predicted uptake for both MMR doses. MMR status groups differed most strongly on preference for single measles, mumps and rubella vaccines (6-9% variance in status explained), previous MMR acceptance/rejection (5-9%), and wishing to protect others through vaccinating one's own child (6-8%). CONCLUSIONS: The measurement instrument is robust on multiple validity and reliability dimensions, and is appropriate for use in research and practice as a tool for designing and evaluating interventions. Parents appear to act in line with their attitudes toward MMR vaccine, though attitudes toward measles infection bore little relation to MMR uptake. This study indicates populations and attitudes to be prioritized in MMR uptake improvement interventions.