Development of a valid and reliable scale to assess parents' beliefs and attitudes about childhood vaccines and their association with vaccination uptake and delay in Ghana.

Background: Parents' attitudes and beliefs in vaccination are important to understand for shaping vaccine acceptance and demand interventions. Little research has focused on developing a validated scale to measure parents' attitudes towards vaccinations in low and middle-income countries; Ghana provided an opportunity develop a caregiver vaccination attitudes scale (CVAS) validated against childhood vaccine compliance.

Methods: We conducted a cluster survey of 373 households with children aged 12-35 months of age from Northern Region, Ghana. Caregivers responded to 22 vaccination behavior and belief survey items and provided the child's vaccination status. In exploratory factor analysis (EFA) to assess CVAS content validity, we used parallel analysis to guide the number of factors to extract and principal axis factor analysis for factor extraction. Reliability of the scale was assessed using McDonald's Omega coefficient. Criterion validity of scale and subscales was assessed against receipt of vaccinations by 12 months of age and vaccination delay, using number of days undervaccinated.

Results: EFA of CVAS responses resulted in removing 11 of 22 survey items due to loadings <0.30 and development of a 5-factor structure with subscales for Vaccine-Preventable Disease (VPD) Awareness, Vaccine Benefits, Past Behavior, Vaccine Efficacy and Safety, and Trust. The 5 factors accounted for 69% of the common variance and omega coefficients were >0.73 for all subscales. Validity analysis indicated that for every unit increase in the parent's scale score, the odds of the child being vaccinated decreased by 0.58 (95% confidence interval [CI]: 0.37, 0.68) and the number of days under-vaccinated increased by 86 (95%CI: 28, 143). The final 3-factor scale included Vaccine Benefits, Past Behavior, and Vaccine Efficacy and Safety.

Discussion: The final CVAS included three factors associated with vaccine compliance in Ghana, although several survey items suggested for use in vaccine acceptance scales were dropped. Replicating this study in several country settings will provide additional evidence to assist in refining a tool for use in routine vaccine acceptance and demand surveillance efforts.