RESUMO
BACKGROUND: This study aimed to develop a tool and assess its validity to measure childhood vaccine related attitudes among parents in a low-income setting. METHODS: We developed a vaccine attitudes scale (VAS) composed of 14 Likert items each with 5 responses ranging from strongly agree to strongly disagree (sum of scores range 14-70). The tool was administered to 901 parents with children 4-12 months of age during a vaccine coverage survey in Sindh, Pakistan. We performed factor analysis with eigenvalues >0.3 for sufficient factor loading and calculated Cronbach alpha for reliability. RESULTS: The mean ± SD score on VAS was 48 ± 3 and Cronbach alpha was 0.61. Factor analysis identified that VAS measured 2 different domains related to the childhood vaccine related attitudes; (1) 10 items related to vaccine perceptions and concerns (mean 40 ± 5.5; Cronbach alpha 0.95) and (2) 4 items related to vaccine preventable disease salience and community benefit (mean 7 ± 3; Cronbach alpha 0.97). The odds of children being unimmunized was 5 times higher among parents who scored high (13-20) as compared with low (<13) on the subscale related to disease salience and community benefit (odds ratio 5.2; 95% CI: 3.6-7.6). The odds of children being unimmunized was 1.5 times higher among parents scoring high (40-50) as compared with low (<40) on subscale related to vaccine perception/concerns (odds ratio 1.5; 95% CI: 1.1-2.2). CONCLUSION: The 4-item scale assessing parental attitudes toward vaccine preventable disease salience and community benefit is sufficiently reliable and can predict vaccine acceptance among parents in low income setting.