Objective To determine the prevalence and risk factors of hepatitis A among children in the region of Tlemcen (north-west Algeria).

Methods 252 children (0.75–8 years) were tested for total hepatitis A virus (HAV) antibodies by ELISA (ETI-AB-HAVK PLUS, DiaSorin). A questionnaire concerning socioeconomic, living and hygienic conditions was completed.

Results The overall prevalence of HAV+ was 12.3% (<2 years 7.1%, 2–5 years 18.9% and 6–8 years 23.7%), with no sex difference (p = 0.053). On univariate analysis, the factors associated with HAV+ (vs HAV−) were: child’s age (mean 3.53 ± 2.2 vs 2.4 ± 2.1 years, p<0.01), mother’s age (mean 34.7 ± 6 vs 31.2 ± 6.1 years, p<0.01), mother’s gestity (mean 3.5 ± 1.8 vs 2.7 ± 1.7, p<0.05), number of persons aged <16 years living in the same housing (mean 3.1 ± 1.8 vs 2.3 ± 1.2, p<0.01), education level of the father less than secondary (83.9% vs 62.6%, p<0.05), child history of jaundice (19.3% vs 1.7%, p = 0.01) and child chronic illness (32.3% vs 5.3%, p<0.01). Multivariate analysis (logistic regression) confirmed specific effects of the number of people <16 years living in the same housing (p<0.05), education level of the father (p<0.05), and the presence of a chronic illness (p<0.0001).

Conclusion This study has shown a dramatic decrease of the age-specific HAV+ prevalence, suggesting a gradual shift from high to intermediate endemicity for hepatitis A in Algeria. Mass vaccination, as recommended by the WHO in areas of intermediate endemicity, should then be considered.