Dramatic reduction in hepatitis B through school-based immunization without a routine infant program in a low endemicity region

While most countries have universal hepatitis B (HB) immunization programs targeting infants, eight of the 10 provinces of Canada initially opted for school-based pre-adolescent or adolescent HB immunization programs [1, 2]. In the province of Quebec (population: ~8.2million; birth cohort ~85,000/year), the publicly-funded HB prevention program commenced with immunization of high-risk individuals in 1983 (including infants in households with known HB carriers), was expanded to include routine pre-natal HBsAg screening in 1989, and further expanded to provide universal grade 4 elementary school-based HB immunization to pre-adolescents 8–10 years of age (yoa) in 1994 [2, 3]. A catch-up HB immunization campaign for high school students was conducted in 1999 [2]. Between 1994 and 2007, students were immunized with a three-dose schedule (0, 1, 6 months) with Engerix-B or Recombivax-HB at pediatric dosage and since 2008 a two-dose schedule of Twinrix-Junior (at 0, 6 months) has been used [4]. This school-based HB immunization program in Quebec has steadily delivered vaccine to ~85 % of students annually and in combination with the catch-up campaign, has provided protection to the vast majority of Quebec students born since 1983 [3].

Evaluation of the Quebec school-based HB immunization program one decade after its implementation showed dramatic effects: reported incidence of acute cases in Quebec decreased by 79 % overall and by 91 % in pre-adolescents and adolescents 10–19 yoa [5]. These trends were similar to those observed in another large Canadian province with an adolescent program [6]. While such early program evaluations showed striking population benefit, it was nevertheless reasoned that a universal infant HB immunization program could prevent residual cases in young children ≤9 yoa, particularly since infants are at highest risk of chronic-HB infection and acute-HB in young children may otherwise be missed because more often asymptomatic [7]. Further, routine three-dose immunization coverage with pediatric pentavalent vaccine (DTaP-Polio-Hib) exceeds 95 % so that switching to the available hexavalent vaccine additionally inclusive of HB could improve coverage and reduce injections related to the school program. Here, we report follow-up analyses using notifiable disease reports to assess potential epidemiologic benefit of a universal infant immunization program after two decades of a pre-adolescent school-based program in Quebec, Canada.

The Quebec provincial registry of notifiable diseases was used to identify confirmed HB cases reported between 1990 and 2013. As per the Canadian national case definition [8], acute HB cases were defined as patients test positive for HBsAg or anti-HBcIgM in the context of compatible clinical history or probable exposure, or an HBsAg-positive newborn of an HBsAg-positive mother. Chronic HB cases were defined as patients with HBsAg or viral DNA detected at an interval of at least 6 months or detection of HBsAg in the absence of anti-HBcIgM and no acute-HB clinical symptoms. Unspecified cases were HBsAg or viral DNA positive individuals for which available information was insufficient to classify them as either acute or chronic. Clinical and demographic information on cases was retrieved from the standard questionnaires used by local public health units to investigate reported HB cases. This surveillance study using denominalized information was legally mandated by the provincial officer of public health of Quebec according to the Public Health Act and research ethics committee approval was not required. Information on epidemiology of hepatitis B in the province of Quebec for the period of 2000–2010 is also available at: http://www.inspq.qc.ca/pdf/publications/1847_Maladies_Vaccination_Declaration_Obligatoire.pdf.

All statistical analyses were performed using the Statistical Analysis System (SAS Institute Inc, Cary, NC, version 9.4). Chi-square tests were used for the assessment of HB trends from 1990 to 2013 and differences were considered significant at p<0.05. The hepatitis B incidence rates and the rates of potentially preventable cases of disease were calculated by dividing the number of new cases by the population by age group throughout each year, expressed as case reports per 100,000 population.

This study using HBV notifiable disease statistics and case reports from a low-endemicity country, reveals the remarkable success of two decades of pre-adolescent immunization coupled with prenatal HBsAg screening in controlling HB not only in adolescents but also in children ≤9 yoa and adults 20–29 yoa. We measured 97 % reduction in acute-HB cases in the population overall since 1990, and an absence of acute cases in children ≤9 yoa since 2010, in adolescents 10–19 yoa since 2007, and in adults 20–29 yoa, formerly with the highest incidence, in 2013. In that context, it is difficult to assert that further program change is likely to bring important epidemiologic benefit.

During the study period, the rate of chronic-HB cases reported also dramatically decreased (by 92 % in children ≤9 yoa and 83 % in those 10–19 yoa). As the vast majority of chronic cases were born in highly-endemic countries, this large reduction is most likely attributable to universal infant immunization programs currently in place elsewhere in most high-endemicity countries, reducing chronic carriage in young immigrants and thereby also the risk of HB transmission during early childhood in Quebec. The potential for prevention of these chronic cases through infant immunization in Quebec would have been limited as most were born in endemic countries and likely infected before arriving in Quebec. Should some immigrant cases have acquired their infection after their arrival in Quebec, as they were up to 19 yoa, they could have been prevented only if an infant program had been in place since the early 1990s. Cases potentially preventable through infant immunization in 2014 are expected to be even lower recognizing no acute cases have been identified in children ≤9 yoa since 2010. Whether such further decrements in disease burden warrant program change ultimately depends upon operational, health care, societal and cost considerations but is less evident based on epidemiologic indicators.

This study has some limitations. It was based on passive surveillance reports of notifiable diseases with imperfect sensitivity likely to underestimate the true incidence of HB, particularly among young children with a greater frequency of mild or asymptomatic acute infection. While diagnostic approaches and clinician testing behaviours in response to patients presenting with acute viral hepatitis are unlikely to have changed substantially over the study period, we cannot exclude some variations in the detection of HB over time. Additionally, HB reporting has been mandatory for physicians and laboratories for several decades [5] and observed trends in acute cases are likely to be robust, reflecting extensive reduction in transmission and risk. While the unspecified case category may include acute cases, this would not change overall conclusions. The number of unspecified cases reported annually in the overall population varied from year-to-year without an evident or statistically significant trend. Unspecified cases were rare in those ≤19 yoa and the upward trend since 2004 was driven mainly by adults 30–39 yoa who were not vaccine-eligible by age. Finally, these results were obtained in a low-endemicity country and should not be extrapolated to moderate- or high-endemicity areas or those without HBsAg prenatal screening. However, as 18 % of the Quebec population is born outside Canada [9], including 54 % of immigrants from Asia, Africa or the Middle East [10], these results may nevertheless be relevant to other jurisdictions with low endemicity but a substantial proportion of immigrants from highly-endemic regions.

In summary, two decades of pre-adolescent school-based immunization coupled with prenatal HBsAg screening has achieved striking reduction in population HB disease burden. Furthermore, the oldest cohorts targeted by catch-up campaigns are now beyond the average childbearing age (30 years) and the age at which the greatest number of births occur (28 years) in Quebec; as such, neo-natal transmission is anticipated to further diminish.