Herd immunity from meningococcal serogroup C conjugate vaccination in England: database analysis

In November 1999, the United Kingdom introduced routine meningococcal serogroup C conjugate vaccination for infants. The vaccine was also offered to everyone aged under 18 years in a phased catch-up programme.¹ The first to be vaccinated were adolescents, and the entire programme was completed by the end of 2000. On the basis of direct protection provided by the vaccine,¹,² this catch-up programme was likely to be cost effective.³

Maiden et al described a 67% reduction (from 0.45% to 0.15%) in the prevalence of nasopharyngeal carriage of serogroup C meningococci in adolescents before and after the vaccination programme.⁴ A fall in meningococcal carriage would be expected to reduce exposure among unvaccinated children and therefore to enhance the effectiveness of meningococcal conjugate vaccine. We present rates of disease in vaccinated and unvaccinated children to provide the first evidence of an indirect effect from meningococcal conjugate vaccine.

Methods and results

Since December 1999 we have investigated the vaccination history of all cases of serogroup C disease confirmed by the meningococcal reference unit of the Public Health Laboratory Service in age groups targeted for immunisation.¹ We collected data on vaccination coverage from immunisation coordinators and departments of child health in England.⁵ Between 1 July 2001 and 30 June 2002, we identified a total of 37 cases in the cohorts targeted for catch-up vaccination, eight (22%) in vaccinated children and 29 (78%) in unvaccinated children. We compared cases in unvaccinated children from each age group in the 2001-2 cohort with those in the same age groups for the period from 1 July 1998 to 30 June 1999. The denominator was mid-1999 population estimates from the Office for National Statistics for the age group, adjusted for the proportion of each cohort vaccinated.

Overall, in the age groups targeted for catch-up vaccination, a reduction of 67% (95% confidence interval 52% to 77%) in the attack rate occurred, with a range of 48% to 80% across the age groups (table). A smaller fall occurred in adults not eligible for vaccination (aged ⩾25 years), for whom the incidence declined by 35% (20% to 49%) from 0.53 (193/36 315 726) to 0.34 (123/36 315 726) per 100 000.

Because of possible underestimation of coverage⁵ we recalculated the attack rates, assuming that coverage was 10% higher in adolescents and 5% higher in other children (based on our experience with other vaccines). The resulting overall estimate was of a 52% reduction (95% confidence interval 30% to 77%) in unvaccinated children; this included a 52% reduction (10% to 74%) in adolescents (where coverage was likely to be least accurate).

We estimated vaccine efficacy (the percentage reduction in attack rate in vaccinated compared with unvaccinated children) by using the same data sources and applying methods described previously.² From July 2001 to 30 June 2002 the attack rate in vaccinated children was 0.09/100 000 (8/9 119 078) corresponding to an overall vaccine efficacy of 94% (86% to 97%). Using the adjusted coverage we estimated an efficacy of 96% (91% to 98%).

Comment

These data show that, in addition to direct protection, meningococcal conjugate vaccine contributes to the control of meningococcal infection by indirect protection, by reducing the attack rate in the unvaccinated population by 67%. These observations may be explained by a natural decline in the incidence of serogroup C disease, although this is unlikely. The reduction in the attack rate is consistent with a reduction in serogroup C carriage rates⁴ and goes against the trends in serogroup C disease before 2000¹ and in serogroup B disease. As adolescents are the only group in which carriage rates have been studied,⁴ these data provide more robust evidence of herd immunity across the whole population. Countries considering introducing meningococcal conjugate vaccine may wish to take account of this indirect protection in the economic evaluation of vaccine policy.

Table.

Attack rate of confirmed meningococcal serogroup C infection in unvaccinated children before and after the launch of the vaccination campaign

Cohort	July 1998-June 1999					July 2001-June 2002
	Date of birth	Cases	Population	Attack rate per 100 000 (95% CI)		Date of birth	Cases	Estimated coverage (%)	Estimated population	Attack rate per 100 000 (95% CI)		% reduction (95% CI)*
Adolescent	1 Sep 1978 to 31 Aug 1981	96	1 818 034	5.28 (4.2 to 6.3)		1 Sep 1981 to 31 Aug 1984	11	66	614 110	1.79 (0.7 to 2.8)		66 (37 to 82)
School years 7-10	1 Sep 1981 to 31 Aug 1985	141	2 546 938	5.54 (4.6 to 6.4)		1 Sep 1984 to 31 Aug 1988	4	86	359 118	1.11 (0.02 to 2.2)		80 (46 to 93)
School years 1-6	1 Sep 1985 to 31 Aug 1991	76	3 911 606	1.94 (1.5 to 2.4)		1 Sep 1988 to 31 Aug 1994	5	87	498 068	1.00 (0.1 to 1.9)		48 (−28 to 79)
Preschool	1 Sep 1991 to 31 Dec 1994	81	2 055 120	3.94 (3.1 to 4.8)		1 Sep 1994 to 31 Dec 1997	6	76	501 449	1.20 (0.2 to 2.2)		70 (30 to 87)
Toddlers	1 Jan 1995 to 31 Dec 1995	41	601 045	6.82 (4.7 to 8.9)		1 Jan 1998 to 31 Dec 1998	2	84	97 369	2.05 (−0.7 to 4.9)		70 (−24 to 93)
Infants	1 Jan 1996 to 28 Jul 1996	24	320 562	7.49 (1.5 to 10.5)		1 Jan 1999 to 28 Jul 1999	1	80	64 112	1.56 (−1.5 to 4.6)		79 (−54 to 97)
Overall		459	11 235 305	4.08 (3.7 to 4.5)			29		2 134 226	1.36 (0.86 to 1.85)		67 (52 to 77)

^*95% confidence intervals were estimated by using the Taylor series method for relative risks.