MMR vaccine is the safest and most effective way to protect against measles, mumps, and rubella. It is recommended in the UK at ages 13 months (MMR1) and 3 years 4 months to 5 years (MMR2),¹ although uptake has decreased following prominent adverse publicity since 1998.^2,3 We analysed MMR uptake across Scotland from its introduction (October 1988) to most recent data (February 2005). This would enable any changes in uptake to be detected, including whether parents are delaying vaccination, and whether there are differences by deprivation, and allow prediction of final MMR1 uptake. These figures would also allow measles susceptibility among nursery and primary school children to be examined, including in small geographical areas, where clustering of non-immunes could be detected.

METHODS

Data on vaccines received by children throughout Scotland are held on the Scottish Immunisation and Recall System (SIRS). On 15 February 2005, this contained over one million entries (n = 1 079 327) for children born between 1 January 1987 and 31 December 2004. Records include date of MMR vaccination, date of birth, sex, postcode sector of residence (e.g. G3 7--), and deprivation category (1 to 7). Deprivation is determined according to the proportion of the population in the postcode sector with access to a car, in overcrowded households, with the head of household in social class IV or V, and in households with unemployed men, using the Carstairs and Morris index.⁴ All children are entered onto the database, at birth or on migration. Measures in place to ensure a high degree of data completeness and accuracy include unique personal identifiers, incentive payments, and data checking by NHS Boards. However, the data are minimum figures for vaccine uptake, as under-recording of data is possible,⁵ and data on single vaccines cannot be entered.

To investigate delayed vaccination and any impact of deprivation, MMR1 uptake, after age 2 years, was examined by deprivation category. MMR1 uptake tends to stabilise for birth cohorts at around age 6 years. Figures for this “final” uptake by birth cohort were predicted by linear regression, based on the most recent data.

To estimate measles susceptibility, we assumed a vaccine efficacy of 90% for one dose and 99% for two doses.⁶ The percentage of susceptible nursery and primary school age children was estimated for individual postcode sectors, as at 1 September 1998 or 1 September 2003 (that is, approximately the beginning of the academic year). The primary cohort included children born between 1 March 1997 and 28 February 1999 (2003) or 1 March 1992 and 28 February 1994 (1998), as these dates are specified in school entry criteria. These children would have been aged 4.5 years to 6.5 years, be expected to be in primary years 1 and 2, and virtually all to have been offered MMR1 and MMR2. The nursery cohort included children born between 1 March 1999 and 28 February 2001 (2003) or 1 March 1994 and 28 February 1996 (1998). These children would have been aged 2.5 to 4.5 years and virtually all in pre-school settings. All in this cohort will have been offered MMR1, and some MMR2.

All statistical calculations were performed using R.⁷ Ethical approval was not required.

RESULTS

The percentage of children who received MMR1 at ages 2, 3, or after 4 years is shown in table 1. The 1999 birth cohort was the first with decreased MMR1 uptake, despite the 1997 birth cohort being the first to be exposed to adverse publicity. There is also evidence of a slight increase in late MMR1 uptake for the 1999 birth cohort onwards. The same overall pattern remains when data are analysed by deprivation (fig 1). However, the greater the deprivation, the greater the tendency for late vaccination, especially for the most deprived (category 7). Those who are more affluent tend to be vaccinated early, or not at all. The predicted “final” MMR1 uptakes for birth cohorts 1999–2001 were all under 95%, but over 90%, including 95% confidence intervals. This is in contrast to birth cohorts 1992–98, which all have “final” MMR1 uptake of around 96%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 MMR uptake after age 2 years, by deprivation category, birth cohorts, 1988–2000.

The percentage of primary children susceptible to measles did not change significantly from 1998 to 2003 and areas with greater than 10% susceptibility tended to be rural, with fewer children (table 2). Susceptibility among nursery children is higher as most have had access to MMR1 only. The median susceptibility in postcode sectors increased significantly over the five years (from 13% to 16.4%) and the percentage of postcode sectors with greater than 20% susceptibility almost tripled (from 8.2% to 23.8%). In 2003, these included areas of higher population density (median 59 children), although areas still tended to have fewer children than those sectors with under 20% susceptibility (median 104 children). Looking at larger areas, of postcode district (e.g. G7 ---), and restricting analysis to districts with over 200 nursery children, in 1998, there were only three districts with greater than 20% susceptibility, but by 2003, this figure had risen to 25, with an excess of deprivation categories 1 and 7.

DISCUSSION

Late uptake of MMR1 has increased slightly for birth cohorts 1999 onwards, but the increase is insufficient to compensate for general declines, which have led to significantly increased measles susceptibility in nursery children.

The study uses the entire immunisation dataset for Scotland and should therefore provide a complete picture of MMR uptake and measles susceptibility. It should be applicable elsewhere in the UK, according to deprivation, as the whole country experienced declines under similar media influences. It is interesting that despite the 1997 birth cohort being the first to be exposed to adverse publicity, it was only cohorts 1999 onwards that had significant uptake declines, suggesting a gradual build-up of its effect, rather than an instant impact.

Increases in measles susceptibility were predicted from decreases in MMR uptake, but this is the first study published to translate these into susceptibility estimates and examine effects, at the level of postcode districts and sectors. Although susceptibility among primary children is acceptably low in most areas, increased susceptibility among nursery children is concerning, particularly in the most vulnerable districts, where a significant outbreak could result should measles be introduced, creating further inequalities in areas of higher deprivation. Outbreaks have already been observed elsewhere in the UK, where uptake has been much lower and for a longer period.^8,9

Factors influencing vaccine uptake have been widely investigated, with previous UK studies finding that MMR1 uptake declined with increasing deprivation,¹⁰ and that living with one parent only^11,12 or being the third or more child was significantly associated with lower uptake,¹² while unemployment was significantly associated with increased uptake.¹¹ However, a more recent study found that parents declining MMR vaccine were more likely to belong to a higher socioeconomic group.⁵ Lower MMR uptake has also been shown for children registered with inducement or single handed practices.¹³ These studies are very informative, but all measured uptake for a range of ages between 19 and 24 months, and did not examine delayed uptake after the age of 24 months, as presented here. The final figures predicted by linear regression are also new.

It is clear that improvements in MMR uptake are desirable. Any patterns associated with deprivation are important in understanding how to maximise uptake within a given social setting. It is most important to increase overall uptake in more deprived areas, but increased late uptake in more affluent areas should also be promoted. We have shown that those who are more affluent tend to make timely decisions as to whether to vaccinate or not, while those who are more deprived experience greater delay. We were unable to distinguish between individual (for example, educational attainment) and area level (for example, health service accessibility) factors that may determine this. Therefore, we are now investigating the association of more specific factors, such as car ownership, unemployment, overcrowding, social class, educational attainment, country of birth, and rurality with measles susceptibility, in a further study. This may clarify whether delays in vaccination are related to a longer process of decision making, or to more non-specific aspects such as accessibility, mobility, or perceptions of the health service. If it is the former, qualitative studies have shown that contact with health professionals and recommended health education material positively contribute to decision making.^5,14 There should therefore perhaps be greater awareness of this potential when working among more deprived communities. The potential for current and future outbreaks of measles, mumps, and rubella is currently being investigated in Scotland by techniques of mathematical modelling. These will further inform possible future interventions, such as initiatives to increase MMR uptake prior to school entry, which is likely to be required to increase MMR1 uptake above 95%, the herd immunity threshold.