Measles vaccination coverage, determinants of delayed vaccination and reasons for non-vaccination among children aged 24–35 months in Zhejiang province, China

Measles is a common infectious disease and is considered as a significant cause of child’s mortality, with more than one million measles deaths reported annually worldwide [1]. Measles can be stopped by vaccination with measles-containing vaccine (MCV), which is one of the most cost-effective public health interventions. Due to the highly contagious nature of measles [2], the World Health Organization (WHO) suggests that over 95% of the target population should be immunized with two doses of MCV to halt measles transmission [1]. Missed individuals, vaccinated individuals with primary vaccination failure or immunity waning are susceptible. Furthermore, the secondary attack rate is around 90% among the susceptible population [1]. As such, the second dose of MCV is critical in resolving the primary vaccination failure or boosting the antibody titers after the immunity waning.

Apart from obtaining the vaccination coverage, timeliness is another important index for the success of control and elimination of measles. It is necessary to get measles vaccinations at the recommended ages which are set based on balance between the risk of infection and the immunogenicity of MCV. Delayed vaccination enlarges the gap between the protection from vaccination and the loss of maternal antibodies, which weakens the herd immunity and increases the risk of outbreaks or epidemics. Vaccination coverage is an indicator widely applied for evaluating the performance of expanded program on immunization (EPI), but in these estimates it is difficult to assess the timeliness as it excludes the precise age of administration.

As the measles elimination activity requires an extremely high level of coverage, planning and implementing an effective measles vaccination program is critical. Since 1980s, the Chinese expert advisory committee on immunization program recommended a two-dose MCV vaccination schedule for all children in routine immunization [3]. In 2004, the schedule of MCV was updated, with the first dose of MCV (MCV1) at 8 months of age and the second dose of MCV (MCV2) at 18 months of age, with a minimum interval of 28 days between two doses. MCV vaccinations are included in the Chinese EPI and provide to all children free of charge.

Though the Chinese government had initiated the measles elimination activity since the beginning of this century as a part of the global vaccine action initiative, the outbreaks of measles occurred periodically in every 2–3 years. In Zhejiang province, the measles incidence rates of 2005 and 2008 were both over 30 per 100,000, which were three times as the provincial average incidence of the last two decades. Outbreaks resulted in 14,317 cases in 2005 and 12,782 cases in 2008, and mostly impacted children under three years of age [4]. According to the estimates of vaccination coverage of MCV among children born from 2005 to 2011 based on Zhejiang provincial immunization information system (ZJIIS), the coverage of MCV1 was 98.6% while the coverage of MCV2 was 82.8% [5], much lower than the threshold of coverage of MCV vaccination for maintaining the herd immunity. Additionally, this study only included the children registered in ZJIIS, which would inflate the vaccination coverage due to a common perception that children not registered were also likely to miss the vaccinations. Another coverage survey which was conducted in 2011 in 18 counties of Zhejiang province showed that the coverage of MCV1 was 92.7% but its timeliness was only 59.3% [6]. The lower coverage of MCV and the poor timeliness were considered as the “significant off-track” to meet measles elimination goals.

In order to improve the coverage of complete two doses of MCV vaccination and its timeliness, Zhejiang provincial center of disease control and prevention (CDC) encouraged all immunization providers to adjust the frequency of vaccination service, promote the health education and social mobilization, and launch a training program towards vaccinators to emphasize the importance of MCV vaccinations. However, we had not systematically evaluated the effects of these initiatives at the provincial level. As such, the 2016 Zhejiang provincial coverage survey was conducted among children aged between 24 and 35 months of age to evaluate (1) whether there was a progress in the coverage rates of MCV since 2011; (2) the timeliness of MCV vaccinations; (3) the reasons for non-vaccination of MCV; and (4) the potential determinants for delayed MCV vaccinations.

For the whole province, the coverage of MCV1 reached the goal of 95% set by WHO while the coverage of MCV2 was still below the threshold for stopping the measles transmission. We found that there was 3% children drop out the MCV2 after receiving the MCV1. Our study also found the disparities among cities which might be masked by the provincial estimate. For example, the coverage rates of MCV1 and MCV2 in Wenzhou were all below 90%, which would put children at the risk of measles infection or outbreak. The inequalities of vaccination coverage had been reported in a previous study from the low- and middle- income counties due to the unbalanced socio-economic development [9].

Our previous studies [4, 7] had already emphasized the importance of measuring timely vaccination since simply focusing the coverage at a given age would overrate the real protection in population. The significance of our point was that the sub-population with the delayed vaccination were posed at the unnecessary risk of diseases. Consequently, outbreaks would occur if the epidemic threshold was surpassed and it might spread much faster when the delayed vaccination was coupled with a low vaccination coverage. Compared with the provincial estimates of a similar survey in 2011, the timeliness of MCV1 vaccination had been improved significantly (76.6% in 2016 vs.59.3% in 2011 survey). Actually, it was prioritized to immune children at the appropriate ages and reduce delays after the 2011 survey. CDCs at province and city levels made great efforts to enhance timeliness, mainly through raising awareness of providers and sending reminders to parents. Timeliness had also been considered as an important indicator for assessing the performance of the local immunization program since 2011, in additional to the traditional indicator like the vaccination coverage. Despite of the encouraging improvement, significant gaps still existed. The coverage threshold of 95% was considered to confer measles herd immunity. However, the timeliness of MCV1 and MCV2 still below 90% in this study. The suboptimal timeliness extended the risk time of measles infection for children. Timeliness of MCV vaccination reported in other settings varied, for example, the delayed proportions ranged from 19 to 78% for MCV vaccination in other developing countries [3, 10, 11]. The delayed MCV vaccination was also observed in Switzerland [12], where children spent on average 266 days susceptible to measles from 6 to 24 months of age, and 1/3 of this time was due to the delayed vaccination. Another concern was that province-wide measles endemics had occurred in the last decade and the proportion of cases under 2 years had increased over time in Zhejiang province [4].

The reasons for non-vaccinations of MCV1 and drop-outs of MCV2 included child’s sickness when the vaccination was due, incompatible schedule of the immunization clinic with working hours, and the overcrowded immunization clinic. Drop-outs were also reported in Kenyan [13], where caregivers’ education level and the distance to the health facilities were considered as the important determinants. These findings indicated obstacles to receiving vaccination service still existed despite of the good geographical and financial accessibilities. There were some potential explanations. For example, parents might have misinformation that children with mild sickness could not be vaccinated. Parents received one or more poor immunization services (such as long waiting time) or time constraints might be unwilling to return. Another possible explanation for the low coverage of MCV2 was that caregivers might be less likely to pay attention to MCV2 which scheduled at 18 months of age, because most of the childhood vaccination was scheduled before the first year of life.

Several determinants of the delayed vaccination of MCV1 and MCV2 were identified in this study. First, high numbers of siblings was associated with the delayed MCV vaccinations. These families might face more resource and time constrains to support more children. Resources and parental attentions would be diverted and childhood immunization might not be prioritized amidst competing other demands as the benefits of vaccination might not be apparent immediately [11, 14]. Second, children with hospital delivery were more likely to be timely vaccinated, which was similar to the results in other settings [15, 16]. We inferred that mothers delivered at hospitals might use vaccination services more frequently as they received the better immunization knowledge and awareness conferred by the obstetricians or midwives. Third, children who had younger mothers were more likely to have the delayed vaccination of MCV2. We assumed that older mothers would have more experience in utilization of health care services, which led to an increase in timeliness. Fourth, higher maternal education level was demonstrated a determinant of the timely childhood vaccination. We assumed that a high education level would help mothers get a better understanding and acceptance of vaccination knowledge in practice, through a better communication with vaccination providers [17, 18]. Fifth, mothers with fixed jobs might not have enough time to spend on the childhood immunization or have time constraints due to the inflexible working hours. Sixth, non-local children were more probably to have the delayed MCV vaccinations as their parents might always face the challenge to survive in a new environment with the higher cost of living and might have difficulties in adapting to a new socio-cultural environment. Lastly, we found children with the poorer economic background were probably to have the delayed MCV vaccinations. Previous studies reported that it could be explained as the inaccessibility which was caused by the indirect costs (such as the transport fee) or the deduction of salary for the work leave to bring children to the immunization clinics [3, 16]. There were more determinants for the delayed vaccination of MCV2 than MCV1. We inferred that parents might take more attention on the vaccination scheduled during the first year of life, which might increase the probability of the timeliness of MCV1.

To our knowledge, this study was the first time to evaluate the coverage, the timeliness and completeness of MCV vaccination in Zhejiang province. The reasons for non-vaccination and the determinants of delayed vaccination would help the decision makers to improve the strategies on the routine MCV vaccination. However, there were still several limitations. First, our analyses excluded the children without any written immunization records. It might influence the internal validity of the results within the target population as the vaccination status of these children might differ from those with the immunization records. However, the exclusion reduced the probability of recall bias and only 4 out of 1386 children without any written records. As such, the information bias could be ignored. Second, we could not controlled all confounders for vaccination coverage because of the data limitations. Thus, the impact from other variables like knowledge, attitude on MCV vaccination could not be evaluated. Third, the results of this study would only represent the situation in Zhejiang province and might not be appropriately generalized to other provinces in China.

To summarize, the results of this study demonstrated that the coverage of MCV had improved though one or two cities had not reached the threshold of halting measles transmission. However, the timeliness of MCV still needs improvement. Thus, we suggest the policy-makers pay more attention to the reasons for non-vaccination and determinants of delayed MCV identified in this study when planning efforts to ensure the high age-appropriate coverage of the MCV.