Measles outbreak in complex emergency: estimating vaccine effectiveness and evaluation of the vaccination campaign in Borno State, Nigeria, 2019

Between 2017 and early 2019, a global resurgence in measles cases and outbreaks has been observed [1, 2]. Four of the six World Health Organization (WHO) geographic regions reported significant increases in the number and size of measles outbreaks. The most affected countries were Madagascar, Ukraine, Philippines and Nigeria. Reported outbreaks worldwide have been attributed to areas of persistently low measles vaccination coverage [1, 2].

Nigeria has experienced repeated outbreaks of measles in recent years due to low routine immunization (RI) coverage. The country introduced measles vaccination into the RI program in 1978 for children aged 9 months. As at 2005, the national level administrative coverage was 32% [3]. Data from the 2018 Demographic and Health Survey (DHS) indicated that only 54% of children 12–23 months had received measles vaccine [4]; far below the herd protection threshold [5]. From January to May 2019, large measles outbreaks affected all 36 states in Nigeria and the Federal Capital Territory (FCT). By the beginning of May 2019, there were more than 28,000 suspected measles cases reported and 89 associated deaths with measles. A majority of cases were reported in Borno State with over 15,000 suspected cases and 75 measles deaths across 37 internally displaced persons (IDP) camps in 22 Local Government Areas (LGAs or Districts). High measles-associated mortality rates have been previously reported from IDP camps, and measles has been one of the major causes of child deaths in humanitarian emergencies [6, 7].

To respond to measles outbreaks in emergencies, the WHO recommends a range of epidemic control activities. These include review of epidemiological and immunization program data to identify the cause (s) of the outbreak, intensive measles surveillance, improving coverage amongst high risk populations and implementing supplementary immunization activities (SIA) in areas not yet affected by the outbreak [8]. In addition, complex humanitarian emergencies pose considerable logistical challenges to the delivery of vaccination services including maintaining appropriate cold chain conditions for vaccine storage. High vaccine effectiveness (VE) is critical to halt outbreaks and prevent ongoing transmission during complex emergencies. Thus, estimating the VE of measles-containing vaccine (MCV) following a large measles outbreak is an essential part of assessing the quality and efficacy of the response [9, 10].

During the 2019 measles outbreak in Borno state, the government of Nigeria through its National Primary Health Care Development Agency (NPHCDA), conducted two phases (March and May) of measles outbreak response immunization (ORI) in the most affected areas of the state. Here we describe the outbreak response and its impact in Borno State by assessing the investigated and confirmed measles cases as well as their vaccination status. To assess the population-level impact, we estimated the age-specific VE for MCV by applying the screening method on these data, and by estimating the vaccination coverage using a post-campaign coverage survey (PCCS).

Increased attacks in Borno State in December 2018 led to a massive population displacement. Subsequently, there were an increasing number of unvaccinated children who moved toward MMC and living in IDP camps in conditions favoring measles transmission. An ORI was essential for improving population immunity and interrupting the large measles outbreak in line with the global measles and rubella control and elimination goals [19].

The State has a history of outbreaks of measles in the last 5 years due to persistently low routine immunization coverage, and has been conducting measles campaigns almost every 2 years as a key strategy for reaching high-risk populations in security challenged/hard-to-reach areas. In November 2015, Borno responded to a measles outbreak targeting 1,255,100 children aged 6 to 59 months for vaccination with MCV. Only 66.7% (837,743 children) of the targeted children were vaccinated during the response. In early 2017, a mass measles vaccination campaign was implemented as part of outbreak preparedness in conflict-affected areas of Borno state. The campaign was implemented in 25 LGAs of the State. According to the PCCS concluded in February 2018, the proportion of children aged 9 months to 59 months who received measles vaccine during the 2017 campaign was 72.2% and, of all respondents, only 52.2% had received measles vaccines before the campaign [20].

During the 2019 ORI, a total of 1,238,181 children were vaccinated with MCV giving an administrative coverage of 98.75%. As administrative coverage reports may not always be accurate, a PCCS was conducted and has indicated an aggregated weighted coverage for the 12 LGAs included in the survey of 85.7%. The 95% coverage objective was not achieved due to the complexity in conducting a vaccination campaign in an emergency setting where insecurity, regular influx of new persons, population displacement, and hard-to-reach areas are extensive. In order to overcome these obstacles, collaboration with the military enabled the vaccination teams to immunize eligible children in security compromised areas. This strategy builds upon the Reaching Inaccessible Children strategy (RIC) conducted by the military and the Reaching Every Settlement (RES) strategy which uses some military support to reach areas that were partially or completely inaccessible to the vaccination teams [17]. In recently accessed territories at the time of the ORI (i.e. camps and communities in Bama, Gwoza, Dikwa, Damboa, Ngala, Kukawa and Monguno LGAs), “hit and run” vaccination activities have been conducted with military escort to provide immunization and basic health care in the camps in these areas. Additionally, in hard-to-reach settlements, the polio eradication platform was used to reach underserved communities with measles vaccination [21].

Data on measles VE estimations in emergency settings in Nigeria and other countries are limited, and, where available, quite dated. During a measles outbreak in refugee camps in Mozambican refugee camps in Malawi, VE was estimated for children less than 5 years using the screening method. The findings of this investigation showed a VE of more than 90% [22]. A two-stage cluster survey of 563 children in famine emergencies in Ethiopia found a low VE of 66.9% in children 9–36 months old. The authors suggested problems with the cold chain or vaccine administration [7]. We have obtained a high estimate of MCV VE among all age groups during ORI in Borno state Nigeria. VE ranged from 87.3% (95%CI: 71.0, 95.2) to 99.5% (95%CI: 98.3–99.9) for children from 9 to 71 months. Measles VE at 9–11 months and in those greater than 12 months of age is expected to range between 84 and 92.5% [5].

Despite the limited storage capacity and poor immunization infrastructures at the lower level, efforts were made to ensure adequate cold chain and vaccine handling. Upon receipt of the MCV in-country, the vaccines were prepositioned in the Zonal cold store in Bauchi State (452 km driving distance to MMC). The vaccines and devices were distributed from the Zonal to the State store 2 weeks before the commencement of the ORI. The State distributed bundled vaccines to the LGA stores ahead of the campaign using cold boxes. Fast-cold chain was used to comply with the vaccine distribution process up to the communities and service delivery points.

This study is subject to some limitations. First, the result of the PCCS survey cannot be generalized to all children in the eligible age group in Borno state. The PCCS coverage results do not include coverage for children in the 14 LGAs that were not included in the survey since they were not part of the ORI and the additional LGA (Biu) dropped from the survey due to insecurity. The survey teams did not visit IDP camps and in addition, as a consequence of the ongoing insecurity, there has been constant displacement and migration of populations within and out of Borno state. The number of children eligible for vaccination at any settlement in Borno state is constantly changing and it is difficult to track movements and how the movements may affect the proportion of vaccinated and unvaccinated children. Limiting the analysis to non-IDP cases may under-estimate the PPV of the case population. Secondly, the vaccination status of children in the inaccessible wards could not be estimated. The final limitation of the PCCS is (due to security challenges) the lag time of 4 months and 2 months between completion of the two phases of the campaign and timing of the survey. This may have influenced survey results as a consequence of recall bias. Only 32% of all interviewed individuals produced their vaccination cards, and 20% did not receive cards or could not recall whether they received vaccination cards. Considering multiple antigens delivered in the RI schedule and periodic ORI and preventive supplementary immunization activities, it is difficult for parents to accurately recall all doses of all antigens received by their child in the absence of widespread retention of vaccination cards. However, as demonstrated by previous immunization surveys conducted in Nigeria, using cards alone actually underestimates coverage, if not all doses are recorded [23].

In addition to the above mentioned limitations, there are some potential biases in all observational VE studies including the screening method [24]. Specifically, bias related with misclassification and definition of cases, and recording of vaccination status. In our study, these biases were minimized by using the WHO standard case definition for measles, and the cases that were investigated and line-listed by the surveillance officer (i.e. DSNO).

Our findings suggest that decreased VE is unlikely to explain the recurrence of measles outbreaks in Borno State. Insecurity resulting in lack of health and immunization services, frequent population displacements and clustering of susceptible children in IDPs are more likely contributors to measles morbidity and mortality in the State [25]. Despite the estimated VE from the ORI activities, sub-optimal RI coverage is insufficient to maintain population immunity and prevent the accumulation of susceptible children that allow outbreaks to occur.

This analysis provides important insight for future vaccination in emergency and humanitarian settings in Nigeria and elsewhere. Data on the target age range for vaccination campaigns and strategies for measles control in response to an outbreak of measles in complex emergencies are available [26‐28]. Adequate VE can be obtained with appropriate cold chain capacity when conducting ORI in situations where insecurity may challenge vaccine conservation. The integration with other health strategies (i.e. polio platform) and the engagement of security personnel increased the number of communities that could be accessed and reduced the time spent in these areas by the vaccination team. As measles has been one of the major causes of child deaths in complex emergencies, regular assessment of local measles epidemiology and population immunity status is recommended to determine the risk of an outbreak [6, 27].