Effectiveness of Community-Based Surveillance and Referral Slips in Reducing Zero-Dose and Under-Immunized Children in Ethiopia

Background: Immunization Challenges in Ethiopia’s Remote Communities

In Ethiopia, routine vaccination coverage for the recommended Expanded Program on Immunization (EPI) vaccines remains low, particularly in remote and mobile communities with limited education, access, and poor infrastructure. Low vaccination coverage has persisted despite the continued implementation of the recommended EPI vaccinations, indicating the need for stronger measures to improve uptake. In 2023, for example, the uptake of the first dose of diphtheria, tetanus, and pertussis (DTP)-containing vaccine was 77% ⁽¹⁾. The country is also home to high numbers of unvaccinated (“zero-dose”) children and under-vaccinated children ^{(2, 3)}. In 2023, an estimated 917,000 zero-dose children were in the country ⁽²⁾. The Ethiopian Mini Demographic and Health Survey 2019 revealed that vaccination coverage among children aged 12–23 months is low, with only 43% of children receiving all essential vaccinations ⁽⁴⁾.

Zero-dose children are often concentrated in marginalized urban areas, isolated rural locations, regions affected by conflict, and regions having lower infrastructure like pastoralist communities. Zero-dose children face a high risk of illness from vaccine-preventable diseases. Therefore, there is an urgent need to close the immunization gap and ensure that vaccines are equitably distributed, particularly in areas where healthcare resources are limited or where geographical, social, or political barriers obstruct access.

This study examined the effectiveness of active community-based surveillance, augmented by referral slips, on reducing zero-dose and under-immunized children in pastoralist/semi-pastoralist communities in Ethiopia.

Research Focus: Addressing Zero-Dose and Under-Immunized Children 

The primary objective was to assess the effect of active community-based surveillance augmented by a referral slip to improve immunization coverage and reduce zero-dose and under-immunized children. 

Study Objectives: Measuring Impact and Cost-Effectiveness 

Key objectives: 

• To assess the effect of the intervention in improving immunization coverage and reducing zero-dose and under-immunized children. 

• To assess the cost analysis of the intervention “active community-based surveillance augmented by a referral slip” to determine the program’s cost-effectiveness. 

• To assess the intervention’s effect in improving respondents’ knowledge, beliefs, and attitudes towards childhood immunization. 

Methodology: Cluster Randomized Controlled Trial Approach

This study included a cluster randomized controlled trial to examine the effectiveness of active community-based surveillance and a referral slip system in boosting immunization coverage among pastoralist/semi-pastoralist communities in Ethiopia. The intervention arm included villages that actively participated in community-based surveillance delivered by community volunteers, enhanced with referral slips. In contrast, the control arm consisted of villages that did not engage in the program but continued receiving routine health services.

Study Location: CORE Group Polio Project (CGPP) Implementation Areas

The study was conducted within the implementation region of the CORE Group Polio Project (CGPP) – Ethiopia. CGPP Ethiopia operates in over 80 districts, known as woredas, in remote and pastoralist regions across six regional states. These districts are home to a total population of 6.2 million people, which includes 218,513 children under the age of one. Each woreda comprises 10 to 20 kebeles, the smallest administrative units, containing between 300 and 1500 households. Most rural kebeles have a health post managed by health extension workers. In the CGPP implementation areas, additional community volunteers carry out essential immunization elements, provide health education, and advise mothers to vaccinate unimmunized children through weekly house-to-house visits. 

These CGPP implementation areas are characterized by their challenging accessibility, encompassing mobile communities in remote regions with fragile infrastructure, often in the country’s border areas (Figure 1). The population in these regions typically has limited educational opportunities with pastoralist culture.  

Figure 1. Map of CORE Group Polio Project intervention areas 

Study Population: Children and Caregivers in Pastoralist Communities

The primary study population was children aged 12 to 24 months living in the kebeles and their mothers/caregivers within the study area. Children living in urban woredas with access to health centers were excluded, including guests and children not permanent residents of the selected kebeles.

Sampling Technique: Randomized Selection of Households

In the first stage, woredas were stratified by regional states. Next, two woredas from each region were selected, with four kebeles from each woreda subsequently selected, using probability proportional to the size of the eligible kebele populations. This sampling resulted in the selection of 12 rural kebeles from three districts. Rural kebeles were further randomized intervention and control arms.

In the final stage, households with eligible children were selected using the EPI method to identify eligible households. In the EPI method, the sampling starts at the center of a village by rotating a bottle and beginning data collection in the first household in the direction of the bottle failed. Data collection continues sequentially for eligible persons until the sample size for the village is achieved ⁽⁵⁾. Each Kebele was divided into four villages, and at each village center, data collectors spun a bottle to determine the direction for selecting the first household, then sequentially selected ten households with eligible children in each village. A total of 40 eligible children from each kebele were included in the study’s treatment and control arms to meet pre-determined sample size requirements. This sampling approach was utilized during baseline and endline surveys.

Study Procedures: Community-Based Surveillance and Referral Slip System

The study team developed a standardized questionnaire to measure the socio-demographics of parents and children, their knowledge, attitudes, and practices related to immunization, and other factors influencing vaccination in the intervention and control areas. The effect of the intervention was assessed by conducting baseline and endline surveys among intervention and control groups. The baseline was conducted before the intervention, while the endline was made at the end of the project after ten months of starting the intervention. Three teams of data collectors, each consisting of 8 enumerators, were deployed to conduct face-to-face surveys. A senior supervisor led each group from a CGPP partner or the Woreda health office.

Following a baseline survey and identifying intervention and non-intervention Kebeles, an intervention was included in the intervention Kebeles. The primary intervention included an active community-based surveillance augmented by a referral slip system. An orientation pamphlet was developed to describe what qualifies a child as zero-dose or under-immunized to guide the community volunteers, the details of activity performed by health extension workers, and the supervising group in the intervention kebeles.

A referral slip was also developed to improve immunization coverage. The referral slip was a card with pictures representing the registration of newborns, unimmunized children, and “defaulters”[1] and carried tailored messages. These messages were pre-tested in different communities before being printed and laminated. As each kebele had five community volunteers, the laminated referral slips came in five colors to differentiate the community volunteers and track who referred community members to health extension workers.

The intervention was as follows: community volunteers conducted house-to-house assessments in their catchment areas and tallied all infants to establish a denominator as a baseline. When a zero-dose, under-immunized, or vaccine-defaulter child was identified, the community volunteers provided persuasive education, and they also gave the child’s caregiver a referral slip to go to the health post for a vaccination. Upon arriving at the health post, the child’s caregiver handed the referral slip to the health extension worker, who provided persuasive education based on the reason for defaulting, administered the necessary vaccines, documented the service, and placed the referral slip in a tickler box. During the monthly review meeting at the end of the month, the health extension workers brought the referral slips and evaluated the five community volunteers’ performance. These meetings included special monitoring to encourage community volunteers to improve their performance.

[1] “Defaulters” refers to individuals who have not completed the immunization schedule.

Data Collection & Analysis: SPSS and Cost Evaluation Methods

Data was collected using the Open Data Kit (ODK) and exported to a server in Excel format. Analysis was undertaken using SPSS with complex sampling design commands. Financial data were collected based on financial assistance given to the intervention areas. Immunization coverage and rates of zero-dose and under-immunized children were then compared between the two groups.

The immunization status of children was determined using three methods. First, information was obtained from EPI cards provided by health posts. If the vaccination card was lost, immunization records were verified through the health facility where the child was vaccinated. Lastly, the study team enquired about key vaccination times, including checking for a BCG scar and whether the child had received polio drops and the measles vaccine around the 9-month mark. Children were classified as zero-dose if they had not received any vaccinations and as under-immunized if they had only been partially vaccinated for their age. Immunization status was determined for children who had lost their immunization cards by linking them with health facilities, asking questions, and examining key vaccination indicators.

Additionally, data related to the financial aspects of the intervention were collected. The study considered direct costs, including training of community volunteers and health extension workers, direct supervision and supplies of intervention materials, and indirect costs, including program management.

Key Findings: Impact on Zero-Dose and Immunization Coverage

The study involved 960 respondents (480 in the intervention group and 480 in the control group). Most caregivers (59.5%) were aged between 15 and 60, with an average age of 28.1 years (Table 1). A majority (82.5%) of caregivers were female. More than half of the participants were completely illiterate (54.7), but a relatively higher proportion of participants in the intervention group had elementary or high school education.

Table 1. Socio-demographic characteristics of caregivers in remote pastoralist and semi-pastoralist communities in Ethiopia

*SD, standard deviation.

Impact of the intervention on zero-dose and under-immunized children

• At baseline, 2.5% of children were zero-dose and 47.3% were under-immunized
• in the intervention group, compared with 21.0% of children who were zero-dose and 36.5% who were under-immunized in the control group.
• At endline, only 1% of children in the intervention group remained zero-dose, compared to 32.1% in the control group.
• Partial immunization was significantly lower in the intervention group (12.5% vs. 40.4%), while full vaccination rates increased to 86.5% in the intervention group, compared with just 27.5% in the control group.

The intervention also proved cost-effective, requiring:

• US$339.00 to reduce zero-dose children by 1%
• US$376.67 to reduce under-immunized children by 1%
• US$178.12 to increase full immunization coverage by 1%

The cost per child to prevent a zero-dose case was US$17.74, while US$19.71 was needed to reduce under-immunization, making the intervention impactful and financially viable (Table 2).

Table 2. Endline and baseline immunization status in intervention and control groups among remote pastoralist and semi-pastoralist communities in Ethiopia

Knowledge, attitudes, and beliefs about childhood vaccination

In the endline survey, mothers/caregivers from the intervention group had significantly higher knowledge than those in the control group (62.3% vs. 27.5%, P < 0.001). Specifically, mothers/caregivers from the intervention sites demonstrated greater awareness in areas such as understanding common side effects of childhood vaccines, knowing the appropriate timing for starting and completing routine immunizations, recognizing the role of vaccines in preventing infectious diseases, and knowing where to access vaccines (Figure 2). Additionally, mothers/caregivers in the intervention group were significantly less likely to refuse to vaccinate their children than the control group (5.0% vs 14.6%, P < 0.001). Both groups had a high proportion of mothers/caregivers who understood that administering multiple vaccines concurrently does not pose significant problems (Figure 2).

Figure 2. Knowledge of mothers/caregivers on routine immunization for children in remote pastoralist and semi-pastoralist communities in Ethiopia

Most mothers/caregivers from the intervention and control groups were generally neutral regarding child vaccination (Figure 3). However, there was a higher level of neutral beliefs around childhood vaccination among the control and intervention groups (60% vs 49.8%), but not much difference between the groups for neutral attitude (47.9% and 44.4%). Additionally, a more significant proportion of mothers/caregivers in the intervention group showed positive beliefs (42.5% vs 18.3) and attitudes (42.7% vs. 30.2%) towards childhood vaccination than the control group. In contrast, a higher proportion of mothers/caregivers in the control group displayed negative beliefs (21.7% vs 7.7%) and attitudes (25.4% vs 9.4%) toward vaccinating children.

Figure 3. Beliefs and attitudes of mothers/caregivers around childhood vaccination among remote pastoralist and semi-pastoralist communities in Ethiopia

Conclusion: Policy Implications and Scalability of Intervention

The use of active community-based surveillance alongside referral slips was effective in reducing the number of zero-dose and under-immunized children in remote pastoralist/semi-pastoralist communities in Ethiopia. The intervention was also cost-effective, costing US$17.74 per child to prevent a zero-dose case, making the intervention both impactful and financially viable. The significant reduction in zero-dose and under-immunized children in the intervention group demonstrates the effectiveness of raising awareness and fostering positive attitudes toward immunization. The findings highlight the need for continued investment in community-based interventions and suggest that similar programs could be scaled up globally as a sustainable approach to enhance immunization coverage.

Recommendations

• Community members should actively engage in vaccination programs.
• Local health providers should adopt similar intervention strategies, utilizing community volunteers and referral slips for house-to-house educational outreach to find vulnerable children and increase immunization rates.
• Continued investment in community-based approaches, including the use of community volunteers, can be effective in improving the uptake of routine immunizations.
• Healthcare providers should directly address attitudes toward childhood vaccines to help boost immunization uptake.
• Policymakers and researchers should collaborate to implement evidence-based strategies that reduce barriers to vaccination, ensuring that more children are fully immunized and public health outcomes are improved.

References

1. World Health Organization. Diphtheria tetanus toxoid and pertussis (DTP) vaccination coverage 2024 [Available from: https://immunizationdata.who.int/global/wiise-detail-page/diphtheria-tetanus-toxoid-and-pertussis-(dtp)-vaccination-coverage?CODE=ETH&ANTIGEN=DTPCV1&YEAR=.
2. UNICEF. Immunization 2024 [Available from: https://data.unicef.org/topic/child-health/immunization/.
3. World Health Organization. Immunization dashboard 2024 [Available from: https://immunizationdata.who.int/.
4. Ethiopian Public Health Institute (EPHI) and ICF. Ethiopia Mini Demographic and Health Survey 2019: Key Indicators 2021 [Available from: https://dhsprogram.com/pubs/pdf/FR363/FR363.pdf.
5. Chao L-W, Szrek H, Peltzer K, Ramlagan S, Fleming P, Leite R, et al. A comparison of EPI sampling, probability sampling, and compact segment sampling methods for micro and small enterprises. Journal of development economics. 2012;98(1):94-107.