Micro-planning for immunization in Kaduna State, Nigeria: Lessons learnt, 2017.

BACKGROUND: The OPV 3 coverage for Kaduna State, 12-23 months old children was 34.4%. The low OPV 3 coverage, due mainly to weak demand for routine antigens and the need to rapidly boost population immunity against the disabling Wild Polio Virus (WPV), led the Global Polio Eradication Initiatives (GPEI) to increase supplemental OPV campaigns in Kaduna State, despite the huge cost and great burden on personnel. The OPV campaigns, especially in high risk (low vaccine uptake, <80% OPV 3 coverage and high vaccines refusal rate) states of northern Nigeria with poliovirus transmission has resulted in overestimated denominators or target population, as the highest ever vaccinated is used to set OPV campaign targets.
METHODS: We utilized a cross-sectional study that assessed the impacts and possible solutions to the challenges of overestimated denominators in immunization services planning, delivery and performance evaluation in Kaduna State, Nigeria. We used both descriptive and quantitative approaches. We enumerated households and obtained the target populations for routine immunization (<1 year), polio campaign (<5 years) and acute flaccid paralysis surveillance (<15 years).
RESULTS: We found a significant difference in mean scores between the micro-planning and supplemental vaccination data on a number of <5 years (M = 102967, SD = 62405, micro-planning compared to M = 157716, SD = 72212, supplemental vaccination, p < 0.05). We also found a significant difference in mean scores between the micro-planning and projected census data on a number of <1 year (M = 26128, SD = 16828, micro-planning compared to M = 14154, SD = 4894, census, p < 0.05).
CONCLUSION: Periodic household-based micro-planning, aided with the use of technology for validation remains a useful tool in addressing gaps in immunization planning, delivery and performance evaluation in developing countries, such as Nigeria with overestimated denominators.

Introduction

Vaccines are one of the most cost effective public health interventions [1], [2], [3], [4], [5], [6]. The global community is at risk of more outbreaks, owing to dwindling resources to health, wars and conflicts, weak vital statistics, the dearth of skilled health manpower and low herd immunity [7], [8], [9]. The emergence and re-emergence of vaccine preventable diseases have put extra pressure on already fragile public health structure in developing countries [6], [9].

Vaccines save lives and confer herd immunity that breaks the transmission of vaccine preventable diseases in communities [10]. Vaccine hesitancy is on the rise in developed and developing countries, making delivery of potent vaccines to at risk population, mostly children difficult [11], [12], [13]. But developing countries, like Nigeria, face more challenges in addition to the low demand for vaccines; weak infrastructure, crippling poverty, inadequate planning and low premium on health [6], [9].

Planning for immunization services depends largely on the availability of reliable demographic statistics, dedicated skilled manpower, and fund [14], [15], [16], [17], [18]. For vaccination to be cost effective the target population must be known [7], [15]. The population of <1 year, <5 years, <15 years, pregnant women and women of child bearing age, obtained from periodic census or surveys are critical in effective planning and delivery of vaccines during routine and campaigns [7], [15], [16].

The last Wild Polio Virus (WPV) isolated in a paralyzed child in Nigeria, was in September 2016, despite robust and vigorous certification standard Acute Flaccid Paralysis (AFP) surveillance [19]. The successes recorded in Nigeria’s effort to interrupt poliovirus transmission is driven largely by aggressive and high quality Oral Polio Vaccine (OPV) Supplemental Immunization Activities (SIAs), robust surveillance and outbreak response, and Routine Immunization (RI) [20], [21], [22]. The OPV 3 coverage for Kaduna State, 12–23 months old children was 34.4% (2017, National Immunization Coverage Survey for Routine antigens, NICS). Caregivers for many reasons do not go to health facilities for routine antigens, including OPV. The low OPV 3 coverage, due mainly to weak demand for routine antigens and the need to rapidly boost population immunity against the disabling Wild Polio Virus (WPV), led the Global Polio Eradication Initiatives (GPEI) to increase supplemental OPV campaigns in Kaduna State, despite the huge cost and great burden on personnel [21], [22], [23]. The OPV campaigns, especially in high risk (low vaccine uptake, <80% OPV3 coverage and high vaccines refusal rate) states of northern Nigeria with poliovirus transmission has led to several campaigns which inadvertently resulted in overestimated denominators. Subsequently, the overestimation of target populations has led to huge financial losses. Planning parameters for vaccines and pluses depend on the highest <5 years vaccinated with OPV in the last three campaigns as the denominator [11], [19], [21], leading to high vaccines wastages from overestimated coverages and and a significant drain of limited human and material resources [6], [16], [23]. Pluses are additional incentive given to children and caregivers after vaccination, for example, sachet milk, sweet, balloon and so on.

The Global Polio-eradication Initiative (GPEI) in Nigeria has initially relied on national census figures and health surveys conducted periodically for planning and evaluation of immunization services [24], [25]. The projected census population of 3% per annum, has not kept pace with the high birth rate of >5% per annum, especially in northern Nigeria with very high fertility rates [26], [27], [28]. The last head to head count in Nigeria was 12 years ago, making it unreliable for effective immunization planning, implementation and performance evaluation [24]. The use of administrative data, is usually lower than the actual population with resultant high (RI) coverages, despite decreasing uptake from household coverage survey and numerous outbreaks of vaccine preventable diseases [7], [23].

The magnitude of overestimated denominators on immunization planning, delivery, and performance evaluation, has not been fully investigated in Nigeria. However, in this study, we compared impact on the effective and efficient planning for RI and SIAs with the use of overestimated denominators in immunization planning versus enumeration of households (<1 year, <5 years and <15 years). Our study aims to verify whether, enumeration of households will lead to clear reduction in vaccine wastage, pluses, and funding in Kaduna State, Nigeria.

Methods

Study design

We utilized cross-sectional study design to compare the impacts of using overestimated denominators versus the approach of household enumeration in immunization service planning, delivery and performance evaluation in Kaduna State, Nigeria. We used both descriptive and quantitative approaches. We enumerated households, <1, <5, and <15 populations in 23 local government areas (LGAs) of Kaduna State.

Kaduna, one of the states in Northern Nigeria, has 23 LGAs and 225 Wards (a ward is the smallest unit of traditional or governmental administration) with a projected total population of 8.6 million. (<1 = 325, 531, <5 = 1, 627, 657 and <15 = 3, 997, 229) [24]. Kaduna State was the defunct capital of the old northern region, and many tribes (Hausa, Fulani, Igbo, Yoruba, and others) in Nigeria, co-exist in the state. It has a vibrant agrarian population and is a centre of education and civilization in northern Nigeria. The infant mortality rate in north west Nigeria from the last Nigeria Demographic Health Survey (NDHS, 2013) was 89 per 1000 live births and <5 mortality rate of 185 per 1000 live births [25]. It is a high risk state of poliovirus transmission with low vaccine uptake, the OPV-3 coverage for Kaduna State, 12–23 months old children was 34.4% and pentavalent-3 coverage of 29.5% (2017, National Immunization Coverage Survey for Routine antigens). [11], [21], [23].

Study coordination

The planning for households and children enumeration was conducted at various levels: national, state, LGA and ward. The National Primary Healthcare Development Agency (NPHCDA), supported by development partners, namely; World Health Organization (WHO), United Nations Children Fund (UNICEF), and others led the planning at the national level. The meetings at the national level raised the consciousness of government and partners to the importance of quality and reliable target population. It set the stage for resources mobilization, and timelines for subsequent activities at the state, LGA and ward levels. The planning activities at state, LGA and ward levels were coordinated by government agencies supported by various development partners, with clear agenda, objectives, timelines and responsible persons.

Microplanning process involves all activities leading to reliable data and rationalization of team workload in delivery of immunization service.

A micro-planning process was undertaken by the team, guided by a Standard Operating Procedures (SOP) developed by WHO Country Office, in collaboration with government and other development partners.

The microplanning process recognised the importance of key stakeholders, traditional leaders, religious bodies, bureau of statistics, the ministry of education and others towards successful enumeration. Advocacy was conducted through visits to traditional institutions at state, LGA and ward levels. The objective of this advocacy was to gain further support and mobilize study participants in the enumeration exercise. A meeting was held with community leaders in Kaduna state, as part of the advocacy efforts. During this meeting, the list of settlements from the SIAs was updated, providing a much-needed platform

Training of personnel

Personnel for the households and children enumeration received an elaborate and qualitative training across the national, state, LGA, and ward levels. The training was conducted in a systematic way, such that, all personnel received training before the commencement of the enumeration. Key officers from the states were trained at the national level, and they in turn at a later date trained others at the state level. The step-down approach guaranteed that trained officers at higher levels trained others at lower levels. The training at various levels was supervised by officers, who were previously trained and are familiar with training checklist. There was the field exercise of the enumeration training. This provided a unique opportunity for all participants to have a shared understanding of the field enumeration process and to critique the entire process.

Walk-through and enumeration of households and children

The enumeration team, made up of a supervisor and a community leader conducted a “walk-through” of settlements within their catchment area, at the ward and kept a record of households and children of various ages (<15 years, <5 years and <1 year) living in them. A catchment area is a predefined work area of a team which is usually a collection of settlements (collection of houses with an identifiable name), while a household was operationally defined as a woman and her children. The old definition of a household as a man, his wives, and children living in the same place and eating from the “same pot” was tinkered with so that no woman or child is left behind (missed or not enumerated). The new definition had the added advantage of precisely identifying noncompliant households, thereby removing the confusion often encountered when a polygamous man will have some wives accepting while others are refusing the vaccine. The enumeration teams were trained to record in their microplanning template a woman and her children as a separate household. Each woman and her children or wards were captured separately from the other women and their children in the situation the man has more than one wife. In addressing the sociocultural context of the environment, only females were engaged as supervisors since males are culturally prohibited to enter other people’s homes. The enumeration team worked for four to seven days, leaving no households or eligible children unaccounted for, in their catchment area. All enumeration teams utilized data tools: enumeration booklets, summary forms, daily implementation plans and micro-planning templates on and off the field to properly document and plan for supplemental immunization campaign.

Sketching of settlement map and border demarcation

The field work allowed enumeration teams to diligently and concurrently draw the map of the catchment area and properly define the catchment area borders with other teams working in the same ward or other wards. The delineation of catchment area borders guaranteed that no settlement or household was missed by teams, one of the key objectives of the enumeration exercise. At the end of the field work, a detailed catchment area map with all the important landmarks and boundaries was developed by the team to guide supplemental immunization activities. The aggregation of the catchment area maps assisted the officers’ in-charge of ward activities to sketch the ward map, and synthesis of the ward maps enabled the LGA to produce the LGA map.

Tracking of personnel

The enumeration teams were tracked in real-time with the use of electronic tracker. The teams’ supervisor was provided with a tracker that once activated can give information on the location of settlements and length of stay in a settlement. The trackers were configured and deployed by e-Health, a company engaged by Bill and Melinda Gates Foundation (BMGF) to support government at various levels to produce reliable maps of settlements. The enumeration supervisors submitted their trackers at the end of each day’s activities, the generated data were analyzed and the feedbacks shared with the enumeration teams and all other supervisors. The feedbacks were very useful in taking on the spot corrective actions, as missed settlements were identified and immediately followed up for enumeration.

Supportive supervision and validation

Supportive supervision at all levels and more especially at the catchment area guaranteed that settlements and households were not missed by enumeration teams. It was also useful in improving the quality of data obtained by the teams during “walk-through”. Enumeration teams received on -the-job-training, technical and moral supports that spurred them to do more despite the enormous demand of the exercise. The supervision was real-time with mobile device enabled with Open Data Kits (ODK) and data uploaded to a server on the availability of internet. The ODK tool is not new and WHO staff are familiar with it and have utilized it for field supervision since 2016. The spontaneity and speed of supervision and data flow reduced bias and deliberate data falsification. The validation of enumeration teams’ work by supervisors gave credibility and quality to the data. Validation requires that a senior supervisor revisits the entire day’s work of a team, from start to end, and compares total households and children with that of the team supervisor, with a decision criterion of ±15%, variation in total households and children as acceptance level. A variance of > ±15% requires validation of more day’s work, if any team in the additional validation has > ±15% the entire team’s work is rejected, and > ±30% variance means poor work and rejection of the team’s work, and the team must redo the work.

The use of social media groups, especially “WhatsApp” made coordination and sharing of findings in the field by supervisors possible and seamless. The sharing of best practices documented by supervisors enriched the working experiences of others and guaranteed that majority of the supervisors were on the same page.

Daily evening review meeting

The field work usually ended by 2–3 pm at the catchment area, and enumeration teams and their supervisors met at the ward to review the progress of work, challenges, successes, and plan for the remaining days’ activities. Ward level evening review meeting reports were further deliberated at the LGA evening review meeting, chaired by the director of primary health care of the LGA. The State also received reports from all the LGAs and feeds the national level on the progress of the implementation, challenges and action point to address identified gaps. The State Emergency Operations Centre (sEOC) made up of government officials and partners co-ordinated the state actions and feedbacks to all levels and also tracked the implementations of action points by agencies and partners.

Data collection and summary

Data tools were developed and utilized in 23 LGAs of the 255 wards of Kaduna State to collect and summarize information obtained from the field. The catchment area data were collated and summarized at the ward level, and transmitted to the LGA. All data from the LGAs were summarized to produce the State data which were transmitted to the national level. We utilized Microsoft Office Window 2010, Excel package for data collection and collation. Data quality checks were conducted at all levels to guarantee validity and reliability of data.

Debriefing

The enumeration exercise ended with debriefing of stakeholders by supervisors who supported the programme. The debriefings at the ward, LGA and state were useful fora to share data, field experiences, best practices and action points. The overall goal of the enumeration exercise was further appraised: reliable denominators (target population), that will guide planning, implementation, and evaluation of immunization of all eligible children in all settlements and households were documented and shared.

Statistical analysis

We analyzed trends in households, <1 year, <5 years and <15 years after the households and children enumeration (micro-planning) in August 2017. We gauged (paired samples t-test) the impact of the micro-planning by assessing the significant differences in mean scores of the micro-planning, supplemental immunization and projected census data of comparable period. Pearson’s coefficient of correlation was used to assess the relationships that exist between the variables (households, <1, <5 and <15).

Ethical approval

We obtained ethical approval from the ethical review committee of Kaduna State Ministry of Health before the commencement of the households and children enumeration (micro-planning). The enumeration teams clearly explained the rationale and objectives of the micro-planning to caregivers and obtained consent before the interview.

Results

The number of households increased to 2,110,630 (24%) after the households and children enumeration (micro-planning) in August 2017 compared to vaccination data of 1,698,711 obtained from the last campaign in July 2017, except in Jaba, Kaduna North, Sanga, and Zango Kataf LGA (Fig. 1). However, the difference in mean scores between the micro-planning (91,767) and vaccination data (73,857) on a number of households, was not significant (M = 91,767, SD = 50,889, micro-planning compared to M = 73,857, SD = 38,499, vaccination, p = 0.195) (Table 1).

Fig. 1

Trend of Households (HHs) in 23 LGAs of Kaduna State during Enumeration of Households and Children (Micro-planning Aug 2017) and Jul 2017 Supplemental Vaccination (SIAs).

Table 1

A comparison of Mean, Standard Deviation and P-value of Households, under 1 year, under 5 years and under 15 years of Micro-planning, Supplemental Immunization Activities (SIAs) and Census Data, 2017.

	Mean	SD	P-value	r
Households
Micro-planning	91,767	50,889	0.195	0.92*
SIAs	73,857	38,499
<1 year
Micro-planning	26,128	16,828	0.004	0.56*
Census	14,154	4894
<5 years
Micro-planning	102,967	62,405	0.01	0.85*
SIAs	157,716	72,212
<15 years
Micro-planning	204,148	120,941	0.30	0.73*
Census	173,793	60,092

Pearson’s coefficient of correlation at P < 0.05.

The number of <1 year increased to 600,939 (84%) after the households and children enumeration (micro-planning) in August 2017 compared to2017 projected census data of 325,531 except in Jaba, Jema’a, Kaduna North, Kaduna South and Kagarko LGAs (Fig. 2). The study revealed a significant difference in mean scores between the micro-planning (26,128) and census data (14,154) on a number of <1 year (M = 26128, SD = 16,828, micro-planning compared to M = 14,154, SD = 4894, census, p = 0.004) (Table 1).

Fig. 2

Trend of <1 year in 23 LGAs of Kaduna State during Enumeration of Households and Children (Micro-planning Aug 2017) and 2017 projected Census Population.

The number of <5 years decreased from 2,368,248 (35%) after the households and children enumeration (micro-planning) in August 2017 compared to vaccination data of 3,627,459 obtained from the last campaign in July 2017 (Fig. 3). The study showed a significant difference in mean scores between the micro-planning (102,967) and supplemental vaccination data (157,716) on a number of <5 years (M = 102,967, SD = 62,405, micro-planning compared to M = 157,716, SD = 72,212, supplemental vaccination, p = 0.01) (Table 1).

Fig. 3

Trend of <5 years in 23 LGAs of Kaduna State during Enumeration of Households and Children (Micro-planning Aug 2017) and Jul 2017 Supplemental Vaccination (SIAs).

The number of <15 years increased to 4,696,152 (17%) after the households and children enumeration (micro-planning) in August 2017 compared to 2017 projected census data of 3,997,229 except in Jaba, Jema’a, Kachia, Kaduna North, Kaduna South, Kagarko, Kaura and Sanga (Fig. 4). However, the difference in mean scores between the micro-planning (204148) and census data (173,793) on a number of <15 years was not significant (M = 204,148, SD = 120,941, micro-planning compared to M = 173,793, SD = 60,092, census, p = 0.30) (Table 1).

Fig. 4

Trend of <15 years in 23 LGAs of Kaduna State during Enumeration of Households and Children (Micro-planning Aug 2017) and 2017 projected Census Population.

Pearson’s coefficient of correlation revealed a moderate to strong significant relationships between micro-planning, supplemental vaccination and projected census data on number of households (r = 0.92, p < 0.05), <1 year (r = 0.56, p < 0.05), <5 year (r = 0.85, p < 0.05) and <15 years (r = 0.73, p < 0.05) (Table 1).

The estimated cost of conducting six (6) OPV campaigns in 2017 decreased from 8,705,902 USD (34%) using July 2017, <5 years population to 5,683,795 USD with the August 2017, <5 years micro-planning population (Table 2) [29].

Table 2

Estimated Cost of OPV Campaigns Using <5 Denominator from Micro-planning Aug 2017.

	<5 years	Cost of Operations @0.38 USD per child	Cost of OPV Vaccines @0.40 USD (vial = 20doses)	No of OPV Campaigns 2017	Total Cost of Campaigns 2017
Jul 2017 IPDs	3,627,459	1,378,434	72,549	6	8,705,902
Micro-planning August 2017	2,368,248	899,934	47,365	6	5,683,795

Discussion

We found a significant decrease in <5 years in 23 LGAs of Kaduna State after the households and children enumeration (micro-planning) in August 2017. We also found a significant increase in <1 year in 18 LGAs of Kaduna State, with exception of Jaba, Jema’a, Kaduna North, Kaduna South and Kagarko LGAs. We found no significant increase in households and <15 years in all the 23 LGAs (Table 1).

The <5 years population, was overestimated in all the LGAs of Kaduna State. [6], [16], [23]. The more the <5 years, the more the vaccines, pluses, and manpower allocated to the campaign [6], [16], [23]. The reduction in <5 years’ population in 23 LGAs of Kaduna State, may lead to fewer vaccines, pluses and vaccination teams needed to vaccinate the target population (Table 2). The performance evaluation of immunization campaign will be driven by quality and reliable data. Based on the findings of this study, the use of denominators that are more accurate will reduce the strains experienced by vaccination teams. The tendency to overestimate the denominators will significantly reduce. Ultimately, there is a need to use data that is more up to date as it will benefit both the government and development partners especially with dwindling GPEI resources (Table 2). Our results are consistent with unpublished findings from Kano and Sokoto States, Nigeria that have conducted the similar micro-planning process in 2017.

The increase in <1 year population in many LGAs of Kaduna State was not unexpected, as reliance on projected census population for over a decade to compute RI denominator is fraught with errors [7], [15], [23] (Fig. 2). The projected census at 3% growth rate bench mark in a developing country like Nigeria that has over 5% annual birth rate, will inevitably lag behind the true figure [24], [25]. The disparity in birth rate across Nigeria’s vast regions makes it difficult to project the population, as projected figures fail to reflect the population dynamics [25], [26], [27], [28]. The micro-planning process was a mini-census and generated a more reliable <1 year population, which is indispensable for RI planning, implementation and performance evaluation [15], [23]. Our results may give meaning to national RI coverage survey (NICS), which has shown abysmal RI antigen coverages in many states of the nation despite high administrative performance (23). It may also explain surveillance data, as measles and yellow fever morbidity and mortality remain on the rise despite high administrative coverages [8], [9], [25]. The findings in 5 LGAs (Jaba, Jema’a, Kaduna North, Kaduna South and Kagarko) were different, as the projected census population was higher than the data from households and children enumeration (micro-planning). The actual reason is not known, but it may be due to lower total population and lower birth rates.

The rise in households, after the micro-planning process was not surprising, due to the redefinition of a household as a woman and her children (Fig. 1). The rise in households was expected, and 19 out of the 23 LGAs recorded increase in households, while the outliers were Jaba, Kaduna South, Sanga and Zango Kataf. The reasons for this deviation is not clear but may be due to lower rates of polygamy in these LGAs. Allocation of vaccination teams during the campaign is guided primarily by households and <5 years population. Hence reliable household data will greatly assist team distribution and rationalization across wards, LGAs, and states. The result is consistent with unpublished findings in Kano and Sokoto States of Northern Nigeria that has conducted the similar micro-planning process in 2017.

The study revealed an increase in <15 years in 15 out of the 23 LGAs of Kaduna State (Fig. 4). Jaba, Jema’a, Kachia, Kaduna North, Kaduna South, Kagarko, kaura and Sanga LGAs recorded reductions in <15 year population and may be due to lower total population and lower birth rates. A reliable <15 years population will guarantee robust AFP surveillance, as the incidence of paralytic polio is commoner in <15 age group [19]. Our results are consistent with unpublished findings in Kano and Sokoto States of Nigeria that have conducted the similar micro-planning process in 2017.

The micro-planning process, though expensive and tedious has generated variables that will serve as reliable denominators in planning, delivery, and evaluation of RI and SIAs in Nigeria. It will save the government and GPEI partners’ enormous resources, as campaigns will be more cost- effective and driven by accurate data, in the face of polio eradication funding gaps.

We recognized limitations in our work. The demand for resources and manpower were very huge, and many partners and government agencies made great sacrifices to see the micro-planning process through. The micro-planning process suffered in some settlements due to the failure of enumeration teams and supervisors to adhere to the standard operating procedures. Some settlements were not adequately supervised due to the dearth of senior supervisors. Insecurity and difficult geographical terrains were key challenges enumeration teams faced during “walk-through’’. The limitations encountered during the enumeration did not affect the overall quality of the outcome, as >99% of all the settlements were visited by teams with tracking devices.

We recommend micro-planning in all pending states in Nigeria, and other developing counties with issues with denominators for planning, delivery, and evaluation of health interventions. The impact of the microplanning exercise in other public health interventions, for example, malaria control, water and sanitation, and childhood de-worming should also be investigated as it is a veritable tool to extrapolate targets for these interventions. We recommend future use of new technologies such as satellite imaging to track settlements and cross-referencing with microplanning data for accuracy for immunization and other health activities. The process should remain government driven, with the active support of development partners, and data generated should guide immunization activities.

Periodic micro-planning remains a useful tool in addressing gaps in immunization planning, delivery and performance evaluation in developing countries, such as Nigeria with overestimated denominators.

Conflict of interest

The authors declare there is no conflict of interest.

Funding

This work received financial support from WHO Country Office.

Authors’ contributions

The study design, methods, and data collection were by GCU, DMM, CK, NL and TIN, while data analysis and discussion were by GCU, DMM, CK, NL, SI, HI, SO, TIN, AA, AM, RD, AMIN, FB, PMD, HS, and NI. All authors read and approved the final manuscript