Preventive Health Service Coverage Among Infants and Children at Six Maternal-Child Health Clinics in Western Kenya: A Cross-Sectional Assessment

Despite the substantial reduction of child mortality in recent decades, Kenya still strives to provide universal healthcare access and to meet other international benchmarks for child health. This study aimed to describe child health service coverage among children visiting six maternal and child health (MCH) clinics in western Kenya. In a cross-sectional study of Kenyan children who are under the age of 5 years presenting to MCH clinics, child health records were reviewed to determine coverage of immunizations, growth monitoring, vitamin A supplementation, and deworming. Among 78 children and their caregivers, nearly 70% of children were fully vaccinated for their age. We found a significant disparity in full vaccination coverage by gender (p = 0.017), as males had 3.5 × higher odds of being fully vaccinated compared to females. Further, full vaccination coverage also varied across MCH clinic sites ranging from 43.8 to 92.9%. Health service coverage for Kenyan children in this study is consistent with national and sub-national findings; however, our study found a significant gender equity gap in coverage at these six clinics that warrants further investigation to ensure that all children receive critical preventative services.


3
Background Maternal-child health (MCH) clinics are the primary source of healthcare in Kenya for infants. These clinics provide services to prevent and treat diseases, including immunizations, vitamin A supplementation, deworming, and growth monitoring, as well as serve as a source of educational knowledge about health-related topics. Preventative service coverage, such as immunizations, vitamin A, and deworming, are reported on national and regional levels within the Kenya Demographic and Health Survey (KDHS), most recently published in 2015 (Kenya National Bureau of Statistics, 2015). However, for individuals working on a community-level, understanding the service coverage estimates at local health facilities helps to provide context for the national-and sub-national-level data.
The objective of this study was to determine the coverage of services and interventions provided to infants and children at six maternal and child health clinics in western Kenya in order to understand service coverage and relevant factors at the subnational level.

Methods
This cross-sectional descriptive study was performed at six MCH clinics located in towns within Bungoma, Nandi, Trans Nzoia, and Uasin Gishu counties located in western Kenya, including Webuye, Mosoriot, Kitale, Burnt Forest, Turbo and Eldoret. In each of these town, there is a primary government-funded clinic, which works in close collaboration with an institutional partnership: Academic Model Providing Access to Healthcare (AMPATH). This institutional partnership is led by Moi University School of Medicine and Indiana University and aims to improve delivery of healthcare services in western Kenya (Einterz et al., 2007). This existing partnership and their close collaboration with the local MCH clinics throughout the region provided the rationale for the selection of the study setting. This study was approved by the ethical committees of both Indiana University and Moi University.
During the period of November 7, 2016-December 7, 2016, a research assistant attended a full pediatric clinic day at six MCH clinics, which are only open during weekdays. While pregnant women may attend the clinic on any day of the week, each clinic dedicated 1-2 weekdays per week to seeing children. Each respective clinic manager determined which pediatric clinic day was attended by the research assistant. Caregivers were recruited for participation at each of these six MCH clinic locations using convenience sampling. A research assistant approached all caregivers meeting the following inclusion criteria: bringing a child under the age of 5 years to be seen and identifying their current location as their primary MCH clinic. Every present caregiver who met inclusion criteria was recruited. A total of 78 caregivers consented and agreed to participate in the study. Only one recruited caregiver declined participation. For each caregiver, the child brought to clinic was also included in the study. Only retrospective data were collected from the children's medical record.
The Mother and Child Health Booklet (MCHB) was the primary data source for this study. The MCHB contains the services and interventions received by the child at the MCH clinic. The MCHB, which is the primary health record for the mother and her child, is brought to every visit with a healthcare provider. Healthcare providers record services and interventions provided at each visit in the MCHB.
Data were collected via brief oral interviews with caregivers and review of the MCHB were utilized. Structured oral interviews captured demographic data, such as their relationship with the child, age, and the presence of other children within the household. Interviews were conducted in either English or Kiswahili. A research assistant was trained to ask questions and to categorize the responses into the pre-assigned answer choices. A free text option was available if the study team member found that available categories were not appropriate. Retrospective data collection was performed to ascertain health services related data for each child. These data were directly collected from the child's MCHB. A research assistant reviewed the MCHB for immunizations, vitamin A supplementation, deworming, and growth monitoring. For immunizations, a research assistant also recorded whether the child received each dose in each vaccine series and time point or visit for vitamin A supplementation, deworming, and growth monitoring. For all health services data, eligibility for health services was verified by referencing the age of the child, presence at MCH clinic, and timing of service delivery.
Following the conclusion of data collection, health services records extracted from the MCHBs and responses from the basic caregiver questionnaires were entered into a Microsoft Excel file. For immunizations data extracted from the MCHBs, variables were created for each vaccine series to determine whether each child was up-to-date for each vaccine, as defined for each vaccine as the eligibility of child receiving every dose in the series based on age. For example, if a child was 12-weeks old and she received the first three doses of the oral polio vaccine (OPV) (at birth, 6 weeks, and 10 weeks), she was determined to be up-todate on OPV despite not receiving the last dose (given at 14-weeks). If a child under the age of 5 years received all vaccines and was up-to-date on all series in the immunization schedule, she was considered up-to-date on all vaccines 1 3 and fully vaccinated for her age. Similar variables were created for vitamin A supplementation, deworming, and growth monitoring of weight. Height was not accounted for in the growth monitoring variable because only 30% of children had any height measurements recorded.

Statistical Analysis
Descriptive statistics were used to analyse the responses from the caregiver questionnaire and the proportion of children who received health services and those up-to-date (having received all services for which they were eligible). These health services included immunizations, vitamin A supplementation, deworming, and growth monitoring. Additionally, a drop-out rate was calculated for the pentavalent vaccine, which is the proportion of children receiving the first dose in the series but not the third if eligible. Pentavalent drop-out rates are a common way to measure access to services and the capacity of health systems (i.e. MCH clinics) to provide services that require multiple visits (Baguune et al., 2017). Pearson's chi-square test was applied to determine significant differences in the proportion of males and females up-to-date on all vaccines, vitamin A supplementation, deworming, and weight measurement. Similarly, chi-square was also performed to determine differences in proportion of children fully immunized for age between households with or without other children at home, those who travelled more than 30 min to clinic and those who travelled less, different methods of transport, and those who reported barriers in accessing services at the clinic. All analyses were conducted using SPSS (version 24) (IBM, 2016).

Results
The background characteristics for both the caregivers and children are presented in Table 1. The majority of caregivers were mothers of the children brought to clinic (98.7%), while one caregiver identified as household help. Nearly half of all caregivers (46.2%) were between the ages of 25-29 years, and 61.5% had more than one child. Among the children included in the study, gender was evenly distributed (50%) and the median age was 15.0 weeks (range = 1-104 weeks). Only four children were over the age of 12 months. MCH clinics in Turbo and Mosoriot had the most participants (n = 17 and n = 16, respectively) followed by Kitale (n = 14), Burnt Forest (n = 11), Eldoret (n = 10), and Webuye (n = 10).
Over 80% (n = 64) of children were first brought to clinic within the first 28 days of life. The most common forms of transportation used to reach clinic were motorbike (38.5%) and walking (37.2%), while the remaining caregivers used matatus, which are privately owned share taxis. The most common reasons for bringing their child to clinic were weight checks (62.8%), immunizations (53.8%), and routine health monitoring (39.7%). Only 20.5% indicated vitamin A supplementation as a reason and 10.3%, listed seeking treatment for illness. When asked what types of services their MCH clinic offered, most caregivers (98.7%) listed growth monitoring, while 48.7% and 37.2% listed immunizations and nutrition, respectively. Only 7.7% listed vitamin A supplementation as a service provided, and two caregivers 1 3 (2.3%) said their MCH clinic offered educational health talks and teaching for mothers.

Immunizations
Individual dose coverage for all vaccines ranged from 84.4 to 100%. The third dose of OPV and second dose of measles vaccines had the highest coverage at 100%, while the first dose of rotavirus vaccine had the lowest (84.4%). Out of 15 total doses of vaccine given, only four doses had coverage below 90% (Fig. 1). The drop-out rate for the pentavalent vaccine was 6.7%. Among all vaccines incorporating a 10-week dose in their series (OPV, pentavalent, PCV, and rotavirus), the dose with the highest coverage was the 10-week for all vaccines, regardless of whether it was the second (pentavalent, PCV, rotavirus) or third (OPV) dose in the series. After accounting for varying child age and eligibility for vaccines, the proportion of children up-to-date on each vaccine was calculated (Table 2). BCG vaccine given at birth had the highest coverage of all vaccines (94.9%), while PCV and rotavirus had the lowest coverage (each 82.8%). The proportions of children up-to-date on pentavalent, OPV, and measles were 92.2%, 85.9%, and 88.9%, respectively. A lesser proportion of females were up-to-date on all vaccines, with the exception of BCG and measles, which had the same coverage as males. Across all MCH sites, Mosoriot was the only site where coverage fell below 80% for any single vaccine. At this site, less than 80% of children were up-to-date on OPV (68.8%), PCV (66.7%), and rotavirus (66.7%).
Slightly less than 70% (69.2%) of children were up-todate on all vaccines and fully vaccinated for their age. The only demographic factor found to be significantly associated with up-to-date vaccination status was gender, as males showed 3.5 times higher odds of being up-to-date for all vaccines compared to females [95% CI 1. 256, 9.936]. More than 8 out of every 10 males (82.1%) were up-to-date on all vaccines compared to just 56.4% of females. Differences in the proportion of children up-to-date on all vaccines were also found between MCH sites (Table 2). Among all MCH sites, Kitale had the highest proportion of children up-to-date on all vaccines and fully vaccinated for their age (92.9%). In Eldoret and Webuye, 80.0% of children were up-to-date on all vaccines, while less than two-thirds were in Turbo (64.7%) and Burnt Forest (63.6%). Mosoriot had the lowest proportion (43.8%).

Vitamin A and Deworming
Only 35.9% (n = 28) of children were eligible for vitamin A supplementation (VAS) ( Table 3). Of these, over threefourths received VAS at 6 months of age, while half received it at 12 months of age. All children received VAS at 18 and 24 months of age. Twenty-two (78.6%) of children received every dose of VAS for which they were eligible. VAS coverage was similar between males (80.0%) and females (77.0%). Turbo and Burnt Forest were the only MCH sites where a child did not receive VAS. Less than half (3/7) eligible children in Turbo received VAS in Turbo, compared to just over half (3/7) in Burnt Forest. Of the six children eligible to receive deworming medicine at 12 months of age, 2 (33.3%) received this medicine. At 18 months of age, none of the three children eligible received deworming. Only one child was eligible for deworming at 24 months, and she received the medicine.

Growth Monitoring
Only 33% (n = 26) of children had a birthweight recorded, while 83.1% of children had a weight measurement recorded at 6 weeks of age. The majority (61.5%) of children had a weight measurement recorded at all MCH visits. Two out of three males (66.7%) had a weight recorded at every visit compared to 56.3% of females. Across MCH sites, Eldoret and Kitale had the highest proportions of weight recording, 100.0% and 90.0%, respectively, while Burnt Forest (22.2%) and Turbo (30.8%) had the lowest. Only 30.8% (n = 24) of children had a height measurement recorded at one or more MCH clinic visits. None had a height measurement recorded at every MCH visit.

Discussion
This study aimed to assess the coverage of health services and interventions provided to infants and young children at six MCH clinics located in western Kenya. In this small cross-sectional study, we found that immunization coverage rates were consistent with regional data in Kenya's national surveys. However, these rates are still below the necessary coverage rates for adequate herd immunity. In addition, we found a gender gap and an inequitable distribution of vaccine coverage across MCH clinics in that males had higher coverage for immunizations and nearly all other services included in the study compared to females.
Nearly 70.0% of children were up-to-date on all vaccines, which is comparable to national estimates in Kenya reported by the most recent KDHS (Kenya National Bureau of Statistics, 2015). The KDHS reports 74.9% of Kenyan children are fully vaccinated. Individual vaccine coverage found in our study is also comparable to national estimates as the difference in coverage is within 5% for each vaccination (Kenya National Bureau of Statistics, 2015). Subnational vaccine Table 2 Number and percent of children up-to-date* on vaccines by gender and MCH site (n = 78) The bolded numbers within the row indicate the total number of individuals who received that specific vaccine *"Up-to-date" is defined as "up-to-date for the child's age"   ( (Berger, 1999). The drop-out between the first and third pentavalent vaccine coverage for all MCH clinics was only 6.7%. This dropout rate is consistent with other studies in Kenya (Maina et al., 2013), and suggests that the six MCH clinics studied likely deliver effective immunization services as a whole. However, national and subnational data sources typically use population-based sampling rather than clinic-based, which was used in this study. Recruiting caregivers and their children from the clinic, a population already accessing services, could explain the higher vaccination coverage found in this study. The vaccination coverage rates for the six MCH clinics combined distort the poor coverage in individual clinics, as vaccination coverage varied significantly from site to site. In Mosoriot, less than half of children were considered fully vaccinated for their age. Further investigation is needed to determine factors related to poor vaccination coverage in certain MCH clinics.
Other studies in similar settings found that wealth (Figueiredo et al., 2016;Kawakatsu & Honda, 2012;Kawakatsu et al., 2015), caregiver knowledge of immunization services (Kawakatsu & Honda, 2012), fewer children in the household (Calhoun et al., 2014), skilled birth attendance (Figueiredo et al., 2016;Ushie et al., 2014), high health worker performance (Kawakatsu & Honda, 2012), literacy (Kawakatsu & Honda, 2012), and parental education (Calhoun et al., 2014) was associated with higher vaccination coverage. In this study, factors such as a child's gender, age of caregiver, number of siblings, type and duration of transportation, and clinic site were assessed; however, only gender was found to be associated with vaccination rates. The use of inferential statistics within this small sample size increases the risk of a Type II error and may have influenced the identification of associations within this study.
The finding that males in our study were more likely to be fully vaccinated than females is not consistent with recent KDHS reports and some studies, which showed no difference in coverage rates between males and females (Kawakatsu et al., 2015;Kenya National Bureau of Statistics, 2015;Mutua et al., 2016). These studies were conducted at the national level (Kenya National Bureau of Statistics, 2015), in an urban setting (Mutua et al., 2016), and in rural western Kenya (Kawakatsu et al., 2015). In our study, we found that males showed over 3.5 times the odds of being fully vaccinated for their age compared to females. A recent study conducted in Nairobi, Kenya also supported the presence of an immunization gender gap, finding that only 65.2% of females were fully vaccinated compared to 73.2% of their male counterparts (Egondi et al., 2015). Similar to the gender disparity found in vaccination coverage, a lesser proportion of females received vitamin A supplementation and growth monitoring. While these differences were not as stark as the gender gap in vaccination coverage, they still raise concern and could indicate that female children do not have the same access to preventive health services as male children. Additional research is needed to ascertain whether caregivers are seeking the same care regardless of their child's gender in order to develop strategies to improve vaccination coverage in this region of Kenya that target female children to achieve equitable coverage. In future studies, the high coverage rate for vaccines and growth monitoring at 10-weeks of age will be considered as a highly compliant visit in MCH clinics. The completion rate of the 10-week dose was the highest among all vaccines in the series. In addition, the largest proportion of children receiving growth monitoring was at 10-weeks of age as well.
Vitamin A supplementation coverage among the children in the study was higher than findings from the most recent KDHS (Kenya National Bureau of Statistics, 2015). According to this KDHS, 67.9% of children ages 6-8 months received vitamin A supplementation compared to 78.6% of children receiving vitamin A supplementation within this study at their 6-month visit. Coverage was similar among males and females within the KDHS as well (Kenya National Bureau of Statistics, 2015). Deworming coverage was lower in the sample than national estimates. Only one-third of children in the present study received deworming medicines in the last 12 months compared to 51.0% estimated in the most recent KDHS (Kenya National Bureau of Statistics, 2015). Due to the youth of this study's sample, it was difficult to accurately measure and compare vitamin A and especially deworming coverage rates to the national levels. The mean age of this sample was just over 5 months and children are not eligible to receive vitamin A supplementation and deworming medicines until 6 and 12 months of age, respectively. Additional data from an older sample of children are needed to provide an estimation of service vitamin A supplementation and deworming coverage.
Primary care facilities, such as the MCH clinics, are crucial for strengthening service care delivery for children and their mothers. Overall, the results of our small cross-sectional study of facilities were relatively consistent with the national sources of population-based data, showing that immunization coverage is not at rate required to ensure herd immunity for diseases such as measles. While these facilitates are charged with ensuring immunization 1 3 coverage for children within their catchment areas, challenges arise in achieve this goal. A 2016 survey of Kenyan health facilities found high rates of missed opportunities for vaccinations within vaccine eligible children attended MCH clinics, due to limited training by staff and issues with vaccination supply (Li et al., 2016). A qualitative study of healthcare facility managers in western Kenya perceived themselves to be overworked and unmotivated to expand immunization access, with low supportive supervision of their clinical staff (Chesoli et al., 2018). While additional resources and support are needed for these facilities to optimize the care they deliver, other methods of expanding service coverage have been used in Kenya and globally. The COVID pandemic has highlighted the importance of community health workers in delivering care to individuals within the community, with an increased numbers of immunization campaigns throughout the country during the pandemic to ensure adequate service coverage (UNICEF, 2021a(UNICEF, , 2021b. Policy makers must utilize a multiprong approach to ensure optimal coverage for critical preventative services, such as immunizations.
A limitation of this study is that it was subject to sampling bias as participants were recruited at the MCH clinic. Only caregivers and children who were already present in the MCH clinic were eligible for inclusion in the study. Recruiting participants at the clinic excludes caregivers and children who do not have access to services, encounter barriers in accessing preventive health services, or are disengaged from care for other reasons. This sampling could potentially lead to an overestimation of service coverage because we were unable to account for the children not receiving any services. Another limitation to this study is the MCH clinics involved in this study were at the primary, governmentfunded health facility located within each of the six towns. Thus, our study would have missed children who attended private MCH clinics for their care. Additionally, recruiting at each MCH clinic site for just one day could potentially lead to a non-representative sample.
Another limitation was that the sampling strategy might be potential source of overestimating service coverage. For instance, it was difficult to determine whether the caregiver was just arriving or just leaving the clinic at the time of recruitment. Further, when reviewing the Mother and Child Health Booklet, it was assumed that all services were provided on the visit. Therefore, systematic bias was possible within the data source.
Other limitations that could lead to inaccurate coverage estimates include the age of the children in the sample and the sampling strategy. Due to the majority of the children in the sample being under 1 year of age, it was not possible to determine vaccination coverage in terms of the number of children receiving all required doses. Few children were old enough to have even been eligible to receive all vaccines, so coverage was defined by receipt of eligible vaccines. This could overestimate vaccine coverage because we were unable to account for potential future missed doses. Finally, the small sample size used in this study restricts the representativeness and generalizability of the results.

Conclusions
Monitoring local coverage of preventive health services for children is a critical component for achieving the Sustainable Development Goals and improving health outcomes and survival among children in Kenya and similar settings in sub-Saharan Africa. We found a significant gender disparity in vaccine coverage, as females in our study had lower odds of being fully immunized compared to males. Further, females were also found to have lower coverage of vitamin A supplementation and growth monitoring. The gender gap found in this study needs further investigation to elucidate what factors are associated with females having significantly lower vaccination coverage than males. This cross-sectional study was small in nature and consequently, limited in its generalizability. However, it is a critical step in understanding how well national and regional data reflect clinic-level data on health service coverage and the factors that impact coverage. This will ultimately get us closer to realizing the Sustainable Development Goals and improving the health of young Kenyan children.
Ethical Approval This study was approved by the ethical committees of both Indiana University and Moi University.