KeMU

health system strengthening. In line with the pillar, a Maternal, Newborn and Child Health (MNeCH) Project was initiated by the Ministry of Health and World Vision Kenya in Bamba Division, Kilifi County. The project's goal was to strengthen health systems at the facility and community levels; thus, improve demand for health services, including Antenatal Care (ANC). One of the strategies used by the project to achieve its goal was to strengthen community health referral system. Although studies conducted in various contexts demonstrate the potential of such referral systems in improving ANC indicators, in Kenya, there was a dearth of academic information regarding the subject. This study was initiated to achieve three objective: determine whether providing maternal health education to pregnant women had any influence on ANC attendance in Mirihini (an intervention community in Bamba Division) and Midoina (a control community in the neighboring Bahari Division); determine whether issuing referral advice forms to patients has any influence on ANC attendance in the two communities; as well as establish whether promoting partner support had a significant influence on ANC attendance in the intervention and control communities. This study adopted the static group control design to assess the influence of community health referral practices on ANC attendance in the intervention and control communities. Primary data were sourced from 246 mothers of children aged below 2 years; male and female parents of children aged 2 to 4 years, as well as Community Health Volunteers (CHVs). Quantitative analysis yielded descriptive statistics and cross-tabulations with Chi-square (x2) tests. Qualitative data were transcribed, described and analyzed systematically to reveal themes and patterns. Maternal health education: In Mirihini, there was no significant association between providing maternal health education on the ideal number of ANC visits and women's achievement of optimal ANC attendance (x2 = 1.423, df=1 & a p-value = 0.233); hence, the null hypothesis (H01) was not rejected. In Midoina, a significant relationship between the two aspects was obtained (x2 = 3.109, df = 1 & p-value = 0.078), which led to rejection of the null hypothesis (H01). Referral documentation: In Mirihini, issuing referral documents to pregnant women significantly associated with their achievement of optimal ANC attendance (x2 = 8.308, df =1 & p-value = 0.004). Consequently, the null hypothesis (H02) was rejected. In Midoina, there was no significant association between two aspects (x2 = 0.823, df = 1 & p-value = 0.185); hence, the null hypothesis (H02) was not rejected. Male involvement: In Mirihini, partner counseling significantly associated with women's achievement of optimal ANC attendance (x2 = 14.175, df = 2 & p-value = 0.001), which led to rejection of the null hypothesis (H03). However, in Midoina, there was no significant relationship between the two aspects (x2 = 3.710, df = 2 & p-value = 0.156); thus, the null hypothesis (H03) was not rejected. The intervention and control communities varied in terms of the influence of maternal health education, issuance of referral documentation and partner counseling, on women's achievement of optimal ANC attendance. The study recommends the need for: improved facilitation of CHVs to deliver more information on maternal health; project officers to strengthen supervisory support to CHVs by engaging with MoH to provide reporting materials and ensure consistency of monthly review meetings. Besides, there is need for MoH to involve more male CHVs to influence changes in knowledge and perceptions towards ANC attendance among men.