Background: Kenya is faced with a triple burden of malnutrition which is multi-faceted with health and socio-economic implications. Huge geographical disparities exist, especially, in the arid and semi-arid lands exacerbated by inadequate resource allocation to the nutrition sector and challenges in multi-sectoral coordination and nutrition governance. UNICEF’s Maternal and Child Nutrition Programme is a four-year (2018–2022) resilience-building, multi-sectoral program focused on pregnant and lactating women, mothers of children under five years and children under five years. The objective of the mid-term evaluation was to establish the relevance, effectiveness, efficiency, and sustainability of the programme. Methods: The field evaluation conducted between June and July 2021, adopted a concurrent mixed-methods approach, where qualitative information was gathered through 29 key informant interviews and 18 focus group discussions (6 FGDs per population group; women of reproductive age, adolescent girls and men). Quantitatively, data were obtained through desk review of secondary data from programme reports, budgets, and project outputs where descriptive analysis was undertaken using Excel software. Qualitative information was organized using Nvivo software and analyzed thematically. Results: The findings provide evidence of the relevance of the Maternal and Child Nutrition Programme II to the nutrition situation in Kenya and its alignment with the Government of Kenya and donor priorities. Most planned programme targets were achieved despite operating in a COVID-19 pandemic environment. The use of innovative approaches such as family mid-upper arm circumference, integrated management of acute malnutrition surge model, Malezi bora and Logistic Management Information Management System contributed to the realization of effective outputs and outcomes. Stringent financial management strategies contributed toward programme efficiencies; however, optimal utilization of the resources needs further strengthening. The programme adopted strategies for strengthening local capacity and promoting ownership and long-term sustainability. Conclusion: The programme is on track across the four evaluation criteria. However, a few suggestions are recommended to improve relevance, effectiveness, efficiency, and sustainability. A formal transition strategy needs to be developed in consultation with multi-stakeholder groups and implemented in phases. UNICEF Nutrition section should explore a more integrated programming mode of delivery through joint initiatives with other agencies under the Delivery as One UN agenda, along the more gender transformative approaches with more systematic involvement of males and females in gender-based discussions.
This midterm evaluation was undertaken after two years of MCNP II implementation and aimed to evaluate the relevance, effectiveness, efficiency, and sustainability of MCNP II based on the Organization for Economic Cooperation – Development Assistance Committee/United Nations Evaluation Group [10] criteria. The evaluation also identified key successes and lessons learned and covered aspects of gender, human rights, and equity sensitivity of the program. The evaluation was executed using a non-experimental concurrent mixed method approach. The quantitative data on key indicators from the programme’s result framework were collected through a desk review of programme reports and relevant documents detailed under the data collection section. Further, The Theory of Change Additional file 1: Annex 1 was used as a guide for the logical relationships between strategies, activities, and the results chain. Before the evaluation, a comprehensive review of processes and approaches was undertaken to understand the strengths and gaps in programme implementation and complement the evaluation findings. The [10] evaluation matrix used included key evaluation questions, sub-questions (probes), primary and secondary key indicators and data sources. A comprehensive mapping of the relevant stakeholders was done to understand their role in the program. Following this, purposive sampling was used to identify stakeholder groups and key informants, involved in the program implementation. Key stakeholder groups included government ministries and departments (Ministry of Health (MOH), Division of Nutrition and Dietetics (DND), Ministries of Education, Livestock, Agriculture and Fisheries, Labour and Social Protection, Treasury and Planning); implementing partners, County Departments of Health, donor agencies and private sector organizations; UNICEF representatives including decision-makers involved in program planning and design and field teams including zonal officers and nutrition support officers (NSOs); and the communities in which the program was implemented. Multi-stage cluster sampling was used to identify counties, sub-counties, and recruitment of participants for the beneficiary field study. Three counties—Kitui, Isiolo and Turkana (Fig. 1), were purposively selected from 13 program counties based on the intensity of MCNP II, levels of malnutrition, UNICEFs investment, livelihood cluster, UNICEFs field presence, partner presence, access and characteristic of the region—arid or semi-arid. Selection of counties and sub-counties Mapping of sub-counties in each of the three selected counties was conducted using Kenya National Bureau of Statistics (KNBS) data. The selection of sub-counties was based on the intensity of MCNP II and performance of Integrated management of acute malnutrition (IMAM) program indicators and Vitamin A supplementation (VAS) coverage. Based on these criteria, Kitui Central (Kitui), Isiolo sub-county (Isiolo) and Turkana Central (Turkana) were selected based on poor programme performance while Mwingi West (Kitui), Garbatulla (Isiolo) and Turkana South(Turkana) (Fig. 1) were selected based on better performance. In each sub-county, key community groups were recruited based on their influence on nutrition and health-seeking behaviour. They included—Community health volunteers (CHVs); Community health extension workers (CHEWs); Community leaders; Mothers of children below 5 years of age; Pregnant and lactating women; Adolescent girls and Fathers/Males/Household influencers. The sample size was determined based on an assumption that saturation of information will be achieved through this sample size. The sample for qualitative design is based on the premix of saturation. Training the teams on data collection tools and evaluation matrix was conducted in two phases: Phase 1: A two-day training for the evaluation team including qualitative researchers and note-takers to undertake key informant interviews with the stakeholder groups and Phase 2, one-day training for the beneficiary field study. In both phased training, a team of six were trained on evaluation tools, probing techniques, evaluation questions for key informants in the community and how to conduct focus group discussions with the beneficiaries, ethical considerations, field-level practicalities, probing techniques and note-taking. UNICEF team and members from the Expert Review Group constituted by the MoH also participated in phase 1 training as observers. the tools were pre-tested with different participants, not part of the study. In both pieces of training, debriefing sessions to discuss the flow of questions, challenges in eliciting responses and probing were undertaken. The training were conducted at the IQVIA Nairobi Office. A comprehensive desk review of key MCNP II programme documents was conducted to understand the project context, key approaches and the results achieved by the programme. The document included programme-level data sets on nutrition indicators, LMIS, nutrition action plans and budgets, MCNP II progress reports and briefs. Through field visits, primary data using semi-structured interview/discussion guides, across two phases, to capture insights from both demand and supply sides were collected. Qualitatively, a total of 167 participants (55 males and 112 females) were interviewed. At the policy and program implementation and oversight level, the study conducted online 29 in-depth interviews through Microsoft Teams with the key informants from the selected key stakeholder groups. At the community level, 18 face-to-face FGDs (6 per county) were conducted with beneficiaries of the program who included women of reproductive age, adolescent mothers and other decision-makers in the family (including men) at convenient levels majorly health facilities Eighteen (18). In-depth interviews with the key informants from the community including community leaders, health workers and community health volunteers (CHVs). were conducted. At the community level, tools translated into Swahili were used. Each KII lasted between 45 -60 min while the FGDs lasted between one and half hours to two hours. All interviews were audio recorded and notes were taken to capture insights. The interviews were sufficient, and saturation was achieved. COVID-19 public health guidelines were observed as provided for by SMART Interim guidance on restarting population-surveys and household level data collection in humanitarian situations during the Covid-19 pandemic” [9]. All interviews were audio recorded and transcribed verbatim. Where necessary, the translations were undertaken. Stringent quality assurance mechanisms were followed to ensure the quality of data and transcripts. The qualitative data were organized, and the Inductive method was used to code and generate themes and sub-themes using NVIVO software. Quantitative data from the secondary datasets was analyzed using EXCEL. Insights were generated for comparative and trend analysis of results and program indicators. Quantitatively the review assessed the changes between the midpoint and endpoint. Analysed budget allocation versus utilization for both UNICEF funding and implementing partners’ contributions.