The study was carried out within Pobè municipality, located in southeastern Benin. Comprising 58 villages and city districts, the municipality is divided into five boroughs. The commune’s population was estimated at 146,835 [9] .

This was a cross-sectional evaluative study carried out in 2021. Two types of target population were included. Primary targets were children aged 0 - 59 months, mothers or babysitters and village community health workers. The secondary targets were the head nurses, the Pobè town hall health focal point, the project’s community facilitator and the Pobè chief medical officer.

The studied variables were composed of a principal component (quality of community-based health interventions to reduce infant and child mortality) and explanatory components: inputs, processes and outcomes (Figure 1) [11] .

For each component, the assessed variable was evaluated as follows: when the practice met the standards, a rating of “1” was assigned. If not, a score of “0” was assigned. The sum of the ratings gave the score obtained, which was expressed as a percentage of the expected score for each component. To assess the main

component, the sum of the scores obtained by the explanatory variables was related to the total expected score.

The percentages obtained for each component were assessed according to a three-level scale: “good, average, insufficient” adapted from Varkevisser’s measurement scale [12] .

If the percentage of achieved score was greater than or equal to 80% of the expected rating, the assessment was “good”. If the percentage of achieved score obtained was between 60% (included) and 80% (excluded) of the expected rating, the assessment was considered “average”. If the percentage of achieved score obtained was less than 60% of the expected rating, the assessment was qualified as “insufficient”.

Data was collected by questionnaire survey, interview using an interview guide (semi-structured interview and focus group), observation using an observation grid, and documentary analysis using a data processing sheet. All these tools were designed by the authors.

The challenge, in this research, is that the “gold standard tool” itself has not been identified. We then appreciate the construct validity based on conceptual frame as there is no criterion for comparison. Since the questionnaire was generated from the conceptual diagram (Figure 1), its construct validity was assumed.

We used Cronbach’s alpha coefficient to appreciate the internal consistency of a set of questionnaire items. Cronbach’s alpha coefficient was 0.81 indicating an acceptable reliability of the questionnaire.

Quantitative data were expressed as mean and standard deviation, whilst qualitative data were expressed as their proportion. Percentages of scores obtained in relation to the expected score were calculated.

The sample size of participants included 46 community health workers, 300 mothers-child aged 0 - 5 pairs, 3 nurses (men), 3 nurses (women), 1 community facilitator from the non-governmental organization Catholic Relief Service, 1 chief medical officer and 1 health focal point of Pobè town hall.

The mean age of the mothers was 28 ± 5.51 years. Among these mothers, 67.7% had never attended school, 60.66% were either farmers or craftswomen, and 93.67% were married. Half (50%) of the community health workers were aged between 21 and 30, and had a level of secondary school education (56.32%). Men predominated, accounting for 65.21% of them.

The characteristics of the surveyed mothers and children aged 0 - 59 months are shown in Table 1. The mean age of the nine other participants including 6 nurses, 1 community facilitator, 1 chief medical officer and 1 health focal point of town hall was 42 ± 4.503 years.

Table 2, shows the inadequacy of the structure (53.28%), linked to the insufficiency inputs other than “malaria control inputs” such as of number of households with improved latrines (42%), number of households with a source of drinking water (31%) and number of households with a household solid waste management system (20.33%).

However, in Table 3, the quality of the process was good (81.80%). It should be noted that the population’s participation in educational sessions was low (data not shown).

Outcomes shown in Table 4 were average (66.33%) linked to weak performance rate of children vaccination (65.33%), exclusively breastfeeding last child (41.33%) using an improved latrine by households (58.67%), recognizing at least two danger signs in children by mothers (56%).

Overall, the quality of interventions was average, with a performance rate of

LLITNs: Long-Lasting Insecticide-Treated Mosquito Nets.

ACT: Artemisinin-based Combination Therapy; RDT: Rapid Screening Test.

73.80%. The weakest aspects of implementation were the “structure of community-based PIHI” and its “results”, which were rated “insufficient” and “average” respectively (Table 5).

Table 5 shows that the quality of community-based health interventions within Pobè municipality was average.

Several evaluative studies of community-based health interventions have been undertaken around the world, though it should be noted that the aspects assessed varied from one study to another. Most published studies have assessed the quality of community health workers’ services by direct observation of clinical encounters and by comparison with an evaluation of reference standards. Thus, at the end of the present study, the quality of interventions was average, with a score of 73.80%. The weakest links in implementation were “non-ILP inputs” and “outcomes”, which scored “insufficient” and “average” respectively. This result is below that of Gahou, who found the quality of care to be good at So-Ava in 2018, with a score of 84.09% [13] . Despite this average performance, the gains made must be maintained and new efforts made to improve children’s health throughout the municipality.

Every intervention’s success depends on the quality of the resources mobilized and the managerial style established. The average performance obtained in the evaluation of interventions in the municipality was mainly due to the acceptance of the community health workers in their locality, to the availability of malaria control inputs and above all to the supervision, ongoing training and motivation of the community health workers.

Gray et al. in his technical folder for the forum on community management of acute malnutrition entitled “community engagement” recognized the need to motivate community health workers but mentioned that it was important to stress that remuneration was not the only solution and that, on its own, it would not be able to maintain staff motivation or the quality of care [21] .

To improve performance, it will be necessary to maintain the gains made by mobilizing additional financial resources through the communal council to ensure the sustainability of activities, but also and above all to encourage greater community involvement. According to Gray et al., it is through a good knowledge of their health needs and their active participation in the provision and management of health services that communities strengthen their ownership of health interventions. Most communities have some form of local organization or structure capable of expressing the needs of the community. Despite this, the process was good, demonstrating the ability of community health workers to work in a context of limited resources. In Malawi, for example, a key factor in the successful scaling-up of community-based care was the creation of village health committees: these were linked to each village health center, and community leaders were put in charge of their management [21] .

In short, all stakeholders will need to work together to find solutions to the shortcomings noted above all in the non-availability of inputs other than “malaria control inputs”, and the low level of participation by the population in educational sessions designed to enable them to adopt behaviors conducive to reducing mortality and morbidity in children under 5.

The limitations of the present study lie in its limited scope. The results can only be generalized to the population of the commune of Pobè. Furthermore, desirability bias is possible since mothers know the best practices to be observed to improve their children’s health. Answers could be modified accordingly. However, the results reported in this community-based study are very useful for the implementation of PIHI.