3.3.2. Self-Help Groups

Self-help group (SHG) members were primarily interested in the delivery of childcare services and suggested they may provide pick up and drop o ff services for children to and from swimming and childcare interventions. Some SHGs were already involved in delivering governmen<sup>t</sup> programs, such as the mid-day meal scheme in schools. They expected to be paid for involvement (Ref 38 in Table S2).

Many households in communities had at least one SHG member, making them well connected. They would be able to support community engagemen<sup>t</sup> activities, such as through organising mothers' meetings and household visits (Ref 39 in Table S2).

Some possible barriers for the engagemen<sup>t</sup> of SHGs were identified. Firstly, many were busy with their family businesses and may have minimal time to be engaged. Secondly, some were concerned about their lower levels of education and stressed the need for comprehensive training. Lastly, some SHGs others faced challenges with the engagemen<sup>t</sup> of all members. The supervision of SHGs also varied and the managemen<sup>t</sup> of SHGs involved in drowning intervention delivery may require a separate system (Ref 40 in Table S2).

Community leaders and grassroots organisation participants noted that SHG members were easier to engage in drowning interventions than Anganwadi centres or ASHAs as they required fewer governmen<sup>t</sup> permissions. However, some SHG members may face restriction from their husbands or families due to cultural constraints on women's mobility and employment (Ref 41 in Table S2).

### 3.3.3. Anganwadi Centres (ICDS Program)

Anganwadi centres were considered possibly suitable for the implementation of childcare supervision and parent engagemen<sup>t</sup> activities. Centres were usually open from 7 a.m. to 9 or 10 a.m. with 20–30 children attending each day. Some centres already provided a limited range of childcare activities, and parents left their children for 1–2 h with the Anganwadi centre.

However, there was grea<sup>t</sup> variability described and observed in the quality of services. Participants reported that many Anganwadi centres only provided food and no childcare services. This may be due to the lack of an appropriately enclosed venue, lack of training for Anganwadi workers and parents' low trust in the centre. In two out of three of the Anganwadi centre observations, the Anganwadi worker did not facilitate any games or activities. Many participants also complained of a lack of educational materials and repeating activities (Refs 42 and 43 in Table S2).

In addition, many venues lacked toilets and water and children were left alone if a child was taken to relieve themselves. Many participants reported that Anganwadi venues did not have enough space for both cooking and childcare activities and were not safely enclosed. A barrier to finding appropriate venues was that the local governmen<sup>t</sup> requested private land to be leased for 50–100 years for the centres, which few people agreed to. Parents also did not always have time to pick and drop their children, especially if the centre was at a further distance from their home (Ref 44 in Table S2).

Anganwadi workers were also burdened with their duties and had limited training. Anganwadi workers had other responsibilities such as conducting surveys for the Department of Health on sanitation and maintaining the registers of children. They were often busy until 12 p.m. after the centre closed at 10 a.m. They also struggled to cook, clean and provide childcare activities at once. Many centres did not have an Anganwadi assistant allocated or regularly attending. Anganwadi workers also reported being unsatisfied with the pay (Refs 45 and 46 in Table S2).

Anganwadi workers were trained when they joined the program, but the training did not cover ECE activities in detail. They were provided limited ongoing support, where meetings with Panchayat officials who oversaw the implementation of ICDS, visits from supervisors and block-level o ffices were infrequent (Ref 47 in Table S2).

Parents also complained that food was of inadequate quantity. Improper food provision meant many parents had lost trust in the Anganwadi centres. Anganwadi workers and community leaders

stated that poor food quality was due to resourcing issues such as insufficient money provided for ingredients amidst rising prices and a lack of water and sanitation in the venues (Ref 48 in Table S2).

Making changes to Anganwadi centres at a local level required permissions from both Health and Women and Child Development representatives at the block level. Block-level representatives (the level of governmen<sup>t</sup> just above Gram Panchayats) are responsible for monitoring program performance. Although Gram Panchayats are responsible for the program implementation of ICDS, they do not have the permission to make operational changes as their targets and delivery requirements are set by State policy and enforced by block-level supervisors. Grassroots organisation and community leader participants noted that engaging block-level representatives may be challenging without higher state-level permissions which may take months to obtain. Grassroot participants had experienced that governmen<sup>t</sup> departments were cautious about giving permissions when liabilities were not clear. These participants stated that running a parallel program for childcare may be easier than using the ICDS (Refs 49 and 50 in Table S2).

A few communities had parallel NGO-run childcare programs which children attended after visiting the Anganwadi centre. These programs were considered of better quality than Anganwadi centres (Ref 51 in Table S2).

### 3.3.4. Other Community Programs

Local youth clubs were identified by many participants as potential implementers of programs. These clubs were organisations run by youth and overseen by Gram Panchayats. They aimed to engage young people in self and community development activities (Ref 52 in Table S2).
