2.2.1. Community-Level Participants

Community men, women and leader participants were recruited through convenience sampling. The partner NGO first approached local governmen<sup>t</sup> bodies for permission to conduct the interviews. Data collectors then entered the communities as recommended by the Gram Panchayats (who are the lowest local governmen<sup>t</sup> body representing a group of villages) and engaged local leaders such as ASHA and Anganwadi workers, who introduced the data collectors to possible participants. Participants were required to be parents and usual residents of the community living there for the past three years. Participants were recruited across all 19 blocks of the Sundarbans to ensure a range of perspectives.

These participants provided insights into community acceptance and perceptions towards drowning interventions, as well as possible barriers and enablers to implementation in the local context. A total of ten IDIs and nine FGDs were conducted with community-level participants, with men and women equally represented.

### 2.2.2. Government Program Participants

Anganwadi workers, SHG members and ASHA workers were approached through purposive sampling after entering communities in which permission for data collection had been granted by local governmen<sup>t</sup> o fficials. Communities had one or two ASHA and Anganwadi workers each, so whoever was available upon contact was scheduled for interview. As self-help group members were found in many households, community members would lead data collectors to the closest home of a member. These participants were included if they were active in their respective programs for a minimum of 6 months within the Sundarbans region.

Observations of SHG meetings and Anganwadi centres were conducted to understand their operations. By policy, one Anganwadi centre is required to serve a population of 1000 people, providing any children aged 3–6 years old with early childhood education activities for two hours each day along with a nutritious meal. Each centre should have one Anganwadi worker and one helper, and usually operates between 7 and 10 a.m. We observed the children who came to the Anganwadi Centres and provided insights into how children interacted with the Anganwadi workers, as well as the ground realities of the program delivery. Observations of SHGs identified decision-making methods and revealed the role of SHGs in the community. Anganwadi centres and SHGs for observations were purposively selected in partnership with a local NGO working with these programmes to cover a range of performance levels. Nine governmen<sup>t</sup> program participants (ASHA workers, Anganwadi workers and SHG members) were interviewed. Two FGDs with SHG members were also held and three observations each were conducted at Anganwadi centres and SHG meetings.

### 2.2.3. Grassroots Organisations

Interviews were conducted with the individuals from organisations working in the child health, education, safety or nutrition in the Sundarbans or other similar rural contexts in West Bengal. This provided insights into the considerations and challenges related to delivering grassroots programs in the Sundarbans. Potential participants were introduced to the researchers by our partner NGO and were required to have oversight over program delivery for at least one year. Three representatives from grassroots organisations were interviewed.

### *2.3. Tools and Transcriptions*

Tools for all IDIs, FGDs and Observations were developed before the commencement of data collection and translated into Bengali. The data collection guides were semi- structured to ensure all domains relevant to research questions were covered.

All participantswere also shown a pictorial presenting theWHO-recommended drowning interventions. Barriers were described as any physical object preventing children's access to water such as playpens, door barriers or fencing. Childcare was described as any group-based supervision in an enclosed space. Swimming lessons encompassed both swim and rescue training skills, and first responder training was described as training adults on how to save children if they fall into water or start drowning.
