Study design
SPRINT-Sen is a population health intervention research (PHIR) study using a mixed-methods design [82]. In line with the Medical Research Council (MRC) and PHIR frameworks for complex interventions, we combine effectiveness evaluation with theory-driven evaluation to generate comprehensive evidence for decision-making [83].
Mixed-methods sequence
We employ a mixed-methods approach to elucidate both the what and the how of intervention effects in a heat-stressed, complex social-ecological system [84]. Figure 2 provides an overview of the mixed-methods design across the three key phases of the study. This multiphase approach supports triangulation, contextual understanding, and the identification of change mechanisms.
The study follows an integrated mixed-methods design structured across three complementary and sequential phases. First, an exploratory sequential phase (Objectives 1 and 2) combines qualitative assessment of lived heat experiences with epidemiological time-series analyses to generate baseline evidence. Second, an embedded design (Objectives 3 and 4) integrates quantitative micro-surveillance with qualitative process evaluation, including realist evaluation, during intervention implementation. Third, an explanatory sequential phase (Objective 4) includes quantitative outcome evaluation followed by qualitative interpretation to contextualize findings. Objective 5 (knowledge translation and policy engagement) is transversal and implemented throughout all phases to ensure continuous stakeholder involvement and policy relevance.
Objective 1 will be addressed through an ecological time-series analysis, using retrospective meteorological data and routine health facility records to examine the association between temperature variations and morbidity outcomes among women and children in the Matam region.
Objective 2 will adopt a qualitative exploratory design, including in-depth interviews and focus group discussions with urban and rural community members, including pregnant and breastfeeding women, health professionals, and local leaders. The aim is to understand the lived experiences of EH, identify barriers and coping strategies, and examine the factors that shape heat vulnerability and resilience.
Objectives 3 to 5 correspond to the three interlinked components of the intervention. The overall intervention will be evaluated using a before-after design. A randomized controlled trial was considered but not retained due to the high risk of contamination, the limited number of independent clusters, budget constraints and ethical concerns related to withholding preventive measures in a high-vulnerability setting. The evaluation will also incorporate a realist evaluation approach to explore the mechanisms through which the intervention produces observed outcomes in specific context.
Objective 3 consists of implementing a prospective micro-surveillance system to monitor environmental exposure and informs preventive responses. Objective 4 involves the co-design, implementation, and evaluation of community-based and health-system preventive strategies. The co-design phase will follow a structured co-design approach, including workshops with mothers and relatives, community actors, and health professionals, guided by intervention mapping principles [85] and behavioural change frameworks [86].
For objective 4, we will conduct four repeated cross-sectional surveys (approximately 1,200 women per round) at key time points: baseline during the hot season, baseline during the cooler season, during intervention implementation, and one year after implementation to assess sustained effects. This quasi-experimental design enables comparison across seasonal contrasts and intervention phases, while accounting for variations in environmental exposure and implementation intensity across sites [87,88].
Objective 5 focuses on knowledge translation and policy engagement and is implemented throughout the project as a cross-cutting component.
Study setting and population
The study will be conducted in the Matam region of northern Senegal. Within this region, four main sites in urban and rural settings, covering nearly 50,000 inhabitants, have been selected in the Matam health district, including one located in the urban municipality of Matam town (see Figure 3 below). These sites were selected to capture environmental diversity based on their relative proximity to the Senegal River (near vs. far), to a health centre, and to allow for the enrolment of at least 300 houses per site. Site selection was conducted using random spatial sampling.
Eligibility criteria
The study population comprises pregnant women and mothers of infants under 12 months old living in selected localities of the Matam district. Women are eligible if they are aged 15 or above, are pregnant at any trimester or breastfeeding an infant under 12 months, and can give informed consent in Wolof, Pulaar, or French. Women who are not present at the time of the survey visit will not be included in that specific survey round. If they are younger than 15 years or are unable to provide informed consent in one of the study languages they will be excluded. The lower age limit of 15 years was selected to reflect the epidemiology of adolescent pregnancy in the region. For participants aged 15–17 years, assent and parental consent procedures will be applied in accordance with ethical guidelines.
Sampling
The study is powered to detect a 5 percentage-point increase in the primary outcome — exclusive breastfeeding for infants aged 6 months (yes/no) — between the baseline and post-intervention surveys. Based on an estimated baseline EBF prevalence of 31% in the region [50], and assuming a total sample of approximately 1,200 breastfeeding women across all four survey rounds, this study will achieve >80% power to detect a 10% absolute increase in EBF at a 5% significance level, accounting for design effects and clustering by site. This sample size calculation applies to the evaluation of the preventive intervention (objective 4) and is based on expected changes in EBF prevalence during periods of extreme heat. EBF will be defined according to WHO guidelines [89], based on maternal report of feeding practices in the previous 24 hours (for children under 6 months), since birth, and retrospectively up to 12 months postpartum. This approach allows for the estimation of both point prevalence and duration of EBF.
Household sampling will be based on a thermal aerial image captured by a drone (AMVIC 3T©), which will generate a thermal emission map of the study areas. To account for variability in household thermal environments, the sampling design incorporates stratification by village and roof type, considered a key determinant of indoor temperature exposure. The sampling frame was established through a prior census combining drone imagery, photointerpretation of orthomosaics to classify roof types (primarily zinc/sheet metal and concrete slab), and ground-truth field verification. Household selection will be proportional to the distribution of roof types observed in each locality. The sampling design therefore follows a multi-stage cluster approach (village and household), with stratification by roof type to ensure that variability in household thermal configurations is adequately represented. Each selected household will be equipped with a thermal sensor and included in the indoor temperature monitoring system.
Within each of the four selected sites, an exhaustive household listing will be conducted. All households located within the defined geographic boundaries will constitute the sampling frame for participant recruitment. To reach the required sample size of 1,200 women per survey round, the sampling frame will be expanded to include additional villages within the district. In total, approximately 1,100 households will be monitored, which is expected to yield 1,200 eligible women per round, as multiple eligible women may reside within the same household.
Intervention description
Co-design approach
The co-design process pertains to Objective 4 and aims to develop and implement preventive strategies to enhance exclusive breastfeeding and support maternal mental health during periods of extreme heat. The intervention will address gaps identified through Objectives 1 and 2, including epidemiological evidence of heat-related maternal and child health burdens and qualitative findings on lived heat experiences, barriers to exclusive breastfeeding, and existing coping strategies. Intervention development will follow the Intervention Mapping framework [90] which will guide the systematic needs assessment, identification of behavioural and contextual determinants, and selection of appropriate behaviour change techniques. The Godin framework for health behaviour change [91] will inform the identification of key determinants such as beliefs, perceived risks, social norms, self-efficacy, and behavioural intentions related to breastfeeding and adaptation practices during extreme heat. Two core elements will be co-designed: awareness and capacity building. The intervention will be co-designed with local stakeholders, community members, and health professionals to ensure sociocultural appropriateness [92]. We will adopt a structured two-step approach to meaningful engagement [93], which includes: (i) defining the collaborative arena to prepare for participation and promote active engagement, and (ii) designing involvement mechanisms that prioritize user perspectives and value critical feedback [93].
To operationalize this approach, we will implement contextually-appropriate participatory tools, such as community dialogue circles or photovoice [94]. Across every study phase, we will systematically track participation with the ‘Involvement Matrix’ [79] to ensure participation remains meaningful. Community members will act as “advisors”, aligned with the study design and sociocultural context [79]. Finally, we will assess the co-design phase using the PROSECO framework [95], which offers a structured approach to evaluating co-design processes in public health interventions.
We will pay close attention to avoid reinforcing guilt, pressure, or the normative image of the “good mother”, as negative effects have been documented in EBF campaigns in both Western countries [96,97] and the Global South [98]. A key priority is to avoid idealised maternal archetypes [99], by respecting rather than displacing mothers’ experiential knowledge [100] and by resisting the transformation of exclusive breastfeeding as a ‘moral imperative’ imposed by Northern or international institutions [100,101]. Mothers’ lived experiences will guide the intervention through a co-design process, addressing gaps between knowledge and practice and positioning breastfeeding as a ‘meeting point’ where different forms of knowledge can productively interface [98]. For maternal mental health outcomes, women’s perspectives will likewise be at the centre of the intervention, and we will explicitly avoid both stigma and the medicalisation of well-being. The intervention will systematically integrate local and cultural practices when developing community-based actions [102,103].
Logic model and strategic elements of the intervention
Figure 4 presents the logic model of SPRINT-Sen intervention, which aims to reduce the impact of extreme heat on exclusive breastfeeding and maternal mental health. The programme theory will be developed using a realist evaluation approach. The intervention will be co-designed with community members and stakeholders and is expected to comprise three interlinked components: (1) a heat-health surveillance and early warning system to monitor temperature exposure and generate alerts; (2) the co-design of community-based and health system prevention strategies, including the development of a health warning system; and (3) knowledge transfer through stakeholder engagement and dissemination. Short (within 1- 2years), mid-term (3-5 years), and long-term (beyond 5 years) outcomes are expected to improve health behaviours, reduced health inequities, and sustainable policy integration at local and national levels. Researcher reflexivity will be embedded throughout the process [104–106]. This reflexive approach will help us critically examine our own positionality (gender, social, cultural, racial) and its influence on the intervention co-design and implementation, ensuring that power dynamics and cultural assumptions are continuously interrogated and addressed [107]. Contextual factors—including environmental, economic, cultural, institutional, and political dimensions—may influence the implementation and outcomes of the intervention.
Data collection and analysis
Data collection
Data will be collected at four strategic time points (Figure 5):
baseline 1– heat season, pre-intervention
baseline 2 – cold season, pre-intervention
implementation – heat season, during intervention
follow-up – heat season, 12 months post-implementation
We use an open cohort, meaning that pregnant or breastfeeding women (up to 12 months postpartum) may enrol at any of the four survey rounds and remain in the study until their infant reaches 12 months of age. The endline survey is conducted 12 months after Measure 3 (during the intervention phase) to enable comparison under similar heat-exposure conditions. For women enrolled during the intervention, follow-up may be extended up to 18 months post-inclusion to assess sustained effects on maternal mental health outcomes (Figure 6).
Participants are grouped based on their timing of enrolment relative to the intervention phases:
Subgroup 1: pre-intervention
Subgroup 2: during intervention (up to 6 months)
Subgroup 3: post-intervention (0 and 12-month follow-up)
Subgroup 1 includes women enrolled during the pre-intervention phase, with an approximately six-month interval before intervention rollout to allow for co-design, preparation, and completion of the second baseline survey. The repeated cross-sectional survey is designed to capture breastfeeding practices and the multifactorial determinants of perinatal mental health outcomes, which are largely transitory, as well as factors influencing heat exposure shaped by seasonal patterns and the built environment. This design enables comparative analysis of outcomes from early pregnancy through the first year after birth.
Objective 1. Describe recent heat trends and heat-sensitive morbidity
To assess the association between ambient temperature and health outcomes in women and chidren, we will conduct an ecological time-series analysis using routine health facility data and meteorological records (2020–present). Temperature exposure will be characterized using indicators such as daily mean, minimum, and maximum temperature, as well as heat indices adapted to or constructed for the local climatic context. Health outcomes will be derived from routine health facility data (e.g., daily counts of consultations in health centres). Associations will be estimated using time-series regression models, including distributed lag non-linear models (DLNM), which allow the estimation of delayed and non-linear effects of temperature exposure. Meteorological data will be obtained from national weather stations (ANACIM) and complemented with satellite data, while health data will be sourced from local health centres. Analyses will assess inter-village inequalities. Spatial analyses will identify geographic disparities in temperature exposure and heat-vulnerability hotspots across the Matam district. We will apply spatial clustering methods and stratify by village characteristics (e.g., roof type, vegetation cover, river access, etc.). Time-series models will be used to assess short-term associations between temperature and specific maternal and neonatal outcomes, including maternal mental health consultations, pregnancy-related complications, and neonatal morbidity, while controlling for seasonality and temporal autocorrelation. Attributable fractions will be calculated to estimate the burden of heat-related illness among women and children. Previous studies have quantified the burden of health outcomes attributable to non-optimal temperatures using distributed lag non-linear models and attributable fraction methods [111–114], although evidence remains limited in African settings [115] and for maternal health outcomes [116].
Objective 2. Assess the consequences of EH and identify health adaptation strategies
Phase 1: Contextualisation, baseline and pre-implementation (Months 0 to 11, Measures 1 and 2)
At baseline, the qualitative data collection will aim to understand the determinants of exclusive breastfeeding, maternal mental health burden, and strategies for resilience to extreme heat among mothers, households, and health professionals. We will employ a multiple-case study approach [117] in which the four selected study sites constitute the analytical cases. Each site represents a distinct socio-environmental and health-system context, allowing comparative analysis of households’ adaptation and resilience strategies during extreme heat events. To analyse the local context, we will deploy multiple data collection methods:
Focus group discussions with community members and health workers
Individual in-depth interviews with key informants
Structured observations of health facilities and household practices
Concept mapping to visualize local understandings of heat-health relationships
To address inherent inequalities within villages, the study will focus on local and individual constraints and how different population sub-groups may deploy differential adaptation strategies. The analytical approach will employ a comparative perspective using heuristic tools to identify configurations and their regularities [118]. We will conduct cross-case analysis to “discover patterns of invariance and constant association” [118]. We will adopt a realist evaluation approach to reconstruct the intervention theory using multiple sources, including the logic model, empirical data and a validation workshop. This process will help clarifying assumptions about mechanisms, expected effects, and contextual factors influencing their activation [119]. Qualitative interviews with mothers and health personnel will explore adaptive behaviours to extreme heat and identify pathways for change. Using context–mechanism–outcome (CMO) configurations, the analysis will examine how contexts shape responses and outcomes [120]. Qualitative findings will be shared with communities through validation workshops to strengthen content validity. Credibility, rigor, triangulation, rich descriptions, and respondent validation will guide data collection and analysis, as recommended for case studies [117].
Table 1: Concepts and their operationalisation in the SPRINT-Sen study.
Concept
| Operationalization
|
Exclusive breastfeeding
| Maternal report of infant feeding in the previous 24 h, coded per WHO/UNICEF criteria as only breastmilk, no other liquids or solids, among infants < 6 months [89]. Questions will be adapted from the DHS module.
|
Heat-related morbidity
| Physical or mental health condition directly caused or exacerbated by elevated ambient temperatures, based on clinical diagnoses recorded in health facility registers (e.g., heat stroke, dehydration, hypertension in pregnancy, acute diarrhoeal infant disease, anxiety, sleep disturbance, etc.)
|
Anxiety
| GAD-7 (Generalized anxiety disorder) score ≥ 10 over the past 2 weeks, indicating moderate or higher generalized anxiety symptoms [108].
|
Sleep quality
| ISI (Insomnia index) score ≥ 15 over the past 2 weeks, indicating poor sleep quality [109].
|
Suicidal risk
| Any positive response to MINI (Mini International Neuropsychiatric Interview) suicidality items C1–C5 (thoughts, plans or preparations) within the past month; C6 documents any lifetime suicide attempt [110].
|
Interpersonal conflict
| Any reported tension between the woman and her household, or with health personnel/at the health centre.
|
Costs of resources and inputs used for the implementation
| Direct and indirect costs of all resources required for the implementation process, including those used for warning systems, from the health-care providers perspective.
|
DALYs
| Disability-Adjusted Life Years attributable to mental health outcomes (anxiety, depression, interpersonal conflict) and heat-related morbidities.
|
A quantitative survey will be conducted in both the hot and cooler seasons to measure the impact of EH on maternal physical and mental health, using standardised instruments and context adapted tools informed by the literature and qualitative findings (Appendix). These data will establish the baseline values for both outcomes (EBF rates and mental health status), and when the survey is repeated in subsequent waves, will allow us to track seasonal variation and detect any changes attributable to the intervention.
The integration of qualitative and quantitative findings will contribute to refining the program theory prior to implementation and better understand the studied phenomenon.
Objective 3. Implement a prospective micro-surveillance system (years 1-2)
A two-year prospective study using satellite remote sensing and aerial data (drones equipped with thermal sensors) will help capture intra-village differences in heat exposures and enable prospective monitoring in connection with the intervention. Temperature monitoring will be conducted both indoors (within households) and at the local level (village or study neighbourhood) and will be cross-referenced with data from the ANACIM weather station located in Matam. Approximately 1,100 households will be equipped with RAK © thermal sensors capable of measuring temperature, humidity, and atmospheric pressure every hour. This frequency is sufficient to capture daily temperature peaks relevant for heat–health analyses while ensuring long-term operational feasibility. These measurements will be transmitted via a LoRaWAN © network to a self-powered server that securely stores the data and transmits it via 4G to a RGPD-compliant hosting service. A dynamic dashboard will provide real-time visualization of the data, displaying maps, tables, and graphs of hourly readings. Temperature and humidity data collected by the household sensors will be used to estimate various indicators of heat exposure. This measure is particularly relevant, as it is the combination of heat and humidity that determines the level of health risk. Alerts will be automatically triggered and sent to project managers whenever predefined thresholds are exceeded. Temperature thresholds will not rely solely on predefined percentiles or standard cut-offs. Locally relevant thresholds will be empirically derived by modelling exposure–response relationships between heat metrics (e.g., temperature and humidity-based indices) and selected maternal and neonatal outcomes. These empirically derived thresholds will be compared with standard heat indicators (e.g., Heat Index, Humidex or Wet Bulb Index) to assess their relative performance and contextual relevance. Sensitivity analyses and internal validation across time periods and sites will be conducted to ensure robustness.
Moreover, dust load will be analysed using data from the ground-based observation network SNO INDAAD (with continuous measurements in Bambey since 2006) and/or from CAMS (Copernicus Atmosphere Monitoring Service) reanalysis products.
These data will enable us to compare multiple heat indicators to better characterise heat impacts, including the concept of high-impact weather events [121], which accounts not only the occurrence of a strong or extreme hazard, but also situations where both exposure to hazard and vulnerability to that exposure are extreme. Furthermore, we will develop indoor heat metrics, tailored to the daily living conditions of populations exposed to sustained high ambient temperatures.
Objective 4. Co-design, implement and evaluate community-based intervention
Phase 2: Intervention implementation (Months 6 to 11, Measure 3)
The community-based intervention will focus on prevention through two core elements: community sensitization to heat-related risks, and capacity strengthening of mothers and health professionals to support exclusive breastfeeding and maternal mental health in the context of extreme heat. The measured variables will align with the dimensions of the Roos et al. (2021) and Chersich et al. (2023) framework and our primary intervention outcome: the rate of exclusive breastfeeding at six months. We will assess breastfeeding practices for infants under 6 months using the French version of the WHO questionnaire, which will be translated into local languages (Wolof and Pulaar), and will inquire about all foods and liquids consumed during the previous 24 hours (day and night) [122]. For infants above 6 months and under 12 months, we will adapt the FeedCat Tool [123] to determine the timing of introduction of non-breastfeeding, capturing both the months at first introduction and the types of complementary foods given. Complementary measures will be collected to comprehensively assess mothers’ knowledge and practice of breastfeeding, including early initiation, frequency of breastfeeding, continuing breastfeeding (6-11months), supplementation practices, and colostrum feeding, as recommended by WHO and studies assessing breastfeeding [124,125]. For pregnant women, we will assess breastfeeding intention using Godin’s framework [91], which will be compared with actual breastfeeding practices through follow-up. These measurements will be conducted at baseline (hot season, and cooler season just before the intervention), during the intervention, and endline.
Mental health will be assessed as a secondary outcome of the intervention, with a focus on acute and subacute manifestations during periods of extreme heat. Standardised and validated tools will be used, including the GAD-7 (Generalized Anxiety Disorder-7) for anxiety, the IPVS-Toolkit for interpersonal conflicts, the Mini International Neuropsychiatric Interview for suicide risk assessment, the WHO-5 (Wellbeing Index) to enable international comparisons. To attribute changes in mental health outcomes to heat exposures, we will combine i) comparisons across four repeated cross-sectional surveys conducted in different climatic periods, and ii) a tailored heat exposure index that integrates indoor and outdoor temperature data with built environment characteristics.
The questionnaires will be adapted based on the qualitative findings and existing literature on recommended actions in extreme heat contexts [126–128]. We will also measure prospective and retrospective acceptability of critical neonatal prevention interventions in periods of extreme heat [129]. To evaluate extreme heat impacts on maternal and child health, the questionnaire will incorporate elements from Nakstad et al. [39] conceptual framework, addressing: mental and physical health status, socioeconomic characteristics, extreme heat risks, knowledge, attitudes, health prevention practices (breastfeeding, danger signs, isolation, etc.) and housing and environmental characteristics. Quantitative variables will be described using means, standard deviations, and medians if the distribution is non-normal.
The intervention evaluation will also assess acceptability, reach, fidelity and implementation. Acceptability refers to participants’ and health professionals’ perceived appropriateness and satisfaction with the intervention components. It will be evaluated using Sekhon et al. framework of acceptability [130] adapted to the Senegalese context [131]. Reach will assess the proportion and characteristics of eligible women and households exposed to the intervention activities. It will be assessed by using both quantitative data (coverage indicators) and qualitative insights on access and inclusion. Fidelity will examine the extent to which the intervention components are delivered as intended, including adherence to planned content and frequency. It will be assessed using the CORE Fidelity Method, embedded within the broader co-adaptation process framework [132,133], to measure the extent to which the intervention is delivered as intended while allowing for necessary contextual adaptations. This is critical for distinguishing between effects attributable to the intervention itself and those attributable to its implementation. Implementation processes will document contextual facilitators, barriers, and adaptations occurring during delivery. It will be explored through the Consolidated Framework for Implementation Research (CFIR) [134] using a qualitative approach adapted to French speaking countries [135]. Finally, the realist evaluation will test the intervention theory formulated during the previous stage, based on initial field insights. By examining how specific contexts trigger or inhibit mechanisms, it will help explain not only whether the intervention works, but also how, for whom, and under which conditions. In addition, we will construct an intensity of implementation variable across sites and over time to capture variation in strength, frequency and reach of the intervention in different villages. This proxy ‘dose’ indicator will support dose-response analysis between implementation intensity and outcomes and mimic a concurrent group evaluation.
The micro surveillance system could also become part of the intervention: sensors would trigger alerts whenever preset temperature thresholds are exceeded, promptly informing the community so they can implement preventive measures.
Phase 3: Post-intervention and follow-up (measure 4)
Intervention effectiveness will be measured using a difference-in-differences approach, comparing outcomes between baseline and final survey. Results will be compared using McNemar’s test for paired measurements. Binary outcomes (exclusive breastfeeding) will be analysed using mixed-effects logistic regression, accounting for village-level clustering and repeated measures within individuals. Maternal mental health outcomes will be measured using validated instruments. Depending on the structure of each scale and the distribution of the resulting scores, outcomes will be analysed as continuous, ordinal, or categorical variables using appropriate mixed-effects models. Repeated measurements collected at baseline, during the intervention, and at 12-months follow-up will be analysed using generalized estimating equations (GEE) for population-averaged effects and mixed-effects models for subject-specific trajectories. Analyses will examine dose-response relationships between intervention intensity and outcomes to explore heterogeneity in intervention effects in across sites. Models will adjust for maternal age, parity, socioeconomic status, and heat exposure intensity. Sensitivity analyses will include multiple imputation for missing data and per-protocol analysis.
Throughout data collection and analysis, the research team will maintain reflexive journals documenting methodological decisions, challenges encountered, and potential biases. This reflexivity will strengthen the credibility of findings and enhance transparency about the knowledge production process [104].
Objective 5. Knowledge transfer
The project aims to generate actionable evidence from scientific knowledge. Thus, it is essential to consider the science of using science [136] to design specific activities that promote research results. All team members will receive knowledge transfer training through specialized Massive Open Online Course (MOOC) [137]. A knowledge transfer plan will be developed with stakeholders’ participation, specifying the nature of knowledge to translate, target audiences, and culturally appropriate dissemination approaches, and its implementation will be qualitatively evaluated [138]. Our primary audiences include: (1) mothers and community members, reached through tools adapted to low-literacy contexts; (2) health professionals, engaged through practice guidelines and training modules; and (3) policymakers at district, regional and national levels influenced through deliberative dialogues. Beyond peer-reviewed publications, we will prioritize proven knowledge translation tools for the climate-health nexus [139].
Anticipated limitations
This study presents several anticipated limitations. The quasi-experimental design may limit causal inference compared with randomized approaches. Exclusive breastfeeding data rely on maternal self-report and may be subject to recall or social desirability bias. Contextual variability and seasonal fluctuations may affect implementation and exposure measurement. Finally, findings may be context-specific to the Matam region. Nevertheless, the mixed-methods design and repeated seasonal measurements aim to strengthen contextual validity and interpretation. Finally, the implementation of community-based interventions may introduce variability in participation and exposure across sites.
Ethics and dissemination
This study protocol received ethical approval from Comité National d’Ethique pour la Recherche en Santé (CNERS) in Senegal (n°0000158/MSAS/CNERS/SP, June 25, 2024), and from the Ethics committee of the London School and Hygiene and Tropical Medicine (LSHTM) (n°31204). Written informed consent will be obtained from all participants, with particular attention to pregnant women and young mothers to ensure voluntary participation and full understanding of the study’s purpose, procedures, and rights. For participants under 18 years of age, written informed consent will be obtained from a parent or legal guardian, in addition to the assent of the adolescent participant, in accordance with national ethical regulations and the requirements of the ethics committee.
Primary risks include potential psychological distress when discussing heat-related health challenges and concerns regarding household-level surveillance data. To mitigate these, we have established: (1) a referral pathway to mental health services; (2) Weather surveillance data collected at household levels will be aggregated without personal identifiers.
Data will be stored on encrypted servers at Huma-num [140] and Institut de la Recherche pour le Développement (IRD) servers, in line with the FAIR principles — Findability, Accessibility, Interoperability, and Reusability —which guide data management [141]. It will also comply with Senegalese data protection law and align with the European Union’s General Data Protection Regulation (GDPR), including requirements for data collection, storage and use, including anonymization, encryption and formal data-sharing agreements [142]. Compliance with GDPR and risk assessment have been validated through a Data Protection Impact Assessment (DPIA) conducted at IRD [143].
Results will be shared via open-access peer-reviewed articles, communications, policy briefs, community workshops, and disseminated to local and national stakeholders. Participants will receive plain-language summaries, and key findings will be integrated into training materials for community health workers and health and environmental plans.