| Site visits and field deployment discussion (May–June 2023) |
| Training of research assistants, community health volunteers, and youth leaders, meeting health facility heads for permissions (March–June 2023) |
| Identification of the right sensors and their deployment (June 2023) |
| Survey with pregnant and parenting adolescents, interviews with health workers, adolescents, and policymakers (not reported in this paper, but findings reported separately), July–September 2023 |
| Journey mapping and ecological momentary analysis on mental health and general health in three participants (June–November 2023) |
| Post‑deployment training and follow‑up (July 2023–May 2024) |
| Continuous testing and feedback (July 2023–June 2024) |
Figure 1
Outdoor air quality sensors at health facilities, showing where they were installed and positioned.
Figure 2
Indoor sensors in households, showing their household infrastructure and sensor placement.
Figure 3
Noise sensors at health facilities, showing their placement.
Table 1
Data collection unit and exposure studied.
| DATA COLLECTION UNIT | EXPOSURE STUDIED | NATURE OF DATA CAPTURED |
|---|---|---|
| Health facility‑level data | Indoor and outdoor air quality Noise pollution sensors | Continuous data captured from June to December 2023 Continuous data was captured from September 2023 to April 2024 |
| Household‑level data | Indoor and outdoor air quality | Failed in capturing data reliably due to numerous challenges |
| Adolescent participant‑level data | Surveys and interviews on exposures, health conditions, mental health, and quality of life covering the Edinburgh Postnatal Depression Scale [32], Kessler’s 10 [33], Cohen’s Perceived Stress Scale [34], Generalized Anxiety Scales‑7 [35], NIH PROMIS family functioning measures [36]. | N = 241 survey data N = 40 qualitative data |
Figure 4
Calendar tracking three months’ activities.
Figure 5
(a) One Monday hourly air quality period, and (b) presents weekly average scores over time.
Table 2
Journey mapping along routes perinatal adolescents take to school.
Figure 6
Trends in ecological momentary tracking of mood, stress, and feelings overtime.
Figure 7
Adaptation and mitigation system actors.
Figure 8
Climate adaptation cycle recommended by the Kenyan government.
Table 3
Areas of concern and efforts needed in the future for the refinement of strategies.
| STRATEGY TYPE | SUB‑COMPONENTS OF THE STRATEGY | STRATEGY ACTION | VALUE ADDED AND OUTCOMES | CHALLENGES ENCOUNTERED AND NEW ONES |
|---|---|---|---|---|
Capacity building
| Building the team’s ability to understand measurement and monitoring of environmental exposures (train the trainer model) Identifying, training, and engaging health workers and lay CHWs who could monitor exposures | The research team needed a better grasp of environmental determinants that impact mental health/perinatal health outcomes. Knowledge of combined programming of mental health with environmental health areas for perinatal adolescents would advance holistic interventions | Learning Collaboration was developed that allowed a deeper sense of bidirectional learning between mental health and environmental teams. The importance of intervening at primary care and using the support of lay CHWs and facility staff for building in environmental health components for perinatal adolescents and other groups. Their own well‑being in that environment became a question of concern apart from delivering community care | Teams had different priorities and expectations. Environmental teams wanted greater assistance with monitoring than anticipated. Having people attend to these issues within facilities at one time when sensors were installed was a challenge. Key people within facilities were busy/unavailable for continuous learning. Facility staff and lay CHWs did not have resources for community mobilization, and the study budget was limited |
Implementation process
| Deployment of sensors Education about their use and training for troubleshooting Engaging adolescents and the whole household/household head to install and maintain household‑level sensors | Identification of appropriate sensors, technical choices for their selection, their right positioning, and additional tools/materials for their deployment took time and discussion. Research team members and health facility staff were identified for basic training on the sensors, additional gadgets (even those for protection), and communication around troubleshooting occurred throughout the period of sensor deployment. Basic environmental health education, engagement, and understanding the practical stressors of adolescents who volunteered their households for the deployment of sensors were actioned | The selection of the location of sensors helped in capturing the right exposure at the right clinics visited by perinatal populations There was a great deal of time spent training different members in understanding the deployment process, the need to monitor and ways to address problems and keep the sensors operational Identifying pregnant and parenting adolescents, and their caregivers to volunteer for the deployment of sensors was critical | There were times when power outages were experienced, and the health facility needed data bundles or internet services or when the sensors were switched off. Unavailability of reliable desktops at the health facilities to download data from noise sensors. The facilities in the Kenyan coastal county of Kilifi were far apart, and it took time to address issues. The staff did not engage too much once the gadget was deployed There were fights in the household over who “owned” the sensors; most households were poor and needed electricity and data support Households expected more financial support that could not be covered by the study. Someone had to be tasked to monitor the site adding to costs |
Integration
| Addressing common factors Identifying environmental determinants of mental health | Combining the environmental and mental health survey for health facility and individual participants to map exposures and their impact on mental health. The study team had to identify areas such as sanitation, air quality, quality of life, food safety and quality, and home environment, which are linked to both environmental exposures and mental health | While combining the domains took time, it was beneficial to see how both areas could be brought together for exploration with perinatal adolescents. Distal and proximal risk factors identification using literature, in‑country research & health/environmental policy‑driven indicators was an exciting learning process | Polishing items for two distinct settings was not easy, Refining indicators for environmental domains specific to two landlocked and coastal ecosystems is a time‑consuming process. Public health awareness of mental health is very low and still poor. Adding environment as a determinant is exciting but a challenging concept—the measurement of these exposures is still under‑investigated, especially in LMICs |
Dissemination
| Consultative meetings on air and noise pollution quality Seeking facility and community feedback | Throughout the study and at the end, consultations were held on the health implications of air and noise pollution and the best practices for addressing these in primary care and community settings We sought feedback and participation from health facility actors to identify | The importance of shared learning and educating health service staff, community members, and the research community was felt in every engagement Community feedback provided new information on life conditions, problems with sanitation, water quality, and access to basic life amenities | Did not have adequate resources to provide in‑depth training or education comprehensively, and to provide ready interventions. Resource and infrastructural issues are beyond the scope of any research, socio‑structural problems like stigma and discrimination of adolescent mothers and those with mental illnesses need a wider national strategy and government investment |
Scale up
| Engagement of stakeholders Willingness to share costs | Bringing together a well‑represented collaborative of multidisciplinary researchers, clinicians, policymakers, and advocates from health and environmental backgrounds Agreement to share costs of sensors provided by UNEP—field deployment by the environmental agency and monitoring by the research team with partners and health facility ownership worked | Learning about the issues of environmental determinants of mental health was deep and informed by multiple perspectives Recognition of total investment needed for change and willingness to take some responsibility per entity, lots of cooperation | Sustained participation and attention towards integrated programming needs more conceptual harmonization and resources to strengthen the partnership. Resources may be needed and identified per entity to bring to the table for joint programming and intervention development, and their impact evaluation |
| Coordinated environmental and health programming at county and sub‑county levels |
| Deployment of air and noise pollution devices needs to also include other key exposures like soil, water, microplastics, etc., within the facility and key community spaces with a data dashboard for research, county/sub‑county health decision‑making, and key highlights shared with the facility |
| Identified mental health indicators, especially around perinatal adolescents, embedded within environmental health indicators |
| Trainer of trainers on a continuous basis for health facilities with lay health workers with strong advocacy to build a department for routine monitoring of exposures and outcomes at the sub‑county level |
| Provision of electricity, internet, and backup support for data to review sensor information |
| Real‑time monitoring feedback to feed into clinical, community, and local policy decision‑making, including infrastructure costing and planning for the feedback process |
| Engaging communities to understand and address sources of pollution and poor air quality and developing guidelines for noise pollution regulation for health facilities |