Developing and Testing Strategies for Improving Cleanliness of Shared Sanitation in Low-Income Settlements of Kisumu, Kenya

Theoretical approach.

The study was guided by the Behavior Change Wheel, a theoretical approach for designing behavior change interventions.¹⁹ The wheel has three rings that detail the stages of designing behavior change interventions. The inner ring requires an understanding of the existing behaviors and identification of what needs to change, the middle ring requires identification of intervention functions that are likely to initiate a change in behavior based on the inner ring, and the outer ring identifies policy categories that support the change.¹⁹ Our earlier manuscript detailed the results from the first stage which entailed characterizing and understanding existing behavior, as well as identification of what needs to change.¹⁷

For development and testing of the interventions, we adopted the Trials of Improved Practices (TIPs) approach. The TIPs approach uses a participatory methodology and aims to design feasible interventions and then test whether these interventions are acceptable.²⁰ Through the TIPs approach, implementers learn from the community about what aspects of the intervention should be promoted or eliminated, the efficacy of the different approaches, and the level of impact the interventions are likely to have.²⁰ Typically, TIPs is conducted after a formative research phase, it requires a small sample size (at least 20 participants), and lasts for a short time (minimum of a week) depending on the behaviors being tested.²⁰ The TIPs approach enables participants to codesign and improve the interventions, which are thus more likely to be adopted by the end users.^21,22 In this study, we used the TIPs approach to develop and test a combination of strategies that promote sustainable compound-level cleaning of shared sanitation facilities.

Context.

This study was conducted in Kisumu, Kenya’s third largest city after Nairobi and Mombasa. The city has a population of approximately 500,000 residents as per the 2019 population and housing census.²³ Approximately 60% of the city’s population lives in low-income settlements which are characterized by overcrowding, inadequate water and sanitation service provision, and poor housing structures.^24,25 These low-income settlements include Nyalenda A and B, Manyatta A and B, Bandani, and Obunga. Land in these settlements is on free-hold tenure, with landowners passing on ownership from one generation to the next.²⁶ Due to population growth, there has been an increase in the demand for housing, and landowners have responded by constructing housing structures for rental purposes. The structures are usually single housing units, with occupants sharing common facilities such as the yard, a water source, and toilet facilities (if available). Over time, some landowners have moved out of the settlements whereas others have continued to stay within the settlements, often residing with the tenants on the same piece of land (called a plot or compound). Compounds where the landowners have moved out are usually occupied by tenants, with some compounds having one tenant designated as a caretaker.²⁵ Sanitation facilities are mainly improved pit latrines shared by an average of eight households.^4,25 Previous studies indicated that majority of these latrines were dirty, mainly because users did not participate in their cleaning.^4,17 The study was conducted in Nyalenda A, which is further divided into four clusters/villages, commonly called units: Western A, Kanyakwar, Dago, and Central units

Developing the intervention strategies.

In development of the strategies, we aimed to develop and test strategies that were driven by the compound members, that could be easily implemented by community leaders and resource persons—such as Community Health Volunteers (CHVs)—and which could be sustained beyond the study period. The strategies were identified through workshops with the research team and stakeholders, and validated through group discussions with community members.

Outcomes of the group discussions.

Table 1 summarizes the rankings from the discussions. The highest ranking strategies were cleaning plans, discussions in the compound, and monitoring. Cleaning plans and discussions in the compound both scored highly among landlords and tenants.

Testing the intervention strategies.

Testing of the intervention strategies was conducted from January to May 2020. The identified strategies were compared with the theoretical approach guiding the study. Previously identified strategies from the qualitative stage included use of restrictions, environmental restructuring, enablement, enhancing education and training, and improving sanitation technologies.¹⁷ These strategies tallied with the newly codesigned strategies, namely: development and use of cleaning plans; restricting against poor use through rules and monitoring; and discussions in the compound as a form of enablement. The proposed package of strategies entailed discussions with compound members, introduction of a cleaning plan and selection of a representative within the compound who would be in charge of monitoring the cleaning of the toilet as well as coordinating the compound meetings. Education and sensitization would be integrated into the compound discussions as part of the intervention package.

To qualify for selection to test the identified strategies, compounds needed to have at least two households and at least one toilet that was shared by the households. All the households in the compound had to be users of the toilet. The toilets needed to have a superstructure that offered privacy, a door that held in place, and a roof. Using these criteria, a CHV from each of the four units within Nyalenda A, together with the research assistants walked through the units and identified eligible compounds that met these criteria. In line with the TIPS recommendation for a small sample size, CHVs and the field staff were asked to recruit a minimum of 10 compounds from each of the units, thus making a total of 40 compounds (to also cater for losses during follow up). During the enrolment, two compounds refused to continue with the study, therefore reducing the sample size to 38 compounds that agreed to participate in testing the strategies. Community Health Volunteers and the research assistants visited the selected compounds to introduce the intervention strategies and encourage the compound members to participate, clarifying that the testing would entail attending meetings and discussions. Data collection during the first three meetings and discussions was done physically in the compounds over a period of 1.5 months. The next three follow-up sessions were done through phone calls to the compound leaders over a period of 2.5 months.

During the first visit and meeting with the compound members, CHVs introduced the research assistants, and the compound members introduced themselves. Once introductions were complete, the research assistant led the discussions by focusing on the use of their sanitation facility, prevailing cleaning practices, availability of a cleaning schedule, challenges faced in keeping the toilets clean, ways of improvement, and the importance of keeping shared sanitation facilities clean. The assistants proposed the strategies, and encouraged compound members to draft a cleaning plan (if they did not have any) and to propose a representative within the compound who would oversee the whole process. Members were encouraged to work together by mobilizing each other, identifying how best the strategies would work for them, and participating in implementing the strategies. At the end of the meeting, the members suggested a convenient date for the next meeting.

The second follow-up visit was held 1 week after the first visit. The CHV team and research assistants visited the compound members again to further discuss their experiences with trying out the strategies. During the meeting, compound members reported challenges encountered, progress made, and suggestions for improvement. Research assistants noted those cases where progress had been made and encouraged the members to work together in improving the cleanliness of the toilets. Where little or no progress had been made, the research assistants still encouraged the compound members to work together and try out the proposed strategies. Again, the members proposed the most convenient date for the next visit.

The third follow-up visit and discussion was done approximately 2 weeks after the second visit, to allow compound members to continue with the strategies without solely relying on the research team. This visit followed a similar format as the second visit, focusing on progress made, challenges experienced, and any outcomes from the discussions and strategies.

During these three visits, assistants collected quantitative information such as the number of compound members who attended, the type of toilets, and the number of cubicles being shared. They also collected qualitative data including the strategies that had been implemented by the compound members, challenges faced, and benefits experienced. Where possible, the assistants observed the cleanliness of the toilets. Compound meetings were only held if a minimum of 50% of all compound members who used the toilet were available. If the members were not available or the numbers were not enough, the meetings were postponed to a time that was convenient for all members.

Three more follow-up sessions were conducted remotely through phone calls due to restrictions in fieldwork activities as a result of the COVID-19 pandemic. The fourth and fifth follow-up visits were held approximately 3 weeks after the preceding visits while the sixth and final follow-up was held 1 month later. Data was collected through phone calls to the compound representatives, and to at least two other compound members. Discussions during the phone calls focused on the progress made and challenges faced.

Ethical procedures.

Ethical approval for the study was obtained from the ethics committee of the Great Lakes University of Kisumu GLUK (Ref: GREC/001/285/2018). A research permit was further obtained from the Kenya National Commission of science and Technology (Ref: NACOSTI/P/18/5546/24979) and from the Kisumu County government. Permission to conduct the study was sought from administrative heads at the local level in the study area.

During the group discussions, field assistants detailed the purpose of the study, progress and findings from previous phases, the purpose of the group discussions, duration of the discussions, actions that would be taken to ensure anonymity and confidentiality such as not using names, and the next steps after the discussions. They were also informed that they were free not to continue with the discussions if they chose to. Discussions began after all questions from the participants had been answered satisfactorily, and participants had given their signed consent to participate in the discussion.

During the first household visits, field assistants provided detailed information sheets to the participants. The assistants explained the details of the information sheet, including the purpose of the study, reasons why the compounds were selected, the study procedures including community dissemination plans, duration of the discussions, expected requirements from the participants, steps to ensure confidentiality, how the data would be used, and contact information of the researchers. They were allowed to ask questions after which they were given a copy of the information sheet to keep. The participants signed two copies of the consent form in the presence of the CHVs. They retained one copy of the consent form, and another was kept by the study team.

Data management and analysis.

Quantitative data from the compound visits was analyzed through descriptive statistics such as means, frequencies, and percentages. Qualitative findings from the group discussions were analyzed the matically using already predefined themes, that is, the intervention strategies. Quantitative and qualitative data were combined during analysis and presented together to provide meaningful explanations for the uptake of the strategies.

Strategy 1: Uptake and attendance of compound-level meetings and discussions.

Furthermore, compound members often reported that they held subsequent meetings (in addition to the meetings we held with them), where they further discussed implementation of the strategies and other matters affecting them. Members who did not attend the discussions were away due to work, travel, or social functions. Whereas both men and women attended the discussions, women were the majority, partly because they were always at home, or were usually involved in cleaning of the toilets.

Strategy 2: Uptake of the cleaning plan.

During the first visit, almost all the compounds reported that they did not have any form of cleaning plan. Similar to results from the earlier qualitative phase,¹⁷ toilets were cleaned by any household on a voluntary basis, or they were cleaned by individuals from specific households (e.g., women who had children). A few (13%) compounds admitted that all households in the compound were involved in cleaning, although without any form of schedule/plan.

By the end of our visits, approximately 66% of the compounds had adopted a cleaning plan (Table 3). This plan was either a written schedule or an informal cleaning plan where households organized themselves according to the order of houses or days of the week. Where households had a written plan, the schedule was often pasted on the door of the toilet or written on a paper and kept by the compound representative. This plan had the names of individuals from all the households or their house numbers. According to these plans, toilets were cleaned on a daily basis, and members often collectively selected their preferred cleaning days. In one compound, households had a weekly cleaning plan before our visits, which they opted to continue with.

Informal plans were agreed upon by members, after they discussed and organized themselves. For example, participants in several compounds reported making verbal agreements during the meetings and assigning each household a specific day to clean the toilet. Two compounds in particular felt that they did not need to write down a formal plan and have it pasted on the toilet because it would be plucked out. In other compounds, especially those with few households, members reported no need for a formally written cleaning plan. Households from such compounds organized themselves and often assigned households to specific days of the week to clean the toilets. Such households frequently reported that they were few in number and each household knew when they were to clean the toilet.

Adoption of the discussions and the cleaning plan led to over 80% of the compounds implementing some form of organized cleaning, compared with the pre-intervention stage where households volunteered to clean toilets (74%). Notably, 47% of compounds had all households involved in the overall management of their sanitation facility. With regards to gender, men were directly involved in cleaning in most of the compounds. When men were not directly involved, compound members agreed to have women clean the toilets while men purchased cleaning materials. In such arrangements, although all households were involved in the overall management of the sanitation facilities, only specific households (especially those where women were often at home) cleaned the toilets, hence the increase from 13% in the pre-intervention stage, to 44% of cases where specific households were involved in cleaning. Before implementation of the intervention strategies, the specific households that were involved in cleaning were those with younger children, with little or no participation by other households.

In such compounds, men were involved in purchasing the cleaning items while women cleaned the toilets. Men were also exempted from cleaning in cases where they were unmarried and/or not always at home. In instances where individuals were not available to clean the toilets, they agreed to exchange their cleaning days with their neighbors, or they asked their neighbors to clean the toilets on their behalf. In one compound for example, a man opted to pay his neighbor to clean the toilet.

Compounds experienced various challenges with the cleaning schedule, mainly its getting torn or plucked from the toilet door (N = 3). Lack of materials to paste the written schedule on the toilet door was also a common challenge. Other challenges arose from the movement of tenants out of the compounds, or their movement to rural areas especially during the COVID-19 pandemic. Consequently, the available households cleaned the toilets on the days when their neighbors were to clean, or they had to rewrite the cleaning plan and only involve the available households.

Strategy 3: Monitoring and taking responsibility.

Before the introduction of the intervention strategies, none of the tenant-only compounds had any individual who was responsible for overseeing cleanliness of the sanitation facilities. Landowners and caretakers often took on this responsibility in compounds where they resided mainly because of their status as landowners or caretakers in the compound. Compound members selected the caretaker (8%), landowner (34%), or one of the tenants (42%) to be responsible for monitoring the progress of the strategies that were being implemented in their compounds. In compounds where the landowners were resident, members selected them to oversee and be responsible for the initiatives, while in compounds without a resident landowner, members selected the caretaker or one of the tenants. Besides being our contact persons within the compound, the selected individuals also followed up with members and reminded them of meetings, they organized the compound members and drafted the cleaning plan, they followed up on any individuals who missed the discussions, and reminded members to clean the toilets. In a handful of the compounds (16%), the members felt that they did not need anyone to act as a monitor, since “they were all adults and could organize themselves.”

Other strategies that were implemented.

In addition to the proposed strategies, compounds implemented additional strategies, which included individual or collective contribution of resources for purchase of cleaning materials and agents (N = 14); purchase of a padlock and locking of the toilets (N = 9); cleaning and slashing grass next to the toilet (N = 1); emptying of filled up pit latrines (N = 1); accompanying young children to the toilet to avoid misuse (N = 1); installing handwashing containers (N = 5); cleaning the bathrooms in addition to the toilets (N = 4); and instituting rules for use of the toilet (such as not dumping diapers in toilets, not allowing households from other compounds to use the toilet, warning and evicting members who do not participate in cleaning) (N = 4).

Reported benefits.

The improved relationships made it easier for compound members to organize themselves and participate in other compound-level activities such as cleaning the compound and barring outsiders from sneaking into and using their toilets. Individuals who previously cleaned the toilets voluntarily without help from their neighbors reported satisfaction and relief from everyday cleaning and from division of work among all households.

Other reported benefits included improved relationships between the CHVs and the community members, as well as the sharing of experiences with neighboring compounds and encouraging them to keep their toilets clean.