Mobility up the sanitation ladder following community-led total sanitation in rural Zambia

Globally, some 2.4 billion (10⁹) people do not have access to improved sanitation with sub-Saharan African representing 29% of the total (United Nations Children's Fund (UNICEF) & World Health Organization (WHO) 2015). Unimproved or inadequate sanitation refers to a broad range of conditions, the most severe being a complete lack of facilities, then ranging to pit latrines in rural and peri-urban areas missing key components such as a lid to cover the hole, smooth and cleanable floors and/or handwashing stations. Inadequate sanitation is a major public health problem as it contributes to the transmission of viral and bacterial, helminths and parasite pathogens (Bartram & Cairncross 2010; Mara et al. 2010; Prüss-Ustün et al. 2014). Improved sanitation prevents the transmission of these pathogens through both the containment of fecal matter as well as the cleansing of dirty hands through hand washing.

In rural Zambia, an estimated 64% of households do not use improved sanitation (UNICEF & WHO 2015). The resultant health and economic impacts of improving sanitation in Zambia are expected to be great as adequate sanitation is positively associated with improved health (Black et al. 2010; Keino et al. 2014). Inadequate sanitation in Zambia is a likely contributor to poor health throughout the country, for example more than 42% of children in rural Zambia are stunted (Central Statistics Office, Ministry of Health & ICF Macro 2014). Furthermore, in Zambia an estimated $194 million per year or 1.3% of the national gross domestic product is lost because of inadequate sanitation (Hutton 2013).

The Government of the Republic of Zambia (GRZ) through the Ministry of Local Government and Housing (MLGH) has adopted community-led total sanitation (CLTS) as an intervention to improve access to adequate sanitation in rural areas. The intervention has been well received across the country (Lawrence et al. 2016), with Chiengi District becoming open defecation free in 2015 (Zimba et al. 2016). CLTS has been thoroughly explained elsewhere (Kar & Milward 2011); in brief CLTS is a departure from supply-driven interventions that aim to provision communities with sanitation and leverages pride, shame, disgust and fear within a community during a ‘triggering’ event to prompt collective action and create demand toward ensuring adequate sanitation in communities (Kar 2005). Following the community triggering, a volunteer community champion monitors the uptake of sanitation access and reinforces collective action of the community.

In Zambia the goal is for every household to have an adequate latrine defined as a latrine with measurable components including: a smooth cleanable floor, a lid to completely cover the hole, a structure that provides privacy and a handwashing station with water and soap or ash. This definition includes a lid to prevent fly vectors from transmitting fecal material out of the latrine and a handwashing station to ensure access to basic hygiene facilities. Herein we use routine monitoring data to describe village coverage increases and decreases in these adequate components (mobility up and down the sanitation ladder) following triggering with CLTS in rural Zambia.

We used program data from DHIS2 to determine the percentage of monthly sanitation household coverage of the following latrine components for each village: a pit latrine in use, a superstructure for privacy, a smooth cleanable floor, a lid to cover the hole, and a handwashing station with water and ash or soap present. A village is a discrete, predefined geopolitical unit determined by a reigning Chief. Programmatically, villages were categorized as 100% adequate sanitation coverage if they had 100% coverage of every component, or not if they did not have 100% coverage of every component. Among villages that achieve 100% adequate sanitation facility coverage, villages may regress from 100% adequate sanitation coverage if during any month the village has <100% coverage of every component.

We examined the mobility up and down the sanitation ladder using time-to-event analyses to determine the time to reach 100% adequate sanitation coverage, time to drop in coverage in one or more of the adequate parameters below 100% and time to regain 100% adequate sanitation coverage. We used a Cox proportional hazard model while accounting for village size (categorized into quintiles), time since CLTS triggering (categorized into quintiles) and wet or dry season. Time since triggering is determined with the first report that is sent by the community sanitation volunteer. The volunteer will conduct the triggering and then collect the first report immediately afterwards. This first report also serves as the baseline sanitation coverage for the village. The months December through April were categorized as wet and the months May through November were categorized as dry in Zambia. We additionally account for province as a covariate in the model and district as a shared frailty. All analyses were conducted in Stata version 13.1.

As of August 2015, 13,688 villages reported into the information system with the first village reporting in July of 2013. Village size ranges from 1 to 6,082 houses with a mean of 45.3 houses. Time since CLTS triggering ranged from 1 month to 26 months with a mean of 14.8 months.

At some point following achievement of 100% adequate sanitation facility coverage, 41.9% of the 4,646 villages fell below 100% coverage of at least one of the latrine components, with handwashing stations followed by lids being the most common components to fall below 100%. Of villages that dropped, 88.4% of them experienced a drop in coverage of a handwashing station, and 77.7% of them experienced a drop in coverage of a lid covering the hole. When isolated, 12.7% of all drops from 100% access to adequate sanitation were solely caused by handwashing stations and 3.3% were solely caused by a drop in coverage of lids (Tables 3 and 4). Figure 3(b) shows that the majority of the drop-off from 100% access to adequate sanitation occurred within 5 months of achieving 100% coverage of all parameters. The time-to-event analysis shows drops from 100% access to adequate sanitation occurred more often with larger villages, villages with more time since CLTS triggering, and during the rainy season (Table 5).

Of the 1,947 villages that attained 100% adequate sanitation facility coverage and then dropped below 100% in one or more of the parameters, 55.3% regained 100% coverage of parameters with most of those doing so within 5 months of dropping and a large portion regaining the month following the drop (Figure 3(c)). Moving back up the sanitation ladder was not associated with village size and there was no apparent trend with time since CLTS triggering (Table 6). Moving back up the sanitation ladder was less likely to occur in the wet season compared to the dry season (hazard ratio = 0.79, 95% confidence interval = 0.70–0.89).

Over the study period villages in rural Zambia moved both up and down the sanitation ladder, with the overall trajectory being upward. More than one third of the villages in the study achieved 100% coverage of all adequate latrine parameters. Achieving 100% adequate sanitation facility coverage did not guarantee retaining such status; 41.9% of the 4,646 villages reaching 100% coverage fell below 100% coverage of at least one adequate latrine component.

Typically, the drops below 100% coverage of the adequate latrine parameters involved the lack of 100% coverage of a handwashing station with water and soap or ash. The Zambian government made the decision to include a handwashing station as one of the adequate latrine components, which is not in the definition of an improved latrine from the WHO and UNICEF Joint Monitoring Program (UNICEF & WHO 2015). From a health perspective, handwashing is known to impact individual health and its inclusion as a component in the definition of an adequate latrine may be beneficial. When estimating community-level access to adequate sanitation analysis, however, the inclusion of a handwashing station as one of the components of adequate sanitation is quite strict. This analysis did not measure how the CLTS intervention may have affected moving up the sanitation ladder from open defecation to an unimproved latrine.

These data highlight that uptake and retention of handwashing facilities and a latrine lid are the most limiting factors regarding achieving 100% adequate sanitation facility coverage in Zambia. Specifically, hand washing alone in a rural, developing setting is attributed to a 47% reduction in diarrheal disease, 48% reduction in intestinal infections and a 59% reduction in shigellosis (Curtis & Cairncross 2003). As for lids, the control of fly vectors greatly reduces the transmission of trachoma (Rabiu et al. 2012). Poorer households take longer to adopt handwashing behaviors (Luby et al. 2004), and having a rudimentary but permanent handwashing station such as a tippy-tap and a lid for the latrine does require some basic goods. Low-cost handwashing facilities such as tippy-taps have positive impacts (Zhang et al. 2013), but we have limited understanding of the financial or resource barriers of these items to the poorest community members. Lifespan of these low-cost interventions could also be a consideration as the need to frequently replace even a few materials could prove to be overly burdensome.

A significant proportion of villages that progress up the sanitation ladder to reaching 100% adequate sanitation coverage will regress back, at least in some regard, down the sanitation ladder. These villages tend to be larger, the regression is more likely during the wet season, and increased monitoring time is associated with the observed drop. Given the association between monitoring time and the drop from 100% adequate sanitation coverage, the drop may seem inevitable, especially for larger villages. These results highlight the need for continued support to achieve and maintain 100% adequate sanitation. Sanitation interventions in that regard may be similar to community-managed water systems that need continued support and intervention to maintain water supply (Harvey & Reed 2007; Foster 2013). The intervention design and programmatic implementation could perhaps ensure sustainability by increased engagement and empowerment of chiefs who are uniquely able to pressure villages to rapidly achieve adequate sanitation coverage and then sustain sanitation coverage over time. In particular, engagement and continued follow-up with chiefs have been shown to increase the probability of a village becoming open defecation free by 22% and the number of new users of sanitation to increase by 30% in this Zambian context (Tiwari et al. in press).

Half the villages experiencing a drop from 100% adequate sanitation coverage were observed to move back up to 100% coverage. The results here failed to find factors associated with moving back up the sanitation ladder, however. Neither village size nor time in the monitoring system were associated with the recovery; we only see that recovery was more likely during the dry season. Further studies need to be carried out to determine the ‘stickiness’ of CLTS, i.e. whether villages can maintain a high coverage of adequate sanitation in the years following triggering. Although not observed in this analysis, the impact of continuous monitoring could lend something akin to a Hawthorne or Observer Effect wherein behavior changes when monitored (Monahan & Fisher 2010). Leveraging this effect, continuous monitoring of village-level access could potentially be a cost-effective method of ensuring sanitation uptake until sanitation becomes a cultural norm. Both the impact of social pressure and the degree of the Hawthorn Effect in sustaining adequate sanitation require additional research but are potential likely tools that could be used by water, sanitation and hygiene (WASH) programming to ensure sustained adequate sanitation uptake.

Currently during triggering CLTS emphasizes the relationship between openly exposed fecal matter (‘open defecation’) and food contamination by placing food next to fresh human feces found within or near the village. Flies then fly from the feces to food and back again, explaining how feces in the environment can contaminate food. Hand washing and lids on latrines are not emphasized as important pieces of sanitation. Because handwashing stations and lids were the principle issues in both villages achieving 100% adequate sanitation coverage and villages dropping from 100%, the CLTS intervention itself could do more to emphasize handwashing and lids on latrines. Triggering could adopt visualizations to highlight the importance of these parameters, for example contaminating a hand with mud as a proxy for human feces could reinforce the need for handwashing stations. These modifications may be appropriate for the Zambian context, where handwashing stations and latrines are components of adequate sanitation, however further research is needed to ensure that a relatively simple intervention aimed to end open defecation is not diluted with too much information and behavior change communication.

These analyses were conducted using routinely gathered monitoring data from the Zambian MLGH (Markle et al. 2017). These data were extensive both in space and time, allowing for a unique analysis that would not be possible via surveys or through other data collection methods. Improving monitoring systems in sanitation will further understanding on how to ensure sustainable sanitation uptake.

In general, great improvements in access to sanitation have been achieved because of the CLTS program in Zambia including the first ‘open defecation free’ district, Chiengi, in sub-Saharan African (Zimba et al. 2016). We now know that achieving 100% coverage of adequate sanitation facilities is often not a static state and that many villages will regress especially in terms of handwashing facilities with water and soap or ash and the presence of lids. These results highlight serious threats to maintaining adequate sanitation in Zambia following CLTS intervention.

As the world pushes toward the sustainable development goals of sanitation for all by 2030, programs, governments and communities must address the behavioral tendency of communities to move both up and down the sanitation ladder. Without continued behavioral change efforts accompanied by routine surveillance of sanitation aspects, we are unlikely to achieve that goal and risk further morbidity and mortality associated with poor sanitation standards.