History of WSPs

The WSP concept originated with performance of routine “sanitary inspections” of water supply systems in the early 20th century. These inspections follow a short, standardized observational checklist to assess risks to public health. They were first promoted by WHO in the 1976 monograph “Surveillance of Drinking-Water Quality” and have since been mentioned in each edition of the WHO guidelines for drinking-water quality (Kelly et al., 2020). Another key predecessor of WSPs was the hazard analysis and critical control points (HACCP) process. In the 1960s, the U.S. Army Laboratories, U.S. National Aeronautics and Space Administration (NASA), and a packaged food company (Pillsbury) teamed up to develop an approach aimed at preventing hazards, rather than testing the safety of a product retrospectively. This food-safety management process was well established by the 1990s (Havelaar, 1994), leading to formation of the International HACCP Alliance. Iceland first adopted national legislation applying the HACCP concept to public drinking-water supplies in 1995 (Gunnarsdottir & Gissurarson, 2008). HACCP and similar International Organization for Standardization (ISO) standards (i.e., 9001, 14001, 22000, 31000) remain in widespread use for food safety, manufacturing, engineering, and other business applications.

Other ideas contributing to preventive management have included failure mode analysis, multiple barriers, continuous quality improvement, total quality management, and quantitative microbial risk assessment (QMRA). The QMRA concept also originated in the food industry and was first proposed for water safety management in the 1990s (Regli et al., 1991). QMRA has been used in various applications, including drinking water, recreational water, and water reuse, and it has been embedded in the WHO guidelines (WHO, 2016a).

In 2004, after a decade-long global consultation process involving more than 500 individuals from approximately 100 countries, the third edition of the WHO guidelines for drinking-water quality carried a global recommendation for WSPs (WHO, 2004). The International Water Association (IWA), a global professional association, echoed this recommendation in the Bonn Charter for Safe Drinking Water (IWA, 2004). The first edition of the WSP manual was jointly released by WHO and IWA in 2009 (Bartram et al., 2009).

Evolution of the Concept Over Time

As WSP application expands around the world and more case studies and research become available, the spatial scale, scope, and expected outcomes continue to evolve.

Regarding scale, WSPs designed at a watershed level have been adapted to serve the needs of urban areas, individual buildings (e.g., hospitals), ships, sanitation systems, and direct potable reuse systems (see Almeida et al., 2014; Goodwin et al., 2015; Mouchtouri et al., 2012; WHO, 2011). The practice of safe wastewater management via sanitation safety plans (SSPs) was introduced by the 2006 WHO Guidelines for the safe use of wastewater, excreta and greywater (WHO, 2006). The SSP concept, which mirrors WSPs, applies the risk assessment and management approach from the point of wastewater generation (e.g., the toilet, sink, or shower) to the final use or disposal (WHO, 2016b). SSPs operate in a less-defined regulatory environment than WSPs, but they also have multiple objectives and stakeholders (e.g., wastewater utility managers, sanitation enterprises, farmers, and community-based organizations; Winkler et al., 2017). They address risks to multiple exposure groups, such as sanitation workers, farm workers, local communities, and consumers who eat or use agricultural products produced using recycled wastewater, biosolids, or fecal sludge.

Water cycle safety plans (WCSPs) have been developed under the framework of the PREPARED research project funded by the European Commission to provide a risk management approach for the entire water cycle, including water supply, wastewater treatment, water reuse, and management of water bodies (Almeida et al., 2014). Subsequently, another European Union (EU) project, DEMOWARE, aimed to develop and promote adoption of a risk management framework for water reuse by integrating elements of WSPs, SSPs, and WCSPs (Hochstrat et al., 2017). As water reuse applications expand in the future and blur the lines between “new” and “recycled” water, different types of water safety and environmental stewardship programs are increasingly likely to be combined.

Regarding scope, numerous parties have proposed add-ons, revisions, or adaptations of the guidance on WSP implementation in the literature and specific country or regional policies (WHO & IWA, 2020). For example, the Techneau project funded by the European Commission developed a set of tools for risk assessment and risk management, including a hazard database, a risk-reduction option database, a decision-support tool integrating asset management and cost optimization, and an integrated and quantitative risk model for comparing risk reduction alternatives (Rosén et al., 2009). Some countries, such as the Netherlands, use a more complex risk assessment and risk management strategy, where one of the key elements is QMRA implementation (van den Berg et al., 2019). In comparison to WSPs, QMRA provides a numerical output of estimated health outcomes from microbial hazards that may support risk management in finer detail than qualitative (e.g., sanitary inspections) or semi-quantitative (e.g., WSP) approaches (WHO, 2016a). Sanitary inspection guidance from the WHO is similarly under review, with the goal of better aligning this low-cost approach, commonly practiced in developing settings, with WSPs (Pond et al., 2020).

In Canada, Bereskie et al. (2017) proposed implementation of WSPs along a timeline akin to plan-do-check-act (PDCA) quality improvement cycles, which are used prominently in the clinical healthcare industry. Some water supply organizations (e.g., in Australia) integrate WSPs with organization-wide business and financial risk planning, for example by having separate business units (a) to propose appropriate technical risk management solutions and (b) to select the most effective solution based on cost, client relations, and environmental impacts (Setty et al., 2019a). In 2019, the Joint Research Center of the European Commission published a document for water utility operators on water security planning; it aimed to provide clear guidance for unexpected water security emergencies and recommended integrating security planning with the WSP approach where possible (Teixeira et al., 2019).

The evolving scope of WSPs has increasingly refocused attention on the hazards posed by global climate change, which is projected to cause more frequent and severe extreme weather events (WHO, 2017a). The potential for WSPs to address social equity has also been highlighted, especially since the 2015 global emphasis on Sustainable Development Goal (SDG) 6 (“Ensure availability and sustainable management of water and sanitation for all”), and is reflected in a targeted guidance manual for equitable WSPs (WHO, 2019). As of 2020, WHO has undertaken a comprehensive review and revision effort to appropriately integrate these and other concepts into WSP practice, with the goal of issuing an updated second edition of the WSP manual. Evaluation approaches are similarly evolving as experience with WSP implementation and outcome measurement expands (discussed further under “Evaluating Impacts”).

Evaluating Implementation

Periodic auditing (or independent assessment) is a vital tool for sustaining and improving the WSP (WHO & IWA, 2015). Third-party review helps to maintain elements of standardization across locations, to demonstrate compliance with regulatory or voluntary standards, and to demonstrate the WSP is complete and accurately recorded as implemented in practice. The WSP manual strongly recommends audits be conducted in a supportive, rather than punitive, manner to reveal areas of potential system-wide improvement (Bartram et al., 2009). A guide to auditing WSPs was developed in 2015 (WHO & IWA, 2015), followed by the WSP audit training package in 2019. Audits can be “internal” when carried out by someone employed or contracted by the water supplier (but independent of the WSP team), or “external” when implemented by a government body or other entity outside of the water supplier. Audits are “formal” when they are used to confirm compliance with regulatory requirements or for organizational quality assurance purposes; “informal” audits are used to provide advice and for learning purposes.

Despite their promotion, global auditing uptake and practices remain inconsistent (Ferrero et al., 2019). The lack of auditor training and certification schemes represents a bottleneck contributing to incomplete or inaccurate WSP implementation. Only a few documented examples of such certification systems are available (e.g., State of Victoria, Australia; WHO & IWA, 2015). Organization of regional, national, or multicountry guidance and certification systems for auditors could improve consistent application of WSP auditing (Ferrero et al., 2019; Schmiege et al., 2020) and produce additional helpful data about how implementation processes and adaptation affect WSP outcomes and impacts (see “Evaluating Impacts”).

Audits may clarify how basic adherence to WSP guidance on a larger scale balances with the need for site-specific adaptations, eliciting potential areas of adjustment for future quality improvement efforts. Quality improvement and implementation tools from related disciplines, such as clinical healthcare, may prove valuable, especially in challenging cases of WSP implementation (Setty, 2019). These tools, such as implementation frameworks, implementation strategies matched to site-specific barriers, timed quality improvement cycling, process theories linking specific inputs to outcomes, and contextual factor checklists, were designed to help move effective evidence-based public health interventions from pilot applications to full-scale application. Some implementation tools may require adaptation (e.g., of terms, definitions, and constructs) to shift from a clinical healthcare context to the water supply industry (Setty et al., 2019b). Although few examples have been applied to WSP implementation to date (e.g., Bereskie et al., 2017; Omar et al., 2017; Setty et al., 2019a), they offer an option for use in WSP research or for addressing recurring or stubborn water safety issues.

Evaluating Impacts

Since 2011, specific WSP impact evaluation indicators have been refined with added research, including a targeted investigation of 99 water supply systems across 12 countries in the Asia-Pacific region (Kumpel et al., 2018). Building on the body of available evidence, practical impact evaluation assessment guidance incorporating climate resilience is under development by WHO to help guide users toward consistent, rigorous methods. Recommended WSP evaluation categories group policy, operational, financial, institutional , climate, water supply, and health indicators (Table 2). Water system managers may opt to supplement, reduce, or adapt the precise evaluation indicators based on their available data, study design, and local contextual factors. Evaluations should ideally be conducted at baseline (prior to WSP implementation) and at regular intervals after WSP initiation. Where baseline data are not collected prior to WSP implementation, baseline conditions can, in some cases, be established retrospectively (Kumpel et al., 2018).

Evaluating Public Health

When implementing WSPs, Summerill et al. (2010) suggested public health should be considered the primary motivator among other legitimate drivers, such as political or stakeholder priorities or financial efficiency. However, the Asia-Pacific impact assessment highlighted that health indicators were particularly challenging to quantify (Kumpel et al., 2018). Since protecting consumers’ health is the core objective of WSPs, the authors called for robust and standardized monitoring methods.

Health impact evaluation for WSPs is challenged by both the quality of health data (Setty et al., 2017) and limited epidemiological study design options (Kumpel et al., 2018). In theory, certain conditions must apply before pathogens can cause a detectable health outcome, such as acute gastroenteritis (diarrhea and/or vomiting). For instance, pathogens must:

Because many instances of drinking water contamination do not meet all of these (or other) conditions, and because acute gastroenteritis is typically self-treated and self-limiting, public health surveillance often detects only a small fraction of cases. Other factors influencing data capture include when healthcare facilities are open, healthcare affordability, temporality in wage or pension distribution, travel in or out of service areas, and ambient temperatures that exacerbate dehydration symptoms (Bounoure et al., 2020; Setty et al., 2018a).

Variability in public health surveillance approaches across locations and over time may hamper detection of significant changes (Kumpel et al., 2018; Setty et al., 2017). Depending on the local health systems and context, fewer than an estimated 2% to 3% of all acute gastroenteritis cases may be identified through hospital-based surveillance systems (Setty et al., 2017). A novel approach used in France since 2010 detects an estimated one third of all cases using algorithm processing of medication reimbursements from the state-run healthcare system (Bounoure et al., 2020). In addition to its underreporting, diarrhea may be chronic (e.g., associated with other health issues or chronic pathogen exposure in environments lacking adequate human and animal sanitation infrastructure), making it difficult to define individual cases or to assign clear causation. Where public health surveillance data are inadequate, household surveys to quantify the incidence of diarrhea are preferred over questionnaires completed by the water supplier or review of existing household data (Kumpel et al., 2018). Still, self-reported data may be affected by recall bias. If possible, surveys should be implemented using appropriate sampling strategies (e.g., stratified random sampling with adequate power to detect statistical trends).

Epidemiological study design for WSP evaluation requires careful attention due to several additional challenges. First, cases of acute gastroenteritis are more commonly attributed to foodborne or person-to-person transmission than to drinking water. The disease burden attributable to drinking-water exposure may be a small percentage of total cases, especially in developed settings (Setty et al., 2017). Second, seasonal trends in diarrheal diseases are also common, with person-to-person viral transmission more common in the winter months or rainy seasons, depending on climate (Ahmed et al., 2013). Microbial infectivity can also fluctuate over time as different strains of pathogens become more or less prominent. Third, water suppliers and healthcare facilities may cover different service areas, with some populations served by multiple, mixed, or overlapping drinking-water systems or sources (Setty et al., 2017).

Robust study designs must take these underlying influences and background levels into account, preferably by comparing a WSP location to a baseline period before the WSP was implemented, or a comparable area nearby that did not undergo WSP implementation. Since WSPs are recommended for all locations to achieve improved drinking-water safety, it may be unethical to randomize assignment of the intervention for long periods, although using stepped-wedge designs or temporarily delaying WSP implementation may provide acceptable research alternatives. To establish causality, controlled prospective study designs (i.e., where the outcome has not occurred when the study starts and subjects are followed over a period) are typically the gold standard, but also require the most resources and expertise. Before–after or interrupted time series (“quasi-experimental”) designs may be more appropriate and practical to provide evidence of association of observed changes with WSP implementation over time (Kumpel et al., 2018; Setty et al., 2018a). Randomized controls (i.e., locations where a WSP is not implemented to understand background changes unrelated to the intervention) may be difficult to identify if the pool of candidate water suppliers is limited, but matched controls can be selected based on characteristics such as water system size, geographic setting, and/or revenue (Kumpel et al., 2018; Setty et al., 2017).

These challenges highlight the need for robust public health systems and regular communication and cooperation between health and environmental (i.e., water management) authorities to support ongoing WSP improvement.

Public Health Impact

Promising studies on public health outcomes of WSPs have been conducted in Iceland, France, and Spain, demonstrating potential for population-level declines in diarrheal disease rates (Gunnarsdóttir et al., 2012; Setty et al., 2017). Given Iceland’s earlier start using WSPs, they produced the earliest evidence of public health impacts (Gunnarsdóttir et al., 2012). Research found that populations served by an unchlorinated groundwater supply with a WSP in place had 14% fewer clinical cases of diarrhea on average compared to pre-WSP conditions. Follow-up research examining chlorinated surface and groundwater suppliers in France and Spain demonstrated a reduction in acute gastroenteritis incidence following WSP implementation at one of three locations examined, corresponding to about a 4% decrease in acute gastroenteritis incidence in the overall population, or 6% among adults (Setty et al., 2017). Although these percentages form a small proportion of the overall diarrheal disease burden, they likely represent a sizable portion of the burden attributable to drinking-water exposures. Using a time-series design at the same locations in France and Spain, Setty et al. (2018a) showed specific risks associated with surface runoff into water supplies and increased source water turbidity were likely mitigated by WSP-related treatment interventions, adding to the evidence that the health changes were associated with WSP implementation. In the Asia-Pacific region, data on the incidence of diarrheal disease were rarely adequate to assess changes (Kumpel et al., 2018). Taken together, the health impact results suggest the WSP approach to site-specific risk management is potentially effective at reducing baseline gastrointestinal illness risk. Evidence regarding the capacity of WSPs to prevent unexpected, low-frequency but high-impact events (e.g., outbreaks) is more challenging to ascertain quantitatively.

Related Benefits

Although positive, neutral, and negative outcomes (e.g., Setty et al., 2017) have been documented at various WSP implementation locations in the primary evaluation categories (Table 2), most literature shows some aspects of improvement within the first few years of programming (WHO & IWA, 2020).

The most common category of improvement was operations and management (at 95% of evaluated sites), followed by infrastructure improvements (at 86% of evaluated sites) in the Asia-Pacific impact assessment (Kumpel et al., 2018). Case studies from other world regions have similarly shown operation and maintenance shifts to be a key driver of WSP-related benefits (Setty et al., 2018b; String & Lantagne, 2016; String et al., 2020; WHO & IWA, 2018). While risk reduction is hard to measure on brief time scales and without random controls, qualitative evidence and managers’ experience following adoption of WSPs has highlighted better control of hazards that were previously overlooked (Kayser et al., 2019).

Water quality and communication among water suppliers and stakeholders similarly show improvement at a majority of WSP locations (Kayser et al., 2019; Kumpel et al., 2018; Setty et al., 2017). Water quality improvements have also been reported, albeit inconsistently, in community-managed water supplies in developing nations (Mahmud et al., 2007; String et al., 2020). Water quality changes may correspond to more consistent achievement of regulatory or voluntary compliance thresholds developed for public health protection (Gunnarsdóttir et al., 2012; Setty et al., 2017); however, these indicators are limited because most provide only a surrogate for pathogen exposure and disease risk.

A 2016 systematic review found limited evidence to support financial improvements in Uganda and Palau, related to reduced monitoring costs and water conservation (String & Lantagne, 2016). Another five-country study enumerated typical startup and maintenance investments in WSPs (Kayser et al., 2019) but did not monetize benefits. Data on financial benefits would be valuable because much of the socioeconomic impact of WSPs could stem from fewer sick days, reduced healthcare expenses, and greater worker productivity, which benefit disadvantaged groups and other industries beyond the water supplier.

Ongoing Challenges

Despite the potential benefits of WSPs, many actors remain undecided about whether to promote or mandate WSP policies. Review of WSP uptake in high-income countries shows policies, support from public health agencies, and context-specific evidence of the feasibility and benefits have played an important role in enabling widespread adoption and implementation (Baum & Bartram, 2018). Conversely, a paucity of context-specific evidence, inadequate training, impediments to knowledge translation, insufficient financial or human resources, or skepticism about return on investment might forestall decision and implementation processes (Amjad et al., 2016; Ferrero et al., 2019; Kayser et al., 2019). Country-specific cultural influences and the political economy, including local community readiness, are also expected to influence WSP feasibility (Kot et al., 2015; Omar et al., 2017). Decision makers may wait for others to furnish local evidence, requirements, or incentives before adopting new programs. Translational research (Setty et al., 2019b) to pilot, evaluate, and effectively replicate novel WSP programming can play an important role in actively enhancing the enabling environment for WSP scale up.

Securing collaboration among the broad range of stakeholders involved in WSPs may be challenging, especially when implementing improvement plans that fall outside the direct authority or management purview of the water utility (Ferrero et al., 2018). Another common challenge relates to initial improvement upon WSP adoption that is not maintained over time (i.e., “backsliding”). WSP implementation may be incomplete, fail to translate to expected outcomes, or inadvertently result in harmful outcomes (Kanyesigye et al., 2019; Setty et al., 2017; String et al., 2020). For instance, acute gastroenteritis rates remained steady or worsened following WSP implementation at some locations (Setty et al., 2017, 2018a, 2018b), where multiple management changes took place concurrently (e.g., improving maintenance of chlorination while also retaining turbid water to augment quantity). Altering disinfection methods to reduce pathogens, if not done cautiously, could at the same time lead to increases in exposure to carcinogenic disinfection byproducts (Setty et al., 2017). Consistent vigilance, buy-in, trialing and adjustment, accessibility of procedures, and ease of use are critical to sustain gains achieved through these programs.

Finally, impact assessment research has been carried out unevenly throughout the world, highlighting research imbalances between the Global North and the Global South. In particular, limited evidence is available from the African continent, with only two studies carried out in rural settings in South Africa (Mudau et al., 2017) and the Democratic Republic of Congo (String et al., 2020) and one countrywide study in urban settings in Uganda (Kanyesigye et al., 2019).