When a Better WASH Score Is Not Yet a Safety Result
This series examines how water, sanitation, hygiene, environmental cleaning, and health-care waste management are measured in health facilities—and what those measurements can honestly tell us about infection prevention. The immediate question is simple but consequential: when a facility receives a better WASH score, has it actually become safer for patients and staff, or has one layer of readiness merely been documented? The answer requires a distinction between a condition that makes safer care possible and an outcome that demonstrates it happened.
WASH is part of the care system, not a decorative facility feature
A hospital cannot deliver safe care with clinical knowledge alone. Water is needed for many routine services and for cleaning. Toilets and sanitation systems have to contain human waste rather than move pathogens into the care environment. Hand-hygiene facilities have to be available where care occurs and where toilets are used. Environmental cleaning reduces opportunities for contamination to persist on surfaces and equipment. Waste management has to separate, treat, and dispose of infectious and sharp materials safely.
These are not interchangeable tasks, and they do not protect patients through exactly the same route. Together, however, they form part of the setting in which infection prevention and control (IPC) is supposed to work. The 2026 WHO/UNICEF Joint Monitoring Programme (JMP) report places WASH in that wider frame: basic services support safe, dignified care, patient safety, maternal and child health, and IPC. A facility without reliable water or safe waste handling faces constraints before a clinician even begins a procedure.
That importance does not make every WASH measure a clinical outcome. A service ladder may record whether an improved water source is on the premises, whether a functional hand-hygiene station exists, whether cleaning protocols and staff training are present, or whether waste is segregated and safely treated. Those observations answer a readiness question: does the facility have a defined condition that should support safer care? They do not directly answer whether workers used the facilities at the right moments, whether the water remained available through an interruption, whether surfaces were actually cleaned, or whether infection incidence fell.
This difference is easy to lose because the words “good,” “basic,” and “improved” sound like conclusions. In measurement, they are classifications. A classification can be useful without being the final result.
The dashed link is the central boundary. A score can tell decision-makers where to look and what to improve. It cannot skip the intermediate links and become proof of infection reduction by itself.
The revealing puzzle in the numbers
A study of Indonesian hospitals makes the distinction concrete. The authors analyzed 532 hospitals from the 2019 national Health Facilities Research dataset. They used available variables as proxies for six WASH-related services—water, sanitation, waste management, hand hygiene, environmental cleaning, and management—and divided the resulting scores into poor, adequate, and good categories. In this study, “good” meant a score above 75 percent.
At first glance, the profile looks reassuring. The good category covered 97.37 percent of hospitals for water services, 84.02 percent for sanitation, 90.60 percent for hand hygiene, 79.32 percent for environmental cleaning, and 80.64 percent for management. But only 13.72 percent of hospitals were in the good category for waste management; 78.57 percent were adequate and 7.71 percent poor. A headline such as “most hospitals had good WASH” would hide the operational problem that the domain profile makes visible. A facility can have water and hand-hygiene availability while still struggling to segregate, treat, or dispose of health-care waste.
The case also shows why a score should be read alongside the conditions in which it was produced. The Indonesian study found differences by ownership, hospital type, level, and region. Government hospitals generally had better results in several domains than private hospitals. General hospitals tended to have stronger sanitation, waste management, and management services than specialized hospitals. Lower-level hospitals had worse conditions across several services. These patterns may reflect differences in budgets, staffing, infrastructure, oversight, and referral roles as much as they reflect WASH management itself.
That is not an argument for ignoring the score. It is an argument for asking what else the score may be carrying. A high score can partly signal a better-resourced institution. A low score can identify a neglected bottleneck, but it may also be entangled with remoteness, hospital size, staffing, and the kinds of patients the facility receives. The score is informative; it is not self-interpreting.
The global picture reinforces the same caution from another angle. The 2026 JMP report estimates that, in 2025, 85 percent of health-care facilities had basic water services, compared with 40 percent for basic sanitation, 72 percent for basic hygiene, 59 percent for basic environmental cleaning, and 71 percent for basic waste management. These are important monitoring estimates, but their evidence bases differ: the covered share of the global population was 52 percent for water and waste, 48 percent for sanitation, 46 percent for hygiene, and 40 percent for environmental cleaning. They describe the availability of defined services, not the percentage of facilities that have prevented infections.
The Indonesian study and the JMP report should therefore be read as complementary evidence, not stitched into a single trend. The former used a national hospital dataset and proxy scoring; the latter uses standardized service ladders and a broader collection of national sources. Their years, samples, thresholds, and definitions differ. A “good” proxy score in one study is not automatically the same thing as “basic” service in the other. Comparing them can reveal why definitions matter, but it cannot prove that the same hospitals improved or deteriorated over time.
Five words that keep the claim honest
The most useful way to read a WASH result is to keep five layers distinct.
Availability asks whether a facility has a relevant source, toilet, station, protocol, or waste system. This is often the layer most visible in a survey. It matters because an absent facility cannot be used. But presence is only the starting condition.
Functionality asks whether the equipment or service works when needed. The JMP report notes that its basic water indicator does not fully capture sufficiency, reliability, or quality. A facility may have water on its premises yet lack enough storage for an interruption. An incinerator may exist yet be unusable or lack fuel. A hand-hygiene station may be installed yet empty of soap or alcohol-based hand rub.
Practice asks what staff, patients, and caregivers actually do. The Indonesian study could assess facilities but not compliance with handwashing or environmental-cleaning protocols. A written procedure does not show that cleaning occurred at the required frequency, that hand hygiene occurred at the relevant moments, or that waste was handled safely under pressure.
Intermediate control asks whether those practices change the environment in the intended direction. This may include effective cleaning, safe waste treatment, reduced contamination, and functioning IPC processes. It is closer to the pathway that could affect infection risk than a checklist is, but it is still not the same as a patient outcome.
Outcome asks what happened to patients: for example, a defined infection measure or maternal deaths attributed to sepsis. The Indonesian study examined the latter association, not a direct measure of all health-care-associated infections. Its results therefore matter to the safety question without settling it.
The distinction also changes how surprising results should be read. In the Indonesian regression, adequate sanitation and adequate or good hand hygiene had positive associations with the proportion of maternal mortality due to sepsis, while environmental cleaning and management showed negative associations at several levels. The authors themselves caution that the sanitation variables captured the availability of particular toilets, not safely managed sanitation, and that the hand-hygiene measures did not show whether people used the facilities properly and frequently. A counterintuitive coefficient is not evidence that hand hygiene is harmful. It is evidence that the measured proxy, the outcome, and the surrounding differences between hospitals do not form a simple causal test.
The study also reports that its model explained only 15.5 percent of the variation in maternal sepsis mortality proportion and 4.9 percent of the variation in patient satisfaction. Patient satisfaction may reflect medical, administrative, and other services rather than WASH alone. These limitations narrow the claim; they do not erase the practical value of finding a waste-management weakness or an environmental-cleaning gap.
The conclusion a score can support
A better WASH score can support several serious conclusions. It can indicate that a facility has moved closer to a defined readiness standard. It can reveal a domain that deserves resources. It can help managers ask whether the service is reliable, supplied, staffed, and used as intended. It can create a common language for comparing facilities when the definitions and denominators are genuinely aligned.
It cannot, on its own, support the stronger statement that infections have been reduced. For that, the evidence must continue through the chain: the service works, the people responsible use it, the environment and IPC processes improve, and a defined patient outcome changes after the intervention. The supplied studies show why each link matters and why the links should not be collapsed. They do not provide a controlled intervention estimate that closes the chain.
That is the question to carry into the next stage of this series: what do WASH measurement frameworks actually count, and which parts of safe operation remain outside their thresholds? Until those boundaries are clear, “better” is a useful signal—but not yet a safety result.