Cryptosporidium oocysts are around four to six micrometres across, are shed in enormous numbers by an infected host, and remain infective in cool water for weeks. They are also highly resistant to chlorine at the concentrations and contact times used in routine drinking water disinfection, a property demonstrated in classical laboratory work comparing disinfectants (Korich et al., 1990). A supply that depends on chlorination alone, without effective filtration or protection of the source, therefore has no reliable barrier against this organism. The outbreak in Milwaukee, in which a very large number of people were affected after oocysts passed through a treatment plant, remains the reference demonstration of what follows from that gap (Mac Kenzie et al., 1994).

Reviews of waterborne protozoan outbreaks worldwide show that Cryptosporidium and Giardia together account for most reported events, and that the proportion attributed to Cryptosporidium has risen as diagnostic capacity has improved (Karanis et al., 2007; Efstratiou et al., 2017). The burden falls most heavily on young children in low and middle income settings, where cryptosporidiosis is a leading attributable cause of moderate to severe diarrhoea and is associated with lasting effects on growth (Kotloff et al., 2013; Checkley et al., 2015).

Rural piped supplies in monsoon climates present a recognisable combination of hazards: shallow intakes or boreholes close to grazing land, intermittent supply with recurrent loss of pressure, distribution pipes laid in shallow trenches that flood, and minimal treatment. Heavy rainfall mobilises faecal material from both human and livestock sources and can carry it into a source or into a network. Investigations of such clusters are often delayed because the first cases present to scattered rural facilities and because the laboratory work needed to confirm a water source is unavailable locally. This article proposes a protocol intended to be workable under those conditions. It is a proposal: it has not been applied to an outbreak in the course of preparing this article, and no results of its use are reported.

Three bodies of work shaped the protocol. The first is the outbreak investigation literature for waterborne protozoa, which has repeatedly shown that the epidemic curve and the spatial distribution of cases identify the implicated supply earlier than any laboratory result, and that control measures should be decided on that basis (Karanis et al., 2007). The second is the laboratory literature on detection and typing. Oocysts are recovered from large volumes of water by filtration, immunomagnetic separation and immunofluorescence microscopy, a sequence standardised in national regulatory methods (US EPA, 2012), while clinical diagnosis has moved from microscopy of stained smears towards antigen detection and nucleic acid amplification (Chalmers and Katzer, 2013). The third is the water safety planning framework, which treats an outbreak as evidence of a failure in a system of barriers rather than as an isolated event (Bartram et al., 2009; WHO, 2022).

Species and subtype identification deserves particular emphasis because it is what distinguishes a plausible waterborne attribution from a demonstrated one. Cryptosporidium hominis is essentially restricted to human transmission, whereas C. parvum circulates in ruminants as well as people, so the species alone narrows the likely source (Hunter and Thompson, 2005). Within a species, sequencing of the gp60 gene resolves subtypes at a level fine enough to link human cases to one another and to an environmental isolate (Alves et al., 2003; Xiao, 2010). Table 1 groups the seven steps of the protocol and gives the record each is intended to generate.

The seven steps are described below in the order in which they would normally begin, although several run concurrently. Figure 1 shows the sequence and the point at which the investigation may need to widen.

The protocol is a proposal and has not been evaluated. Three questions would need answering before it could be recommended as routine. The first is feasibility: whether a district team without a specialist laboratory can complete steps one to four within the period during which evidence is still recoverable, and what the realistic turnaround is for referred specimens. The second is yield: how often the typing step changes the conclusion reached from the descriptive work alone, since a step that rarely changes a decision is hard to justify where resources are scarce. The third is retention, because the value of the final step depends on whether the revised water safety plan is still being followed at the next flood.

Two cautions apply to any use of the protocol. Negative findings in water carry very little evidential weight, and an investigation that treats them as exculpatory will reach the wrong conclusion. And a cluster detected in a rural supply is likely to be a fraction of the true number of infections, because mild cases do not present and many facilities do not test for protozoa unless asked. The protocol is designed to identify and interrupt a source, not to measure the size of an outbreak.

A suspected waterborne cryptosporidiosis cluster can be investigated with modest means if the work is sequenced sensibly: verify, describe, inspect, sample, type, act and report, with control measures taken on descriptive evidence rather than held back for laboratory confirmation. The protocol set out here collects that sequence into a form that a district team can follow and states the output expected at each step. Its practical value is an open question, and testing it during real investigations, with attention to timing and to the yield of the typing step, is the natural next piece of work.

To be declared by the author(s) before publication.

To be declared by the author(s) before publication.