SECTION I
HEALTH RISKS FROM WATER AND NEW CHALLENGES FOR THE FUTURE
Waterborne outbreaks of cryptosporidiosis
Epidemie di cryptosporidiosi trasmesse con le acque
Rachel M. Chalmers
Cryptosporidium Reference Unit, Public Health Wales, Singleton Hospital, Swansea, UK
ABSTRACT
Water is the most commonly reported vehicle of transmission in Cryptosporidium outbreaks. While mains drinking water quality is highly regulated in industrialised countries, treated recreational water venues remain highly variable and these have emerged as important settings in the transmission of cryptosporidiosis. Epidemiological investigations of outbreaks benefit from supplementary microbiological evidence and, more recently, the application of molecular typing data to link isolates from cases to each other and to suspected sources. This article documents how waterborne Cryptosporidium outbreaks are identified and reported, how such outbreaks have acted as drivers of regulatory change, and some of the recent developments in the detection and investigation of these outbreaks and their spread, especially the application of molecular typing assays.
Key words: Cryptosporidium, genotyping, recreational, drinking, water.
RIASSUNTO
L'acqua è il più comune veicolo di trasmissione di epidemie dovute a Cryptosporidium. Mentre la qualità delle acque potabili distribuite da acquedotti è fortemente regolamentata nei paesi industrializzati, la qualità delle acque trattate di ambienti ricreativi chiusi è fortemente variabile e questi ambienti si sono rivelati importanti nella trasmissione della criptosporidiosi. Le indagini epidemiologiche sulle epidemie traggono beneficio dalle evidenze microbiologiche e, più recentemente, dall'applicazione dei dati di tipizzazione molecolare per collegare i ceppi isolati con i casi d'infezione e con le fonti di contagio sospette. Quest'articolo documenta come vengono individuate e notificate le epidemie di Cryptosporidium trasmesse con le acque, come tali epidemie hanno agito da guida per migliorare la normativa, ed alcuni recenti sviluppi nella rilevazione e nelle indagini di queste epidemie e della loro diffusione, in particolare l'applicazione dei saggi di tipizzazione molecolare.
Parole chiave: Cryptosporidium, genotipizzazione, acqua ad uso ricreativo, acqua potabile.
INTRODUCTION
Human infection with the protozoan parasite Cryptosporidium causes the gastrointestinal disease cryptosporidiosis. Of the 25 or so species currently recognised, 15 have so far been reported in humans, of which some are established as human pathogens: C. parvum, C. hominis (which are the most commonly detected species in human cryptosporidiosis worldwide) and C. meleagridis are supported by human infectivity and clinical outcome data from feeding trials in adult volunteers [1- 3], C. cuniculus (formerly the rabbit genotype) caused a drinking waterborne outbreak in the United Kingdom (UK) [4], and C. felis and C. canis were associated with diarrhoea in children in a shanty town in Peru [5]. Dose response studies have shown similar ranges for some C. parvum isolates compared with C. hominis, and small numbers (< 10) of parasites ingested can cause disease [1, 2]. Other Cryptosporidium species are rarely reported human infections, or have never been found in humans, and many are considered adapted to farmed animal or wildlife hosts [6] (Table 1).
Transmission is by the faecal-oral route, from either humans or animals, depending on the Cryptosporidium species; for example, C. hominis has a human infection cycle while C. parvum also has susceptible animal hosts causing mainly gastrointestinal disease in young ruminants. The natural host for C. cuniculus is the European rabbit (Oryctolagus cuniculus) [13] and for C. felis, cats and C. canis, dogs. Although C. meleagridis was originally identified in farmed turkeys [14] current distribution and risk factors for human acquisition are not known; many cases report no contact with birds, the parasite species has a wide host range and other bird-restricted species are not considered a threat to human health. Although some Cryptosporidium spp. are highly infectious person-toperson, it is the parasite's ability to survive in the environment and its resistance to chlorine disinfection that support transmission via drinking and recreational waters, and other vehicles such as food. Table 2 shows the human risk factors for acquisition of Cryptosporidium spp. and how these relate to settings where outbreaks have occurred.
Symptoms of cryptosporidiosis, which usually occur between 2 to 12 (usually 5 to 7) days after ingestion of oocysts (the transmissive stage of the life cycle), include watery diarrhoea, abdominal pain, nausea and/or vomiting, low grade fever and malaise, and may last for up to three weeks during which time apparent recovery may be followed by temporary recurrence [15]. Oocysts may continue to be shed in faeces for many days after symptoms have ceased. About 10% cases may be hospitalised. Treatment is by supportive therapy to prevent dehydration; there is no licenced specific treatment in the EU, although nitazoxanide is licenced by the United States (US) Food and Drug Administration for immunocompetent patients over 1 year of age. Severely immunocompromised patients, especially those with T-cell immune deficiencies, risk chronic or intractable disease and infection of sites other than the gastrointestinal tract [16]. Immune reconstitution is important in these patient groups, as treatment modalities are otherwise undefined, and prevention of infection through risk reduction is paramount. For example, since the late 1990's, the Department of Health in England advised that those with compromised T-cell function should boil all drinking water (including bottled water) [17]. However, whether, given improved drinking water quality and reduction in Cryptosporidium risk since then [18], this permanent blanket advice remains necessary there is currently under review. Also of great concern are the longer term sequelae of infection in both the general population and in malnourished children in whom, even following asymptomatic infection, reduced cognitive function and failure to thrive have been reported (reviewed by Putignani and Menichella [19]). Even in non-immune compromised people, there is some evidence to suggest there may be different long term health effects depending on infecting species: for example, those cryptosporidiosis patients infected with C. hominis (but not C. parvum) were more likely to report joint pain, eye pains, headaches and fatigue in the two months following infection than controls [20]. The relationship between clinical outcome in terms of severity of acute disease and long term sequelae requires further investigation.
Although people of any age can become infected, most cases of cryptosporidiosis are reported in children under 5 years largely due to intestinal tract immaturity and lack of mucosal immunity [20]. In non-industrialised countries, the peak incidence is in infants less than 1 year old with adult cases rarely identified, and in industrialised countries mainly in 2-4 year olds with a second, smaller peak in adults of child-rearing age. This is likely due to differences in general hygiene and feeding practices for infants, and age-related immunity generated by repeated exposure (for example, low-level exposure through drinking water) in older age groups. However, any immunising effect would come at a high risk to the health of the most vulnerable sectors of the population: young children and immunocompromised patients. The epidemiology of human cryptosporidiosis globally has been reviewed most recently by Putignani and Menichella [19].
The clinical problems of cryptosporidiosis described above contributed to the inclusion of the parasite in the World Health Organisation's (WHO) Neglected Diseases Initiative in 2004 [21]. Further, large waterborne outbreaks have highlighted the clinical and economic importance of Cryptosporidium. Waterborne cryptosporidiosis outbreaks are more commonly reported than outbreaks involving other vehicles; up to the end of 2010, a total of 185 outbreaks had been identified and reported globally [22, 23], contrasted with less than 20 foodborne outbreaks [24]. This is only partly because of the features of Cryptosporidium favouring waterborne transmission, as some of these also favour the foodborne route:
- multiple hosts for some human pathogenic species, for example, C. parvum (mainly humans and young ruminants);
- ubiquitous distribution (Cryptosporidium spp. occur worldwide);
- large numbers of oocysts (1010), the transmissive stage, are shed by susceptible hosts during acute infection;
- oocysts are shed containing fully infective sporozoites no secondary hosts or maturation conditions are required;
- transport vectors may provide further distribution of oocysts within or to the acquatic environment;
- small size of oocysts (4-6 µm for human-infective species) means they may pass between sand grains in filter beds, although other forces including sedimentation and adsorption also interplay during filtration, and application of a coagulant (flocculant) improves removal;
- oocysts can be discharged in sewage effluent in significant numbers;
- oocyst are robust and can survive for months in cool, moist environments; they also survive chlorination and;
- small numbers of ingested oocysts can cause disease [25].
Furthermore, the large scale of some drinking waterborne cryptosporidiosis outbreaks (the Milwaukee outbreak in 1993 involved an estimated 403 000 cases of cryptosporidiosis [26]) led to greater emphasis on monitoring and intervention of water supplies, and greater awareness and investigation of water as a transmission vehicle, including recreational waters such as swimming pools. There is a perception of cryptosporidiosis as a waterborne disease, perhaps at the expense of investigation in other possible routes and investment in their interventions. However, where chlorine disinfection is widely used to treat drinking water, Cryptosporidium is indeed one of the most common waterborne pathogens. For example, in a review of 89 waterborne outbreaks of infectious intestinal disease (IID) involving 4321 cases in England and Wales, Cryptosporidium was the causative agent in 69% [27]. In recreational waters, Cryptosporidium is the leading microbial cause of outbreaks in both the UK and USA [27, 28]. Despite global distribution of Cryptosporidium, reports of waterborne outbreaks are weighted heavily towards industrialised countries in the continents of Australia (especially New Zealand), Europe (especially UK and Ireland) and North America [19, 23]. This distribution may reflect national variations in surveillance, reporting, monitoring, and investigation of cases and outbreaks, as well as highlighting risk factors such as intensive stocking of farmed animals, environmental contamination, weather conditions and events, discharge of sewage effluent into drinking water sources, and the use of surface water sources. The pathways and host, parasite and environmental factors that determine the risk of infection, and thus public health outcomes, are shown in Figure 1, and how they contribute to waterborne Cryptosporidium outbreaks has been reviewed previously [for example, see 19, 22, 29]. Risebro and colleagues used a fault tree analysis to examine the contribution and interpretation of events in outbreaks in the EU occurring between 1990 and 2005 [29]. Of 31 protozoal drinking water outbreaks, 29 were Cryptosporidium, and most of these outbreaks were attributed to chronic filtration failures or livestock and rainfall in the catchment, contributing concurrently in 11 outbreaks. Interestingly, the most recent outbreaks in the UK have been attributed to human or wildlife sources (Table 3). It is now possible, using molecular methods, to differentiate the Cryptosporidium species found in water samples and supplement catchment data to track the source of contamination to humans, farmed animals or wildlife and to assess the level of risk posed to public health beyond that previously possible from oocyst counts alone.
The aim of this article is to document how waterborne Cryptosporidium outbreaks are identified and reported, how such outbreaks have acted as drivers of regulatory change, and some of the recent developments in the detection and investigation of these outbreaks, especially the application of molecular typing assays.
IDENTIFICATION OF CRYPTOSPORIDIUM OUTBREAKS
Waterborne cryptosporidiosis outbreaks are identified through various surveillance systems, including passive and active morbidity reporting such as syndromic surveillance, laboratory reporting, sentinel surveillance systems, drug purchase or prescribing data, and media broadcasts. Poor water quality results, incidents or treatment failures will alert providers and authorities to a potential or existing outbreak. For example, following a water quality incident in the East Midlands, England in 2008 (Table 3; outbreak number 08/278), when oocysts were detected in treated water with a long history of non-detects in routine monitoring, syndromic surveillance data collected under the Qsurveillance scheme showed a significant increase in general practictioner consultations for diarrhoea and gastroenteritis in the week of the incident in the water distribution areas, compared with no increase in the unaffected areas [30]. Of the 33 cases of laboratory confirmed cryptosporidiosis identified during the outbreak investigation, 23 were C. cuniculus, the outbreak species (see below) [4] but QSurveillance data estimated an excess of 422 diarrhoea cases during the outbreak, an increase of about 25% over baseline weekly levels [30]. Thus, syndromic surveillance described the extent of cryptosporidiosis in the general population and provided reassurance that there was no further widespread impact.
The symptoms of cryptospordiosis are non-specific, so laboratory detection of the parasite, usually in stools, is required to confirm the infection. This is either by microscopy with prior staining using tinctorial, fluorescent or immunofluorescent stains, or immunoassay-based test kits, or more recently molecular assays. Routinely, detection in both clinical diagnostic and water testing laboratories, is of the genus: species identification can only be done by specific molecular "genotyping" assays, and is not usually performed by primary diagnostic laboratories. Not all diarrhoea patients will seek medical attention, have a stool sample taken or a Cryptosporidium test applied, and laboratory practice is varied. A request for "ova, cysts and parasites" will not include a test for Cryptosporidium: where not routinely sought, this must be specified on the request form. Cryptosporidium diagnosis is statutorily notifiable in only a few countries and cross-country comparisons are hampered by this [31]. For example, Cryptosporidium is included in Directive 2003/99/EC of the European Parliament and the Council of the European Union (EU), and cryptosporidiosis is therefore a notifiable disease within the EU with laboratory-confirmed case data collected through the European Surveillance System (TESSy). In 2009, reports were provided by 21 out of 31 EU and European Economic Area/European Free Trade Association countries; 8016 cases were reported across 13 countries and zero cases were reported by eight countries, an overall case rate of 2.7 per 100 000 population. The highest confirmed case rate was reported in Ireland (10.0 per 100 000 population) followed by the UK (9.3 per 100 000) and Belgium (4.1 per 100 000) [32]. Thus, Cryptosporidium is under diagnosed and underreported, but to varying extents. Even at a local level, testing and reporting practice is variable [33]. One study of IID in the UK has estimated the reporting ratio (i.e. the ratio of disease rates in the community and presenting to general practice relative to the rate of reported diagnoses to national surveillance) to be 8.2 (95% CI 2.1 to 31.7), estimating the annual number of cases in the community to be 43 834 (95% CI 11 393 to 168 655) [34]. This equates to an estimated annual incidence of 69.5 cases per 100 000 UK population, and the authors acknowledge the study may have underestimated Cryptosporidium rates as the case definition for acute gastroenteritis excluded cases of duration of illness over two weeks.
Surveillance data for England and Wales have shown that about 10% of reported cases of cryptosporidiosis are part of identified and reported outbreaks at a variety of settings (Table 1) [35], using the following definitions of an outbreak:
- an incident in which two or more people experiencing a similar illness are linked in time or place; or;
- a greater than expected rate of infection compared with the usual background rate for a place and time.
In investigations involving water, an incident may be defined as a suspected, anticipated or actual event involving microbial or chemical contamination of food or water. The Health Protection Agency (HPA) Centre for Infections (formerly the Communicable Disease Surveillance Centre) and local authorities in England and Wales have conducted structured surveillance of outbreaks of IID since 1992 [36]. Outbreaks are classified, according to the robustness of epidemiological and microbiological evidence, as definite, probable or possible defined from the following criteria:
a. the pathogen found in patients was also found in water samples;
b. documented water quality or treatment failure;
c. significant result from analytical epidemiological study, demonstrating association between water and illness;
d. suggestive evidence that outbreak is water related from a descriptive epidemiological study, excluding obvious alternative explanations;
Which are combined to indicate the strength of association:
strong = a+c or a+d or b+c;
probable = b+d or a only or c only;
possible = b only or d only [37]. Outbreaks were summarised bi-annually until 2006 in the Communicable Disease Report, and are now reported in the Health Protection Report (www.hpa.org.uk).
A similar classification is in use in the USA where the Centers for Disease Control and Prevention (CDC), the Environmental Protection Agency (EPA), and the Council of State and Territorial Epidemiologists have collaborated on the Waterborne Disease and Outbreak Surveillance System (WBDOSS) for collecting and reporting data on waterborne disease outbreaks since 1971 for drinking water and 1978 for recreational water [28]. Two criteria must be met for a health event to be defined as a waterborne disease outbreak: 1) two or more persons must be linked epidemiologically by time, location of exposure to water, and illness characteristics, and 2) the epidemiologic evidence must implicate water or volatilization of water-associated compounds into the air surrounding the water as the probable source of illness. Outbreaks are classified according to the strength of both epidemiologic and clinical laboratory data, and environmental data implicating water as the vehicle of transmission:
Class 1 = epidemiologic and clinical laboratory data and environmental data are provided and are adequate;
Class 2 = epidemiologic and clinical laboratory data are provided and are adequate, but environmental data is not provided or are inadequate;
Class 3 = epidemiologic and clinical laboratory data are provided but are limited, and environmental data are provided and are adequate;
Class 4 = epidemiologic and clinical laboratory data are provided but are limited, and environmental data is not provided or are inadequate.
These classes, first delineated in the 1989-1990 surveillance report [38], have been updated to reflect the increasing use of molecular characterization of pathogens both in clinical specimens and environmental samples; molecular data that link people who had an identical water exposure are considered adequate to support a Class I or Class II assignment, and molecular data that link at least one person to the implicated water exposure now are considered adequate water quality data to support a Class I or Class III assignment [28]. The use of molecular data in the investigation of particular incidents and outbreaks is described below. However, as with many pathogens, the Cryptosporidium contamination event may have passed by the time of sampling and/or recognition of an outbreak. For Cryptosporidium the incubation period between ingestion and the onset of illness is long (up to two weeks) and the interpretation of microbial and water treatment process data requires careful interpretation. Differences in Cryptosporidium monitoring strategies for drinking water in the UK and USA are explored further below.
DRINKING WATERBORNE OUTBREAKS AS DRIVERS OF REGULATION
Although the first human cases of cryptosporidiosis were reported in 1976 [39, 40], the first outbreak linked to drinking water was identified in 1984, at Braun Station, Texas, USA [41]. Over 200 individuals were involved, when sewage contaminated a groundwater supply. The next outbreak was on an even larger scale: in 1987 an estimated 13 000 people were affected at Carrollton, Georgia, USA when mains water became contaminated from a surface supply during a period of operational irregularities at the conventional water treatment works [42]. In both cases the drinking water met existing water quality standards, based on WHO Guidelines for Drinking Water Quality, focussing on monitoring E. Coli. At the end of 1988 and into 1989, over 500 individuals were affected in Swindon and Oxford, England, when oocysts in contaminated surface source water broke through the conventional mains water treatment and in to supply [43]. These early outbreaks startled both public and water industry perception of drinking water in industrialised nations; filtration and chlorine disinfection were commonplace and assumed to control waterborne disease, and people no longer commonly became sick through the mains water supply. In 1989, Cryptosporidium was not even considered in the US Environmental Protection Agency (EPA) Surface Water Treatment Rule to control Giardia and viruses or in the companion Guidance Manual for Compliance with the Filtration and Disinfection Requirements for Public Water Systems, published in 1990. The 1993 outbreak in Milwaukee focussed the work initiated to understand the sources, routes of transmission, detection and prevention of spread of the parasite.
In addition to work seeking improvements in methods to detect, and water treatment to control Cryptosporidium, the outbreaks in the 1980's and early 1990's prompted regulatory agencies to develop rules for public health protection through monitoring, removal or inactivation. This has been approached differently in USA and UK.
In the USA, Cryptosporidium monitoring was first required under the Information Collection Rule (ICR) of 1997, requiring surface water sources of supplies to more than 100 000 people to be monitored for Cryptosporidium oocysts, Giardia cysts and viruses for 18 months. Sources with > 1000 of either protozoan per 100 L, or viruses > 100 per 100 L, required final water monitoring. A total of 93% of 5829 surface water samples were reported as nondetects for Cryptosporidium. However, results based on testing small sub-samples and extrapolation of counts to the original volume are now considered unreliable. During the ICR period, a standard method was published by the EPA, stipulating the acceptable sampling filter types, elution and oocyst recovery processes and microscopical detection procedures for enumeration of oocysts, and, critically, that the oocyst count be expressed per volume of water sampled. In 1999 the method was validated for simultaneous detection of Cryptosporidium oocysts and Giardia cysts, currently published as EPA Method 1623: Cryptosporidium and Giardia in Water by Filtration/IMS/FA, 2005. Method 1623 now supports promulgation of EPA's Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR), which is the current regulation for all public water systems that use surface water or ground water that is under the direct influence of surface water, finalised in 2006. This rule requires monitoring of source waters to determine the level of treatment required for Cryptosporidium reduction by removal or disinfection. Mean oocyst counts, based on a two year, monthly sampling programme, are used to classify ("bin") supplies in one of four categories and determine the extent of treatment required, if any, above conventional full treatment. Systems classified in higher bins must provide additional water treatment to further reduce Cryptosporidium levels by 90 to 99.7 percent (1.0 to 2.5-log), depending on the bin, using treatment and management options in a "microbial toolbox". All unfiltered water systems must provide at least 99 or 99.9 percent (2 or 3-log) inactivation of Cryptosporidium, depending on the results of their monitoring. Suitable removal is through filtration provided by granular media, cartridge filters or membranes, and approved disinfectants effective against Cryptosporidium are chlorine dioxide, UV, and ozone. Systems must conduct a second round of monitoring six years after completing the initial round to determine if source water conditions have changed significantly. The EPA estimates that full compliance with the LT2ESWTR will reduce the incidence of cryptosporidiosis by 89 000 to 1 459 000 cases per year, with an associated reduction of 20 to 314 premature deaths. Additional expected benefits include reduced exposure to other pathogens, such as Giardia, that can co-occur with Cryptosporidium.
In the UK, the Oxford and Swindon outbreak in 1987-1988 led to the establishment of the group of experts, chaired initially by Sir John Badenoch and latterly by Professor Ian Bouchier. A series of reports, published in 1990, 1995 and 1998 set out what was known about Cryptosporidium, its occurrence in the environment, its importance as a waterborne infection for humans, the outbreaks and lessons learned from them, the treatment requirements for Cryptosporidium deficiencies in detection and enumeration methods, the need for methods to establish oocyst viability, characterise the species present and their infectivity for humans, the risks from groundwaters infiltrated by surface waters, and established good epidemiological practice in investigation of outbreaks and the need for clarification in the role and function of incident and outbreak control teams and their members. A correlation between outbreaks and inadequacy occurring in elements of the multi-barrier approach (source, treatment, distribution, monitoring and response) was identified in the third report (available online at www.dwi.defra.gov.uk/research/bouchier/index.htm). This was initiated following an outbreak in north west London and Hertfordshire in 1997 when 345 cases were reported following contamination of a groundwater source, a type previously considered to present a low risk from Cryptosporidium, by infiltration of surface water containing oocysts [44].
Although the legal standards for all EU member states are set out in the European Drinking Water Directive 1998, and are themselves informed by the WHO guidelines on drinking water quality, national standards are also set. The first, and only, regulatory requirements for Cryptosporidium in finished water anywhere in the world were introduced in England and Wales in 1999 as part of the Water Supply (Water Quality) (Amendment) Regulations 1999 and incorporated in to the Water Supply (Water Quality) Regulations 2000 in England and 2001 in Wales. The key driver for these regulations was the lack of admissible evidence to prosecute the water company associated with an outbreak in 1995 in Torbay, south west England, when 575 cases occurred associated with a lowland river with direct abstraction and bankside infiltration of unfiltered water [45, 46]. The regulatory requirement mandated water undertakers to conduct risk assessments with respect to Cryptosporidium on all water treatment works, considering the source water, catchment characteristics and treatment provided. Sites with a "significant risk" classification, had to treat the water to ensure an average of less than 1 oocyst in 10 L of treated water supplied, measured by continuous sampling of at least 40 litres of water per hour. Compliance was demonstrated by continual monitoring and reporting of results to the regulator, the Drinking Water Inspectorate (DWI), unless all particles > 1µm were continuously removed. The initial risk assessment reported in the DWI's annual report in 2001 (http://dwi.defra.gov.uk) identified 332 sites as being at significant risk, of which 158 were works treating surface waters and 174 were groundwater abstractions. Some sites were decomissioned and others subjected to improvement. In 2004, the DWI reported that none of the 146 307 samples taken between 2000 and 2003 exceeded the treatment standards, with most samples below 0·02 oocysts/10 L.
Improvements in drinking water quality driven in part by the England and Wales regulations, and by long-term water company investment programmes, led to a substantial reduction in both Cryptosporidium cases, especially in the first half of the year, and in reported mains drinking water outbreaks [18, 47] at a time when swimming pool outbreaks appeared to increase in number (see below). Only four outbreaks linked to public water supplies have been recorded in England and Wales since 2000; one in 2002, two in 2005 and one in 2008, all with strong evidence for association with drinking water ( Table 3). In November and December 2002 an outbreak involving 31 laboratory confirmed cases caused by C. hominis was reported in a population of 158 558 in South East England served by a mixture of water from a groundwater source and a surface water-treatment plant at significant risk, and where the continuous monitoring samples never exceeded treatment standards (Table 3, outbreak 02/1701) [48]. A second outbreak of 140 laboratory confirmed cases caused by C. hominis in the same area occurred between September and November, 2005 and once again, oocyst counts were below the treatment standard (Table 3, outbreak 05/552). The recognition that even small numbers of oocysts detected in drinking water can cause outbreaks contributed to the revocation of the treatment standard in the amended water quality regulations, The Water Supply (Water Quality) Regulations 2000 (Amendment) Regulations 2007. Furthermore, an outbreak in north west Wales in 2005, involving 218 cases of cryptosporidiosis (again, C. hominis) was linked to a surface water supply derived from a sparsely populated catchment. The treatment works, despite the absence of effective treatment or barriers in the catchment to remove Cryptosporidium, was not continually monitored as it had not been given a significant risk assessment [49]. The outbreak was controlled in the short term by a notice to boil drinking water, which was in place for 9 weeks until a UV treatment plant could be installed, although oocyst disinfection (inactivation) was not permitted by the regulations at the time, which focussed on oocyst removal. The elimination of the treatment standard in the 2007 regulations permitted application of disinfection such as UV for the control of Cryptosporidium in water supplies. Never-the-less, drinking water remains a risk factor for cryptosporidiosis in England and Wales [18].
As demonstrated above, outbreaks of cryptosporidiosis have been reported through drinking water that met WHO guideline microbiological standards and/or the Cryptosporidium treatment standard imposed in England and Wales. Subsequently, a preventive, risk based approach, derived from the food industry [50], in the form of the requirement for a water safety plan [51], now complements microbiological guidelines, and is therefore incorporated in further amendments to the regulations in 2010 in England and Wales as comprehensive risk assessments. A water safety plan is a systematic inventory of all hazards (including Cryptosporidium), an evaluation of the significance of these hazards and of the efficacy of control measures taken, and spans source water catchment, treatment and distribution of water supplies. The risk assessment is supported by testing and enforcement. However, one of the key differences between the USA and UK approaches remains: the former is historically based on source water monitoring for Cryptosporidium to inform subsequent levels of treatment required and the latter has the historical legacy of an emphasis on final water monitoring. Regulations supporting the Drinking Water Directive and water safety plan approach require raw water monitoring to identify risks to deterioration of raw water quality; there is no list of parameters for this purpose as it is up to the water company to assess the risks and monitor and treat accordingly.
SWIMMING POOL OUTBREAKS AND LACK OF REGULATION
Although the WHO Guidelines for safe recreational waters [52] provide a basis for standards, in swimming pool settings, in contrast with drinking water, there is a lack of legislation. The WHO guidelines, which are currently under revision, provide an authoritative referenced review and assessment of the health hazards associated with swimming pools, their monitoring and assessment, and activities available for their control through education of users, good design, construction, operation and management, and address a wide range of types of hazard, including water quality, physical hazards (leading to drowning and injury), contamination of associated facilities and air quality. In the USA, state and local govermnents establish and enforce regulations for protecting recreational water from contaminants but no federal agency has authority over treated recreational water and, apart from legislation to prevent entrapment, no minimum design, construction, operation, disinfection, or filtration standards exist. Swimming pool codes are enforced by individual state and local public health agencies but there is variation in regulation, compliance and enforcement. In the EU, the Bathing Water Directive 2006 sets out quality standards for natural waters designated for bathing but this does not cover treated waters. In the UK, the publication Swimming pool water. Treatment and quality standards for pools and spas [53] provides authoritative guidance and is viewed as best practice. Therefore, in a court of law, swimming pool operators would be, and indeed have been, prosecuted under the Health and Safety at Work etc. Act 1974, and the Management of Health and Safety at Work Regulations 1999, for causing an outbreak by failing to follow this guidance [54].
Globally, recreational waterborne outbreaks of cryptosporidiosis are reported slightly more commonly than drinking waterborne outbreaks: in the seven years between January 2004 and December 2010, out of 120 reported waterborne Cryptosporidium outbreaks, 46% were associated with drinking water and 54% with recreational waters [23]. However, this analysis belies an apparent increase in swimming pool related Cryptosporidium outbreaks seen, for example, in the UK and USA. Between 1992, when surveillance for waterborne IID outbreaks began [36] and 2011, there were 56 outbreaks linked to swimming pools and 32 linked to public water supplies. However, just 29 (34%) of the swimming pool outbreaks were reported in the first ten years (1992 to 2001) compared with 56 (66%) in the following ten years (2012 to 2011) (Gordon Nichols, personal communication of provisional data). In the USA, Cryptosporidium outbreaks at recreational water venues also increased from 29 in 2005-6 to 60 in 2007-8 [28]. While the emergence of Cryptosporidium may be a continuing factor, other contributing factors may include changes in detection, investigation, and reporting of waterborne disease outbreaks driven by improved resources for the WBDOSS and strengthening of the outbreak response [28]. It is likely that recently improved understanding of risks associated with swimming pools and their investigation, has contributed to the apparent rise in outbreaks at these settings.
APPLICATION OF MOLECULAR ASSAYS FOR SPECIES DETERMINATION AND INVESTIGATION OF THE PROPAGATION OF OUTBREAKS
Molecular detection offers some advantages over standard methods such as EPA Method 1623, which are based on oocyst counts by immunofluorescence microscopy, such as detection of small numbers of organisms and potential for genotyping leading to species determination, this has not yet been adopted for operational or regulatory monitoring of drinking water. This is because only those oocysts containing sporozoites, and thus DNA, will be detected and so far, no reliably quantitative molecular method has been validated to replace oocyst counts. However, without assays determining the species/genotype, viability or infectivity, all oocysts detected by microscopy must be assumed to present a public health risk (Figure 1). Viability and infectivity assays have been reviewed recently by Kothavade, who highlights the difficulties in applying these to the small numbers of oocysts often present in water samples, lack of interlaboratory trials and validation and difficulties in interpretation of the data [55]. In contrast, some genotyping assays have been standardised and applied to counted oocysts from microscope slides; thus both sets of data are collected: the oocyst count and the species, improving the data for assessment of risk to public health.
Cryptosporidium genotyping may be undertaken using material fixed and stained on microscope slides generated during Cryptosporidium monitoring by extracting Cryptosporidium DNA and applying conventional polymerase chain reaction (PCR) of the small subunit ribosomal RNA (ssu rRNA) gene prior to sequencing of amplicons (the benchmark method) or, more recently, using real-time PCR platforms (Figure 2). Other methods, such as laser capture microscopy and reverse line hybridization are at developmental stages, and loop-mediated isothermal amplification have yet to be validated independently [55]. The benchmark assay is a specialist test requiring equipment and skill outwith the scope of most routine diagnostic or detection laboratories. The large genetic diversity in Cryptosporidium ssu rRNA sequences from source waters makes analysis complex. For example, an up to date reference data-base to compare accurate DNA sequences and subsequent phylogenetic analysis with proper interpretation are required, coupled with up to date knowledge of host-parasite relationships [56]. Rueker and colleagues have provided a detailed description and discussion of the sequence and phylogenetic analysis process, especially important in the identification of environmental Cryptosporidium isolates [56], providing a benchmark process. An ongoing Water Research Foundation (WaterRF) programme aims to further refine the benchmark assays and develop and validate simplified, rapid, assays for routine use in water testing laboratories for the differentiation of human pathogenic species (for this purpose defined as C. hominis, C. parvum and C. meleagridis) from animal-infective species (WaterRF 4284) (http://waterrf.org) (Figure 2). Although the simplified assay, as proposed, is not likely to differentiate C. cuniculus, this species will be included in the C. hominis detections and may be differentiated by sequencing of PCR products, which is required for definitive results (Figure 2).
Although there is no standard method for subtyping Cryptosporidium species, sequence analysis of the gp60 gene is informative for C. parvum and C. hominis to a certain extent and sequence data can be compared readily [57]. Although the epidemiology of gp60 subtypes was reviewed by Xiao in 2010 [57], new variants are frequently identified. The utility of the analysis has been demonstrated during the investigation of both zoonotic [58] and anthroponotic transmission, as illustrated in the investigation of waterborne outbreaks described below. However, single locus typing will under-estimate diversity, and a standardised, validated, internationally accepted, multi-locus scheme is required for epidemiological investigations of each species [59]. How subtype variation relates to virulence and pathogenicity is unclear; in fact, only putative virulence factors have been identified so far, although gp60 subtypes have been associated with varying severity of illness [60]. Despite this, genotyping has proved to be beneficial to the epidemiological investigation of cryptosporidiosis, and extracted DNA can be stored long-term and re-tested retrospectively.
In non-outbreak situations, the data from typing sporadic cases of cryptosporidiosis provides a baseline against which trends can be identified and changes monitored. Since the improvements in water quality influenced by the 2000/1 regulations in England and Wales, the spring peak identified as being caused by C. parvum declined substantially [61]. Depite this, drinking water remains a risk for illness caused by C. parvum, as identified by Lake and colleagues in a case-control study design to investigate wider environmental and socioeconomic risk factors for human cryptosporidiosis in England and Wales [62]. In contrast to the decline in C. parvum cases in the spring, there has been no reduction in late summer/early autumn cases of C. hominis which predominates at this time of year. This is especially obvious in 2012, and recreational waters are implicated [63]. More swimming pool related outbreaks are caused by C. hominis than C. parvum, although both have occurred in the same outbreak (Table 3), and may reflect multiple episodes of contamination.
When applied to Cryptosporidium isolates from catchment studies or routine drinking water monitoring, genotyping data can also further refine human health risk assessment by differentiating human pathogenic from solely animal-associated species and supports the other information gathered under drinking water regulatory frameworks. For example, a one year survey of Cryptosporidium oocysts detected in the Scottish Water Routine Cryptosporidium Monitoring Programme identified the species or genotypes present in 62.5% of 1042 oocyst-positive slides [64]. A high diversity of Cryptosporidium species and genotypes was present in source (no. = 456) and treated drinking (586) waters, with 2 or more in 16.9% samples; human-pathogenic species were present in fewer samples than non-pathogenic species. In source waters, C. andersoni (which is host-adapted to cattle) was most frequently identified (25.2% samples) followed by C. parvum (11.2%) and C. ubiquitum (7.2%). In drinking waters, C. ubiquitum was most frequent (12.6% samples), followed by C. parvum (4.3%) and C. andersoni (4.1%). In a long-term study of source waters in the agriculturally intensive South Nation River catchment in Ontario, Canada, Cryptosporidium species associated with livestock made up 39% of the total molecular detections, compared with wildlife associated species and genotypes accounted for 55% and C. hominis and C. parvum 1.6%, indicating a small risk to humans [56]. A study in Australia, where recreational activities within 2 km of drinking water sources is prohibited, compared Cryptosporidium species in recreational water and non-recreational water catchments [65]. This revealed a predominance of C. hominis in recreational water catchments which allowed swimming and camping, compared to non-recreational catchments which had a lower prevalence of Cryptosporidium and all the samples genotyped were C. parvum. Increasing population was strongly correlated with an increase in the prevalence of Cryptosporidium recreational catchments, and the data support government policy limiting activities to the outer catchment [65]. The predominance of host-derived species in some catchments, contributes to effective evaluation and selection of management practices to reduce Cryptosporidium contamination in source waters.
Retrospective analysis of clinical isolates can be enlightening, identifying, for example, that the Milwaukee outbreak was caused by C. hominis [66], where previous uncertainty over the origin of the oocysts identified cattle, slaughterhouses, and human sewage as potential candidates [26]. Further analysis of isolates collected in the Torbay outbreak [46] and north west London and Hertfordshire outbreak [44] established that these were also caused by C. hominis [67]. Even where species identification has been undertaken, further analysis of isolates from cases may be required to investigate outbreaks. Retrospective analysis of C. parvum isolates by multilocus fragment typing enabled more accurate mapping of outbreak-related cases to water supply zones [68]. Some nine years after the Milwaukee outbreak, C. hominis isolates, further analysed for variation in the gp60 gene, were identified as indistinguishable from the outbreak isolate, year round in Milwaukee wastewater indicating continuing, stable transmission of human cryptosporidiosis in the city [69]. Analysis of C. parvum isolates from patients linked to a swimming pool outbreak in Stockholm, Sweden in 2002, at three loci (gp60, TP14 and hsp70) identified that two separate outbreaks had occurred simultaneously, as the genotypes segregated people using an outdoor pool from those that swam only in the indoor pool [70].
Latterly, genotyping methods have been applied in "real time" during incident and outbreak investigations and have been demonstrated to assist in attributing sources and establishing correct interventions. In the UK, isolates from clinical cases are genotyped routinely to the species-level, and in some outbreaks additionally from suspected sources (Table 3). For example, during investigation of a drinking waterborne outbreak caused by C. hominis in north west Wales in 2005 (outbreak number 05/790), isolates found in clinical cases were indistinguishable by sequence analysis of the ssu rRNA gene and the gp60 gene from isolates in the surface water source and in the treated water supply to the affected area, adding strength to the epidemiological evidence for the association with drinking water [49,71]. Although a diversity of Cryptosporidium species was identified in the catchment, the outbreak C. hominis subtype, IbA10G2, was only found in source waters under the influence of sewage contamination, and in the tap water, highlighting the need for water companies to throughly document wastewater inputs in surface waters [72]. During a water quality incident in the East Midlands, England in 2008 (Table 3, outbreak number 08/278), isolates in treated water were identified as C. cuniculus before any human illness was identified, and were matched subsequently by sequencing the ssu rRNA, hsp70 and gp60 genes to those from a dead rabbit found in a chlorine contact tank and 23 ensuing human cases [4]. This outbreak established C. cuniculus as a human pathogen. In two out of three drinking waterborne outbreaks in Northern Ireland in a single year, the confirmation of C. hominis re-directed the investigations away from farmed animals and towards raw sewage contamination in one outbreak and wastewater in the other, illustrating how genotyping isolates from cases can also eliminate alternative suspected sources [73].
Secondary spread from initial outbreaks at swimming pools, for example, through other swimming pools and community settings such as nurseries can lead to community wide outbreaks, as has been shown in Japan [74] and in 2007 in Utah, USA [75]. The ability of Cryptosporidium to survive for over 10 days even in properly chlorinated water, the protracted incubation period (commonly 5 to 7 days) and possible diagnostic delay which prolongs the time between infection and epidemiologic link with the source of the outbreak contribute to this, as does behaviour of cases who continue to swim while ill. They can introduce the protozoan to other recreational water venues, as was shown in the Utah outbreak where an estimated 20% of cases swam while ill with diarrhoea and identified approximately 450 potentially contaminated recreational water venues [75]. Subsequent to the Utah outbreak, increases in cases in 2007 were seen in neighbouring states of Colorado, Idaho, New Mexico, and Iowa. Subtyping 57 C. hominis isolates at the gp60 gene identified the same, previously rare subtype (IaA28R4) in 40 (70%) samples [76]. Unfortunately, none of the Utah samples were available for typing, and it is impossible to know whether this increase was related. This would have been a powerful investigation in to the spread of cryptosporidiosis and highlights the importance of the availability of isolates for further investigation.
Thus, the benefits of genotyping in outbreaks are:
- characterising the outbreak in terms of the cause of the cryptosporidiosis;
- linking cases with each other and monitoring their spread, or excluding cases with non-related isolates;
- linking cases with isolates in suspected sources, or excluding other sources;
- providing supporting data for further investigation of the source of contamination;
- providing additional evidence for the strength of evidence for the association with water;
- providing new data for prevention and control of further outbreaks.
However, there are additional challenges to genotyping most especially from environmental samples, requiring mitigation strategies:
- sensitivity, as small numbers of oocysts may be present on slides and in the pesence of potential PCR inhibitors;
- specificity, as there is difficulty in selecting target gene sequences that are both conserved within all Cryptosporidium species but different or absent in other genera;
- not all isolates from environmental samples are typable, even when present in large numbers, for unknown reasons;
- in concurrent Cryptosporidium populations, it may be difficult to identify multiple sequences present, especially of minor genotypes which may be underdetected;
- genotypes have been found in source waters which have no known host; there is a need for more information on Cryptosporidium shedding by most host species;
- a limited amount of Cryptosporidium DNA may be available for subtyping from environmental samples;
- assays are time consuming, specialised and costly.
CONCLUSIONS
Despite the knowledge gained since the first drinking waterborne outbreaks, there are still gaps that need to be filled. Waterborne outbreaks have occurred where small numbers of oocysts have been detected, and conversely, high numbers of oocysts do not necessarily lead to increased disease [76]. This reflects the multifactorial dynamics of waterborne disease, including human behaviour (water consumption) and immunity of the exposed population, as well as the potential infectivity of the oocysts for humans, determined by their species and viability (Figure 1). These present analytical and interpretive challenges to public health and quantitative microbial risk assessment. Nevertheless, early detection, investigation and appropriate control of outbreaks can reduce their impact, and are facilitated by molecular methods to establish the relationship between isolates from cases and suspected sources.
Acknowledgements
I extend my thanks to Stephen Hadfield, Cryptosporidium Reference Unit, for his helpful comments on the manuscript and to George Di Giovanni, University of Texas-Houston School of Public Health, for providing information about the WaterRF program.
Conflict of interest statement
There are no potential conflicts of interest or any financial or personal relationships with other people or organizations that could inappropriately bias conduct and findings of this study.
References
1. Okhuysen PC, Chappell CL, Crabb JH, Sterling CR, DuPont, HL. Virulence of three distinct Cryptosporidium parvum isolates for healthy adults. J Infect Dis 1999;180:1275-81. http://dx.doi.org/10.1086/315033
2. Chappell CL, Okhuysen PC, Langer-Curry R, Widmer G, Akiyoshi DE, Tanriverdi S, Tzipori S. Cryptosporidium hominis: experimental challenge of healthy adults. Am J Trop Med Hyg 2006;75:851-7.
3. Chappell CL, Okhuysen PC , Langer-Curry RC, Akiyoshi DE, Widmer G, Tzipori S. Cryptosporidium meleagridis: infectivity in healthy adult volunteers. Am J Trop Med Hyg 2011;85(2):238-42. http://dx.doi.org/10.4269/ajtmh.2011.10-0664
4. Chalmers RM, Robinson G, Elwin K, Hadfield SJ, Xiao L, Ryan U, et al. Cryptosporidium rabbit genotype, a newly identified human pathogen [letter]. Emerg Infect Dis 2009;15:829-30. http://dx.doi.org/10.3201/eid1505.081419
5. Cama VA, Bern C, Roberts J, Cabrera L, Sterling CR, Ortega Y, Gilman RH, Xiao L. Cryptosporidium species and subtypes and clinical manifestations in children, Peru. Emerg Infect Dis 2008;14:1567-74. http://dx.doi.org/10.3201/eid1410.071273
6. Xiao L and Feng Y. Zoonotic cryptosporidiosis. FEMS Microbiol Med Microbiol 2008;52:309-23. http://dx.doi.org/10.1111/j.1574-695X.2008.00377.x
7. Elwin K, Hadfield SJ, Robinson G, Crouch N, Chalmers RM. Cryptosporidium viatorum n. sp. (Apicomplexa: Cryptosporidiidae) among travellers returning to the United Kingdom from the Indian Subcontinent. Int J Parasitol 2012;42:675-82. http://dx.doi.org/10.1016/j.ijpara.2012.04.016
8. Ryan U, Power M. Cryptopsoridium species in Australian wildlife and domestic animals. Parasitol 2012;20:1-16. http://dx.doi.org/10.1017/S0031182012001151
9. Kvác M, Kvetonová D, Sak B, Ditrich O. Cryptosporidium pig genotype II in immunocompetent man. Emerg Infect Dis 2009;15(6):982-3. http://dx.doi.org/10.3201/eid1506.07621
10. Rasková V, Kvetonová D, Sak B, McEvoy J, Edwinson A, Stenger B, Kvác M. Case of human cryptosporidiosis caused by Cryptosporidium tyzzeri and C. parvum presumably transmitted from wild mice. J Clin Microbiol 2012. [Epub ahead of print]. http://dx.doi.org/10.1128/JCM.02346-12
11. Robinson G, Elwin K, Chalmers RM. Unusual Cryptosporidium genotypes in human cases of diarrhoea. Emerg Infect Dis 2008;14:1800-2. http://dx.doi.org/10.3201/eid1411.080239
12. Xiao L, Hlavsa MC, Yoder J, Ewers C, Dearen T, Yang W, Nett R, Harris S, Brend SM, Harris M, Onischuk L, Valderrama AL, Cosgrove S, Xavier K, Hall N, Romero S, Young S, Johnston SP, Arrowood M, Roy S and Beach MJ. Subtype analysis of Cryptosporidium specimens from sporadic cases in Colorado, Idaho, New Mexico, and Iowa in 2007: widespread occurrence of one Cryptosporidium hominis subtype and case history of an infection with the Cryptosporidium horse genotype. J Clin Microbiol 2009; 47(9):3017-20. http://dx.doi.org/10.1128/JCM.00226-09
13. Robinson G, Wright S, Elwin K, Hadfi eld SJ, Katzer F, Bartley PM, et al. Re-description of Cryptosporidium cuniculus Inman and Takeuchi, 1979 (Apicomplexa: Cryptosporidiidae); morphology, biology and phylogeny. Int J Parasitol 2010;40:1539-48. http://dx.doi.org/10.1016/j.ijpara.2010.05.010
14. Slavin D. Cryptosporidium meleagridis (sp. nov.). J Comp Pathol 1955;65:262-6. http://dx.doi.org/10.1016/S0368-1742(55)80025-2
15. Hunter PR, Hughes S, Woodhouse S, Syed O, Verlander NQ, Chalmers RM, et al. Sporadic cryptosporidiosis case-control study with genotyping. Emerg Infect Dis 2004;10:1241-9. http://dx.doi.org/10.3201/eid1007.030582
16. Hunter PR, Nichols G. Epidemiology and clinical features of Cryptosporidium infection in immunocompromised patients. Clin Microbiol Rev 2002;15:145-54. http://dx.doi.org/10.1128/CMR.15.1.145-154.2002
17. CMO Update 23. Cryptosporidium in water: advice to the 25 immunocompromised. A communication to all doctors from the Chief Medical Officer (August 1999).
18. Lake IR, Nichols GL, Bentham G, Harrison FCD, Hunter PR, Kovats RS. Cryptosporidiosis Decline after Regulation, England and Wales, 1989-2005. Emerg Infect Dis 2007;13(4):623-5. http://dx.doi.org/10.3201/eid1304.060890
19. Putignani L, Menichella D. Global distribution, public health and clinical impact of the protozoan pathogen Cryptosporidium. Interdiscipl Perspect Infect Dis 2010. http://dx.doi.org/10.1155/2010/753512
20. Hunter PR, Hughes S, Woodhouse S, Raj N, Syed Q, Chalmers RM, Verlander NQ, Goodacre J. Health sequelae of human cryptosporidiosis in immunocompetent patients. Clin Infect Dis 2004;39:504-10. http://dx.doi.org/10.1086/422649
21. Savioli L, Smith H, Thompson A. Giardia and Cryptosporidium join the "neglected diseases initiative". Trends Parasitol 2006;22:203-8. http://dx.doi.org/10.1016/j.pt.2006.02.015
22. Karanis P, Kourenti C, Smith H. Waterborne transmission of protozoan parasites: a worldwide review of outbreaks and lessons learnt. J Water Hlth 2007;5:1-38.
23. Baldursson S, Karanis P. Waterborne transmission of protozoan parasites: review of worldwide outbreaks - an update 2004-2010. Water Res 2011;45:6603-14.
24. Robertson L J, Chalmers RM. Foodborne cryptosporidiosis: is there really more in Nordic countries? Trends Parasitol (in press). http://dx.doi.org/10.1016/j.pt.2012.10.003
25. Medema G, Teunis P, Blokker M. Deere D, Davison A, Charles P, Loret JF. Risk assessment of Cryptosporidium in drinking-water. Geneva: WHO; 2009.
26. MacKenzie WR, Hoxie NJ, Proctor ME, Gradus MS, Blair KA, Peterson DE, Kazmierczak JJ, Addiss DG, Fox KR, Rose JB, David JP. A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply. New Engl J Med 1994;331(3):161-7.
27. Smith A, Reacher M, Smerdon W, Adak GK, Nichols G, Chalmers RM. Outbreaks of waterborne infectious intestinal disease in England and Wales, 1992-2003. Epidemiol Infect 2006;134(6):1141-9. http://dx.doi.org/10.1017/S0950268806006406
28. Centers for Disease Control and Prevention. Surveillance for waterborne disease outbreaks and other health events associated with recreational water United States, 2007-2008 and surveillance for waterborne disease outbreaks associated with drinking water. United States, 2007-2008. MMWR 2011;60(No. RR-12):1-75.
29. Risebro HL, Doria MF, Andersson Y, Medema G, Osborn K, Schlosser O, Hunter PR. Fault tree analysis of the causes of waterborne outbreaks. J Water Health 2007;05(Suppl. 1): 1-18. http://dx.doi.org/10.2166/wh.2007.136
30. Smith S, Elliot AJ, Mallaghan C, Modha D, Hippisley-Cox J, Large S, Regan M, Smith GE. Value of syndromic surveillance in monitoring a focal waterborne outbreak due to an unusual Cryptosporidium genotype in Northamptonshire, United Kingdom, June-July 2008. Euro Surveill 2010;15(33): pii=19643.
31. Semenza J, Nichols G. Cryptosporidiosis surveillance and water-borne outbreaks in Europe. Euro Surveill 2007; 12(5):120-4.
32. European Centre for Disease Control and Prevention. Annual epidemiological report. Reporting on 2009 surveillance data and 2010 epidemic intelligence data. Stockholm: 2011. Available from: www.ecdc.europa.eu/en/publications/.
33. Chalmers RM, Campbell B, Crouch N, Davies, AP. Clinical laboratory practices for the detection and reporting of Cryptosporidium in community cases of diarrhoea in the United Kingdom, 2008. Euro Surveill 2010;15:pii519731.
34. Tam CC, Rodrigues LC, Viviani L, Dodds JP, Evans MR, Hunter PR, Gray JJ, Letley LH, Rait G, Tompkins DS, O'Brien SJ. On behalf of the IID2 Study Executive Committee. Longitudinal study of infectious intestinal disease in the UK (IID2 study): incidence in the community and presenting to general practice. Gut 2012;61(1):69-77. http://dx.doi.org/10.1136/gut.2011.238386
35. Nichols G, Chalmers R, Lake I, Sopwith W, Regan M, Hunter P, Grenfell P, Harrison F, Lane C. Cryptosporidiosis: a report on the surveillance and epidemiology of Cryptosporidium infection in England and Wales. London: Drinking Water Inspectorate; 2006.
36. Wall PG, DeLouvois J, Gilbert RJ, Rowe B. Food poisoning notifications, and outbreaks where do the statistics come from and what do they mean? CDR Wkly 1996;6:R93-100.
37. Tillet HE, DeLouvois J, Wall PG. Surveillance of outbreaks of waterborne disease: categorizing levels of evidence. Epidemiol Infect 1998;120(1):37-42. http://dx.doi.org/10.1017/S0950268897008431
38. Herwaldt BL, Craun GF, Stokes SL, Juranek DD. Waterborne-disease outbreaks, 1989-1990. MMWR 1991;40 (No. SS-3):1-21.
39. Meisel JL, Perera DR, Meligro C, Rubin CE. Overwhelming watery diarrhea associated with a Cryptosporidium in an immunosuppressed patient. Gastroenterology 1976;70:1156-60.
40. Nime FA, Burek JD, Page DL, Yardley JH. Acute enterocolitis in a human being infected with the protozoan Cryptosporidium. Gastroenterology 1976;70:592-8.
41. D'Antonio RG, Winn RE, Taylor JP, Gustafson TL, Current WL, Rhodes MM, Gary GW, Zajac RA. A waterborne outbreak of cryptosporidiosis in normal hosts. Ann Intern Med 1985;103:886-8.
42. Hayes EB, Matte TD, O'BrienTR, McKinley TW, Logsdon GS, Rose JB, Ungar BLP, Word DM, Pinsky PF, Cummings ML, Wilson MA, Long EG, Hurwitz ES, Juranek DD. Large community outbreak of cryptosporidiosis due to contamination of a filtered public water supply. New Engl J Med 1989;320:1372-6. http://dx.doi.org/10.1056/NEJM198905253202103
43. Richardson AJ, Frankenberg RA, Buck AC, Selkon JB, Colbourne JS, Parsons JW, Mayon-White RT. An outbreak of waterborne cryptosporidiosis in Swindon and Oxfordshire. Epidemiol Infect 1991;107:485-95. http://dx.doi.org/10.1017/S0950268800049189
44. Willcocks L, Crampin A, Milne L, Seng C, Susman M, Gair R, et al. (1998). A large outbreak of cryptosporidiosis associated with a public water supply from a deep chalk borehole. Comm Dis Publ Health 1998;1:239-43.
45. Barrell RA, Hunter PR, Nichols G. Microbiological standards for water and their relationship with health risk. Comm Dis Pbl Hlth 2000;3(1)8-13.
46. Harrison SL, Nelder R, Hayek L, Mackenzie IF, Casemore DP, Dance D. Managing a large outbreak of cryptosporidiosis: how to investigate and when to decide to lift a "boil water" notice. Commun Dis Public Health 2002;5(3):230-9. http://dx.doi.org/10.1017/S0950268807008503
47. Lloyd A, Drury D. Continuous monitoring for Cryptosporidium -a novel approach to public health protection. Water Sci Technol 2002;46(11-12):297-301.
48. Neira-Munoz E, Okoroa C, McCarthy ND. Outbreak of waterborne cryptosporidiosis associated with low oocyst concentrations. Epidemiol Infect 2007;135(7):1159-64.
49. Mason BW, Chalmers RM, Carnicer-Pont D, Casemore DP. A Cryptosporidium hominis outbreak in North-West Wales, UK, associated with low oocyst counts in treated drinking water. J Water Health 2010;8(2):299-310. http://dx.doi.org/10.2166/wh.2009.184
50. Deere D, Stevens M, Davison A, Helm G, Dufour A. Management strategies. In: Fewtrell L, Bartram J (Eds.). Water quality: guidelines, standards and health. Assessment of risk and risk management for water-related infecious disease. London: IWA Publishing; 2001.
51. World Health Organization. Water Safety Plans: managing drinking-water quality from catchment to consumer. Geneva: WHO; 2005.
52. World Health Organization. Guidelines for safe recreational waters. Volume 2. Swimming pools and similar water environments. Geneva: WHO; 2006.
53. PWTAG. Swimming pool water. Treatment and quality standards for pools and spas. 2d. Diss: Greenhouse Publishing Limited; 2006.
54. Verma A, Bolton FJ, Fiefield D, Lamb P, Woloschin E, Smith N, Mccann R. An outbreak of E. coli O157 associated with a swimming pool: an unusual vehicle of transmission. Epidemiol Infect 2007;135(6):989-92. http://dx.doi.org/10.1017/S0950268807007947
55. Kothavade RJ. Potential molecular tools for assessing the public health risk associated with waterborne Cryptosporidium oocysts. J Med Microbiol 2012;61(Pt 8):1039-51. Epub 2012 May 24. http://dx.doi.org/10.1099/jmm.0.043158-0
56. Ruecker NJ, Matsune JC, Wilkes G, Lapen DR, Topp E, Edge TA, Sensen CW, Xiao L, Neumann NF. Molecular and phylogenetic approaches for assessing sources of Cryptosporidium contamination in water. Water Res 2012;46:5135-50. http://dx.doi.org/10.1016/j.watres.2012.06.045
57. Xiao L. Molecular epidemiology of cryptosporidiosis: an update. Exp Parasitol 2010;124:80-9 http://dx.doi.org/10.1016/j.exppara.2009.03.018
58. Cacciò SM, Sannella AR, Mariano V, Valentini S, Berti F, Tosini F, Pozio E. A rare Cryptosporidium parvum genotype associated with infection of lambs and zoonotic transmission in Italy. Vet Parasitol 2012 Aug 20. [Epub ahead of print]. http://dx.doi.org/10.1016/j.vetpar.2012.08.010
59. Robinson G, Chalmers RM. Assessment of polymorphic genetic markers for multi-locus typing of Cryptosporidium parvum and Cryptosporidium hominis. Exp Parasitol 2012;132: 200-15. http://dx.doi.org/10.1016/j.exppara.2012.06.016
60. Bouzid M, Hunter PR, Chalmers RM, Tyler KM. Cryptosporidium pathogenicity and molecular virulence factors. Clin Microbiol Rev (in press).
61. Chalmers RM, Elwin K, Thomas AL, Guy EC, Mason B. Long-term Cryptosporidium typing reveals the aetiology and species-specific epidemiology of human cryptosporidiosis in England and Wales, 2000 to 2003. Eurosurveillance 2009;14(2)2009b;pii=19086.
62. Lake IR, Harrison FCD, Chalmers RM, Bentham G, Nichols G, Hunter PR, Kovats RS, Grundy C. Case-control study of environmental and social factors influencing cryptosporidiosis. Europ J Epidemiol 2007;22:805-11.
63. HPA. Late-summer seasonal increase in cryptosporidiosis recorded. Health Protect Rep 2012;6(41). Available from: www.hpa.org.uk/hpr/archives/2012/news4112.htm#smmrcrpt.
64. Nichols RAB, Connelly L, Sullivan CB, Smith HV. Identification of Cryptosporidium Species and Genotypes in Scottish Raw and Drinking Waters during a One-Year Monitoring Period. Appl Environl Microbiol 2010;76(17);5977-86. http://dx.doi.org/10.1128/AEM.00915-10
65. Loganthan S, Yang R, Bath A, Gordon C, Ryan U. Prevalence of Cryptosporidium species in recreational versus non-recreational water sources. Exp Parasitol 2012;131(4):399-403. Epub 2012 May 17. http://dx.doi.org/10.1016/j.exppara.2012.04.015
66. Sulaiman IM, Xiao LH, Yang CF, Escalante L, Moore A, Beard CB, Arrowood MJ, Lal AA. Differentiating human from animal isolates of Cryptosporidium parvum. Emerg Infect Dis 1998;4:681-5. http://dx.doi.org/10.3201/eid0404.980424
67. Patel S, Pedraza-Díaz S, McLauchlin J, Casemore DP, Outbreak Control Team South and West Devon 1995, Incident Management Team and Further Epidemiological and Microbiological Studies Subgroup North Thames 1997. Molecular characterisation of Cryptosporidium parvum from two large suspected waterborne outbreaks. Commun Dis Public Health 1998;1:231-3.
68. Hunter PR, Wilkinson DC, Lake IR, Harrison FCD, Syed Q, Hadfield SJ, Chalmers RM. Microsatellite typing of Cryptosporidium parvum in a waterborne outbreak. J Clin Microbiol 2008;46:3866-7. http://dx.doi.org/10.1128/JCM.01636-08
69. Zhou L, Singh A, Jiang J, Xiao L. Molecular surveillance of Cryptosporidium spp. in raw wastewater in Milwaukee: implications for understanding outbreak occurrence and transmission dynamics. J Clin Microbiol 2003;41(11):5254-7. http://dx.doi.org/10.1128/JCM.41.11.5254-5257.2003
70. Mattsson JG, Insulander M, Lebbad M, Bjorkman C, Svenungsson B. Molecular typing of Cryptosporidium parvum associated with a diarrhoea outbreak identifies two sources of exposure. Epidemiol Infect 2008;13:1147-52. http://dx.doi.org/10.1017/S0950268807009673
71. Chalmers RM, Robinson G, Elwin K, Hadfield SJ, Thomas E, Watkins J, Casemore D, Kay D. Detection of Cryptosporidium species and sources of contamination with Cryptosporidium hominis during a waterborne outbreak in north west Wales. J Water Health 2010;8(2):311-25. http://dx.doi.org/10.2166/wh.2009.185
72. DWI 2005 Information Letter 17/2005. Cryptosporidiosis in England and Wales. Available from: http://dwi.defra.gov.uk/stakeholders/information-letters/2005/17_2005.pdf.
73. Glaberman S, Moore JE, Lowery CJ, Chalmers RM, Sulaiman I, Elwin K, Rooney PJ, Millar BC, Dooley JSG, Lal AA, Xiao L. Three drinking-water-associated cryptosporidiosis outbreaks, Northern Ireland. Emerg Infect Dis 2002;8:631-3. http://dx.doi.org/10.3201/eid0806.010368
74. Ichinohe S, Fukushima T, Kishida K, Sanbe K, Saika S, Ogura M. Secondary transmission of cryptosporidiosis associated with swimming pool use. Jpn J Infect Dis 2005;58(6):400-1.
75. Center for Disease Control Prevention. Communitywide cryptosporidiosis outbreak. Utah, 2007. MMWR 2008;57:989-93.
76. Hunter PR. Advice on the response from public and environmental health to the detection of cryptosporidial oocysts in treated drinking water. Commun Dis Public Health 2000; 3:24-7.
Address for correspondence:
Rachel M. Chalmers
Cryptosporidium Reference Unit
Public Health Wales, Singleton Hospital
, Swansea SA2 8QA, UK
E- mail: rachel.chalmers@wales.nhs.uk
Submitted on invitation.
Accepted on 24 September 2012.