Ch 6 - Discussion
Canoeist Pathogenic Illness Guide


6.1 - Risk Assessment
6.2 - Risk Management
6.3 - Water Quality Standards

Forward to Chapter 7 - Conclusions and Recommendations

Chapter 6 - Discussion

6.1 - Risk Assessment

6.1 a) Overall Risk

A study of the epidemiological surveys reported in Chapter Two shows a wide variability in the detailed procedures and results of each individual project. There is also the problem that many surveys are based on swimming activity rather than canoeing, and often on salt water, (although Fewtrell's work on canoeing, deliberately using the same procedures as the bathing studies, did confirm similarities in outcome).

Despite this variability there are certain overall themes and trends which are remarkably consistent:

It may be helpful to compare these risks against other risks, especially in sport. There are an average 200 deaths from drowning in the UK each year, and there have been 46 deaths on one Scottish Mountain, Ben Nevis, in the last ten years. Many indoor or school based sports also cause injury, with rugby and football causing a low but steady stream of limb fractures, concussions, cartilage and soft tissue injuries and even spinal injuries.

For comparison with less serious risks and symptoms, it is worth noting that a number or reports noted the difficulty of dealing with confounding factors in the "non-bathing" risks such as fast food. To a canoeist the risk of gastrointestinal symptoms from a burger or camp-site cooking may be of the same order of magnitude as the risk from canoeing activity. Another helpful, if non scientific, comparison may be to compare a less than ten per cent risk of diarrhoea, vomiting, headache and dehydration from a canoeing event, to the greater than 50% risk of diarrhoea, vomiting, headache and dehydration after the consumption of large amounts of alcohol.

Many people choose to take all of the risks listed above, for a variety of benefits. As an overall assessment, therefore, white water canoeing in sewage contaminated water produces an increased risk of gastrointestinal and other symptoms. By comparison with many other risk factors both in sport and in general life these symptoms are mild, non-life threatening and of short duration.

Whilst people must be allowed to make an informed choice about whether they wish to undertake any activity with an element of risk, a decision to go canoeing on sewage contaminated waters, even certain waters outside bathing water standards, would not appear to be irrational or excessively risky.

6.1 b) The risk from specific activities

Fewtrell's research has added more detail to this overall statement in two areas.

Firstly she has shown that there are different grades of water contact activity. White water canoeing is a high contact activity with a high (relative) risk of illness symptoms, whilst marathon canoe racing has a lower risk, even on more polluted waters. This may be surprising to some canoeists, as marathon canoe racing is not a dry sport. In the close confines of a race there is extensive spray flung up by canoe paddles into canoeist's faces. It is conceivable that there could be as much water spray swallowed in a marathon as in a white water canoeing session. Nevertheless the research would appear to show that this factor can be discounted. Low water-contact canoesport, including marathon racing and placid water touring, shows little quantifiable health effect, even when canoeing on waters contaminated with sewage to well above the EC Bathing Water Directive standards.

Fewtrell also confirms the evidence of the bathing water studies in showing that canoeing health effects are not limited to gastrointestinal symptoms. Other symptoms reported include headaches, ear and eye infections and coughs. Non gastrointestinal symptoms are often more prevalent in the studies, although as they are often less traumatic to the infected person than diarrhoea or vomiting, such symptoms seem to be causing less public concern.

Lee's study at Holme Pierrepont has also added the possibility of giving advice to canoeists about how to manage or minimise risk.

Lee showed firstly that drinking or eating whilst wet (ie before showering or washing your hands) increased the relative risk of symptoms. He also reported that those who had made many previous visits to Holme Pierrepont had a much smaller chance of showing symptoms.

Note by Simon Dawson. There are two published documents relating to this research project at Holme Peirrepont. The Lee 1997 reference is an academic paper written mainly by John Lee (PHLS) and Keith Neal (epidemilogy). There is also the "National Watersports Center Water Quality - Management Report". This 1997 document is a risk assessment study relating to HPP water quality. Copies of the document (or it's up to date successor) should available on request from HPP.

The patterns of activity of the participants in this study does cause a problem for the statistical analysis, in that many of the activities are interlinked. People who were infrequent or first time users of the slalom course will often tend to spend all day on the course, and be more likely to eat whilst wet. They might be less experienced and more likely to swim, or be engaged in rafting or white water recreational paddling with a high chance of getting wet or sprayed with water. Frequent visitors would by contrast tend to be high level competitors who visit for a short one hour sessions, not eat, not swim and not capsize or get sprayed with water.

Within these patterns of activity there is the possibility of confounding. If somebody who eats whilst wet shows a higher risk of symptoms, is it due to the eating, or is it affected by the high possibility of such an on-course eater being a non-local, beginner who would spend many hours on the course, and have a much higher chance of swimming.

Despite these queries, the advice issued as a result of the report, on hygiene and hand-washing facilities, is sensible and helpful.

It is interesting to consider the possibility of some form of disease resistance in frequent visitors to the course. Lee reports that such frequent visitors have an activity pattern which excludes risk factors such as eating or swimming. This may well be a contributory factor. There are many anecdotal reports however of other canoeists who appear to have a high level of resistance to infection. It has been suggested by canoeists that this is due to acquired immunity, although this is unlikely as the range of pathogens is constantly changing, and acquired immunity to one species of pathogen does not automatically confer immunity against others.

One other explanation may be a form of natural selection based on an individual's resistance to disease. In a class of beginner canoeists each individual will have a different level of resistance to infection by water-borne pathogens. Those beginners who frequently go on to develop illnesses may decide to give up the sport and try another, whilst those who do not develop symptoms might be more likely to stay in the sport of canoeing.

Whatever the reason, it may be possible to assume that a group of expert canoeists will contain a higher proportion of people who have a high level of resistance to water-borne pathogens than a group of beginners, or a group of the general population.

If a canoeing site had poor water quality temporarily due to high water levels, it might be possible to close the site to beginners, but allow experienced paddlers onto the site on the grounds of increased disease resistance, and a better ability to make valid decisions on the levels of risk such paddlers are willing to accept for themselves.

6.2 - Risk Management

Whilst the overall risk to canoeists from pathogenic microorganisms may be low and acceptable, it is sensible to find various means of managing the risk, so that steps may be taken to reduce the risk still further.

6.2 a) Estimating the risk from water-borne pathogens using Indicator Organisms

In the study of pathogenic microorganisms in recreational water, the question is whether the presence such an indicator species in water is a broad measure, which will only indicate the possibility of sewage contamination, and therefore the potential for some form of infectious disease, or whether the use of indicator species can be taken further, and whether a knowledge of the exact level of indicator species concentrations can be used to accurately predict the likelihood or probability of various forms of infection.

The World Health Organisation defines a "health effect water quality indicator" as:

a microbial, chemical or physical agent, substance or quality, which indexes the potential risk of infectious disease coincident with man's use of the aquatic environment for recreation or the production of food. (WHO 1977)

Theoretically the best indicators are those showing the highest degree of correlation with associated health effects. WHO has noted that the use of indicators should be based on epidemiological studies, and that indicator organisms should:

  1. be consistently and exclusively associated with the source of the pathogens.
  2. be present in sufficient numbers or quantities, without proliferation or somatic/genetic changes to provide a reasonable estimation of the presence of pathogens and the potential existence of a health risk.
  3. approach the resistance to disinfectants and environmental stress, including that resulting from toxic materials deposited in the aquatic environment, of the most resistant pathogen potentially present at significant levels in the source.
  4. be quantifiable in environmental samples by reasonably easy and inexpensive methods, and with considerable accuracy, precision and specificity.

In the natural environment where each different microorganism will have it's own particular life cycle and life span, each organism will be differently affected by changes in climate, temperature, water-borne pollutants, sunlight, nutrient levels and predators. It remains a question as to how constant will be any relationship between the concentration in the water of a pathogenic organism, or an epidemiologically determined risk of infection, and the presence or concentration in water of any associated indicator species. It may be possible to determine a statistically valid relationship between an epidemiologically proven risk and an indicator species in one set of environmental conditions. A change in one single environmental factor, such as sunlight or temperature, could then change the morbidity or distribution of the indicator species or pathogen, and destroy the validity of the relationship.

In the current state of development, the use of indicator organisms would appear to be helpful in long term studies, assessing mean seasonal trends. They are of use for the Water Companies and Regulatory Authorities who are attempting to devise long term policies to manage the quality of UK watercourses, or to sports managers who are looking at proposals to site a canoeing facility in a certain area.

On a short term basis, however, the massive localised temporal and spatial variability means that indicator organisms cannot be used by canoeists on routine basis to assess the exact risk from any particular site on that day.

Their only value may be to a site manager who uses occasional routine checks as a confidence measure, to detect unexpectedly high readings which may indicate defects, mechanical breakdowns or other unusual events.

6.2 b) Estimating the risk from water-borne pathogens using Flow Modelling

An alternative method of estimating the risk on a short term basis may be to find some way of modelling pathogen levels based on what is known about pathogen production and disinfection in the various parts of the river and waste management system. The source and fates of river pathogens can be summarised by Figure 4. which shows the net inflow and outflow which would affect any watercourse.

Taken in it's entirety this model is so complex that it precludes the possibility of accurate quantitative, predictive modelling. By using much of the information from Chapter Four it is however possible to produce an overall qualitative model. This discussion relates mainly to typical English mature lowland rivers.

In average summer conditions with very low rainfall, high temperatures and clear skies, storm water related inputs will tend to zero, and disinfection rates will be high, both in the sewage works and in the river, due to high sunlight levels and high water temperatures. Overall pathogen levels in the river will be relatively low. In many areas of the UK, lowland rivers will pass EC Directive Bathing Water Standards under such conditions.

During average winter conditions many factors will change. Disinfection rates will drop in the river and sewage works due to lower sunlight and lower water temperatures. Rainfall may produce increased runoff from urban and rural areas which will flush pathogens into the river. Increased flow through the sewage works in combined sewerage systems will tend to drop sewage disinfection efficiencies. To counter these increases, higher river flow rates will increase the rate at which bacterial and viral loadings will be flushed away. The net effect is typically to raise mean indicator levels to a level 2-5 times mean summer levels.

After a heavy storm event, a transient will start, bringing in changes to water flows and pathogen levels.

During a winter storm the pathogen input rates will rise due to sewage works overflow, and to storm water flows from urban and rural areas. This may cause a build up of pathogens in the river. Outflows due to natural disinfection may be low, but as the river flow rate increases, bacterial loadings will be to be flushed rapidly downstream.

As the storm continues, pathogen inputs may drop dramatically once the fields and storm culverts have been flushed clean. If river flows remain high, however, it is possible for the high pathogen outflows to be maintained as pathogens continue to be flushed downstream. Eventually lower than average pathogen levels may be found in the river. A winter storm transient, therefore, might start at the average winter bacterial rates, increase as the rain starts, and then decrease to lower than average at the tail end of the transient.

A summer storm after a long dry period may start at a low average bacterial loading, but there will be a very high pathogen input as the sediments built up in culverts and pipes over a long dry period are flushed into the river. Unless the storm is very heavy, pathogen outflows will be much lower, as the river flow will be lower and the bacteria will not be flushed rapidly away. Natural disinfection is a longer term process which will have little effect in the timescale of this storm transient.

Summer storms therefore have the potential to cause much higher pathogen loadings than winter storms, and for these loadings to increase very rapidly from a low initial level after quite low amounts of rain.

The four levels of risk are summarised in table 10 below.

AVERAGE SUMMER High natural disinfection. High sewage works efficiency . No storm water.
LOW RISK. Probably EC Bathing water Directive Standard
AVERAGE WINTER Lower natural disinfection. lower sewage worke efficiency. MEDIUM RISK
WINTER STORM Storm Water input. Sewage work overflow. BUT - high dilution, no sediment build up in storm drains MEDIUM TO HIGH RISK
SUMMER STORM Storm water input. Sewage works overflow. Little dilution. High sediment build up in storm drains. HIGH RISK

Table 10. Estimate of variation relative risk levels of canoeing with season and weather

Within this model of low, medium, high and very high risk: for average summer, average winter, winter storm and summer storm respectively; there is the question of what forms of canoeing activity have an acceptable risk for each water level.

In the summer there need be no restrictions, and although the risk may be higher in winter, the risk assessments show that canoeing activities in average winter conditions need not be a problem for beginners or for experienced canoeists as long as information is given about the level of risk, and people are given access to sensible advice about hygiene.

In a winter storm with high water levels it is normally inappropriate to take beginners onto the water due to the danger from flood water and poor weather conditions, rather than the risk of infection. Experienced canoeists might choose to go onto the water to enjoy the added excitement of flood conditions. If so it would be sensible to avoid canoeing in the early stages of the storm, but to get afloat after the peak of the flood when conditions might be quite clean.

In a summer storm, after a long dry period, it is possible to get exceptionally high levels of pathogens in the water, and it may be advisable even for experienced canoeists with a high level of disease resistance not to go canoeing. Anecdotal reports from canoeists at Holme Pierrepont suggest that local paddlers follow a rule of not going canoeing in the first floods of the autumn each year.

The risk at Holme Pierrepont is managed by shutting the course to canoeists and rafters during winter and summer storm events, and allowing free access during winter and summer average conditions after offering advice to all users on the risk of infection and hygiene measures. Whilst these measures may be sensible and appropriate for normal use, there are certain high profile competitive events where there may be pressure not to close the course. In 1992 a Slalom World Cup event was held in July after two days of heavy rain with high water flows. It may be sensible to carry out further risk assessment studies during winter and summer storms, to allow informed decisions to be made about the risk of holding such competitions during storm events.

There are some suggestions in the literature that whilst gastrointestinal symptoms are commonly associated with sewage related pathogens; skin, eye and ear irritations and infections may sometimes be more strongly associated with animal related pathogens which are flushed into the water during storm events, from both urban and rural areas. This has two implications.

Research projects into the risk to canoeists during storm events should carefully consider the choice of indicator species and the symptoms to study. A reliance on the standard faecal indicators may prove ineffective due to the inability of these methods to distinguish human from non-human (and possibly non disease causing) bacteria from soil and animals.

In addition, other pathogens, not related to normal faecal flora, may cause disease which is not indexed by the normal faecal indicators. Leptospirosis is but one example of a bacteria which gains access to humans through broken skin and the mucous membranes, which shows most infectious activity during high water levels in late summer and autumn, and which is not indexed by tests for E. Coli or Faecal Streptococci.

For similar reasons canoeists travelling to areas with little human faecal pollution should still consider the risk from microbial pathogens. If a river has rainfall run-off from vegetated or forested areas with high animal populations, sensible precautions might include covering cuts and scratches with waterproof dressings, and careful medical treatment of eye, ear and skin infections.

6.3 c) Risk estiimation on Reservoirs and Lakes

In a reservoir or lake the techniques of using natural disinfection to provide clean water for recreational use is well established and understood. Inputs can be controlled by excluding flow into the lake, or by ensuring that only clean water flows in. A risk will only arise if sewage contaminated water enters the lake, perhaps due to a flood, overflow or mechanical breakdown. Experience has shown that routine monitoring of the lake to detect such contamination is a sensible precaution.

Although sewage related bacterial or viral pollution in lakes can be managed and poses a very low risk, the threat from blue green algae remains. Apart from leptospirosis, cyanobacteria are the only organisms with a realistic potential to cause fatal disease to canoeists in the UK. Although advice is given to canoeists about Weil's disease in the British Canoe Union Yearbook, no advice is given about blue green algae. Such advice might be helpful bearing in mind the potential seriousness of any infection.

6.3 - Water Quality Standards

Although water quality standards are set for coastal bathing waters through use of the EC Bathing Water Directive, there are no inland recreational waters in the UK which have been nominated as bathing waters. Other European Nations do however apply Bathing Water Directive standards to inland waters.

It would be helpful to apply such standards to inland recreational waters to ensure that the development of inland waters for recreation is not lost amongst all the other priorities for water usage when Environment Agency or Water Company policies are being set. The Statutory Water Quality Objectives for Water Contact Activities provide a means of setting water quality standards, and sports governing bodies should lobby for their introduction.

When it comes to achieving standards once they have been set, decisions on how best to achieve desired disinfection rates will be complex and varied. The optimum measure in a particular area might be to control pathogens in storm water flows, to fit disinfection tertiary treatment to a sewage works, or simply to upgrade the efficiency of primary and secondary treatment in an existing but obsolete sewage works. Whilst UV disinfection systems will not be the optimum answer in every case, their rapidly improving design and availability will make the disinfection of sewage effluent an increasingly cost effective option.

Forward to Chapter 7 | Return to Top of Page