Health and the Environment: Critical Pathways O · • Hazardous Products Act (HPA):prohibits the...

Health and the Environment

A Policy Nexus

Healthy Environments,Healthy People

Measuring the Impact

Populations and Pathways

Living Conditions and Disease

Taking Action on Vehicle Emissions

About Health Surveillance

Understanding StatisticalSignificance

Mark Your Calendar

3

Health and the Environment: Critical Pathways

Over the years, public health and environmental policies aimed at

reducing environment-related disease have contributed to signifi-

cant improvements in the health status of Canadians. Despite

these advances, however, health problems as a result of exposure to envi-

ronmental contaminants remain a serious concern for many Canadians.

This is especially true for certain populations who, because of their stage

of physical development, or their living and working conditions, are at

greater risk than the general population. A growing concern for many

Canadians is the cumulative effect of long-term exposure to low doses

of environmental contaminants.

While “the environment” in its broadest sense has important influences

on human health, this issue of the Health Policy Research Bulletin focuses

on the health impacts of the “physical environment.” More specifically,

it explores the range of research and evidence required to effectively

assess and manage environmental health risks including, for example:

• developing environmental health indicators and ensuring that appro-

priate surveillance systems are in place

• identifying potential environmental threats and assessing the associated

health risks

• determining the pathways and mechanisms of exposure and identify-

ing potential points for intervention

In this Issue

I S S U E 4 , O C T O B E R 2002S T R E N G T H E N I N G T H E P O L I C Y – R E S E A R C H C O N N E C T I O N

5

9

13

15

19

25

27

32

HEALTH POLICY RESEARCH BULLETIN — Issue 42

Health Policy Research BulletinThe opinions expressed in these articles, including interpretationof the data, are those of the authors and are not to be taken asofficial statements of Health Canada.

This publication can be made available in alternative formatsupon request.

Permission is granted for non-commercial reproduction pro-vided there is a clear acknowledgment of the source.

Published under the authority of the Minister of Health.

© Her Majesty the Queen in Right of Canada, represented bythe Minister of Public Works and Government ServicesCanada, 2002

ISSN 1496-466 X ISSN 1499-3503 (Online)

Editing, design and layout by Allium Consulting Group Inc.

Publications Mail Agreement Number 4006 9608

Return if undeliverable to:Health Canada2750 Sheffield Road, Bay #1Ottawa, ON K1B 3V9

About the Health Policy Research BulletinHealth Canada’s Health Policy Research Bulletin is pub-lished three times a year. The Bulletin is part of a largerpolicy research dissemination program designed toenhance Health Canada’s policy-relevant evidence base.

A departmental steering committee guides the devel-opment of the Bulletin. The committee is chaired byCliff Halliwell, Director General of the Applied Researchand Analysis Directorate (ARAD) of the Information,Analysis and Connectivity Branch. The ResearchManagement and Dissemination Division (RMDD)within ARAD coordinates the Bulletin’s developmentand production. RMDD would like to thank the steeringcommittee members for their contributions, as well asNancy Hamilton and Linda Senzilet, Managing Editors,Jaylyn Wong, Assistant Editor, and Marilyn Ryan,Production and Distribution. Special thanks go to theGuest Editor of this issue, Ray Edwards, DirectorGeneral of the Policy and Planning Directorate,Healthy Environments and Consumer Safety Branch.

We welcome your feedback and suggestions. Pleaseforward your comments and any address changes [email protected] or phone (613) 954-8549 or fax(613) 954-0813. Electronic HTML and PDF versions of theBulletin are available at: http://www.hc-sc.gc.ca/arad-draa

Our mission is to help the people of Canadamaintain and improve their health.

Health Canada

Environmental Health Legislation

Health Canada is involved in administering, in wholeor in part, 20 pieces of health-related legislation. Anoverview of some of the key legislation aimed atprotecting and promoting environmental health isprovided below. A more complete list of relevantlegislation can be found at: http://www.hc-sc.gc.ca/english/about/acts_regulations.html

Responsibility of Environment Canada and Health Canada• Canadian Environmental Protection Act (CEPA):

governs pollution prevention and protection ofthe environment and human health, all within thecontext of sustainable development goals

Responsibility of Health Canada• Hazardous Products Act (HPA): prohibits the

advertising, sale and importation of hazardousproducts

• Food and Drugs Act (FDA): ensures the safety offood, drugs, cosmetics and therapeutic devices

• Pest Control Products Act (PCPA): governs theimportation, manufacture, sale and use of pesticides

Responsibility of Environment Canada• Canadian Environmental Assessment Act (CEAA):

ensures all new projects with federal involvementinclude an environmental impact assessment,including an assessment of human health impacts

Depending upon the evidence, options for managing

the risks associated with potential environmental

threats may vary from minimum-level interventions

(e.g., increasing public awareness) to maximum-

level interventions (e.g., legislation banning the

release of a substance into the environment). A case

study on developing regulations aimed at limiting

the sulphur content in gasoline provides a good

illustration of how research can be used to help

manage environmental health risks.

http://www.hc-sc.gc.ca/english/about/acts_regulations.html

http://www.hc-sc.gc.ca/english/about/acts_regulations.html

http://www.hc-sc.gc.ca/iacb-dgiac/arad-draa/english/rmdd/bulletin/bulletin.html

3Issue 4 — HEALTH POLICY RESEARCH BULLETIN

This issue of the Bulletin examines the relationship between twoimportant areas — health and the environment. What is a good

starting point for understanding this relationship?

Health and the environment are two very important, intertwined areasof policy. From a health policy perspective, our starting point is a recog-nition that the “environment” is one of several determinants of health.Our health — in fact, our very survival — depends on the environment,from the air we breathe, to the water we drink and the food we eat. Whenany of these is threatened, human health is compromised. This is thestarting point from which health departments work to identify, reduceand prevent environment-related health risks.

When you say that the environment is a determinant of health,how would you define “environment?”

In its broadest sense, environment refers to the physical, social, culturaland economic attributes of our surroundings. While all of these haveimportant influences on human health, the “physical environment”has been identified within the determinants of health model as one of12 interrelated health determinants. So, for our discussions here, we’llfocus on the impacts of the physical environment on human health.However, regardless of what we understand the environment to be —air, water, soil, trees, the biota and so on — human health depends on awell-functioning environment and harmonized ecosystems.

Recognizing that the physical environment isa determinant of health, in what ways does it

affect our health?

The physical environment can influence ourhealth through the direct impacts of naturallyoccurring substances within the environmentand also as a consequence of our individualand collective interactions with the environ-ment. As a society, we use the natural resourceswithin our environment for purposes of eco-

nomic and social development. Consequently,we have to consider the primary, secondary and

tertiary impacts of development on human health,most of which stem from — in the terminology of

the ’80s and ’90s — a polluted environment.

QThe following article is based

on an interview with Rod

Raphael, Director General

of the Safe Environments Programme,

Healthy Environments and Consumer

Safety Branch, Health Canada. The

interview was conducted by Nancy

Hamilton, a Managing Editor of the

Health Policy Research Bulletin.

&&EnvironmentHealthA Policy Nexus

QIncome and

Social Status

SocialSupport

NetworksEducation

Employmentand WorkingConditions

PhysicalEnvironments

Biology andGenetics

PersonalHealth

Practices andCoping Skills

HealthyChild

Development

HealthServices

and SocialServices

SocialEnvironments

Gender

Culture

Determinantsof Health

Determinantsof Health

Q


Health and the Environment: A Policy Nexus

How can we measure the impact of the envi-ronment on human health?

Being able to measure the impact of the environ-ment on human health is an important public policychallenge and one that requires a much greaterinvestment than we have given it so far. We need tounderstand the risks and benefits associated withvarious courses of action so we can make policydecisions that favour human health. One way todo this is to identify variables/indicators or proxyvariables/indicators that can help us understandhow human health and systems are affected by ourinteractions with the environment. We’ve madeprogress in some areas. For example, with respect tothe relationship between air quality and cardiorespi-ratory diseases, we’ve been able to measure trendsusing some rough indicators of illness and disease,including mortality and morbidity analyses (seearticle on page 9). But, we have a long way to go!We often focus on acute situations and fail torecognize the underlying chronic dimensions thatmust be taken into account.

How can research inform policy developmentwith respect to health and the environment?

Policy development in the area of health and theenvironment is evidence-based. We use the scientificmethod to understand the ecological processes thatare being disturbed and the actions that can be takento prevent or mitigate their impacts or to avoid themin the first place. This is a monumental researcheffort involving scientists in Health Canada andother government departments, as well as those inacademic institutions and private corporations.

A critical issue relates to the precautionaryprinciple and the role that research plays in thedevelopment of public policy. Applying the precau-tionary principle within a risk management frameworkmeans that, in situations where serious or irreversiblehealth threats exist, action is justified even in theabsence of full scientific certainty. For example, in thearea of children’s environmental health, we may notknow all of the pathways by which environmentalfactors affect the developing physiology of a child,and yet we make decisions to act based on rudimen-tary information with the belief that society expectsus to protect children’s health from environmental

insults. It’s important to remember that the availabilityof evidence is not the sole determinant of policy action.Social values also play an important role.

What is the current status of the evidence baseon the effects of the physical environment on

human health?

While we’ve made progress, the evidence base inmany areas is still fairly new, especially with respectto the cumulative effects of long-term exposure toenvironmental change. For example, an evidencebase is only now emerging about the endocrine-disrupting effects of chemicals that have enteredthe environment as a result of 20th century devel-opment (see article on page 5). In other areas, theevidence base is more mature. For example, we knowthere is tremendous diversity within the humanpopulation, including an underlying genetic diversitythat can lead to predispositions to certain types ofreactions to environmental factors. Consequently,some people are more sensitive than others to factorswithin the environment. It is also clear that we gothrough phases of heightened sensitivity during ourlives, from early childhood through to adolescence,during pregnancy and as we age (see article onpage 13).

There is also a growing body of research exploringthe pathways of exposure by which the environment-health link works. We know, for example, that one’shealth risk to a substance in the environment dependson more than individual sensitivity; it depends onhow concentrated the substance is and the pathwaysor mechanisms of exposure. Understanding suchpathways is critical to eliminate or control theassociated health risks.

As a result of exciting developments in an areathat might be called “molecular epidemiology,” we’rebeginning to understand more about the subtle,but long-lasting effects at the level of the genome.While evidence of these chronic, potentially cross-generational impacts is building, it is not at thesame level as evidence of acute impacts of short-acting toxic materials. Nevertheless, improvedsurveillance techniques at the population level and ourgrowing capacity to analyze surveillance data at themolecular level are contributing to a rapidly growingevidence base in this promising new area.

Q

Q

Q


In this overview of the link

between the environment and

health, the authors focus on the

role of the physical environment as a

determinant of health. They trace the

evolution of knowledge about how

elements of the physical environment

affect health and introduce some

key considerations in assessing and

managing environmental health risks.

The Environment — A Determinant of Health

“The environment is everything that isn’t me.” (Einstein)

This statement not only underscores how pervasive “the environment”actually is, it also alludes to the difficulty of establishing causal linksbetween specific elements in the environment and their impact onhealth. According to the Canadian Environmental Protection Act(CEPA), the federal government’s key environmental protectionlegislation, the environment includes “. . . the components of theEarth, including: the air, land and water; all layers of the atmosphere;all organic and inorganic matter and living organisms; and theinteracting natural systems that include components referred to inthe latter.”1

From a human health perspective, the term environmental healthis even broader, encompassing all aspects of human health, disease andinjury that are determined by factors in the environment. Thesefactors include the direct pathological effects of chemical, physicaland biological agents, as well as the health effects of the broad physicaland social environment (e.g., housing, urban development, land useand transportation, industry and agriculture).2 This concept of envi-ronmental health — as established by the World Health Organization(WHO) — is illustrated in Figure 1.

The concept of health has also expanded over the years, beginningwith the WHO’s 1948 assertion that “health is not merely the absenceof disease but a state of complete mental, physical, emotional andspiritual well-being.” A significant break with past thinking, this broad-er understanding of health continued to evolve, first making inroadsin Canada with the publication of A New Perspective on the Health ofCanadians.3 This report marked the first time a major governmentdocument acknowledged that the influences on health extended beyondhealth care to include human biology (genetic factors), living habits(lifestyle) and the environment.

Anthony W. Myres, PhD, Safe Environments Programme, Healthy Environments and Consumer Safety Branch,Health Canada, and Katherine Betke, School of Applied Biology (Co-op Program), University of Victoria

= HealthyPeople

HealthyEnvironments


Healthy Environments, Healthy People

The Changing Nature of Environment-Related Disease While a holistic approach to public health may berelatively new, the connection between health and theenvironment — in the narrower sense of clean air andwater, and safe and nutritious food — has long beenknown. Advances in these areas have been largelyresponsible for the significant improvements in publichealth in Canada over the past century. However, manyof these environmental health concerns are still seriousproblems among Canada’s First Nation communities(see article on page 15).

Despite recent advances, Canadians are worriedabout the impact of the environment on their health.One of the main reasons is the perception that envi-ronmental health problems are increasing and adverselyaffecting their health now. Almost two thirds ofCanadians believe their health is affected by the envi-ronment, and a similar proportion view environmentalpollution as posing the greatest threat to future gener-ations (compared to 9 percent who viewed “wars andconflict” as the greatest threat).4

This raises an interesting paradox since, by all theestablished measures of public health (e.g., infantmorality, life expectancy), Canadians have never beenhealthier. Yet, serious problems remain. For example,the incidence of children’s asthma, which is certainlyexacerbated by poor air quality, has increased fourfoldsince the 1970s.5

The Chemical RevolutionOne growing environmental concern is the prolifer-ating use of chemicals in industry, agriculture andconsumer products — a chemical revolution that issaid to rival the significance of the IndustrialRevolution. While many industrialized societies nowdepend on chemicals to maintain their standard ofliving, there are increasing concerns about relatedhealth hazards, particularly long-term exposure tolow levels of chemicals and the adverse effects on thedeveloping foetus, infant and young child (see articleon page 13).

The 1962 publication of Rachel Carson’s SilentSpring first drew public attention to environmental

Figure 1: Environmental Influences on Health

Ambientair quality

Surfacewaterquality

Hazardouschemicals

Radiation

Naturalhazards

Waste

Insect-bornediseasesRoad

safety

Crimeand

violence

Childlabour

Drugabuse

Foodsecurity

Foodsafety

Nutritionand diet

DrinkingwaterHousehold

chemicals

Wastedisposal

Housingquality

Indoor airquality

WatersupplyNoise

Sanitation

AMBIENT ENVIRONMENT

NEIGHBOURHOOD

HOME

AMBIENT ENVIRONMENT

NEIGHBOURHOOD

HOME

pollution caused by the indiscriminate use of pesticides.6

This work and the later environmental movements itspawned, helped galvanize governments in Canadaand the United States to establish departments of theenvironment and to enact legislation to protect theenvironment.

The Federal Response Environmental legislation in Canada is a shared respon-sibility between the Ministers of the environment andhealth, with the Environment Minister maintainingoverall administrative responsibility. The country’sfirst environmental protection legislation was the1974 Environmental Contaminants Act, a forerunnerto the more comprehensive Canadian EnvironmentalProtection Act (CEPA), 1988, which was significantlyamended in 1999. Under CEPA, the government hassuccessfully put into place controls on environmentalhazards — for example, phasing out the use of ozone-depleting substances, the emission of dioxins andfurans from pulp mills using chlorine bleaching andthe use of lead and sulphur in gasoline (see article on page 19).

As outlined on page 2, Health Canada has additionalresponsibility for a wide range of legislation designedto protect and promote the health of Canadians. Fourof these Acts are of particular importance for protectingpublic health from chemical risks, either by controllingthe level of hazardous substances in a product, or bycontrolling substance emissions at their source.

Sustainable Development: LinkingEnvironment, Health and Economy The link between health, the environment and theeconomy came into international focus in 1987 whenthe Bruntland Commission defined sustainable develop-ment as “that which meets the needs of today withoutcompromising the ability of future generations to meettheir own needs.”7 In 1995, the federal governmentofficially adopted both pollution prevention and sus-tainable development as elements of its policy agenda.By 1997, all federal departments were required todevelop and implement sustainable developmentaction plans demonstrating the social, economic andenvironmental factors taken into consideration in theirdecision making.8 An overarching federal strategy onsustainable development is currently in development.

Risk Assessment and RiskManagement

“There is no safety without risk.” (Wildavsky)

Recognizing that a completely risk-free environmentis not an achievable goal, governments need a system-atic decision-making process to determine the level ofrisk that is acceptable both to the environment and topublic health. A number of frameworks have beendesigned over the years, all of which include the follow-ing core steps: identify the hazard; characterize andquantify the risk; develop options for controllingthe risk; implement the control measures (risk man-agement); and evaluate. Health Canada’s currentapproach to decision making incorporates thesesteps.9 It should be noted that, although this decision-making process appears simple, it includes a numberof important constraints, such as the need to:10,11

• extrapolate from evidence derived at high doses todetermine risk at lower doses

• extrapolate from animal data to human risk andfrom past or current data to future generations

• allow for the effects of exposure to complex mixtures of chemicals and their interactions

• define and value the quality of life

The Role of Science “When you cannot measure it . . . your knowledge

is of a meagre and unsatisfactory kind.” (Lord Kelvin)

The goal of hazard identification, risk assessment andrisk management is to identify a course of action thatis not only scientifically sound, but also cost effectiveand integrated. In other words, risks are reduced whiletaking into account key social, cultural, ethical, political,economic and legal considerations. Although scientificinput is important, establishing an unequivocal cause-effect relationship is a long and difficult process, andthere are always uncertainties. As a result, discussionsof environmental and health protection over the pastquarter century have given increasing attention to anapproach that is guided by the “precautionary principle.”

The most widely-accepted definition of the precau-tionary principle is one endorsed at the Earth Summitin Rio in 1992. The Rio Declaration states that: “Where





there are threats of serious or irreversible damage, lackof full scientific certainty shall not be used as a reasonfor postponing cost-effective measures to preventenvironmental degradation.” The CEPA has adoptedthis definition and it is now the explicit duty of theGovernment of Canada to implement the principle.

Similarly, Health Canada’s first SustainableDevelopment Strategy recognizes the need to applythe precautionary principle as a means of preventingserious and irreversible impacts on human health.8

This does not mean taking action in the absence ofevidence but, rather, ensuring that the quality andthe weight of the evidence is taken into accounteven though it falls short of scientific “proof.” Nordoes it mean automatically assuming the “worst case”scenario, an approach that may result in inefficienciesor wasting scarce resources that might otherwise beused to address more pressing problems.

Clearly, it is possible for people to become exposedto any substance in the environment. The question is:“What is the risk that exposure will cause harm?”(risk assessment). The corollary is: “What can be doneabout it?” (risk management). The harm that is doneto human health is a function of the inherent toxicity(hazard) and the degree of exposure (i.e., the concen-tration or dose and the length of time exposed).

Hazard is a property of the chemical or substance,susceptibility is a property of the organism beingaffected by the chemical (in this case, human beings)and exposure is a property of the intervening envi-ronment (i.e., the environment, from the point wherethe chemical is released and along its pathway tohumans). Risk is a measure of both the hazard and the

probability of its occurrence (see Figure 2). To illustrate,consider crossing the ocean in a boat — either anocean liner or a row boat. The hazard (of drowning)remains the same, but the risk is considerably reducedin an ocean liner because the probability of exposure,compared to a row boat, is minimal.

Over 400 years ago, Paracelsus observed: “All thingsare poisons, for there is nothing without poisonousqualities. It is only the dose that makes a thing a poison.”This presents a profound truth — overexposure to anysubstance can threaten health. Because the level ofharm depends on the level of exposure, it is a contin-uing challenge to determine through risk assessmentthe level of exposure to particular substances — includ-ing food, water, air and consumer products — that willresult in increased risks to health. Risk managementoptions cover the full spectrum, from minimal inter-vention (e.g., enhancing public awareness) tomaximum intervention (e.g., a ban).

Future Challenges — Finding theRight BalanceScience will continue to play the primary role inestablishing the evidence base for decision making,particularly with respect to identifying environmentalhazards and their links to health outcomes. Togetherwith Health Canada, the Canadian Institutes of HealthResearch (CIHR) is taking a leadership role in devel-oping a national research agenda on the environmentalinfluences on health that is aimed at strengthening theresearch base over the next 10 to 15 years. However,the need for additional research does not necessarily

mean postponing action. The British med-ical statistician, Sir Austin Bradford Hill,12

who wrote a classic text on causality anddisease, highlights the challenge of achiev-ing an appropriate balance:

“All scientific work is incomplete. All sci-entific work is liable to be upset or modifiedby advancing knowledge. That does notconfer upon us a freedom to ignore theknowledge we already have, or to postponethe action that it appears to demand at agiven time.”

Click here for references.@

Risk

Hazard Probability

AdverseHealth Effects

Exposure Susceptibility

Figure 2: What is Risk?


IntroductionIndividuals, corporations and governments have come to rely oneconomic indicators such as the Gross Domestic Product (GDP), theinflation rate and the unemployment rate to help guide their decisionmaking. Spurred by the widespread use of such economic indicators,international organizations and governments at various levels haveinitiated a range of projects focussing on the development of indicatorsin the social and environmental fields (see also “Who’s Doing What?”on page 23). To ensure that these new indicators serve the variedneeds of researchers and policy makers, they must meet a number ofimportant criteria.

What are Indicators?Environmental indicators are important measures of phenomenathat may pose a threat to the natural environment as well as to livingorganisms. Environment Canada’s Indicators and Assessment Officedefines indicators as statistics or parameters that, tracked over time,provide information on trends in the condition of a phenomenon.

Environmental indicators are key statistics that represent or sum-marize a significant aspect of the state of the environment. They focuson trends in environmental changes, as well as on the stresses causingthese trends, how the ecosystem and its components are responding tothe changes, and what society is doing to prevent, reduce or amelioratethe stresses. A significant challenge in developing useful environmentalindicators lies in achieving a balance between scientific accuracy andsimplicity. Often, the difficulty lies in aggregating complex data into simplesummary indicators, particularly if there is no overarching conceptualframework governing the relevant indicators. Environmental indicatorslike the Air Quality Index (AQI) in the box on page 10 provide only

indications of health risks, as nodirect links between indicatorsand health have been established.Therefore, environmentalhealth indicators (EHIs) arebeing developed to measurethe relationship between theenvironment and health.

Environmental Health Indicators

(EHIs) are important tools

for assessing the impact of

environmental factors on human

health. Health Canada is currently

collaborating with other federal

departments to develop a set of

EHIs for use in Canada. This article

examines the need for such indicators

and presents a framework for

navigating the complexities of the

indicator development process.

Sabit Cakmak and Sheryl Bartlett, Safe Environments Programme, Healthy Environments and Consumer SafetyBranch, Health Canada, and Paul Samson, recently with Environment Canada

Environmental Health II nn dd ii cc aa tt oo rr ss



EHIs: Indicators of a RelationshipEnvironmental health indicators are measures ofhealth status attributable to the physical environment.Because people can relate to many of them (e.g., indi-cators of mortality, disability), EHIs are often betterunderstood than environmental indicators. EHIs arealso more amenable to aggregation because establishedmethodologies exist for weighting different health states(e.g., losing lung capacity versus getting cancer). Aswell, EHIs serve as useful adjuncts to environmentalindicators, as they can provide “macro” corroborating(or not) evidence about environmental developments— for example, air quality is improving and air quality-related health problems are diminishing.

Air pollution is an example of an environmentalproblem that has been identified as having health con-sequences. Bell and Davis1 recently reassessed data onthe health consequences of a lethal fog episode duringthe winter of 1952 in London, England. They estimatedthat 12,000 deaths occurring from December 1952through February 1953 were due to acute and persistingeffects of smog (Figure 1 shows the effect of pollutionon mortality during December 1952). Air pollutionlevels for the period were 5 to 19 times above current

(i.e., 2002) regulatory standards and guidelines, aswell as levels in some rapidly developing regions ofthe world.1 Since that study, similar associations havebeen found in studies of other metropolitan areas —for example, ambient levels of particulate matter havebeen linked to daily mortality rates for the period1986-94 in Toronto, Canada (see Figure 2).

How Are EHIs Developed?In developing EHIs, consideration must be given tothe nature of the relationship between a factor in theenvironment and its impact on health. In this regard,an analytical framework is essential in classifying EHIsalong the cause-effect continuum.

Frameworks: A Necessary Starting Point In designing EHIs, a major concern is the ability tolink the impact of the environment to health status,ideally as a cause-effect relationship. Unlike the aboveexamples, however, information may be available oneither exposure or health status, but not on both.In addition, the links between exposure and healthstatus may be tenuous and, as a result, findings mustreflect this uncertainty. For these reasons, it is impor-tant to use an analytical framework in developingEHIs, so that indicators may be classified along thecause-effect continuum.

Several such frameworks have been proposed, mostof which are derived from the Pressure-State-Response(PSR) framework. For example, the World HealthOrganization (WHO) has expanded the PSR frame-work to capture the crucial linkages involved in assessingenvironmental health. The resulting approach, entitledthe Driving Forces, Pressure, State, Exposure, Effects,Action (DPSEEA) model, provides a broad startingpoint from which a set of specifically tailored nationalindicators can be developed.2 While the model offers a“big picture view,” it also focuses attention on indi-vidual components — state, exposure and effects —that are key to determining the environment-healthrelationship.

The issue of air quality illustrates how the DPSEEAFramework and its elements were used to construct anEHI linking air quality and health. Following are the“core” indicators representing the respective elementsof the framework as they relate to air quality:

• urban density/sprawl and the volume and type ofroad traffic (driving forces)

The Air Quality Index (AQI), which relates pollu-tants to National Ambient Air Quality Objectives(NAAQO), is an example of an environmentalindicator that strives to achieve a balancebetween scientific accuracy and simplicity. TheAQI is based on measures of pollutants thathave adverse effects on human health and theenvironment, including sulphur dioxide, nitrogendioxide, ozone, total reduced sulphur com-pounds, carbon monoxide and suspendedparticles measured as the coefficient of haze.Every hour, the concentration of each of thesepollutants at a particular site is converted to anumber on a common scale or index. The valuefor each pollutant is called a sub-index and thepollutant with the highest value determines theAQI for that time period, at that site. The lowerthe AQI, the better the air quality.

An Environmental Indicator



• emissions of air pollutants (pressure)

• ambient concentrations of air pollution, i.e, concen-tration of PM

2.5, PM

10, ozone, etc. (state)

• population exposure to pollution in excess ofmaximum acceptable levels (exposure)

• population mortality/morbidity due to respiratoryor cardiorespiratory disease (effect)

• agreements, initiatives, and programs (action)

Choosing Themes and IssuesEHIs are currently being developed on the following broadthemes: air and atmosphere; water and aquatic systems;land and land cover; and food and food products. WhileEHIs are useful to consider on their own, they are onlyone part of the broader picture of sustainable develop-ment. To be effective, indicators need to: be built aroundclear, specific goals; consider the ambient physical andsocial environment; and be embedded within a sustainabledevelopment context.3

Note: In Figure 1, day zero on the x-axis represents an episode day of high air pollution in December 1952.

Note: In Figure 2, day zero on the x-axis represents the average episode day of high air pollution in 1986-94 in Toronto, Canada.

40.440.240.039.839.639.439.239.038.838.6

Tota

l dea

ths

per

day

4240383634323028262422201816

Parti

cula

te m

atte

r (µ

g/m

3 )

-2 0 2 4 6 8Time relative to episode day

Deaths

Pollution

Figure 2: Effect of Pollution on Mortality, 1986-94, Toronto, Canada

Figure 1: Effect of Pollution on Mortality, December 1952, London, England

-6 -4 -2 0 2 4 6 8Time relative to an episode day in December 1952

500

450

400

350

300

250

200

Tota

l dea

ths

per

day

2,500

2,000

1,500

1,000

500

0 Parti

cula

te m

atte

r (µ

g/m

3 )

Deaths

Pollution



Criteria for Selecting IndicatorsMuch has been written about the criteria for selectingindicators. For example, the OECD’s foundation workon indicators4 discusses both technical/scientificand user/policy elements as criteria. Scientific-basedcriteria include data availability and sustainability,validity, representativeness, reliability and the abilityto be broken down into other variables. User-basedcriteria for indicators include the feasibility of accessand the relevance of the indicator to those affected.Indicators should also be scientificallysound, robust, easily understood, sensitiveto the changes they are meant to represent,measurable and capable of being updatedregularly.5

An Example of an EHI: Making the Air Quality-Health LinkThe primary motivation for reducing airpollutant levels is to protect populationhealth. However, the Air Quality Index(AQI) cannot actually tell the magnitudeof the impact of exposure to air pollutantson health. Burnett et al.6 have proposedan environmental health indicator thatmakes a direct link between air qualityand health. This indicator measuresimprovements in population health, basedon reductions in ambient fine particu-late matter over time. It is a function ofseveral factors — including temporal changes in site-specific ambient concentrations and the relationshipbetween those concentrations and daily mortality orhospital admissions rates for heart and lung problems.The new indicator, which is based on a methodologyused in a number of epidemiological studies, can bedetermined for a single location, or at the regional ornational level, and can also be expanded to includeseveral pollutants. It is even possible to extend themethodology so that the annual number of deathsattributed to fine particulate exposures can be trackedover time across Canada.

EHIs: How Can We Use Them?The potential negative impact of the environment onhuman health has long been a policy concern. Thereis widespread agreement that access to valid andrelevant information about local and national health

impacts of environmental hazards is key to developingand monitoring policy in this area. EHIs offer a concise,effective and easily understood way of making suchinformation available to public health agencies, decisionmakers and health professionals. Within this context,EHIs serve an important function in several key areas:

Identification of hazards/risks: EHIs can be used tomonitor environmental health hazards, thereby helpingto identify and investigate potential links betweenenvironmental factors and health effects.

Decision making and policy development:EHIs can provide input into the decision-making process by monitoring healthtrends in the context of environmentalexposure and risk factors.

Setting and evaluating program objec-tives: Specific EHIs can be used to setprogram objectives that can then betracked and re-evaluated over time —these performance measures can help inassessing the effects of various policiesand interventions.

Accountability and environmentalhealth reporting: Federal departmentsand agencies play an important role inproviding information to Parliament andCanadians. In the past, performancemeasurements have been based largelyon subjective assessments rather than on

objective measures of the outcomes of various actions.Indicators provide a level of transparency in environ-mental reporting that was not previously possible.

Moving ForwardEnvironmental health indicators show significantpromise as a means of credibly and transparentlyconnecting science and policy. However, complexenvironmental health issues may require aggregatingsets of indicators into indices. Researchers will needto respond to these needs by moving a step further todevelop easily understandable indices for use in decisionmaking by government policy makers, business leadersand members of the public.


To be effective, indicators

need to: be built around

clear, specific goals; consider

the ambient physical

and social environment;

and be embedded

within a sustainable

development context.

Critical Pathways


In Northern CommunitiesMany people living in the Canadian Arctic rely on a diet comprisedprimarily of fish and wild game. As a result, they may be exposed torelatively high levels of contamination, due to a cumulative buildup ofcontaminants higher in the food chain. The Northern ContaminantsProgram (NCP), which operates under the aegis of Indian andNorthern Affairs Canada, addresses the issue of food contamination innorthern diets. One of the principal tenets of the NCP is that risks mustbe balanced against benefits. Within this context, the current consensusis that the known nutritional, social and cultural benefits of consuming“country foods” (i.e., from fish and wildlife harvesting) outweigh thehealth risks of contaminants in these foods.1 Health Canada plays a leadrole in the NCP’s health program.

A number of other environmental issues have particular implicationsfor people living in Canada’s North. For example, the Arctic climate actsas a condenser, creating a “sink” for pollutants such as Persistent OrganicPollutants (POPs — e.g., DDT, PCBs) that are conveyed by long-rangetransport. To address this issue, Canada played a lead role in developingthe UN Convention on Persistent Organic Pollutants and was the firstcountry to ratify the Convention.2

Throughout the Life CycleResearch has also shown evidence of heightened sensitivity to environ-mental hazards at certain stages of human development:

Infancy and ChildhoodBecause their tissues and organs are undergoing rapid cellular develop-ment, infants and very young children are more vulnerable than adultsto environmental hazards.3 While there is some debate about whetherchildren are always more susceptible to chemical toxicity than adults,4

there is no doubt that infants have unique characteristics (metabolism,exposure patterns, behavioural features) and cannot be consideredsimply as “little adults.”5 Breast milk is one pathway through whichchildren are exposed to environmental contaminants. As some contam-inants can build up in breast milk (from those stored in the mother’sfat deposits), breastfed infants may be exposed to high doses of

Safe Environments Programme, Healthy Environments andConsumer Safety Branch, Health Canada

In assessing environmental risks to

health, it is important to recognize

that people are not all alike, either

in their level of exposure or their ability

to detoxify. While the goal of public

health is to protect the health of all,

some subpopulations may be more at

risk than others. Understanding the

impact of environmental hazards on

specific populations can result in

concrete steps towards identifying,

preventing, reducing and eliminating

various health risks.

Anthony W. Myres, PhD

VulnerablePopulations


Vulnerable Populations — Critical Pathways

contamination. However, the unique benefits ofbreastfeeding — nutritional, immunological, psycho-logical and emotional — far outweigh the risks fromcontaminants, a conclusion that has been endorsed byhealth professionals and governments worldwide.6

Many of the major disorders confronting Canadianchildren today are chronic, disabling conditionssometimes referred to as the “new pediatric morbidity.”Besides injuries and obesity, the new pediatric mor-bidity includes increases in asthma, disorders ofendocrine and reproductive development and neuro-developmental dysfunction. Evidence shows thatchemicals in the environment may be contributingfactors to these conditions.7 Because healthy early childdevelopment is a prime determinant of later health,there has been an enormous international effort to makechildren’s environmental health a priority.8-12

AdolescenceAdolescence is another “window of vulnerability,” dueto the rapid growth rate and the surge of hormonesrelated to sexual development. Of partic-ular concern are endocrine-disruptingsubstances (EDS), including a variety oforganic and inorganic pollutants, thathave the potential to interfere with hor-mones controlling growth, development,reproduction and the function of theimmune and central nervous systems.While there is no conclusive evidence ofhuman health risks due to current levelsof EDS exposure, adverse effects havebeen documented in fish and wildlife, aswell as on laboratory animals. Studiessuch as these suggest a number of possibleeffects on human health, including earlieronset of puberty, altered development of male andfemale reproductive tracts and reduced sperm pro-duction.13-15 Animal studies indicate that these effectsare likely a consequence of earlier, possibly in utero,exposure.

The Reproductive YearsEnvironmental hazards can also affect fertility andpregnancy outcomes, as measured by rates of infertility,spontaneous abortion, chromosomal anomalies,pre-term delivery, low birthweight and stillbirths. Lowdosage exposure of pregnant women to some toxicsubstances may affect neurophysiological and other

facets of fetal development at critical stages, resultingin later learning disorders or other conditions.16

Very little research has been done on the sex/gender-specific effects of exposures to toxicants such aspesticides and metals.17 However, men and womenare known to have different vulnerabilities to theeffects of EDS. For women, these may be related to ahigher proportion of fatty tissue, and to hormone-regulated cycles and processes of menstruation,pregnancy and menopause. In men, EDS exposuremay contribute to “male reproductive syndrome,”as manifested in lower sperm counts, reproductivebirth anomalies and testicular cancer.18 It appearslikely that such clinical outcomes are due to eventsoccurring very early in development.19,20

The Senior YearsOver the past century, medical science has made itpossible for people to live much longer. Ironically,longer life also means greater exposure to the harmfuleffects of the environment. As a result, seniors face

particular health risks, not only due to thecumulative effect of environmental haz-ards, but also to the declining “marginof safety” in key organ systems such as thecardiorespiratory and immune systems,and to the presence of pre-existing healthconditions.

The effects of this combination of ageand disease-related risks are particularlyevident in seniors’ greater susceptibility toair pollution. Aging commonly reduces aperson’s “maximum oxygen uptake,” withthe result that seniors can find it difficultto meet their daily oxygen requirements.21

In such cases, even low levels of pollutantscan have a critical effect. Research also shows that theprevalence of chronic obstructive pulmonary disease(COPD) increases with age;22 lungs and airways thatare compromised by COPD are more susceptible tothe effects of both indoor and outdoor air pollution.

A special thanks to the following people for their input into this article: Sari Tudiverand Monica Mavrak, Women’s Health Bureau, Health Policy and CommunicationsBranch; Louise Plouffe, Tye Arbuckle and Barbara Sérandour, Centre for HealthyHuman Development, Population and Public Health Branch; and Priya Raju, SafeEnvironments Programme, Healthy Environments and Consumer Safety Branch,Health Canada


Over the past century,

medical science has made it

possible for people to live

much longer. Paradoxically,

longer life also means

greater exposure to

the harmful effects of the

environment.


What is Shigellosis?While cardiovascular disease and cancer may strike fear in the heartsof many Canadians, most people probably don’t know anythingabout a potentially fatal disease called shigellosis. Shigellosis is anacute, bacterial disease characterized by diarrhea, fever and nausea.People may become infected by ingesting food or water contaminatedwith shigella bacteria, or through fecal-oral, person-to-person spread.The disease has an incubation period of one to three days and lastsan average of four to seven days. Treatment in Canada usually includesoral rehydration salts and sometimes antimicrobial medications.Thus, case fatality in Canada is fortunately quite low.

Diarrheal diseases are the second leading cause of mortalityworldwide among children less than 5 years of age. Shigella is oneof five organisms that cause the majority of pediatric diarrhea casesin almost all geographic regions. It is the most significant cause ofbloody diarrhea in the world and is responsible for nearly all episodesthat are clinically severe or fatal.1

Shigellosis in CanadaAll provinces and territories have legislation in place to ensure thatcases of shigellosis are reported to public health authorities. In 1999,the reported incidence rate of shigellosis in Canada was 3.6 per100,000 people. Children aged 5-9 years experienced the highest rateof any age group, at 14.5 per 100,000. Overall rates were higher inthe Prairie provinces than in other parts of Canada, with the highestrates in Manitoba (14.0 per 100,000).2

About the DataWorking in partnership with First Nations and provincial healthauthorities, the First Nations and Inuit Health Branch (FNIHB) ofHealth Canada provides primary health services and public healthprograms in First Nations communities. Cases of shigellosis and othernotifiable diseases occurring in First Nations communities are reportedto FNIHB regional offices. Notifiable disease data for 1999 werecompiled at the national level, analyzed and compared to 1999 data for

Michael Clark

Low rates of diarrheal diseases in

developed nations such as Canada

reflect enormous improvements in

hygiene and infrastructure during the

past century. Despite this progress, the

recent Escherichia coli outbreak in

Walkerton is a reminder that some

communities are still vulnerable to

diarrheal diseases linked to the environ-

ment. A large outbreak of the diarrheal

disease shigellosis also occurred recently

in Ottawa, due to a contaminated food

source. Although shigellosis outbreaks

of this magnitude are not common in

most Canadian communities, many

First Nations communities experience a

disproportionate burden of this disease

as a result of a range of environmental

conditions.

Primary Health Care and Public Health Directorate, FirstNations and Inuit Health Branch, Health Canada

First Nations Communities

Shigellosisand


Shigellosis and First Nations Communities

the entire Canadian population. As well, trend data ofreported cases in three regions and hospital separa-tion data in two regions were obtained from FNIHBregional offices and computerized hospital dischargefiles, respectively.

A Disproportionate BurdenThe First Nations on-reserve population for whichshigellosis cases were reported represents 1.1 percentof the Canadian population. However, 23 percent ofall reported shigellosis cases and 47 percent of casesamong children aged 0-14 years occurred in the FirstNations on-reserve population in 1999. The reportedincidence rate of 74.1 per 100,000 among First Nationscommunities for that year was 26 times higher thanthe non-First Nations rate of 2.8 per 100,000.

The vast majority of First Nations cases (93.6 per-cent) in 1999 were reported in Alberta, Saskatchewanand Manitoba. Reported incidence rates were consis-tently higher among First Nations communitiesthan in the non-First Nations populations of theseprovinces during the late 1990s (see Figure 1). Hospitalseparation rates for shigellosis were also higheramong First Nations in Saskatchewan and Manitobathroughout the decade. Over 80 percent of FirstNations shigellosis patients hospitalized in the twoprovinces during that period were children aged 0-14

years. An epidemic in Manitoba during the early 1990saffected more than half of the First Nations communi-ties in that province.

Most Canadians escape shigella infection duringchildhood, while those who do become infected aregenerally exposed as a result of travel in high-riskcountries or during common-source outbreaks, suchas the recent food-borne outbreak in Ottawa. In devel-oping nations, where epidemics of shigellosis arecommon, the majority of morbidity and mortalityoccurs among children. Notably, the age distribution of shigellosis cases in the Canadian First Nations pop-ulation is very similar to what is generally observed indeveloping countries.

Figure 2 shows 1999 age-specific incidence rates inthe First Nations and non-First Nations populations atthe national level. The highest rate was reported amongFirst Nations children aged 1-4 years (250 per 100,000).Eighty-six percent of First Nations cases in 1999 occurredamong children aged 0-14 years, while 30 percent ofnon-First Nations cases were reported in this age group.

Environmental LinksResearchers have identified important links betweenshigellosis and a number of factors in the environment,including sewage disposal methods, water supply sys-tems and housing conditions.

Figure 1: Comparison of Reported Shigellosis Incidence Rates in the First Nations and Non-First Nations Populations of Three Provinces, 1994-98

*Note: Data points represent three-year moving incidence densities (years represent mid-year of three-year periods).

1994 1995 1996 1997 1998

300

250

200

150

100

50

0

Rate

per

100

,000

Alberta First Nations

Manitoba First Nations

Saskatchwan First Nations

Non-First Nationsfrom the three provinces



Sewage DisposalHumans are the only significant reservoir of shigellabacteria. To be a source of infection, water and foodmust be contaminated with human feces containingshigella. As a result, communities with inadequatesystems for sewage disposal are at increased risk forshigellosis. People living in households lacking sewageremoval systems must sometimes use indoor pailsfor toilets. Children may then be exposed to infectionif feces and diapers are disposed of in the yard.Communities with home sewage systems may alsobe at risk for shigellosis if sewage disposal systems areimproperly constructed. Shigellosis cases have beenlinked to sewage backing out through basement drainsand poorly constructed surface disposal systems.Families using wells may also be at risk if septic sys-tems are located near the well or well water source.

In many cases, sewage systems in First Nationscommunities do not meet provincial design andinstallation standards. The proportion of FirstNations households with adequate sewage disposalincreased from 79 percent in 1990 to 94 percent in2000.3 While this is a significant improvement, manypeople in areas where shigella is endemic still do not have access to adequate sewage disposal. In 1999,22 percent of housing units in First Nations commu-nities in Manitoba lacked modern plumbing (an indoortoilet and an assured supply of running water).4

Water SupplyFollowing the recent E. coli outbreak in Walkerton,concerns regarding the quality of available drinkingwater have been raised throughout the country.While the importance of ensuring a safe supply ofdrinking water should not be underestimated,research has shown that simple access to enoughwater for daily washing with soap (water quantity)is a far greater determinant of childhood diarrhealdisease than water quality.1 Access to a sufficientwater supply for basic hygiene is a problem in manyFirst Nations communities — a concern one mightexpect to find in developing countries but not inCanada.

In communities lacking a water delivery system,families must fill drums from a standpipe or at lakesand rivers. Others may have water delivered by truckto barrels in their houses or to household cisterns.These methods of water supply limit the amount ofwater available for hand washing, which can increasethe risk of fecal-oral person-to-person spread ofshigella bacteria. A study of Manitoba First Nationscommunities found that shigellosis rates were sixtimes higher in communities with truck-to-barrelwater delivery than in communities with piped sys-tems. The association between the type of waterdelivery and the incidence of shigellosis remainedsignificant in a multivariate analysis of risk factors.5

Figure 2: Age-Specific Shigellosis Incidence Rates in the Canadian First Nations and Non-First Nations Populations, 1999

250

200

150

100

50

0

Rate

per

100

,000

First Nations Non-First Nations

< 1 1-4 5-9 10-14 15-19 20-24 25-29 30-39 40-59 60+Age group



Crowded HousingOvercrowded housing conditions increase the contactrate between individuals and the risk of person-to-person spread of many communicable diseases.Shigellosis is the most communicable of the bacterialdiarrheas. Only 10 to 100 viable organisms must beingested for the disease to occur in an individual.6

Forty percent of people who are exposed to a case ofshigellosis may be infected through person-to-persontransmission and attack rates may be much higheramong children.7 Shigellosis rates in the Manitoba FirstNations population have been shown to increase withrising household densities.5 Overcrowded living condi-tions is a problem in many First Nations communities— the average housing density among First Nationson-reserve dwellings is 0.7 persons per room, comparedto the Canadian average of 0.4 persons per room.

Combatting ShigellosisThe high rates of shigellosis among First Nationspeople and the impact of this disease on children’shealth are unacceptable. Short- and long-term strate-gies must be implemented to combat shigellosis andits environmental determinants.

Short-Term PreventionIn communities at high risk for shigellosis outbreaks,health promotion can play an important short-termprevention role. Health promotion messages shouldpromote careful hand washing with soap and water,proper food handling practices, breastfeeding ofinfants and boiling water for infant formula. Childrenwith shigellosis should be excluded from daycarecentres until their stool samples test negative for thebacteria. Emphasizing good personal hygiene andsanitary food preparation techniques is particularlyimportant for those preparing food for large culturalevents, such as pow wows. Parents living in commu-nities with inadequate sewage disposal systems shouldbe educated to dispose of feces and diapers away fromareas frequented by children. Health promotion andeducation initiatives can be carried out in communitymeetings, workshops and radio programs and bydistributing pamphlets in First Nations communities.

It is also important to carry out inspections of publicfacilities, such as daycare centres and long-term carefacilities, to identify factors that may lead to disease.Environmental assessments of water supply and sewage

systems can identify problems of contamination. Thesefunctions are mainly the responsibility of public healthauthorities in First Nations communities, such asFNIHB and First Nations transferred health authorities.Medical officers of health, environmental healthofficers, community health nurses and communitystakeholders are all important in conducting assess-ments and implementing effective corrective action.Early detection of shigellosis cases and rapid interventionrecently prevented a large outbreak in one AlbertaFirst Nations community.

Long-Term PreventionLong-term strategies include those directed at combat-ting the environmental determinants of shigellosis.Measures that will contribute to a long-term reductionin disease burden are described briefly below.

Continued on page 26

Unfortunately, these solutions are beyond the controlof public health authorities, although they mustadvocate for changes when links between environ-ment and disease are found. Responsibility for findinglong-term solutions is shared among the Departmentof Indian and Northern Affairs (INAC), First Nations

Measures for the Long-Term Preventionof Shigellosis in a Community

ensure that the water supply is adequate for daily washing with soap

provide sewage disposal and treatment systems

ensure houses and sewage systems meetappropriate standards for design and installation

build septic systems at an adequate distance and downhill from wells and well water sources

introduce measures to limit overcrowded living conditions


A Three-Part ProcessGenerally speaking, air quality mitigation strategies use a varietyof tools to predict the potential for such strategies to deliver airquality improvements and associated health benefits (see Figure 1).

The sulphur in gasoline review, the most detailed of these effortsconducted to date, was designed in three stages — process design,fact finding, and the development of options and recommenda-tions. In the first stage, government and industry representativesdeveloped an overall plan for the process, while in the second stage,three expert panels gathered evidence in the following issue areas— atmospheric modelling,1 health and environmental impacts,2

and industry cost and competitiveness.3,4 In the third stage, gov-ernment regulators built on the results of the fact-finding stage todevelop regulations for sulphur levels in gasoline for Canada.

Building the EvidenceIt is widely recognized that sulphur is a poison for many chemicaland catalytic processes. Evidence indicates that sulphur at levelspresent in Canadian gasoline prior to regulation significantlyreduces the capability of the automotive catalytic converter. As aresult, cars emit much more sulphur dioxide, particulate matter,nitrogen dioxide, sulphates and other smog-forming pollutantsthan they would using gasoline with lower sulphur levels. Thesulphur-in-gasoline process used this knowledge as its startingpoint for determining the implications of establishing a regulationconcerning sulphur levels in gasoline.

The fact-finding stage of the process focussed on a number ofscenarios, including six potential regulation levels for sulphur ingasoline, ranging from 350 to 30 parts per million (ppm). Theexpert panel on atmospheric modelling used the scenarios toproject changes in tailpipe emissions and the resultant effectson air quality. The panel on health and environment, in turn,used these projected changes to estimate impacts on human andenvironmental health.

While the most significant changes in air quality were projectedfor sulphur dioxide and sulphate, significant reductions were also

Due to concerns about the contribution

of vehicles to poor air quality, the

Canadian government, in conjunction

with the petroleum industry, motor vehicle

manufacturers, provincial governments and

public interest groups, undertook a detailed

analysis of the impact of sulphur levels in

gasoline on air quality and health, and the

implications for Canada’s fuel industry. Up

until the late 1990s, sulphur levels in gasoline

were not regulated and, as a result, varied

widely across the country according to crude

oil quality and refinery capabilities. Sulphur

in gasoline contributes directly to tailpipe

emissions and atmospheric pollution but, more

importantly, poisons catalytic converters.

Catalytic converters are installed on all modern

automobiles and reduce a wide range of gaso-

line combustion products that are toxic and

contribute significantly to smog. The authors

explore how the evidence that emerged from

this multi-stakeholder process provided the

basis for legislation regulating the sulphur

levels in gasoline.

Barry Jessiman and Rick Burnett,Safe Environments Programme, HealthyEnvironments and Consumer Safety Branch(HECSB), Health Canada; and Paul deCivita, Policy and Planning Directorate,HECSB, Health Canada in Gasoline and Other Fuels:

The Case for Action (and Inaction)


Sulphur in Gasoline and Other Fuels

predicted in several other pollutants, including carbonmonoxide, nitrogen oxides, volatile organic compoundsand, to a small extent, ozone. The primary source ofdata was epidemiological literature focussing onsulphate as the primary pollutant and providingquantitative estimates of its effects. A key element inthe expert panels’ deliberations was the presumptionthat the sulphate signal captured in the publishedliterature was indicative of the type of emissionsbeing reduced, rather than being the causative agent.In fact, panel members acknowledged the possibilitythat using this approach would underestimate thetrue benefits of reducing sulphur levels in gasoline,since this literature would only partially capture theeffects of the reduction of the other pollutants(i.e., the “soup” of chemicals to which people areexposed and which cause effects).

At the same time, however, they felt that this literature provided the most credible approach, as itcontained sufficient material to provide some estimateof the health effects that would be avoided due toreductions, including premature mortality, chronicrespiratory disease, cardiac and respiratory hospital-ization, emergency room visits, asthma symptoms,restricted activity, and acute respiratory symptomsand lower respiratory illness in children.

As part of the fact-finding phase, a monetary valu-ation of the avoided effects was conducted, based onhealth economics literature. The panels also calculated

the actual costs of retrofitting refineries to meet thevarious proposed sulphur standards, as well as theprojected impact of these standards on the competitive-ness of the industry both within Canada and abroad.

For the most stringent option considered (i.e., 30ppm, the current standard in California), it was deter-mined that the cumulative health impacts avoidedover a 20-year period would include: 11 million newcases of croup and pneumonia; five million days ofasthma or other restricting illnesses; 100,000 newcases of child and adult chronic bronchitis; 9,000emergency or hospital admissions; and over 2,000cases of premature mortality. While this optionreduced only a small percentage of total mortality(annual mortality in Canada is approximately240,000), hospital admissions, emergency room visits,chronic bronchitis and other endpoints, it appearedto be a cost-effective measure. The associated eco-nomic value of avoiding these health effects (basedprimarily on monetary values associated with prema-ture mortality and illness costs) was estimated to begreater than $6 billion over 20 years. The costs toindustry for complying with the proposed standards(based on new capital costs and ongoing operatingcosts) were estimated to be under $3 billion over 20 years. It is important to note that, because of datalimitations, health benefits could only be calculatedfor the seven largest Canadian cities (representingabout 40 percent of the national population), while

Figure 1: General Process Chart for Conducting Air Quality Benefits Analysis

1 2 3 4 5 6

Selectpollutionreductionstrategy

Model changesin emissions

Model changesin ambient air

concentrationsof pollutants

Model changesin humanexposure

Modelchanges in

health effects

Model value ofhealth benefits

and societalcosts

Estimate uncertainty in each step Conduct sensitivity analysis

Integrating Air Pollution Science and Policy — How to Evaluate a Mitigation Strategy



refinery costs were calculated for the entire Canadianindustry. For this reason, the expert health panelcharacterized the benefits as an underestimate ofthe true total.

Regulating ActionBased on these findings, the Government of Canadachose the most stringent standard of 30 ppm. Of thesix options studied, this standard resulted in thegreatest difference between benefits and costs (seeFigure 2). It is interesting to note that, while the healthbenefits were directly proportional to sulphur reduc-tions, the costs to industry were not. Small reductionsin sulphur could be gained by relatively minor inter-ventions in most refineries, while more ambitioustargets (approximately 150 ppm) required significantinvestment to install complex technologies. Havinginstalled these technologies, further reductions in sulphurlevels do not require additional expensive technology,only increased operating costs. As a result, costs werenot linear with reducing sulphur levels and, combinedwith other (often refinery-specific) factors, resultedin the most stringent level providing the greatest netbenefits. Factors that played a key role in the decisionincluded the effects on catalytic converters, the directhealth benefits and reasonable costs.

The new regulation, as embodied in the CanadianEnvironmental Protection Act (CEPA), requires aphase-in period with a 150 ppm (average) standardthat started in July 2002, with the 30 ppm standardto be achieved by January 2005. This regulation willbring Canadian sulphur levels into line with currentJapanese and California standards, as well as withproposed US and European standards.

In the process leading up to the regulation’s devel-opment, the industry indicated that up to six of theeighteen existing Canadian refineries might have toclose, and that supply shortages and significant pricehikes in southern Ontario would result if the 30 ppmoption was chosen. After the regulation was passed,the industry lobbied strongly for an amendment thatwould allow individual refining companies to elimi-nate the phase-in period beginning in 2002 and moveforward the date of the 30 ppm requirement by oneyear. After detailed refinery-by-refinery analysis, thegovernment concluded that this strategy would resultin fewer health benefits than the existing regulationbecause moving forward the 30 ppm date by oneyear would not fully compensate for eliminating thephase-in period. Based on this further scientificanalysis, the government declined to amend theregulation. The industry has since indicated that they

Cos

ts or

ben

efits

(in

billi

ons

of $

CD

N 1

994)

Figure 2: Costs and Benefits of Sulphur-Reduced Gasoline (in terms of net present value; year 2000)

Costs Benefits Net benefits

30 ppm 100 ppm 150 ppm 200 ppm 250 ppmSulphur level

7

6

5

4

3

2

1

0

*Net benefits = total benefits minus total costs.

Costs and benefits estimated for the years 2001-2020.

Costs estimated country wide; benefits for seven cities including: Toronto,Montréal, Vancouver, Halifax, Winnipeg, St. John and Edmonton.



do not expect to close any refineries and that there areno technological or economic constraints in meetingthe regulation.

Lessons Learned: Moving ForwardAs part of ongoing efforts to reduce the impacts offossil fuel combustion (e.g., home heating oil, diesel,heavy fuel oil), the federal government is consideringa series of sulphur-related measures. However, it isimportant to note that the experts involved in thesulphur in gasoline panel were greatly persuaded bythe multi-pollutant reductions afforded by reducingsulphur, which allowed the automotive catalyticconverter to perform at its design level. This is a keyconsideration in attempting to extrapolate the sulphurin gasoline results (even qualitatively) to other situa-tions. For example, Canada has announced aregulation that will require greatly reducedemissions from diesel engines and requirevery low sulphur levels in diesel. Thisregulation will require the installationof catalytic converters and particletraps (both of which require very lowsulphur fuel). As was the case withgasoline, the significant sulphur reduc-tions will enable technologies thatreduce a number of pollutants (nitro-gen oxides, particles, etc.). While thespectrum of pollutants is quite differentand the direct application of the sul-phur in gasoline results might bequestioned, there is nonetheless a com-pelling case for introducing theregulation and the promise of healthbenefits as a result of doing so.

However, when the discussionmoves to other fuels, such as light fueloil (LFO), heavy fuel oil (HFO) andeven coal, the supposition that reduc-ing sulphur will result in health benefitsis questionable. Currently, there is a movement toimplement regulations that will reduce sulphur levelsin LFO and HFO on the expectation of health bene-fits. However, since a technological aspect (i.e.,catalytic converters or particle traps) is not included,the reductions in emissions will be confined to sul-phur dioxide and sulphate, with no expectation ofreductions in nitrogen oxides, carbon monoxide,

ozone or volatile organic compounds. This factorlimits the applicability of the logic and rationaledeveloped and followed by the sulphur in gasolinestudy described above.

On the other hand, sulphate is a particle, a factthat would appear to provide a justification for thesemeasures as particulate matter (PM) has been declaredtoxic under the CEPA. However, the strategy comesinto question when one considers the basis for labelingPM as toxic. The evidence is almost exclusively basedon large-scale epidemiological studies (the same typeof study used by the health effects panel), which havefound a significant and consistent association betweenambient PM and mortality. The consistency of thefindings from these studies has convinced health andenvironmental agencies (including Health Canada,the US Environmental Protection Agency and the

WHO) to support actions to reduce PM in general.It must be noted, however, that the PM

referred to in the above studies is a chemicallycomplex mixture and part of a “soup” ofother air pollutants (e.g., carbon monoxide,nitrogen dioxide, ozone, etc.). When a riskmanagement strategy reduces many ofthese constituents, there is a reasonablelevel of confidence that the causative agentsare being addressed. Where one specificcomponent is selectively reduced, there isless confidence of a benefit.

To summarize, sulphur reductions willenable technology that reduces a spectrumof air pollutants in the case of gasoline anddiesel. For other sulphur reduction strate-gies, only one or, at best, two pollutants willbe reduced. Current PM research is focusedon finding the factors that give PM itstoxicity. Such research, given time, will givemuch greater focus to risk managementstrategies and potentially deliver muchgreater health benefits for the money spent.Extrapolating the results of the sulphur in

gasoline analysis beyond a reasonable degree maylead to misdirection of risk management resourcesand the development of a potentially false sense ofprogress in combatting the population health impactsof poor air quality.


Canada has announced

a regulation that will

require greatly reduced

emissions from diesel

engines and require very

low sulphur levels

in diesel.


Who’s Doing What?Who’s Doing What?

National Environment and HealthResearch AgendaThe Canadian Institutes of Health Research (CIHR)are collaborating with Health Canada, EnvironmentCanada and other stakeholders to set a national researchagenda on environmental influences on health. Theagenda aims to strengthen research, guide fundingdecisions, and encourage the development of multi-disciplinary, multi-sectoral research projects andinnovative funding partnerships. CIHR developed adraft discussion paper for a national conference heldon September 13-14, 2002 (available at: http://www.cihr-irsc.gc.ca/institutes/ihdcyh/index_e.shtml).

Surveillance and IndicatorsA number of organizations report on various aspectsof the health-environment relationship:

• In March 2002, the Environics ResearchGroup conducted a poll for HealthCanada entitled “Air Pollution and ItsImpact on Health” with Canadian healthcare professionals. The study addressedtopics such as perceived effects of airpollution on health and perceived seri-ousness of air-related health problemsin comparison to other health problems.For more information, [email protected]

• The World Health Organization (WHO)publishes various reports on environmentand health, such as Environmental HealthIndicators: Framework and Methodology(available at: http://www.who.int/environmental_ information/Information_resources/on_line_documents.htm).

• The Health Effects Institute (HEI) is an independ-ent, non-profit corporation that aims to providehigh quality, impartial and relevant research on thehealth effects of pollutants from motor vehiclesand other sources in the environment. A collabora-tive initiative of the US Environmental ProtectionAgency and industry, HEI has published over 100research reports (available at:http://www.healtheffects.org).

• The Health and Environment Group of the Centrehospitalier universitaire de Québec (CHUQ) bringstogether professionals and researchers from manydisciplines to research topics related to health andthe environment. For more information aboutCHUQ, see: http://www.chuq.qc.ca/oms/en/mission/mission.htm

• The International Joint Commission (Canada/USA)held a Conference on Environmental HealthSurveillance in Québec City in October 2000.Conference papers focussing on developing envi-ronmental health indicators are available at:http://ottserver1.ottawa.ijc.org/hptf

• The goals of the Environmental and OccupationalHealth Surveillance Working Group are to: identifyrelevant surveillance networks and systems inCanada; assess needs and opportunities to strengthencapacities; and recommend initiatives and priorities

for Canada. The Working Group reports to theFederal/Provincial/Territorial Committee onEnvironmental and Occupational Health, whileHealth Canada’s Healthy Environments andConsumer Safety Branch (HECSB) coordinatesthe Working Group’s efforts in collaborationwith its partners. For more information, [email protected]

Women, Health andEnvironments The Centres of Excellence for Women’s HealthProgram, which is funded by Health Canada’sWomen’s Health Bureau, has released a number of research papers on the subject ofwomen, health and environments (available at:http://www.cewh-cesf.ca/).

Who’s Doing What? is a regular column of theHealth Policy Research Bulletin that looks at

key players involved in generating policy researchwithin a specific theme area. Because the theme of“Health and the Environment: Critical Pathways”covers such a broad spectrum, the current columnfocuses on stakeholders active in the specific policyresearch areas addressed in this issue of the Bulletin.

htt

p:/

/ww

w.c

ew

h-

ce

sf.c

a/

mailto:[email protected]

http://www.who.int/environmental_information/Information_resources/on_line_documents.htm

http://www.healtheffects.org

http://www.chuq.qc.ca/oms/en/mission/mission.htm

http://ottserver1.ottawa.ijc.org/hptf


http://www.cewh-cesf.ca/

http://www.cihr-irsc.gc.ca/institutes/ihdcyh/index_e.shtml


Who’s Doing What?

Children’s Environmental Health (CEH) The effect of the environment on children’s health hasbeen the subject of several recent workshops andother initiatives of note:

• In May of 2000, the Five Natural ResourcesWorking Group on CEH, established by HealthCanada, organized a national workshop to identifypriorities and opportunities for interdepartmentalcollaboration. A report on the workshop is availableat: http://www.durable.gc.ca

• Health Canada and Environment Canada cohostedthe Canadian CEH Research Workshop in Ottawain March 2002. Among other objectives, the work-shop aimed to develop a Canadian CEH researchagenda. The program is available at: http://www.hc-sc.gc.ca/pphb-dgspsp/cehs-esm/wkshop_e.html

• The Commission for Environmental Cooperation(CEC, Mexico/USA/Canada), the InternationalJoint Commission and the Pan American HealthOrganization are currently preparing a joint projecton measuring children’s environmental health. Theproject is part of the CEC CooperativeAgenda for Children’s Health and theEnvironment in North America (see:http://www.cec.org).

• In 2001, the Canadian Institute ofChild Health and the US Children’sEnvironmental Health Networkcohosted a Global Forum on CEH. Theresulting Joint Declaration on Children’sEnvironmental Health outlined signifi-cant threats to the health of the world’schildren and identified areas in whichimmediate action was needed (availableat: http://www.cich.ca/postglobal.htm).

Food and Water SafetyA number of important initiatives are underway to address issues related to food andwater safety:

• Health Canada’s Healthy Environ-ments and Consumer Safety Branch(HECSB) recently collaborated withEnvironment Canada and severalother partners to deliver an interna-tional conference on water safety in

Ottawa on September 23-25, 2002. The objectiveof the conference, entitled Drinking Water Safety:A Total Quality Management Approach, was toprovide a forum for innovative approaches tosuch issues as ecosystems and health indicators ofwater quality and science and policy guidelinesfor drinking water safety (available at:http://www.neram.ca/Pages/events/events.htm).

• The Institute of Infection and Immunity of theCanadian Institutes of Health Research (CIHR) istaking the lead in establishing the Canadian Coali-tion for Safe Food and Water. Designed to promotea coordinated approach to research funding in thisarea, the Coalition comprises 17 partners, includ-ing federal government departments and agencies,as well as industry and professional bodies. Formore information, E-mail [email protected] [email protected]

• The Toxic Substances Research Initiative (TSRI) isa $40 million program that is comanaged by HealthCanada and Environment Canada. Launched in1998, TSRI’s primary goal is to increase the level

of knowledge about toxic sub-stances and their adverse effects.Research synopses are available at:http://www.hc-sc.gc.ca/tsri

Health ImpactAssessment Health Canada has developed adraft Canadian Handbook onHealth Impact Assessment (HIA)promoting the integration ofHIA into environmental impactassessment (EIA). According tothe Handbook, including the keydeterminants of health in an EIAframework is a cost-effectivemethod for integrating the healtheffects of development projects,programs and policies into thedecision-making process. TheHandbook’s three volumes are targeted at health professionals,environmental assessment practi-tioners and the public (available at:http://www.hc-sc.gc.ca/oeha).

The objective of the

conference, entitled Drinking

Water Safety: A Total Quality

Management Approach, was to

provide a forum for innovative

approaches to such issues as

ecosystems and health indicators

of water quality, and science and

policy guidelines for drinking

water safety.

http://www.durable.gc.ca

http://www.hc-sc.gc.ca/pphb-dgspsp/cehs-esm/wkshop_e.html

http://www.cec.org

http://www.cich.ca/postglobal.htm

http://www.neram.ca/Pages/events/events.htm



http://www.hc-sc.gc.ca/tsri

http://www.hc-sc.gc.ca/oeha


Surveillance: What Is It?

Elizabeth Stratton, Centre for Surveillance Coordination, Populationand Public Health Branch, Health Canada, and Pierre Gosselin,Institut national de santé publique du Québec

Health surveillance is the ongoing, systematic use ofroutinely collected health data to guide public healthaction in a timely fashion.1 Surveillance processesinclude the collection of data, the integration, analysisand interpretation of that data into surveillanceproducts, and the dissemination of the surveillanceproducts to those who need to know. Surveillancehas the following key attributes: it generally involvesthe collection of data in a continuous fashion; it ispopulation based; and it produces information andanalytical products.

An essential component of both these definitions is the notion ofongoing data collection of eithernew or existing data. New data arecollected prospectively for the pur-pose of surveillance, such as withsurveys that focus on specific condi-tions and/or risk behaviours. Forinstance, Health Canada’s CanadianTobacco Use Monitoring Survey(CTUMS) generates new data onsmoking behaviours in regularcycles. New data are also collectedon nationally notifiable diseases(conditions where there are legislatedmandates for reporting — from alocal public health region to theprovince or territory). A numberof communicable diseases are nation-ally notifiable by virtue of theirpotential for serious populationhealth consequences (e.g., measles,

meningococcal infections and sexually-transmitteddiseases). Data are immediately recorded for theseconditions as a part of routine public health diseaseprevention and control practice.

Existing data are retrieved for surveillance from avariety of available sources. Examples include surveysand databases established for purposes other thansurveillance, such as hospital-based administrativedatabases, disease registries and vital statistics. Chronicdisease surveillance relies heavily on existing datasources. The National Diabetes Surveillance System,for example, uses administrative data originating inprovincial/territorial jurisdictions to provide nationallevel surveillance information on diabetes (http://www.diabetes.ca/Section_Professionals/index.asp).

Surveillance does not stop at data collection. Regard-less of whether new or existing sources are used, thedata are then analyzed and transformed into measuresdescribing population health. For example, howmuch disease is in the population (prevalence); howmuch newly-occurring disease is in the population(incidence); and, for some diseases, what proportion ofthe population has received an immunization againstthe disease (coverage rate).

How Are SurveillanceData Used?To be useful in protecting health,surveillance data must be collectedfrequently, with rapid turnaroundfrom raw data to analyzed surveil-lance information indicating unusualor unexplained occurrence or patternof illness. Outbreaks of food-borneillness, for example, need to bedetected and reported quickly sothat control measures can be putinto place. As such outbreaks canoccur within hours of exposure, thesystem must have “real time” detec-tion capabilities. Clusters of eventsin time are not the only indicatorsof health problems — surveillancemust also be capable of detectingrare and unusual events where

Using Canada’s Health Data is a regular column ofthe Health Policy Research Bulletin highlighting

some of the methodologies commonly used in analyz-ing health data. This issue examines the role ofhealth surveillance and discusses how surveillancedata are created, analyzed and used.

Surveillance provides the

means to monitor health

and environmental

outcomes. By monitoring and

comparing disease occurrence

and environmental risk over time,

the impact of exposures and

interventions can be measured.

http://www.diabetes.ca/Section_Professionals/index.asp


Using Canada’s Health Data

there may be only one or two cases. Such casesmay not be close in either space or time, but mayindicate emerging problems. An example is thedetection of extremely rare cancers associated withoccupational exposures.

As previously noted, Health Canada receivesinformation on a series of notifiable diseases.Surveillance for these diseases is conducted at thelocal, regional and provincial/territorial levels, withcase level reporting by condition, location, time,gender and age group only at the national level —individuals cannot be identified.

The urgency and completeness of reporting alsovary depending on the specific condition. Measlesis an example of surveillance where each case mustbe reported to local public health offices as close to“real time” as possible, to permit effective publichealth follow-up on each case. For many other dis-eases, the aim of surveillance is to detect andrespond to trends in disease activity. Influenza, forexample, is consistently under-reported; it is nei-ther practical nor necessary to count each case inorder to obtain a trend for influenza activity in thepopulation.

Click here for references.

Health Canada produces and distributes a variety

of timely surveillance information:

• Disease Surveillance Online (http://www.hc-

sc.gc.ca/pphb-dgspsp/dsol-smed/) provides

information on notifiable diseases and

chronic diseases.

• Infectious Diseases News Brief (http://www.hc-

sc.gc.ca/pphb-dgspsp/bid-bmi/dsd-dsm/nb-

ab/index.html) provides national and

international information (provisional and

confirmed) about communicable disease

incidents and issues.

• Two periodicals that are also available online

are Canada Communicable Disease Report

(http://www.hc-sc.gc.ca/pphb-dgspsp/

publicat/ccdr-rmtc/) and Chronic Disease in

Canada (http://www.hc-sc.gc.ca/pphb-

dgspsp/publicat/cdic-mcc/index.html).

community leaders and the FNIHB. It is not likelythat a single environmental intervention will be ade-quate to rid First Nations communities of shigellosisor developing countries of common diarrheal dis-eases. Sufficient access to water, adequate sewagedisposal systems, reductions in crowded living condi-tions and good personal hygiene are all needed toprevent shigellosis.8 These actions will also help toprevent other infectious diseases with similar socio-economic and environmental determinants, such ashepatitis A and tuberculosis.

It is important to emphasize that not all FirstNations communities experience a disproportionateburden of shigellosis. The epidemiology of the disease

varies dramatically between regions and improvedsanitation has made a significant difference in manycommunities where shigellosis was once common.Preventive programs must be maintained and intensi-fied in communities where shigellosis cases continueto be reported.

The author would like to thank the FNIHB Health Data Technical Working Group,Marion Perrin and Wadieh Yacoub of Alberta Region, and Suzanne Martel ofManitoba Region for providing shigellosis data and outbreak reports. The authorwould also like to acknowledge Saskatchewan Health and Manitoba Health for contributing hospitalization data.


Shigellosis, continued from page 18

@

@

Surveillance Information Products

http://www.hc-sc.gc.ca/pphb-dgspsp/dsol-smed/

http://www.hc-sc.gc.ca/pphb-dgspsp/publicat/ccdr-rmtc/

http://www.hc-sc.gc.ca/pphb-dgspsp/publicat/ccdr-rmtc/

http://www.hc-sc.gc.ca/pphb-dgspsp/publicat/cdic-mcc/index.html

http://www.hc-sc.gc.ca/pphb-dgspsp/bid-bmi/dsd-dsm/nb-ab/index.html


Did You Know? is a regular column of the HealthPolicy Research Bulletin examining aspects of health

research and data that may be subject to misconcep-tions. In this issue, we examine some of the considerationsthat should be taken into account when interpretingwhether or not research results are statistically significant.

The statistics examined in this column are quitedifferent. Because studies can consume considerabletime and resources, researchers must invariably limittheir investigations to samples of the targeted popu-lations. As a result, they are only able to produceestimates of the parameters for which they assume atrue value exists. Any measurement based on a sam-ple will differ from the true value by some amount asa result of random processes or chance. This is themain reason that the results of opinion polls areaccompanied by phrases such as “with a three per-centage point margin of error, 19 times out of 20.”

Therefore, analysts need tools to determine thelikelihood that a conclusion drawn from a sample is

true. This way of generalizing resultsfrom a sample to the entire popula-tion is called inferential statistics.Going back to the example describedabove, this means testing whetherthe lower measured concentrationof the toxin is truly the result of acleaner technology or whether itcould be due to some random factor.Statistical significance is used todemonstrate that an effect did notoccur by pure chance, but is morelikely the result of a particular rela-tionship between variables.

Did You

A Journey into Statistical Significance

Martin Ducharme, Applied Research and Analysis Directorate,Information, Analysis and Connectivity Branch, Health Canada

Consider the following scenario: a team of researchersis studying the discharge of pollutants into water by aparticular industry to determine whether anew technology is cleaner and thus betterfor the environment than the system cur-rently in use. They know that the averageconcentration of a toxin observed in watersurrounding a sample of facilities using thecurrent system is 4.0 mg/l. For a sample offacilities using the new technology, theaverage concentration of the same toxin insurrounding waters is 3.2 mg/l. Should theyconclude that the new system is better for theenvironment? Should they recommend thatthis new technology be used in the facilitiesstill using the standard system?

Answering these questions may not be assimple as it appears. This column explores the con-cept of statistical significance and explains its mostimportant elements. The discussion is intended toprovide readers with a better understanding of someof the statistical terminology they may encounterwhile reading research reports.

Statistical Inference, HypothesisTesting, P Values . . . What Do They Mean?For many people, the notion of statistics is limitedto averages or the numbers used to describe how welltheir favourite teams or players are faring. In fact, theseare part of a field called descriptive statistics — num-bers that deal with the presentation, organization,summarization and, hence, the description of data.

A bit like a criminal suspect

who is considered innocent

until proven guilty, the

observed results of a study

are first considered to be

the same as what might

have occurred as a result of

chance alone.

Hypothesis TestingA bit like a criminal suspect who is consideredinnocent until proven guilty, the observed results ofa study are first considered to be the same as whatmight have occurred as a result of chance alone.This step is called the null hypothesis and it usuallystates that a variable has no effect on another, or thattwo or more variable distributions are no differentfrom one another. The null hypothesis is always themost restrictive and its complement is called thealternative hypothesis. For the pollutant dischargeexample, the null hypothesis is that the toxin levelis no different with the new technology than it iswith the standard system (i.e., it is equal to 4.0 mg/l),while the alternative hypothesis is that the toxinconcentration is different.


Did You Know?

P ValuesThe most common way to test if the null hypothesisholds true is to look at the probability of the observedoutcome under that hypothesis. For the exampleabove, this would mean looking at the probability ofobserving a concentration of 3.2 mg/l when the trueconcentration is really 4.0 mg/l. This probability iswhat is called the P value and it is formally defined asthe probability of observing a test statistic as extremeas or more extreme than the one actually observedwhen the null hypothesis is true. The smaller the Pvalue, the smaller the probability of the observedoutcome under the null hypothesis.

The result is said to be statistically significant whenone has sufficient confidence to rule out the possibilitythat it might have occurred according to the nullhypothesis. The smaller the P value, the more confidentone can be in ruling out the null hypothesis.

Analysts use threshold values to determine the sta-tistical significance of a result. As an example, a resultcould be considered statistically significant when theprobability of observing it under the null hypothesisis smaller than 5 percent (i.e., a P value of less than0.05). The threshold value is called the significance leveland it is often expressed as alpha (α). There are no rulesof thumb in determining the signif-icance level, but it is usually fixed at1 percent, 5 percent or 10 percent.

A P value smaller than the deter-mined significance level means thatthe observed event is sufficientlyunlikely under the null hypothesisthat the latter can be rejected andthe result is said to be statisticallysignificant. Although it is notexplicitly stated, there is a tacitpresumption that the alternativehypothesis provides a more reason-able explanation for that sameevent. A P value higher than thesignificance level means that thenull hypothesis cannot be ruled outwith confidence and the result issaid to be non-statistically signifi-cant (note that one never acceptsthe null hypothesis, only rejects itor fails to reject it). In the currentexample, using a significance level

of 5 percent, one would need to find a P value smallerthan 0.05 to reject the null hypothesis of no differencein the concentration level of the toxin.

Confidence IntervalsFinally, another way to see whether a result is significantor not is to build confidence intervals that wouldinclude 1-alpha percent (100 percent minus the significance level) of the observations. If the valueof the null hypothesis remains outside the interval,the result is said to be statistically significant and thenull hypothesis is rejected; otherwise, the result isnon-statistically significant and one fails to reject thenull hypothesis. For instance, if the 95 percent confi-dence intervals around the 3.2 mg/l measure excludethe value of 4.0 mg/l, the null hypothesis is rejected.

The Limits of P Values and Statistical SignificanceThe reporting of P values to determine the statisticalsignificance of research results has become widespreadbecause of their ease of use and the fact that moststatistical software packages automatically produceP values for each estimated parameter. However,many articles published in educational or statisticaljournals have criticized the use of P values becausethey are often misinterpreted or misused.

One of the most widespread criti-cisms is that P values and statisticalsignificance say nothing about themagnitude or the practical signifi-cance of the results. It is thereforepossible for an effect of little practicalimportance to achieve a high degreeof statistical significance, as it ispossible for an important effect tobe missed because a model lacks the statistical power to detect it at agiven level of significance. However,statistical significance and practicalsignificance should be viewed ascomplementary concepts rather thanas competing ones. Since resourcesare limited, it’s important to knowabout both the likelihood and themagnitude of the impact beforeinvesting in a new initiative.


statistical inference

hypothesis testing

P value

null hypothesis

alternative hypothesis

significance level

practical significance @


Did You Know?

■ Technologies 1, 2 and 3 are statistically significant atthe 5 percent level. The P values for these technologiesare smaller than 0.05 and their 95 percent confidenceintervals exclude the value of zero.

■ With regard to confidence intervals, only the secondprocedure seems to be of any practical significance,while the results for the third and fifth procedures areunclear because their respective confidence intervalsinclude values both above and below 0.5 mg/l.

Suppose there are five new technologies reported to reduce the concentration of a specific toxin. The

results of a study on the effects of these technologies are provided in the table below. Note that the

measure reported is the observed reduction in the concentration compared to the standard system.

Using a 5 percent significance level, try to determine which technologies have a statistically significant

impact on the concentration of the toxin. Which ones have an impact of practical importance if it has

been established that any reduction of less than 0.5 mg/l would not have a significant impact on the

ecosystem and the health of the population? Should any of these technologies be recommended as a

cleaner and healthier substitute to the standard system?

0.1 < 0.0001 (0.08-0.12)

1.8 0.0034 (0.61-2.99)

0.8 0.0010 (0.40-1.20)

0.1 0.5065 (-0.20-0.40)

1.8 0.2330 (-1.18-4.78)

Observed 95%Technologies reduction P value confidence

mg/l interval

1

2

3

4

5

■ Only the second technology has both statisticaland practical significance. This technology istherefore cleaner for the environment and healthierthan the standard system currently in use. However,other studies might be required before this tech-nology becomes the new standard. As an example,one might want to conduct a cost-benefit analysisto see whether the benefits of this new technologyare worth its costs.

Answers to “Test Your Knowledge”

Test Your KnowledgeTest Your Knowledge

Health Canada Research Forum: From Science to PolicyOn November 18-19, 2002, scientists and researchersfrom across Health Canada will come together inOttawa to network and showcase their work andachievements. Sponsored by Health Canada’s HealthResearch Secretariat in the Office of the ChiefScientist, the departmental research conference willinclude discussions organized around three broadthemes (contaminants in food, air and water; chil-dren’s health; and genomics and health), while postersessions will showcase the full range of research andscience initiatives conducted by Health Canada. Thedepartment’s partners will also be invited to attendand learn more about the department’s science andresearch activities. For more information, [email protected]

Funding for Health Policy ResearchA strategic, targeted contribution program ofHealth Canada’s Applied Research and AnalysisDirectorate (ARAD), the Health Policy ResearchProgram (HPRP) generates a range of extramuralpolicy-relevant research designed to meet theneeds of the department. HPRP supports researchand development projects, policy-relevantprojects, workshops, seminars and conferences,and federal/provincial/territorial health researchpartnerships. Watch for upcoming requests forproposals in Health Canada’s priority areas onthe ARAD website (http://www.hc-sc.gc.ca/iacb-dgiac/arad-draa/english/rmdd/funding1.html).

Infrastructure for Communicableand Chronic Disease SurveillanceThe Health Surveillance Working Group (HSWG)has a mandate from the Advisory Committee onHealth Infostructure (ACHI) to identify waysand means of enhancing health surveillance in

Canada. Health Canada’s Centre for SurveillanceCoordination (CSC) provides secretariat support tothe HSWG.

On direction from the HSWG, four breakthroughstudies/papers on health surveillance have beenundertaken and are now available from the CSC:

• Data Definitions and Standards for NationalNotifiable Disease Reporting

• Data Definitions and Standards for NationalImmunization Records Network

• National Surveillance for Chronic Disease in Canada— Charting a Path Forward

• Situational Analysis for Chronic Disease SurveillanceSystems and Networks in Canada

More information is available at: http://www.hc-sc.gc.ca/pphb-dgspsp/csc-ccs/ or by calling (877) 430-9995

Biotechnology Surveillance ProjectHoused in Health Canada’s Centre for SurveillanceCoordination (CSC), the Biotechnology SurveillanceProject (BSP) is developing a national surveillancesystem to monitor potential late health effects onhumans of biotechnology products regulated in

Canada. The BSP’s areas offocus include post-market surveillance of bio-engineeredvaccines and therapeutics, andpost-market surveillance ofgenetically modified foods.

Recent work in the area ofpost-market surveillance ofgenetically modified foodsincluded a Global EnvironmentalScan detailing internationalefforts in this area and identi-fying global experts in the field.A follow-up to this work, aninternational publication onA System Dynamics Approachto Assessing the EconomicImplications of Post-MarketSurveillance of GeneticallyModified Foods, was completed


oteworthy

New and Noteworthy is a regular column of theHealth Policy Research Bulletin highlighting

“up and coming” policy research in the health field.

http

://w

ww

.hc-

sc.g

c.ca

/iac

b-dg

iac/

arad

-dra

a


http://www.hc-sc.gc.ca/pphb-dgspsp/csc-ccs/

http://www.hc-sc.gc.ca/iacb-dgiac/arad-draa/english/rmdd/funding1.html


New and Noteworthy

in Sepember 2002. The First Interna-tional Conference on Post-MarketSurveillance of Genetically ModifiedFoods, hosted by Health Canada onOctober 16-17, 2002, offered a uniqueopportunity to position Canada as aglobal leader in fostering internationalknowledge sharing and collaborationrelated to post-market surveillance ofgenetically modified foods. More infor-mation is available at: http://www.hc-sc.gc.ca/pphb-dgspsp/csc-ccs/

Mobilizing Population HealthThe population health approach isaimed at maintaining and improvingthe health of all Canadians, as well asreducing inequalities among popula-tion groups. A key focus of HealthCanada’s Population and Public HealthBranch (PPHB), the approach directshealth improvement interventionstoward broad, systemic determinantsof health, many of them outside thetraditional health care system.

Case Studies of the Regional Mobili-zation of Population Health — FinalReport presents the findings from sixinitiatives undertaken by PPHB regionaloffices across Canada, including across-case analysis focussing on lessonslearned. The results are intended toinform practice in the field and to helpmarket the population health approachto key decision makers in health policyand planning. The document is avail-able on the Population Health websiteat: http://www.hc-sc.gc.ca/hppb/phdd/case_studies/index.html

Levels of Service in Prenatal Nutrition Programs A new database provides one-stop electronic access toproject-level data and descriptive information aboutOntario projects funded under the Canada PrenatalNutrition Program (CPNP). Compiled by the HealthyChild Development Team, Population and Public

Health Branch, Health Canada(Ontario Region), the databaseallows program consultants toexplore and compare individualprojects and to identify projectsthat deviate from Ontario-widenorms. Two reports integrate theinformation contained in thedatabase with research literatureon “best practices” in prenatalprograms.

In the next phase of the project,upcoming research on best practicesin prenatal programs will be com-bined with information containedin the database to develop corestandards for Ontario CPNPprojects. For more information,contact: [email protected]

Putting the PopulationHealth Approach intoActionThe Social Planning and ResearchCouncil of BC (SPARC) recentlyreceived funding from the Popu-lation Health Fund to analyze andsynthesize lessons learned in puttingthe population health approachinto action in community-basedprojects. The result is a documententitled Creative Spice: Learningfrom Communities about Puttingthe Population Health Approachinto Action. Eleven projects focus-sing on a range of populationsand health issues shared theirexperiences with SPARC. Of

particular interest to policy makers are the discus-sions about the need to educate a broader array ofaudiences and to expand community capacity build-ing and the length of community-based initiatives.The project was sponsored by the BC/Yukon RegionalOffice of the Population and Public Health Branchof Health Canada. The document is available on thePopulation Health website at: http://www.population-health.com

Akey focus of Health

Canada’s Population

and Public Health

Branch, the population health

approach directs health improve-

ment interventions toward broad,

systemic determinants of health,

many of them outside the tradi-

tional health care system.

http://www.hc-sc.gc.ca/pphb-dgspsp/csc-ccs/

mailto: [email protected]

http://www.hc-sc.gc.ca/hppb/phdd/case_studies/index.html

http://www.population-health.com


October 20-24, 2002Santiago, Chile

http://www.paho.org

October 23-25, 2002Ottawa, Ontario

http://www.policyresearch.gc.ca

October 24-26, 2002Halifax, Nova Scotia

http://iareh.usask.ca/meetings/brochure.pdf

November 6-8, 2002Ottawa, Ontario

http://www.statcan.ca/english/conferences/symposium2002

November 7-9, 2002Dallas, Texas

http://www.appam.org

November 8, 2002Vancouver, British Columbia

http://www.chspr.ubc.ca/events.htm

November 18-19, 2002Ottawa, Ontario

E-mail: [email protected]

November 22-23, 2002Vancouver, British Columbia

http://www.cdnhomecare.on.ca/e-conference.htm

November 29-December 1, 2002Toronto, Ontario

http://www.socialjustice.org/conference/index.html

December 1-4, 2002Ottawa, Ontario

http://www.taylorandassociates.ca

The Pan American HealthOrganization’s HealthPromotion Forum in theAmericas

2002 National Policy ResearchConference on Future Trends:Risk

Health Research in Rural andRemote Canada: MeetingChallenges, CreatingOpportunities

XIXth InternationalMethodology Symposium andWorkshops: Modelling SurveyData for Social and EconomicResearch

24th Annual ResearchConference of the Associationfor Public Policy Analysis andManagement (APPAM)

Centre for Health Services andPolicy Research: 15th AnnualHealth Policy Conference

Health Canada ResearchForum: From Science to Policy

12th Annual NationalCanadian Home CareAssociation’s Conference

Social Determinants of HealthAcross the Life-Span: CanadianPerspectives

Third National Conference onTobacco or Health

Promoting health in the Americasthrough follow-up on the MexicoDeclaration, presenting experi-ences and good practices, andnetworking

Extending the understanding ofrisk in the Canadian context

Various themes, such as communityhealth and Aboriginal health

Subject areas of interest includenational statistical accounts andevaluation of social programs

What if . . .? Assessing the publicpolicy and management implica-tions of social science research

Genetic testing: Help, hope or hype

Contaminants in food, air andwater; children’s health; and genomics and health

Maximum impact: Home care’srole in health care reform —ideas, information, implementationand impact

A comprehensive examination ofissues within a social determinantsof health framework

Science and policy in action

What When Theme

Mark Your Calendar

http://www.paho.org

http://www.policyresearch.gc.ca

http://iareh.usask.ca/meetings/brochure.pdf

http://www.statcan.ca/english/conferences/symposium2002

http://www.appam.org

http://www.chspr.ubc.ca/events.htm

mailto: [email protected]

http://www.cdnhomecare.on.ca/e-conference.htm

http://www.socialjustice.org/conference/index.html

http://www.taylorandassociates.ca


References

References References for “Healthy Environments,Healthy People” (p. 5)1. Government of Canada.. Canadian Environmental

Protection Act, Statutes of Canada, 1999. (In forceMarch 31, 2000.)

2. United States Department of Health and HumanServices. Healthy People 2010. Washington, DC:DHSS, November 2000.

3. Lalonde M. A New Perspective of the Health ofCanadians. Ottawa: Government of Canada, April1974.

4. The Environmental Monitor, July 2001. 2001-2.

5. Health Canada, Bureau of Reproductive and ChildHealth. Measuring Up. Ottawa: Minister of PublicWorks and Government Services, 1999: 30-1.

6. Carson R. Silent Spring. Boston: Houghton Mifflin,1962.

7. World Commission on Environment andDevelopment. Our Common Future. Oxford, UK:Oxford University Press, 1987.

8. Health Canada. Sustaining Our Health: HealthCanada’s Sustainable Development Strategy. Ottawa:Health Canada, November 1997.

9. Health Canada. Health Canada Decision-MakingFramework for Identifying, Assessing and ManagingHealth Risks. Unpublished report, August 1, 2000.

10.Tong S, Lu Y. Major Issues in the EnvironmentalHealth Decision Making Process. EnvironmentalHealth 1999; July/August: 33-5.

11.Schwartz E, Corvalán C. Decision Making inEnvironmental Health. In Briggs D, Corvalán C,Nurminen N (Eds.), Linkage Methods forEnvironment and Health Analysis. Geneva,Switzerland: World Health Organization, 1996.

12.Hill AB. The Environment and Disease:Association or Causation? Proceedings of the RoyalSociety of Medicine 1965; 58: 295-300.

References for “Environmental HealthIndicators” (p. 9)1. Bell ML, Davis DL. Reassessment of the Lethal

London Fog of 1952: Novel Indicators of Acute andChronic Consequences of Acute Exposure to AirPollution. Environmental Health Perspectives 2001;109 (Suppl 3): 389-94.

2. Corvalán C, Briggs D, Kjellstrom T. Developmentof environmental health indicators. In Briggs D,Corvalán C, Nurminen M (Eds.), Linkage Methodsfor Environment and Health Analysis. Geneva,Switzerland: World Health Organization, 1996.

3. VanLeeuwen JA, Waltner-Toews D, Abernathy T,Smit B. Evolving models of human health towardan ecosystem context. Ecosystem Health 1999; 5(3):204-19.

4. OECD. OECD Core Set of Indicators forEnvironmental Performance Reviews. EnvironmentMonographs, No. 83. Paris: OECD, 1993.

5. Eyles J, Furgal C. Indicators in EnvironmentalHealth: Identifying and Selecting Common Sets.Paper presented at the Conference onEnvironmental Health Surveillance, October 10-12,2000, Québec City.

6. Burnett RT, Cakmak S, Bartlett S et al. MeasuringProgress in the Management of Ambient Air Quality:The Case for Population Health. Presentation atJoint EC, CIHI, Health Canada Workshop onDeveloping Public Health Indicators, October 25-26, 2001, Ottawa: 2001.

References for “Vulnerable Populations —Critical Pathways” (p. 13)1. Indian and Northern Affairs Canada. Canadian

Arctic Contaminants Assessment Report. ISBN 0-662-25704-9. Ottawa: 1997.

2. United Nations Environment Programme.Stockholm Convention on Persistent OrganicPollutants. Stockholm, Sweden: 2001 (available at:http://www.pops.int).

3. Chance GW, Harmsen E. Children are Different:Environmental Contaminants and Children’sHealth. Canadian Journal of Public Health 1998; 89(Supp 1): S9-S13.

http://www.pops.int


References

4. Charnley G. Protecting the Children: RiskAssessment, Risk Management and Children’sEnvironmental Health. Policy Study 283. ReasonPublic Policy Institute: 2001 (available at:http://www.rppi.org).

5. International Life Sciences Institute. Similaritiesand Differences Between Children and Adults:Implications for Risk Assessment. Washington, DC:ILSI Press, 1992.

6. Canadian Paediatric Society, Dieticians of Canada,Health Canada. Nutrition for Healthy Term Infants.Statement of the Joint Working Group. 1998 (available at: http://www.hc- sc.gc.ca/hppb/ child-hood-youth/cyfh/homepage/nutrition/ index.html).

7. Berkowitz GS, Wolff M, Matte T, Susser E,Landrigan P. Rationale for a Perspective CohortStudy of Environmental Exposure and ChildhoodDevelopment. Environmental Research February2001; 85(2): 59-68.

8. Children’s Environmental Health II: A GlobalForum for Action. Joint Declaration on Children’sEnvironmental Health. Washington, DC: 2001(available at: http://www.cich.ca/postglobal.htm).

9. International Conference on EnvironmentalThreats to the Health of Children. The BangkokStatement: A Pledge to Promote the Protection of Children’s Environmental Health. Bangkok,Thailand: 2002 (available at: http://www.who.int/peh/ceh/Bangkok/bangkokconf.htm).

10.North American Commission for EnvironmentalCooperation. Cooperative Agenda for Children’sHealth and the Environment in North America.2002 (available at: http://www.cec.org).

11.Health and Environment Ministers of theAmericas. Ministerial Communiqué. Ottawa:March 4-5, 2002 (available at: http://www.ec.gc.ca/international/regorgs/docs/english/hema_Comm_e.htm).

12.World Health Organization. Children’s Health and Environment: A Review of the Evidence. InTamberlini G, Von Ehrenstein OS, Bertollini R(Eds). WHO Europe and European EnvironmentAgency: April 2002 (available at: http://www.who.dk/eprise/main/WHO/Progs/CHE/Home).

13.Colborn T, Dumanoski D, Myers J. Our StolenFuture. New York: Dutton, 1996.

14.Steingraber S. Living Downstream. New York:Addison-Wesley, 1997.

15.Foster WG. Endocrine Disruptors and Developmentof the Reproductive System in the Fetus andChildren: Is There Cause for Concern? CanadianJournal of Public Health 1998; 89 (Supp 1): S37-S41.

16.Quang LS, Woolf AD. Children’s UniqueVulnerabilities to Environmental Exposures.Environmental Epidemiology and Toxicology 2000;2: 79-99.

17.Vahter M, Berglund M, Akesson A, Liden C.Metals and Women’s Health. EnvironmentalResearch March 2002; 88(3): 145-55.

18.De Vesa SS, Blatt WJ, Stone BJ, Miller BA, TaroeRE, Fraumeni JF Jr. Recent Cancer Trends in theUnited States. Journal of the National CancerInstitute 1995; 87: 175-82.

19.Skakkebaek NE, Rajpert-De Meyts E, Main KM.Testicular dysgenesis syndrome: an increasinglycommon developmental disorder with environ-mental aspects. Human Reproduction 2001; 16:972-8.

20.Benedict JC, Borgeest C, Silbergeld EK, Flaws JA.Environmental Influences on the FemaleReproductive System: Susceptible Tissues Uniqueto Women. Environmental Epidemiology andToxicology 2000; 2: 91-7.

21.US Environmental Protection Agency. Air QualityCriteria for Carbon Monoxide. Research TrianglePark, NC: National Centre for EnvironmentalAssessment, Office of Research and Development,2000.

22.Viegi G, Scognamiglio A, Baldacci S et al.Epidemiology of chronic obstructive pulmonarydisease (COPD). Respiration 2001; 68(1): 4-19.

References for “Shigellosis and First NationsCommunities” (p. 15)1. World Health Organization. The Evolution of

Diarrhoeal and Acute Respiratory Disease Control atWHO: Achievements 1980-1995 in Research,Development and Implementation. Geneva,Switzerland: World Health Organization, 1999.

2. Health Canada. Notifiable diseases annual summa-ry, 1999. Canada Communicable Disease Report2001; 27S6: 1-147.

http://www.rppi.org

http://www.cich.ca/postglobal.htm

http://who.int/peh/ceh/Bangkok/bangkokconf.htm

http://www.cec.org

http://www.ec.gc.ca/international/regorgs/docs/english/hema_Comm_e.htm

http://www.who.dk/eprise/main/WHO/Progs/CHE/Home

http://www.hc-sc.gc.ca/hppb/childhood-youth/cyfh/homepage/nutrition/index.hml


References

3. Indian and Northern Affairs Canada. Basic Depart-mental Data, 2000. Ottawa: Minister of PublicWorks and Government Services Canada, 2001.

4. Martens P, Bond R, Jebamani L, Burchill C, RoosN, Derksen S et al. The Health and Health Care Useof Registered First Nations People Living in Manitoba:A Population- Based Study. Winnipeg: ManitobaCentre for Health Policy, 2002.

5. Rosenburg T, Kendal O, Blanchard J, Martel S,Wakelin C, Fast M. Shigellosis on Indian reserves inManitoba: Its relationship to crowded housing, lackof running water, and inadequate sewage disposal.American Journal of Public Health 1997; 87: 1547-51.

6. Chin J (Ed.). Control of Communicable Diseases:Manual, 17th Edition. Washington, DC: AmericanPublic Health Association, 2000.

7. Hardy AV, Watt J. Studies of the acute diarrheal dis-eases. XVIII. Epidemiology. Public Health Report1948; 63: 363-78.

8. VanDerslice J, Briscoe J. Environmental interven-tions in developing countries: interactions andtheir implications. American Journal of Epidemiology1995; 141: 135-44.

References for “Sulphur in Gasoline andOther Fuels: The Case for Action (andInaction)” (p. 19)1. Joint Industry/Government Study Sulphur in Gasoline

and Diesel Fuels. Atmospheric Science Expert PanelReport. Available from: Inquiry Centre, 351 St. JosephBoulevard, Hull, Québec K1A 0H3.

2. Joint Industry/Government Study Sulphur in Gasolineand Diesel Fuels. Health and Environmental ImpactExpert Panel Report. Available from: Inquiry Centre,351 St. Joseph Boulevard, Hull, Québec K1A 0H3.

3. Joint Industry/Government Study Sulphur inGasoline and Diesel Fuels. The Costs of ReducingSulphur in Canadian Gasoline and Diesel Phase III,March 1997. Available from: Inquiry Centre, 351 St. Joseph Boulevard, Hull, Québec K1A 0H3.

4. Joint Industry/Government Study Sulphur inGasoline and Diesel Fuels. Competitiveness andViability Impact on Canadian Refining Industry ofReducing Sulphur in Canadian Gasoline and Diesel1997. Available from: Inquiry Centre, 351 St.Joseph Boulevard, Hull, Québec K1A 0H3.

References for “Using Canada’s Health Data”(p. 25)1. Hockin J. Surveillance. In Breslow L (Ed.),

Encyclopedia of Public Health. New York:MacMillan Reference USA, 2002.

Additional Reference:

Health Surveillance Working Group. Canadian HealthInfostructure Health Surveillance Tactical Plan. Finalreport to the Advisory Committee on Health Info-structure. February 2001.

Additional References for “Did You Know?”(p. 27)Dallal GE. The Little Handbook of Statistical Practice(available at: http://www.StatisticalPractice.com).

Davidson R, Mackinnon JG. Estimation and inferencein econometrics. New York: Oxford University Press,1993: 78-81.

Hopkins WG. A new view of statistics. Internet Societyfor Sport Science, 2000 (available at:http://www.sportsci.org/resource/stats/).

Last JM. A Dictionary of Epidemiology, 4th Edition.New York: International Epidemiological Association,Oxford University Press, 2001.

McLean JE, Ernest JM. The Role of StatisticalSignificance Testing in Educational Research. Researchin the Schools 1998; 5(2): 15-22.

Norman GR, Streiner DL. Biostatistics — The BareEssentials, 2nd Edition. Hamilton: B.C. Decker, 2000:42-55.

http://www.sportsci.org/resource/stats/

http://www.StatisticalPractice.com

Health and the Environment: Critical Pathways O · • Hazardous Products Act (HPA):prohibits the...

Documents

Transcript of Health and the Environment: Critical Pathways O · • Hazardous Products Act (HPA):prohibits the...