An adaptable Health Impact Assessment (HIA) framework for assessing health within Environmental Assessment (EA): Canadian context, international application

Abstract

One of the most widely used approaches for assessing environmental effects of large-scale projects is Environmental Assessment (EA). Recently, there has been a focus on including broader health impacts as part of the EA process. One of the tools available to achieve this is Health Impact Assessment (HIA). In order to address the issue of developing a consistent and transparent method for HIA, an assessment framework was developed with the intention of: (1) ensuring that the framework can be used as a stand-alone process and when integrated with EA; (2) applying language to closely align with EA processes; and, (3) devising a system for evaluating overall impact when a multitude of determinants are considered. The Assessment Framework is presented along with a decision matrix to help to determine potential significance of health outcomes. It also provides a process for characterization of effects and identifies whether outcomes are significant. By using an HIA Framework, a well-known yet underutilized tool can effectively address health issues within the EA process, both in a Canadian context and internationally.

Keywords:

• Health Impact Assessment
• HIA
• Environmental Assessment
• health assessment
• assessment framework
• impact assessment

1. Introduction

Assessing the impacts that arise from major infrastructure developments is a complex undertaking that has resulted in a multitude of assessment approaches and methodologies. Arguably, the most established and well-recognized of these approaches is Environmental Assessment (EA), which attempts to characterize and evaluate potential impacts arising from large-scale development projects. Depending on the jurisdiction, EA is often referred to as environmental impact assessment (EIA), Environmental Impact Statement (EIS), or Environmental Impact Report (EIR); however, in general the approaches across North American jurisdictions are similar. In Canada, this process is overseen by the Canadian Environmental Assessment Agency (CEAA) and can be conducted either at the federal or provincial level (CEAA 2016). The EA process is based on the identification of potential environmental impacts and the evaluation of those impacts with the intention of mitigating potentially significant negative effects. With some exceptions, EAs do not explicitly consider the positive aspects of project implementation, other than potential monetary benefits resulting from project revenues and employment opportunities. The decision regarding whether to proceed with a proposal is based on a combination of assessment results, proposed impacts and mitigation measures, and priorities concerning development and economy. Overall, the current EA process is based on ‘the Government’s plan for responsible resource development to modernize the regulatory system and allow for natural resources to be developed in a responsible and timely way for the benefit of all Canadians’ (CEAA 2012).

The EA process is regulated through the Canadian Environmental Assessment Act with the following mandate: ‘the Government of Canada, the Minister, the Agency, federal authorities and responsible authorities, in the administration of this Act, must exercise their powers in a manner that protects the environment and human health and applies the precautionary principle’ (CEAA 2012). Under this legislation an Environmental Assessment (EA) should (CEAA 2016):

•	Identify potential adverse environmental effects;
•	Propose measures to mitigate adverse environmental effects;
•	Predict whether there will be significant adverse environmental effects, after mitigation measures are implemented; and,
•	Include a follow-up program to verify the accuracy of the assessment and the effectiveness of the mitigation measures.

Increasingly, there has been a focus on including overall impact on health as part of the EA process, outside of chemical exposures evaluated in Risk Assessments. One of the tools available to achieve this is Health Impact Assessment (HIA). The potential for HIA as a useful decision-making tool is being realized both nationally and internationally. In 2010, Health Canada published a series of guidance documents on incorporating HIA into the EA process; however, this document is not typically applied in Canadian EA practice.

In order to provide a consistent framework for integration of health into EA, an HIA methodology is proposed. The HIA process consists of several steps including: screening; scoping; assessment; recommendations; reporting; monitoring; and evaluation (Ross et al. 2014). These steps are well aligned with the EA process, suggesting that a common approach is within reach (Figure 1). Despite the consensus around the steps involved in HIA, there is still considerable variation in the specific methods applied. This has resulted in a process that is highly adaptable but lacks any real consistency (McCallum et al. 2015). Considering the broad acceptance of the EA process, it is imperative that the HIA be tailored to work within this existing framework, when necessary.

Figure 1. Steps of the EA and HIA processes (Ross et al. 2014; CEAA 2016).

Guidelines released by the CEAA described the EA process that is used in Canada, but does not discuss evaluation of health impacts outside of chemical exposures considered in Human Health Risk Assessment (CEAA 2016). Health Canada has released a report on significant information for environmental assessments, by pointing to specific areas of expertise that Health Canada can provide when requested to review EA (Health Canada, 2010). These areas include:

•	air quality effects;
•	contamination of country foods;
•	drinking and recreational water quality;
•	radiological effects;
•	Electric and Magnetic Fields effects;
•	Noise effects;
•	Human Health Risk Assessment (HHRA) and risk management;
•	Federal Guidelines (air, water, soil);
•	Toxicology (multimedia); and,
•	First Nation and Inuit health.

Despite this list of potential areas for which to provide health information in an EA, the document is limited to ‘biophysical’ aspects of health and does not include consideration of the social determinants of health. However, there are EAs that consider social and economic aspects of projects (i.e. Socio-Economic Impact Assessments), but they are largely limited to estimating aspects like the number of jobs that will be created or the amount of revenue to be received, and not necessarily how these issues influence human health and well-being:

While SEIA tends to focus on the avoidance of adverse impacts, SEIA also provides a forum for planning how to maximize the beneficial impacts of a proposed development. Beneficial impacts can include: (i) a better standard of living due to increased access to employment, business opportunities, training and education (ii) greater access to and from a community and (iii) increased funding to improve social infrastructure and cultural maintenance programs. (MVEIRB 2007)

Oftentimes, EAs also look at cultural impacts, but again these are typically not evaluated through a health lens. In rare cases where health is considered, it is typically very disjointed with no cohesive framework that ties it together in one comprehensive section. Although it is not common practice, ‘Health Canada suggests that all information relevant to human health be documented in one section of the environmental assessment…’ (HC, 2010). Even if there were a section dedicated to health outcomes, there is currently little to no direction or methodology provided with respect to how to consistently and thoroughly evaluate health effects within the EA process. The effort to develop an HIA framework is timely, given that the Canadian government recently appointed an expert panel to review the EA processes. The Minister of Environment and Climate Change is specifically mandated to review the EA process to introduce new and fair processes that lead to ‘robust oversight and thorough environmental assessments’ and also to ensure that ‘decisions are based on science, facts and evidence and serve the public’s interest’ (Government of Canada 2016).

A report outlining recommendations for how to improve the federal EA process in Canada was issued by the Expert Panel to the Minister, and was made publicly available in April 2017. The report identified five pillars of ‘next-generation’ Impact Assessment (IA) including: (i) environment; (ii) health; (iii) social; (iv) economic; and, (v) cultural considerations. This represents a fundamental shift in the concept of EA (or IA) in Canada, with the identification of a wide range of factors, including human health, that can arise from development projects:

A matter that was heard resoundingly from Canadians was the need for an EA process to move beyond the bio-physical environment to encompass all impacts, both positive and negative, likely to result from a project. The many presenters who raised this suggested that social issues, economic opportunities, health impacts and cultural concerns should be considered. (EPR, 2017)

One of the issues arising from the inclusion of health in EA is the fact that there are a multitude of physical, social, economic, and cultural influences that can directly or indirectly affect health and well-being. In order to decide what to scope into a health assessment, whether under the EA process or in a stand-alone HIA, a comprehensive scoping process should be applied. Once a list of health determinants has been justifiably selected via scoping, the assessment step of the HIA commences. It is during the assessment step that there is the most inconsistency and often a lack of transparency in HIA, with implicit judgment calls affecting the assessment process. It is largely dependent on the practitioner and/or HIA team to decide how to proceed with evaluating determinants and how to draw conclusions based on their findings. There are several examples of HIAs where an assessment is carried out with little to no detail regarding how the determination of impact was completed (McCallum et al. 2015). In many cases, HIAs fail to even identify whether the proposed project or policy will have a negative or positive impact on health overall. This leaves decision-makers and other stakeholders with the responsibility of interpreting results and drawing their own conclusions, which may not be an optimal process, especially if they have no background in environment or public health. It also makes HIA difficult to communicate with stakeholders and members of the public, who look to experts in the environmental health field to identify and disseminate conclusions regarding potential effects to their health and well-being.

In order to address the issue of developing a consistent and transparent method of carrying out HIA, an assessment framework was developed. This framework was created with the intention of: (1) ensuring that the framework is applicable when using HIA as a stand-alone process and when integrated with EA; (2) using language that aligns with EA processes; and, (3) devising a system for evaluating overall impact when a multitude of determinants are considered.

2. Approach

An assessment framework was developed to provide additional transparency and consistency to the HIA process in such a way that is complementary to existing EA practice in Canada and internationally. The relevant literature, key objectives, development process, framework steps, and results are provided and discussed in the sections below.

2.1. Review of relevant literature

The following HIA framework is intended to build upon existing literature on HIA and health in EA, both within Canada and internationally. In order to do this, consideration of the relevant peer-reviewed and gray literature on the subject was reviewed to provide context and a foundation for the development of specific methodologies.

Integration of health into impact assessment is not a new concept; however, the specifics around when, and how to adequately assess health endpoints remains unclear. In Canada, the EA process under the Canadian Environmental Assessment Act (CEAA 2012) allows for Health Canada to ‘review EA documentation on a project-specific basis, examining the baseline assessment and predicted environmental impacts of a project that may affect human health through multiple pathways of exposure’ (Government of Canada 2015). Despite some specified focus areas (e.g. HHRA, noise, electromagnetic fields, etc.), there is little guidance about when and how to assess health in EA, and specifically what constitutes adequate consideration of health impacts resulting from projects.

Interestingly, with the recent commissioning of the Expert Panel to review and provide recommendations for how to improve upon CEAA (2012), health was a major consideration for the panel and the Canadian public and was identified as one of the five pillars of IA in their report released in April. One submission to the panel included a report outlining ‘The Need for Health Impact Assessments to be Integrated into all Federal Environmental Assessment Processes,’ which was a collaboration among several health organizations and health professionals across Canada (Expert Panel Submission 2016). In this submission, the importance of factors beyond just physical and chemical hazards is stressed, pointing to HIA as a necessary and integrative framework for holistic assessment of health. This idea of integrating HIA into EA and IA processes is echoed in the WHO publication ‘Health in Impact Assessments: Opportunities not to be missed’ (Fehr et al. 2014). This report concludes that ‘there is a need to ensure that the health consequences of proposed actions are predicted and understood in a reliable, transparent way, based on the available evidence’ (Fehr et al. 2014). In a response paper from the Government of Canada (2017) they stated that they are considering: ‘broadening the scope of assessment to include environmental, economic, social and health to support more holistic and integrated decision-making in areas of federal jurisdiction.’

Several articles have discussed the effectiveness of integrating health into various forms of impact assessment (McCaig 2005; Fischer et al., 2010; Douglas et al. 2011; WHO 2017). In a review of health considerations in Strategic Environmental Assessment from (SEA) several countries (Austria, the Czech Republic, Germany, the Netherlands, and the United Kingdom), the authors found that while physical aspects of health were adequately covered in all assessments, the social and behavioral aspects were considered to a lesser extent (Fischer et al. 2010). Similarly, inclusion of health within SEA was considered by Fischer (2013) through development of a conceptual model for integrating health determinants (biophysical, social, economic, and behavioral) into SEA practice. However, the paper does note that there are trade-offs with integration of certain factors and that at times distinct approaches, conducted in parallel but with unique frameworks, may be warranted.

Further, in an article by McCaig (2005), the author discusses the various issues around integration of health into EA. Based on the current context, the article describes the way in which HIA is most likely to be integrated into existing EA practice in Canada, a process that is continuing to gain traction over time:

Government policy requiring environmental assessment has predominantly been applied at the project level in Canada and evidence suggests that if health impact assessment is fully integrated into an environmental assessment framework, impacts to human health will be considered at this level. However, HIA has only rarely been applied at a policy level, which indicates that impacts to human health are less likely to be considered at this level. The incorporation of HIA into legislated integrated environmental assessment framework requires compliance by practitioners and proponents. As these become more familiar with HIA they will become more supportive of its inclusion in assessment processes. (McCaig 2005)

Overall, the literature identifies the need for a clear and transparent method of integrating health into the EA process, in a consistent manner that addresses not only the biophysical impacts of projects but also includes consideration of the social determinants of health. Several articles have pointed to HIA as an optimal framework for inclusion of health in EA; this paper expands on this concept with the development of a specific methodological framework that can be integrated within EA itself or can be conducted in parallel, outside of the EA process.

2.2. Identifying objectives

Incorporating HIA into the EA process is a logical progression that could improve both processes by addressing current limitations. In order to accomplish this successfully, the HIA framework needs to complement the typical EA approach. A review of common EA practices in Canada (CEAA 2016; 2015; 2012; Health Canada, 2010) was conducted to identify key elements that would make the HIA framework fit into an EA. Several key factors were identified including:

•	Language consistency: In order for an HIA framework to be able to fit within the EA process, the language used as part of the HIA process should be consistent with, or at least complementary to that used in the EA process.
•	Overall approach: The assessment framework needs to be robust so that it can either be stand-alone or be integrated into an EA process. This means following a similar set of steps and taking a complementary approach to that used in EA, without relying on an EA to have the framework function effectively.
•	Determination of significance: A vital component of the EA process is in determining significance of individual health outcomes, so in order to allow for HIA to complement this process while adding to the rigor of the HIA practice itself, a determination of significance for each potential health outcome must be achieved within the assessment step.
•	Developing conclusions: In addition to determining significance of individual health outcomes, it is vital that the HIA come to a conclusion regarding the overall impact on health. Therefore, an approach that allows the individual outcomes to be summarized and evaluated is key to the EA process, while simultaneously improving transparency of the HIA process.
•	Mitigation considerations: All EAs consider inclusion of mitigation measures. Similarly, HIAs include recommendations to minimize negative and enhance positive impacts. These complementary concepts can be integrated so that the processes can be easily combined, when necessary.

In identifying these key elements, specific objectives were established to guide the framework development. These included: (1) aligning the HIA framework with an EA approach; (2) using consistent language; and, (3) devising a method for coming to an overall conclusion regarding potential impacts on human health. Drawing from these objectives, the framework was developed to allow for an HIA to stand-alone or be an integrated into the EA process.

2.3. Framework development

In order to fulfill the objectives identified for the Assessment Framework, several key considerations were made, with the first being to align the overall assessment approach with existing EA practices. The expectation was that this could be achieved in such a way as to be applied to HIAs being completed alone or alongside an EA. According to CEAA (2015), the recommended approach to determining whether a project is expected to cause significant effects consists of the following three stages: (I) Determining whether environmental effects are ADVERSE; (II) Determining whether environmental effects are SIGNIFICANT; and, (III) Determining whether environmental effects are LIKELY.

Additionally, the guidance around conducting EA uses specific language for making these determinations. For example, determining significance in an EA requires characterizing magnitude, reversibility, frequency, and duration of effect. Thus, the Assessment Framework was created to align with the established EA practice in Canada, and should be transferable/applicable to EA processes in other jurisdictions. This resulted in the following steps for the Assessment Framework:

•	Step 1: Determining if the effect is adverse or beneficial;
•	Step 2: Characterizing effects (based on the magnitude, reversibility, frequency, and duration of impacts) for each health determinant;
•	Step 3: Determination of significance for each health determinant;
•	Step 4: Identifying uncertainties; and,
•	Step 5: Summary and overall conclusions of effect on health.

These five steps constitute elements of the HIA Framework. The first three steps directly align with the EA process for identifying whether effects are adverse, significant and likely. The fourth step considers the level of uncertainty associated with the evaluation of each of the health determinants. Uncertainty is a key concept that is consistently used within EA practice and is especially important in cases where data gaps exist. This is something that has largely been overlooked in HIA practice, but would greatly improve the quality, rigor, and transparency of the process overall. The fifth step includes a method for consolidating the assessment results by plotting significance against uncertainty for each health determinant. The purpose of this step is to summarize the process and come to a determination of the overall effect on human health. The Assessment Framework, along with a detailed description of each step, is provided below.

3. The assessment framework

The following steps outline the Assessment Framework for integrating HIA into EA, and provide instructions on how it should be interpreted and applied.

3.1. STEP 1: identifying adverse or beneficial effects

In the first step of the Assessment Framework, the objective is to identify any effects that are adverse or beneficial. Health outcomes that are either ‘not applicable’ or ‘neutral’ without mitigation or intervention are to be excluded from consideration. One key difference between EA and HIA is that HIAs consider beneficial aspects of the proposal rather than just negative or adverse effects. Therefore, determinants that could potentially have a beneficial effect (directly or indirectly) on health should also be included in the assessment. Identifying whether potential effects might be adverse or beneficial should be made based on available project information and, where information is lacking, practitioners should take a conservative approach and include those determinants whose effects are not fully known. It is essential to acknowledge that different cultural values, judgment calls and paradigms may affect the characterization of effects. Making such values explicit will ensure transparency, sensitivity, and inclusiveness of process.

Q1 (answer for each determinant scoped into the HIA): Is the effect potentially adverse or beneficial to human health?

If YES, proceed to step 2.

If NO, then the effect must be neutral (or not applicable) with no positive or negative influence on human health. The effect must be considered neutral (i.e. no health impact) without the implementation of any mitigation measures.

3.2. STEP 2: characterizing effects

In Step 2, the potential effects (adverse and beneficial) are characterized by the following factors: Magnitude; Frequency; Duration; and, Reversibility. The definitions of each of these factors were developed to promote transparency and consistency in the HIA process (Table 1). If the HIA is stand-alone, then the characterization should be completed assuming that no mitigation measures are in place. Conversely, if the HIA is being completed as part of or alongside an EA, then the evaluation should include consideration of required mitigation measures as outlined in the EA document, where applicable. Ideally, the mitigation measures will have been finalized in the EA prior to conducting the HIA. Alternately, mitigation measures can be identified and revised as a direct result of the HIA process, resulting in an iterative process to ensure protection of health and the environment through mitigation of negatives and enhancement of positives. It should be recognized that such iterative processes, while extending timelines, can increase the effectiveness of the HIA and its recommendations.

Table 1. Definitions for characterizing effects.

Magnitude: what is the severity of the effect on human health?
Low	The effect is minor and does not pose a hazard/benefit to health; health status will not change from baseline
Medium	The effect is detectable and poses a minor to moderate hazard/benefit to health; health status could change from baseline
High	The effect is severe and poses a major hazard/benefit to health; health status will change from baseline
Reversibility (and/or Adaptability): is the effect reversible; how resilient is the community to this type of change; are they able to adapt?
Reversible	The effect is reversible (effect stops once exposure removed) and people will be able to recover or adapt to the change with relative ease, may require support. For positive effects, the improvement is temporary
Irreversible	The effect is not reversible (effect continues once exposure is removed) people are not likely to recover or adapt to the changes, even with additional support. For positive effects, the improvement is permanent
Frequency: how often is the effect expected to occur?
Low	The effect occurs rarely, if ever
Medium	The effect may occur occasionally
High	The effect occurs on a continuous basis
Duration: what would be the duration of the effect, if it were to occur?
Short-Term	A short-term (acute) effect, lasting from days to weeks
Long-Term	A long-term (chronic) effect, lasting from months to years
Geographic Extent: what area will be affected and/or what is the geographic boundary of assessment?
Varies	The geographic extent is a site-specific determination of the amount of area that is anticipated to be impacted. It can include a specific study area (e.g. 5 km radius around the proposed project), a political boundary (e.g. city limits), or be based on topological or ecological features (e.g. watershed)
Timing
Varies	Timing involves temporal fluctuations that can potentially impact health and well-being outcomes (e.g. seasonal changes affecting land/resource use)

Each health determinant should be characterized using the best available information and data. Once characterized, follow the decision matrix (Figure 2) to determine the level of significance and proceed to Step 3.

Figure 2. Decision matrix I: characterizing effects.

Part A: Determine baseline health conditions.

Part B: Assess potential changes (positive/negative) as a result of project activities.

The magnitude, reversibility, frequency, and duration are used to characterize each determinant, which is the first step in determining significance. These factors are all included in the Assessment Framework as components contributing to whether a potential impact is significant or not significant.

‘Geographic extent’ can describe various areas of study including project sites, and locations of different scale (i.e. local, regional, national, and global scale) (CEAA 2015). In cases where impacts need to be assessed on a number of different scales, the full Assessment Framework should be applied to each area. For example, air quality impacts may vary, depending on the scale. Local impacts may be significant, whereas regional or global impacts may be insignificant. Assessing impacts based on differing geographic extents can also help to inform HIA recommendations.

Additionally, timing is considered when assessing impacts in EA and can be applied in certain situations for HIA. In EA, timing typically refers to considerations that could alter an environmental or ecological outcome, depending on temporal aspects such as breeding season, migration times, etc. This can be applied to HIA in situations where timing can potentially influence health and well-being outcomes. For example, ‘seasonal aspects of land and resource use and whether timing is related to Aboriginal spiritual and cultural considerations’ may be relevant issues (CEAA 2015). Other aspects of timing that can be considered in HIA include timing of agricultural practices, seasonality of recreational activities, school attendance for children, and others.

In some cases, the HIA practitioner may disagree with the effect characterization dictated by the decision matrix. It is acknowledged that although the matrix has been designed to apply to a wide range of scenarios, it may not adequately fit all potential outcomes or reflect all stakeholders’ values or perceptions of priorities. Therefore, in some cases it may be appropriate to come to an alternate conclusion on effects. In these cases, it is possible to deviate from the decision matrix; however, a clear and comprehensive justification must be provided.

3.3. STEP 3: determining significance

Although the effect characterization is made based on the above factors (i.e. magnitude, reversibility, frequency, duration), the determination of significance must also include the likelihood of occurrence. For example, a potentially significant impact that is certain to occur would be significant; however, a potentially significant impact that is extremely rare may not be deemed to be significant in terms of overall anticipated health effects.

For each determinant, one should characterize the likelihood of the impact, based on the definitions provided below and the available project information regarding probability of occurrence (Table 2). Then one can proceed to the final step of the decision-matrix to determine significance (Figure 3). Depending on the results of both the effect characterization (Step 2) and likelihood (Step 3), the final determination of significance for each health determinant will fall along a spectrum. This was the chosen approach to account for scenarios where an effect might fall in between two options, leaving room for HIA practitioners to select and justify the most appropriate outcome.

Table 2. Definitions for characterizing likelihood.

Likelihood: What is the probability of the impact occurring?
Unlikely	The impact is anticipated to occur rarely, if ever. This classification is appropriate for those situations where impacts are not zero but they are limited to very rare occurrences, catastrophic events, or highly unlikely system failures
Possible	The impact may occur, but the probability is less than 50%
Probable	The impact will likely occur, the probability is greater than 50%

Figure 3. Decision matrix II: determining significance.

In making a final determination of significance, it was considered important to acknowledge that the significance of an impact is not always black and white, given the importance of judgment calls in the process, which is why a spectrum of significance was developed. Instead this framework, while attempting to make a determination of significance in coordination with EA practices, identifies the level of significance using subheadings that indicate positive (+) or negative (−) effects. This approach was undertaken in order to better inform the decision-makers and to be more transparent in making a determination of significance. It is important to note that the term ‘significance’ as it is applied here is different than mathematical significance for which there can be no sub-classification and where something is definitively calculated to be significant or not, without further interpretation.

To illustrate how varying ‘levels of significance’ have been applied in this Assessment Framework, consider the example of an injury resulting from a car accident. Would one conclude that the injury is significant in terms of a health impact or not? Before answering this question, one might want to know more about the extent of the injury. For example, if the person’s injuries consist of a few bruises, that may not be considered significant from a health perspective. Conversely, if the accident resulted in broken bones, bleeding, organ damage, paralysis, and/or death, it would definitely be considered significant in terms of the negative effects on health. If one considered all of the potential effects in a range between those two extremes, one would be inclined to classify them as significant or not, depending on the extent of the effect. Presumably, one would classify any negative health outcome as significant, even though there is a threshold beyond which one might argue that a negative health effect is present but minor enough to be considered insignificant. These sorts of considerations inform the basic principle behind how the Assessment Framework deals with significance, because when it comes to health, there are many potential outcomes with highly varied levels of severity and different interpretive frameworks.

This same approach to determining the level of significance can be applied for positive health outcomes as well. For example, if one considers the positive impact of exercise on health and well-being, it is likely that one would agree that any level of physical activity is beneficial, but at what point does the health benefit become significant? As with many of the factors that can influence health, there are varying degrees of positive impact. For example, if a person exercises for one hour per week, it may be better than no exercise at all, but it is certainly not as beneficial as one hour a day. Additionally, the type and intensity of the physical activity may contribute to the scale of the outcome (minor to major health benefits). Again, these concepts are embedded into the way the HIA Assessment Framework was developed to account for a wide array of potential positive and negative impacts and outcomes, with respect to health and well-being.

Consequently, aligning HIA with EA should allow for a determination of significance, while acknowledging that not all positive and/or negative impacts are created equal. The way this determination of significance should be interpreted is to first look at whether the effect is significant or not, and then look at the subcategory. The subcategories range from minor (−/+) to moderate (− −/+ +) to major (− − −/+ ++). This additional information around the so-called ‘level of significance’ will help decision-makers interpret the assessment more accurately and provide additional transparency that would be lacking without this differentiation. The significance level can also help to inform the HIA recommendations, mitigation measures, and any follow-up monitoring that may be required.

In some cases, the HIA may be carried out in parallel to the EA, but not be integrated into the process. In these cases, determining a level of significance (−/+ to − − −/+ ++) may be at the discretion of the practitioner; however, in cases where the HIA must be integrated into a prescribed EA process, a final determination of ‘significant’ or ‘not significant’ may be required. This final determination should be justified through the collective evidence and data relied upon in the HIA, while taking into consideration the level of uncertainty in the assessment.

3.4. STEP 4: identifying uncertainties

One vital aspect of the EA process that has been included in the HIA Assessment Framework is the notion of identifying uncertainties. This concept is widely used in EA practice when evaluating potential environmental impacts of proposed projects:

The determinations must take into account uncertainties. All project EAs involve some level of uncertainty, and observed results will often deviate, to some degree, from predictions made in the EA. Uncertainty could be related to a number of factors such as: project design and components, baseline environmental conditions, VC response, effectiveness of mitigation, overall scope of effects, and natural and human causes of accidental events. (CEAA 2015)

In identifying the level of uncertainty (or the level of confidence) associated with the evaluation of effect and subsequent determination of significance, the practice becomes more transparent, acknowledging the limitations of the process. Applying this concept to HIA allows for the same level of transparency, which is vital to a process that is highly dependent on the professional judgment of its practitioners.

In this step of the Assessment Framework, the level of uncertainty (or the level of confidence) associated with the evaluation of health impacts (positive and negative) and significance is determined. Using the definitions provided (Table 3), a level of uncertainty is assigned to each of the health determinants. This evaluation should be based on the quantity and quality of data and information (i.e. quality of sources, completeness of information, model validity, etc.) that has been used to make these determinations.

Table 3. Definitions for characterizing uncertainty.

Uncertainty: what is the level of uncertainty associated with the evaluation of health impacts and determination of significance for each of the health determinants?
Very High	A very high level of uncertainty (i.e. very low level of confidence) is associated with the determination of significance as a result of major data or information gaps and/or a very low level of confidence associated with the assessment methods and approach
High	A high level of uncertainty (i.e. low level of confidence) is associated with the determination of significance as a result of moderate data or information gaps and/or a low level of confidence associated with the assessment methods and approach
Medium	A medium level of uncertainty (i.e. moderate level of confidence) is associated with the determination of significance as a result of minor data or information gaps and/or a moderate confidence associated with the assessment methods and approach
Low	A low level of uncertainty (i.e. high level of confidence) is associated with the determination of significance as a result of negligible data or information gaps and/or a high level of confidence associated with the assessment methods and approach
Very Low	A very low level of uncertainty (i.e. very high level of confidence) is associated with the determination of significance as a result of no data or information gaps and/or a very high level of confidence associated with the assessment methods and approach

Wherever possible, multiple practitioners involved in the HIA team should conduct the uncertainty analysis for each determinant; where discrepancies are found; a discussion leading either to (I) consensus or (II) an ‘average’ of the two classifications should be completed to reduce subjectivity.

3.5. STEP 5: overall determination of effect

Once the health determinants included in the HIA have been evaluated, and the significance and level of uncertainty have been established, the results should be plotted to assist in making an overall determination of effect (Figure 4). By plotting the significance (x-axis) against the level of uncertainty (y-axis) for each determinant, a summary of health effects can be created. This summary can then be used to weigh the positive and negative outcomes, including consideration of uncertainties, helping to reach a final conclusion surrounding overall health impacts. Although there is some subjectivity in the weighting of results, this method of summarizing and analyzing collective results provides a more transparent method of developing conclusions, especially when a multitude of determinants are involved. It can also facilitate a determination of overall significance, if this is a requirement of the EA process.

Figure 4. Revised HIA framework (Step 5): overall determination.

In some cases, multiple determinants of health will have the same ‘score’ in terms of significance and level of uncertainty. When this situation occurs, the plotted points can be scaled to reflect the number of determinants with that score. For example, all single points will be the same size, but when two (or three) determinants have the same score, the plot will be twice (or three times) the size. This will provide a more accurate visual representation of the actual results, and may assist with weighing the outcomes to make a final conclusion regarding health effects.

4. Discussion and conclusions

The Assessment Framework was developed to adhere to the language and approach used commonly in EA practices in Canada. Although such practices differ slightly among different jurisdictions around the world, the framework is intended to allow for adjustment and adaptation by HIA practitioners who wish to use this complementary approach. Additionally, although the framework was developed with the objective of allowing it to be incorporated into the EA process, it was also intended to allow for use in stand-alone HIAs. This approach was taken in order to facilitate the transition between conducting HIAs as optional stand-alone reports, to having them integrated as part of larger, regulated impact assessment processes, which appears to be a promising new direction for the practice of HIA. However, the primary objective of developing this Assessment Framework was to further promote the application of tools to HIA practice with the goal of improving consistency in assessment approach and transparency of the methods applied.

The application of the framework, including the evidence relied upon for the HIA as well as the use of mitigation measures, will vary depending on several factors, including whether an EA is being completed concurrently. In this case, there will be a vast amount of data and information that can be drawn from the EA and used to facilitate evaluation of health determinants in the HIA. This can be a major benefit to both processes, since EAs are often criticized as not supporting sufficient consideration of human health impacts. Additionally, the HIA benefits from the wealth of evidence and availability of pre-established mitigation measures that are required under regulation if the project goes forward. Since any measures identified in the EA would be legally enforceable, the HIA can be based on a scenario where those measures are already in place. In situations where the HIA is stand-alone, the assessment would have to be conducted assuming no mitigation. However, similar to instances of EA, if there are cases where potential impacts are found, mitigation can be developed to address any issues (i.e. reduce negatives and enhance positives) and then the assessment can re-visit these scenarios to determine the effectiveness of the proposed measures. In this way, combining both the HIA and EA can ensure that people and the environment are protected through their collective recommendations and mitigation measures.

Another consideration in developing this framework was to ensure that it could be applied to all levels of assessment. Typically, HIAs follow one of three types: rapid, intermediate, and comprehensive (or detailed). These assessment types fall along a spectrum that includes the amount of effort and time required for the HIA, the amount of evidence and data that will be used, stakeholder engagement considerations, and the level of detail within the evaluation itself. This framework, while attempting to streamline the HIA process and promote transparency, can be applied to all levels of assessment. The one major difference with a rapid as opposed to a comprehensive HIA, would likely be the amount of evidence upon which the effect characterization is based. Therefore, the level of uncertainty with respect to the assessment and determination of significance may be higher in these cases. This is especially true in cases where the HIA is limited to identifying and discussing potential health issues at a high level but does not provide for the available time or budget to conduct a thorough evaluation or to rely on any site-specific modeled or measured data.

There may be instances where the framework does not perform in a manner that is in accordance with the practitioner’s professional opinion. In such cases, should the practitioner choose to deviate from the framework’s decision matrix, a sound justification should be provided. This is especially true in cases where the framework indicates that there is a potentially significant effect on health and well-being, but the practitioner’s assessment finds the health outcome to be not significant. Since framework development was an iterative process, aimed at finding a balance between adaptability for a wide range of situations, and providing a consistent and transparent method of assessment, it is possible that the decision matrix will not fit all possible scenarios. The Assessment Framework is currently being tested on several real-world projects both within an EA framework and as a stand-alone HIA process. In these test scenarios, the usability of the framework, applicability of the definitions, and consistency of results are being evaluated. In cases where any of these factors are found wanting, the framework can be further revised to improve upon the existing approach.

Overall, the Assessment Framework was developed in a way that achieved the goals of (1) aligning with an EA approach; (2) using consistent language; and, (3) devising a method for coming to an overall conclusion regarding potential impacts on human health. The one aspect of the Framework that is still largely conceptual is the final step that combines the significance and level of uncertainty for each health determinant, in order to provide a summary of results and inform an overall conclusion on the health implications of a proposed project. This aspect of the framework would benefit from being tested on a real-world HIA scenario involving a large number of health determinants. As discussed, it is acknowledged that the interpretation of the results in making a final determination of potential impacts on health is somewhat subjective and often relies on judgment calls, especially if conducted by a single individual. However, plotting the results in this manner will enhance transparency within the HIA process and improve the practice overall, regardless of whether there are different ways to interpret the outcomes. By continuing to promote transparency within HIA by developing tools with this objective in mind, it has the potential to become a widely used framework for addressing health impacts, both within the EA process and beyond.

Funding

This work was supported by the Natural Science and Engineering Research Council (NSERC) [grant number 452703].

Disclosure statement

Dr. McCallum works for an Environmental Health consulting firm that does work in the field of environmental and health sciences.