Background

The direct and indirect impacts of climate change on human health continue to be observed globally, and these wide-ranging impacts are projected to continue to increase and intensify this century [1, 2]. The direct climate change effects on health include rising temperatures, which increase heat-related mortality and morbidity [3,4,5], and increased frequency and intensity of storms, resulting in increased injury, death, and psychological stressors [2, 6,7,8]. Indirect climate change impacts on health occur via altered environmental conditions, such as climate change impacts on water quality and quantity, which increase waterborne disease [9,10,11,12,13]; shifting ecosystems, which increase the risk of foodborne disease [14,15,16], exacerbate food and nutritional security [17, 18], and change the range and distribution of vectors that cause vectorborne disease [19, 20]; and place-based connections and identities, leading to psycho-social stressors and potential increases in negative mental health outcomes and suicide [6, 8]. These wide-ranging impacts are not uniformly or equitably distributed: children, the elderly, those with pre-existing health conditions, those experiencing lower socio-economic conditions, women, and those with close connections to and reliance upon the local environment (e.g. Indigenous Peoples, farmers, fishers) often experience higher burdens of climate-health impacts [1, 2, 21]. Indeed, climate change impacts on human health not only are dependent on exposure to climatic and environmental changes, but also depend on climate change sensitivity and adaptive capacity—both of which are underpinned by the social determinants of health [1, 22, 23].

The inherent complexity, great magnitude, and widespread, inequitable, and intersectional distribution of climate change impacts on health present an urgent and grand challenge for the health sector this century [2, 24, 25]. Climate-health research and evidence is critical for informing effective, equitable, and timely adaptation responses and strategies. For instance, research continues to inform local to international climate change and health vulnerability and adaptation assessments [26]. However, to create evidence-based climate-health adaptation strategies, health practitioners, researchers, and policy makers must sift and sort through vast and often unmanageable amounts of information. Indeed, the global climate-health evidence base has seen exponential growth in recent years, with tens of thousands of articles published globally this century [22, 25, 27, 28]. Even when resources are available to parse through the evidence base, the available research evidence may not be locally pertinent to decision-makers, may provide poor quality of evidence, may exclude factors important to decision-makers, may overlook temporal and geographical scales over which decision-makers have impact, and/or may not produce information in a timely manner [29,30,31,32,33,34,35,36,37].

Literature reviews that utilize systematic methods present a tool to efficiently and effectively integrate climate-health information and provide data to support evidence-based decision-making. Furthermore, literature reviews that use systematic methods are replicable and transparent, reduce bias, and are ultimately intended to improve reliability and accuracy of conclusions. As such, systematic approaches to identify, explore, evaluate, and synthesize literature separates insignificant, less rigorous, or redundant literature from the critical and noteworthy studies that are worthy of exploration and consideration [38]. As such, a systematic approach to synthesizing the climate-health literature provides invaluable information and adds value to the climate-health evidence base from which decision-makers can draw from. Therefore, we aim to systematically and transparently create a database of articles published in academic journals that examine climate-health in North America. As such, we outline our protocol that will be used to systematically identify and characterize literature at the climate-health nexus in North America.

Methods

This protocol was designed in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) Guidelines [39, 40] and presented in accordance with the PRISMA-P checklist.

Research questions

Research on climate change and human health encompasses a diverse range of health outcomes, climate change exposures, populations, and study designs. Given the breadth and depth of information needed by health practitioners and decision-makers, a variety of research questions will be examined (Table 1).

Table 1 Examples of research questions guiding the climate change and health in North America literature review
Full size table

Search strategy

The search strategy, including the search string development and selection of databases, was developed in consultation with a research librarian and members of the research team (SLH, AC, and MDA). The search string contains terms related to climate change [41, 42], human health outcomes [1, 25, 43, 44], and study location (Table 2). Given the interdisciplinary nature of the climate-health nexus and to ensure that our search is comprehensive, the search string will be used to search five academic databases:

  1. 1

    CINAHL® will be searched to capture unique literature not found in other databases on common disease and injury conditions, as well as other health topics;

  2. 2

    Web of Science™ will be searched to capture a wide range of multi-disciplinary literature;

  3. 3

    Scopus® will be searched to capture literature related to medicine, technology, science, and social sciences;

  4. 4

    Embase® via Ovid will be searched to capture a vast range of biomedical sciences journals; and

  5. 5

    MEDLINE® via Ovid will be searched to capture literature on biomedical and health sciences.

Table 2 Search strategy for CINAHL® to identify published articles reporting on climate change impacts on human health in North America published after 2013 (see Appendix 1 for search strategies for Web of Science™, Scopus®, Embase® via Ovid, and MEDLINE® via Ovid)
Full size table

No language restrictions will be placed on the search. Date restrictions will be applied to capture literature published on or after 01 January 2013, in order to capture literature published after the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (which assessed literature accepted for publication prior to 31 August 2013). An initial test search was conducted on June 10, 2019, and updated on February 14, 2020; however, the search will be updated to include literature published within the most recent full calendar year prior to publication.

To explore the sensitivity of our search and capture any missed articles, (1) a snowball search will be conducted on the reference lists of all the literature that meet the inclusion criteria and (2) a hand search of three relevant disciplinary journals will be conducted:

  1. 1

    Environmental Health Perspectives, an open access peer-reviewed journal that is a leading disciplinary journal within environmental health sciences;

  2. 2

    The Lancet, a peer-reviewed journal that is the leading disciplinary journal within public health sciences; and

  3. 3

    Climatic Change, a peer-reviewed journal covering cross-disciplinary literature that is a leading disciplinary journal for climate change research.

Citations will be downloaded from the databases and uploaded into Mendeley™ reference management software to facilitate reference management, article retrieval, and removal of duplicate citations. Then, de-duplicated citations will be uploaded into DistillerSR® to facilitate screening.

Article selection

Inclusion and exclusion criteria

To be included, articles must evaluate or examine the intersection of climate change and human health in North America (Fig. 1). Health is defined to include physical, mental, emotional, and social health and wellness [1, 25, 43, 44] (Fig. 1). This broad definition will be used to examine the nuanced and complex direct and indirect impacts of climate change on human health. To examine the depth and breadth of climate change impacts on health, climate change contexts are defined to include seasonality, weather parameters, extreme weather events, climate, climate change, climate variability, and climate hazards [41, 42] (Fig. 1). However, articles that discuss climate in terms of indoor work environments, non-climate hazards due to geologic events (e.g. earthquakes), and non-anthropogenic climate change (e.g. due to volcanic eruptions) will be excluded. This broad definition of climate change contexts will be used in order to examine the wide range and complexity of climate change impacts on human health. To be included, articles need to explicitly link health outcomes to climate change in the goal statement, methods section, and/or results section of the article. Therefore, articles that discuss both human health and climate change—but do not link the two together—will be excluded. The climate-health research has to take place in North America to be included. North America is defined to include Canada, the USA, and Mexico in order to be consistent with the IPCC geographical classifications; that is, in the Fifth Assessment Report, the IPCC began confining North America to include Canada, Mexico, and the USA [45] (Fig. 1). Articles published in any language will be eligible for inclusion. Articles need to be published online on or after 01 January 2013 to be included. No restrictions will be placed on population type (i.e. all human studies will be eligible for inclusion).

Fig. 1
figure 1

Inclusion and exclusion criteria to review climate change and health literature in North America

Full size image

Level 1 screening

The title and abstract of each citation will be examined for relevance. A stacked questionnaire will be used to screen the titles and abstracts; that is, when a criterion is not met, the subsequent criteria will not be assessed. When all inclusion criteria are met and/or it is unclear whether or not an inclusion criterion is met (e.g. “unsure”), the article will proceed to Level 2 screening. If the article meets any exclusion criteria, it will not proceed to Level 2 screening. Level 1 screening will be completed by two independent reviewers, who will meet to resolve any conflicts via discussion. The level of agreement between reviewers will be evaluated by dividing the total number of conflicts by the total number of articles screened for Level 1.

Level 2 screening

The full text of all potentially relevant articles will be screened for relevance. A stacked questionnaire will also be used to screen the full texts. In Level 2 screening, only articles that meet all the inclusion criteria will be included in the review (i.e. “unsure” will not be an option). Level 2 screening will be completed by two independent reviewers, who will meet to resolve any conflicts via discussion. The level of agreement between reviewers will be evaluated by dividing the total number of conflicts by the total number of articles screened for Level 2 (Fig. 2).

Fig. 2
figure 2

Flow chart of screening questions for the literature review on climate change and health in North America

Full size image

Data extraction and analysis

A data extraction form will be created in DistillerSR® (Appendix 2) and will be tested by three data extractors on a sample of articles to allow for calibration on the extraction process (i.e. 5% of articles if greater than 50 articles, 10% of articles if less than or equal to 50 articles). After completing the calibration process, the form will be adapted based on feedback from the extractors to improve usability and accuracy. The data extractors will then use the data extraction form to complete data extraction. Reviewers will meet regularly to discuss and resolve any further issues in data extraction, in order to ensure the data extraction process remains consistent across reviewers.

Data will be extracted from original research papers (i.e. articles containing data collection and analysis) and review articles that reported a systematic methodology. This data extraction will focus on study characteristics, including the country that the data were collected in, focus of the study (i.e. climate change impact, adaptation, and/or mitigation), weather variables, climatic hazards, health outcomes, social characteristics, and future projections. The categories within each study characteristic will not be mutually exclusive, allowing more than one response/category to be selected under each study characteristic. For the country of study, Canada, the USA, and/or Mexico will be selected if the article describes data collection in each country respectively. Non-North American regions will be selected if the article not only collects data external to North America, but also includes data collection within Canada, the USA, and/or Mexico. For the study focus, data will be extracted on whether the article focuses on climate change impacts, adaptation, and/or mitigation within the goals, methods, and/or results sections of the article. Temperature, precipitation, and/or UV radiation will be selected for weather variables if the article utilizes these data in the goal, methods, and/or results sections. Data will be extracted on the following climatic hazards if the article addresses them in the goal, methods, and/or results sections: heat events (e.g. extreme heat, heat waves), cold events (e.g. extreme cold, winter storms), air quality (e.g. pollution, parts per million (PPM) data, greenhouse gas emissions), droughts, flooding, wildfires, hurricanes, wildlife changes (including changes in disease vectors such as ticks or mosquitos), vegetation changes (including changes in pollen), freshwater (including drinking water), ocean conditions (including sea level rise and ocean acidity/salinity/temperature changes), ice extent/stability/duration (including sea ice and freshwater ice), coastal erosion, permafrost changes, and/or environmental hazards (e.g. exposure to sewage, reduced crop productivity).

Data will be extracted on the following health outcomes if the article focuses on them within the goal, methods, and/or results sections: heat-related morbidity and/or mortality, respiratory outcomes (including asthma, chronic obstructive pulmonary disease), cardiovascular outcomes (including heart attacks or stroke), urinary outcomes (e.g. urinary tract infections, renal failure), dermatologic concerns, mental health and wellness (e.g. suicide, emotional health), fetal health/birth outcomes and/or maternal health, cold exposure, allergies, nutrition (including nutrient deficiency), waterborne disease, foodborne disease, vectorborne disease, injuries (including accidents), and general morbidity and/or mortality. Data on the following social characteristics will also be extracted from the articles if they are included in the goal, methods, and/or results sections of the article: access to healthcare, sex and/or gender, age, income, livelihood (including data on employment, occupation), ethnicity, culture, Indigenous Peoples, rural/remote communities (“rural”, “remote”, or similar terminology must be explicitly mentioned), urban communities (“urban”, “city”, “metropolitan”, or similar terminology must be explicitly used), coastal communities (use of “coastal”, or similar terms must be explicitly mentioned), residence location (zipcode/postal code, neighbourhood, etc.), level of education, and housing (e.g. data on size, age, number of windows, air conditioning). Finally, data will be collected on future projections, including projections that employ qualitative and/or quantitative methods that are included in the goal, methods, and/or results sections of the article.

Descriptive statistics and regression modelling will be used to examine publication trends. Data will be visualized through the use of maps, graphs, and other visualization techniques as appropriate. To enable replicability and transparency, a PRISMA flowchart will be created to illustrate the article selection process and reasons for exclusion. Additionally, qualitative thematic analyses will be conducted. These analyses will utilize constant-comparative approaches to identify patterns across articles through the identification, development, and refinement of codes and themes. Article excerpts will be grouped under thematic categories in order to explore connections in article characteristics, methodologies, and findings.

Quality appraisal of studies included in the systematic scoping review will be performed using a framework based on the Mixed Methods Appraisal Tool (MMAT) [46] and the Confidence in the Evidence from Reviews of Qualitative Research (CERQual) tool [47]. This will enable appraisal of evidence in reviews that contain qualitative, quantitative, and mixed methods studies, as well as appraisal of methodological limitations in included qualitative studies. These tools may be adapted to include additional questions as required in order to fit the scope and objectives of the review. A minimum of two reviewers will independently appraise the included articles and discuss judgements as needed. The findings will be made available as supplementary material for the review.

Discussion

Climate-health literature reviews using systematic methods will be increasingly critical in the health sector, given the depth and breadth of the growing body of climate change and health literature, as well as the urgent need for evidence to inform climate-health adaptation and mitigation strategies. To support and encourage the systematic and transparent identification and synthesis of climate-health information, this protocol describes our approach to systematically and transparently create a database of articles published in academic journals that examine climate-health in North America.