Health effects of flooding in Canada: A 2015 review and description of gaps in research

Abstract

Worldwide, floods are the most common natural disaster and produce a broad array of health impacts. In Canada, it is difficult to quantify all of the health impacts associated with flooding. This data gap is salient because of the increasing risks floods are posing to society as a result of climate change. This paper reviews the epidemiological evidence of flood-related health effects, and the Canadian susceptibility to these effects. The health impacts from flooding range from mortality, drowning and other injuries to hypothermia, mental health impacts, deterioration of elderly/patients who required emergency transportation, homelessness and transmission of contagious diseases and others. The Canadian population includes a range of socio-economic levels, demographics and pre-existing illness. Additionally, many Canadians live and work in flood plains or on shorelines where floods are a hazard. There is, therefore, variable susceptibility to flood events in the population. Susceptibility to these health impacts can be buffered by the population’s adaptive capacity and external support. To ensure the safety and health of the Canadian population, government and non-profit organizations can decrease the burden of flooding by increasing adaptive capacity, emergency response and a focus on preventative measures. Vulnerable populations who require extra support to avoid negative health impacts from flooding, such as First Nations populations, need further attention. Recommendations are offered for addressing the health-related effects of flooding.

À l’échelle mondiale, les inondations constituent la catastrophe naturelle la plus répandue qui soit et entraînent un vaste éventail d’impacts sur la santé. Au Canada, il est difficile de quantifier tous les impacts sur la santé associés aux inondations. Il serait important de combler cette lacune statistique en raison des risques accrus que posent les inondations pour la société, du fait du changement climatique. La présente communication examine les preuves épidémiologiques des effets sur la santé liés aux inondations et la susceptibilité canadienne à ces effets. Les impacts sur la santé sont variés et englobent la mortalité, la noyade et d’autres blessures, l’hypothermie, les répercussions sur la santé mentale, la détérioration de l’état des personnes âgées ou des patients transportés d’urgence, l’itinérance et la transmission de maladies contagieuses et autres. La population du Canada est caractérisée par une gamme de niveaux socioéconomiques, de facteurs démographiques et de maladies préexistantes. De plus, bon nombre de Canadiens vivent et travaillent dans des plaines inondables ou dans des zones littorales où les inondations représentent un danger. Par conséquent, au sein de la population, la susceptibilité aux inondations peut varier. Toutefois, la susceptibilité à ces impacts sur la santé peut être tempérée grâce au soutien externe et à la capacité d’adaptation de la population. Afin d’assurer la sécurité et la santé de la population du Canada, le gouvernement et les organismes sans but lucratif peuvent alléger le fardeau des inondations en augmentant la capacité d’adaptation ainsi que les interventions en cas d’urgence et en mettant l’accent sur les mesures de prévention. Une plus grande attention doit être accordée aux populations vulnérables ayant besoin d’un soutien supplémentaire pour éviter les répercussions néfastes des inondations sur la santé, notamment les Premières Nations. Des recommandations sont formulées afin de contrer les effets des inondations sur la santé.

Introduction

Floods are defined by the National Weather Service (2009) as “the condition that occurs when water overflows the natural or artificial confines of a stream, river, or other body of water, or accumulates by drainage over low-lying areas.” Floods can be devastating events and are currently the most common worldwide natural disaster (Guha-Sapir et al. 2015). As certain types of floods become more common, flood-related damages are predicted to increase due to global warming (Westra et al. 2014; Guha-Sapir et al. 2015). In Canada, storm activity is likely to cause contamination of drinking water supplies as surface contaminants and storm-sewage overflows are flushed into water sources. These run-off events are predicted to increase in both number and severity with climate change (Charron et al. 2004; Ford et al. 2006).

The International Disaster Database has recorded 22 flooding events in Canada since 2000 (Guha-Sapir et al. 2015). These 22 events, occurring between 2000 and 2015, resulted in 18 deaths and 126,474 people impacted (Table 1). Canada has a low average death rate per flood in comparison to other countries, approximately one person per flood (Guha-Sapir et al. 2015). Among the estimated 126,474 people affected by flooding, however, the associated morbidity has not been quantified. Injury, fecal-oral diseases, vector-borne diseases, rodent-borne diseases, impacts of mold, various mental health issues, maternal health issues, hypothermia, deterioration of physically unstable persons, and homelessness have all been reported as general health impacts from flooding in the international literature (Centers for Disease Control and Prevention [CDC] 2000; Lowe et al. 2013; Dancause et al. 2015). It is important to realize that various researchers may define a flooding event differently and therefore report a different number of flooding events for an area. Therefore, even though the International Disaster Database has 22 recorded flooding events in Canada between 2000 and 2015, other papers such as Buttle et al. (2016, this issue) have indicated over 60 significant flooding events in Canada during this time period. This shows that although the International Disaster Database may record the most damaging floods, it may not be a true representation of the number of overall flooding events. Some of the sources used by Buttle et al. (2016, this issue) listed multiple floods for the same flooding event; this also leads to confusion about the definition of a flood event.

Table 1. List of Canadian floods from Guha-Sapir et al. (2015); EM-DAT Database (2000–2015).

Start date	End date	Location	Deaths	People affected	Damage (US $)	Disaster number	Type of flood
9 June 2002	12 June 2002	North Western Ontario				2002-0404	Riverine flood
15 February 2003	21 February 2003	Bagder (Newfoundland)		1100		2003-0087	Riverine flood
31 March 2003	3 April 2003	Colchester, Nova Scotia	2	170	10,000	2003-0231	Riverine flood
4 August 2003	5 August 2003	Central Quebec		400	10,000	2003-0417	Flash flood
17 October 2003	26 October 2003	British Columbia	2	800	15,000	2003-0511	Riverine flood
28 March 2004	11 April 2004	Manitoba		1000		2004-0146	Riverine flood
13 July 2004	16 July 2004	Ontario, Alberta and Quebec			200,000	2004-0346	Flash flood
27 September 2005	29 September 2005	Newfoundland		600		2005-0591	Riverine flood
7 June 2005	1 July 2005	Alberta and Saskatchewan	4	5000	357,000	2005-0728	Riverine flood
18 April 2006	24 April 2006	Newfoundland		200		2006-0203	Riverine flood
12 April 2006	20 April 2006	Saskatchewan		1000		2006-0206	Riverine flood
22 April 2006	31 May 2006	Fort Albany (Ontario)		2000		2006-0224	Riverine flood
26 May 2006	2 June 2006	Aklavik (Northwest Territories)		300		2006-0261	Riverine flood
20 May 2006	28 May 2006	British Columbia	1	400		2006-0269	Riverine flood
6 June 2007	11 June 2007	British Columbia		500		2007-0165	Riverine flood
10 April 2008	24 April 2008	Ontario		400		2008-0170	Riverine flood
26 April 2008	1 May 2008	James Bay area (North Ontario)		2000		2008-0264	Riverine flood
1 April 2009	9 April 2009	New Brunswick		200		2009-0189	Riverine flood
4 April 2011	31 May 2011	Manitoba and Saskatchewan	5	2000	800,000	2011-0131	Riverine flood
25 March 2012	30 March 2012	Western New Brunswick		1500	34,000	2012-0068	Riverine flood
20 June 2013	27 June 2013	Calgary (Southern Alberta)	4	100,000	5,700,000	2013-0190	Riverine flood
4 August 2014	4 August 2014	Burlington (Ontario)		6904	129,000	2014-0319	Flash flood

Assessing how the Canadian population has been negatively affected by flood-related health outcomes requires an understanding of the vulnerability of the Canadian population (Hajat et al. 2005). The vulnerability of a population to a natural disaster is defined as “[a population’s] capacity to anticipate, cope with, resist and recover from the impact of a natural hazard” (Hajat et al. 2005, 18). A population that can readily adapt to climate stressors is less vulnerable to the impacts of climate change, and is sometimes referred to as resilient. Vulnerability to natural disasters is dependent on socio-economic considerations, demographics and concurrent illness (Dutta et al. 2005). Within Canada, demographics, socio-economics and pre-existing illnesses are highly varied (Statistics Canada 2011); this creates a challenge when measuring a population’s vulnerability to flooding as these variables interact with each other in complex ways.

The impacts and vulnerabilities related specifically to First Nations reserves are not addressed separately in this review. Little information is available in the peer-reviewed literature about this important area of concern in Canada, leading to significant data gaps. Thompson et al. (2014), in their account of the Lake St. Martin First Nation 2011 flood event and subsequent dislocation, describe some of the marginalization and other factors that lead to particular vulnerability in many First Nations communities, particularly in flood-prone Manitoba. Clearly, more attention and action are needed for this population. Floods also can create new vulnerable populations due to the increased chronic health effects of flooding, such as injury, homelessness and post-traumatic stress disorder (Ramin and Svoboda 2009), and the degradation of living spaces. Robust health care and social-support system response is required in order to respond to the range of public health challenges that arise from both the immediate and chronic health impacts post flood (Runkle et al. 2012).

An often-challenging post-flood concern is the need for increased “surge capacity” required after large disasters (Runkle et al. 2012). Surge capacity refers to the provision of medical care to a large number of people during the acute response (Runkle et al. 2012). Health facilities themselves are sometimes rendered useless by floods when they are most needed, and more resilience to threats from flooding and other disasters is needed in the health system infrastructure (Schneider and Rousseau 2005). This paper aims to review the epidemiological evidence of flood-related health effects, and to briefly describe the Canadian population’s susceptibility to these effects.

Methods

The objective of this paper was to identify published manuscripts of studies examining health-related flooding outcomes in Canada and to provide a review of this literature. An electronic search was completed for the years 2005 to 2015 using PubMed Central, with the following search terms included: (1) Canada AND (2) flood AND (3) public health AND (4) risk AND (5) disease. An additional broad electronic search for newspaper articles was completed for the years 2002 to 2013, using Canadian Newsstand with the following search terms: (1) Canada AND (2) flood AND (3) health. The years used for searching Canadian Newsstand correspond with recent major Canadian flood incidents. Additional studies were identified via a manual review of the reference lists of the previously identified studies. Newer related studies accessed via Web of Science, which included non-disease-related morbidity including injury, were also reviewed. Studies were deemed eligible for inclusion if they met the following criteria: the study was (1) peer-reviewed and (2) evaluated or reviewed health effects or susceptibility relating to flooding. Health effects of flooding were drawn from worldwide literature with particular attention to Canada. The focus on Canada and the Canadian health care system was also applied when exploring a population’s susceptibility to flooding.

Results

A large range of flood-related health effects were identified in the literature. The main health effects noted in the last decade are listed and briefly summarized below (see Table 2). The results also include information on Canada’s susceptibility to negative flood-related health outcomes.

Table 2. Summary of primary health impacts from flooding.

Health hazard	Examples of health impact	References
Mortality	Drowning, other asphyxiation	Lowe et al. 2013
	Non-specific, all-cause	Bennet 1970; Waite et al. 2014
Injury	Fractures, contusion, lacerations	Du et al. 2010; Doocy et al. 2013
Contamination of drinking water	Acute gastrointestinal illness	Vollard et al. 2004
Vector borne disease	West Nile virus infection	Ahern et al. 2005
Zoonotic disease	Leptopirosis, hook worm	Kovats and Akhtar. 2008
Mental illness	Post-traumatic stress disorder	Azuma et al. 2014
Prenatal stress	Low-birth weight	Brand et al. 2006
	Childhood obesity	Dancause et al. 2015
Hypothermia	Impaired neurologic function, coma	Lowe et al. 2013
Respiratory tract irritation	Asthma exacerbation	Du et al. 2010; Hulin et al. 2012

Health effects

Worldwide, flooding causes significant morbidity and mortality. There have been 539,811 deaths related to flooding from 1980 to 2009 (Doocy 2013). The study of the 1968 Bristol, UK, flood found the first year post-flood had a 50% increase in all-cause mortality among those whose homes had been flooded (Bennet 1970). Mortality was most pronounced in the first three months in those aged 45 to 64 years. Causes of immediate mortality from flooding include drowning, heart attack (especially if over 60 years of age), physical trauma, electrocution, carbon monoxide poisoning (Hajat et al. 2005; Waite et al. 2014), motor vehicle accidents, and asphyxiation from mud and other debris slides (Lowe et al. 2013). Between 1959 and 2005 in the US, over 60% of all flood-related fatalities were attributed to motor vehicles (Ashley and Ashley 2008).

Injuries

One major source of morbidity from flood events is injuries. An estimated 0.2 to 2% of flood survivors require urgent medical care (Noji 1995). These include common injuries from debris or falls such as fractures, contusions and lacerations, as well as electrical injuries, burns and – perhaps most important in developed countries – injuries from motor vehicle accidents during evacuation and rescue (Du et al. 2010).

Infectious diseases

Infectious diseases related to flooding can be characterized into three main groups: fecal transmission, vector transmission and zoonotic transmission. Post-flood increases in water-borne illness are common. Fecal matter and other contaminants from human and animal waste are often mobilized during floods and enter drinking water supplies (Batterman et al. 2009). These illnesses include poliomyelitis (Ahern et al. 2005), shigellosis (Schwartz et al. 2006), acute enteric infections (Sur et al. 2000; Vollaard et al. 2004), hepatitis A and E (Ohl and Tapsell 2000), norovirus, rotavirus and campylobacter infections (Cann et al. 2013). Children, the elderly and others with compromised immune function are particularly vulnerable to acute gastrointestinal infections (AGI); a study by Lowe (2013) found that if a child’s home or yard was flooded they had a 1.9 times greater risk of developing AGI. Lowe (2013) also found that people over the age of 50 had a higher risk of AGI post-flood. Boil water advisories (BWA) are often put into effect after a flood to decrease the transmission of infectious diseases for the public water supply (Lowe 2013). Additionally, water reserves may be more heavily treated (usually chlorination) than normally, as was done after the BC Pemberton flood of October 2003 (Postmedia Network Inc. 2003).

Some vector-borne diseases such as West Nile (Ahern et al. 2005), Valtice fever (Cann et al. 2013) and lymphatic filariasis (Bockarie et al. 2009) have been seen to increase post flood. Zoonotic transmitted diseases also increase during flooding because of compromised drinking water systems and altered patterns of contact between people and animals (Ahern et al. 2005). Zoonotic diseases with observed increases post-flood include: leptospirosis, hookworm infection (Kovats and Akhtar 2008), rabbit-borne tularemia, (presenting in hemorrhagic manifestations and shock) and hantavirus pulmonary syndrome (Ahern et al. 2005; Cann et al. 2013).

Psycho-social impacts

A large portion of post-flood morbidity relates to the psycho-social stress from experiencing a flood event and subsequent loss, disruption and reduced sense of security (Carroll et al. 2009) Additionally, the realization of the dangers of living in a floodplain has a negative impact on mental health (Tapsell and Tunstall 2008). Many studies have noted the increase in post-traumatic stress disorder (PTSD), anxiety and depression following flooding events (Hajat et al. 2005; Azuma et al. 2014). Tunstall et al. (2006) found possession of adequate flood insurance, characteristics of flood event-type of property damages, family/community support and socio-demographics were important variables affecting the mental health status of a post-flood population. Tunstall et al. (2006) studied residents from 30 different locations in England and Wales and found that two thirds of the study population demonstrated mental health problems post flood. Heo et al. (2008) noted a depression prevalence of 17%, and a PTSD prevalence of 22%, in Korean post-flood agricultural communities. Stanke et al. (2012) highlight the impact of secondary stressors in prolonging mental health issues post flood, while Hajat et al. (2005) and Azuma et al. (2014) noted even six months post flood, PTSD rates were significantly higher in their study population than in the general population. Stanke et al. (2012) concluded that further studies are needed to properly understand mental health issues following successive flooding, and the mental health issues of children and first respondents post flood. There has been no direct evidence linking suicide rates to flooding (Ahern et al. 2005); however, there is a link between flooding and increased alcohol abuse and dependence, particularly among those first responders, residents and health care workers with previous psychiatric problems (Weisler et al. 2006; Rowe and Liddle 2008). Detrimental cognitive and developmental disorders can also arise from the environmental transport of toxic chemicals, including pesticides and heavy metals, into waterways (Boxall et al. 2008).

Birth outcomes

Recent evidence has emerged linking flooding to negative maternal health and birth outcomes. Prenatal exposures can affect growth of a fetus and thereby have a lifelong influence on chronic disease risk. Some of this risk may be mediated through maternal stress. It has been shown that antenatal stress/anxiety has an effect on the fetus lasting at least until middle childhood (O’Conner et al. 2002; Brand et al. 2006). Low birth weight is associated with adulthood risk of chronic diseases such as high blood pressure, insulin resistance, glucose intolerance and hyperlipidaemia (Reynolds et al. 2001). Dancause et al. (2015) found prenatal stress from flooding increased child obesity when the child reached 2.5 to 4 years of age. The level of increase in obesity was associated with greater maternal hardship during the flood event.

Mold and other dampness-related microbial growth

Mold and microbial growth can be a significant respiratory hazard post flood, particularly with exacerbation of asthma and chronic obstructive pulmonary disease (COPD; Durrie 2005; C. Brown et al. 2006; Du et al. 2010; Hulin et al. 2012). Solomon et al. (2005) studied the dispersion of bioaerosols post hurricane/flood in the fall of 2005 in New Orleans and found dangerous levels of mold (up to 645,000 spore/m³) inside the homes examined. Additionally, Emerson et al. (2015) state that fungal abundances were measured to be three times higher in flooded homes than non-flooded homes. Emerson et al. (2015) also note that fungal communities may continue to be problematic even after humidity has returned to normal levels and remediation has removed visible flood damage. Therefore, all personnel involved in post-flood cleanup should wear the appropriate personal protective equipment and use the appropriate exposure mitigation techniques in order to prevent or reduce adverse respiratory impacts.

Other post-flood health outcomes that have been noted in this review include hypothermia, animal bites, deterioration of moved/displaced patients, homelessness (Lowe 2013), respiratory/nasal symptoms (Azuma et al. 2014), increase in seizures (Kasteleijn-Nolst Trenité 2012), leukemia, lymphoma, spontaneous abortion (Janerich et al. 1981), melioidosis (A. C. Cheng et al. 2003), and various effects of chemical contamination (possibly polycyclic aromatic compounds; Balluz et al. 2001; Hall et al. 2012).

Health risks associated with post-flood cleanup activities

Mortality and morbidity from flooding may go unreported when they occur in the post-recovery/delayed phase of the event (Waite et al. 2014). Factors affecting the severity of a flood include the speed of the flood waters, the area covered and the depth of the flood, and the limitations of advanced warning systems. Bennet (1970) found a statistically significant increase in mortality from malignant diseases occurred after flooding, with an attribution of between nine and 21 deaths in the year following the flood, although this was more difficult to interpret than direct effects. Considering the latency of most malignancies, this finding remains unexplained. Finally, one of the most common causes of flood-related deaths in North America over the last decade is evacuation-related motor vehicle accidents (Du et al. 2010); however, Canada has a low relative flood-related mortality rate, averaging one death per flood between 2005 and 2015 (Guha-Sapir et al. 2015). This low Canadian flood-related mortality rate is probably due to both the characteristics of the floods and good immediate disaster response.

Post-flood exposures during clean-up are also a source of morbidity. For example, Canadians previously have used hypochlorate bleach to clean their homes post flood; this is problematic because bleach does not effectively remove mold, and it promotes additional hazards from the irritant effects of chlorine exposure on respiratory epithelium (Trotter 2013). Canadian public health agencies have issued warnings to curtail the post-flood practice of bleach use for clean-up (Durrie 2005). Another post-flood home danger is asbestos. People may come into contact with asbestos from insulation in older homes and buildings that have become waterlogged and can deteriorate and release fibers when dried. Although asbestos is less of a risk when wet, it is still dangerous and homeowners must be wary when repairing flood-damaged homes that have not been tested for asbestos. Post flood, housing and other buildings are often hazardous. In addition to risk of injury from structural damage, floods may expose people to previously hidden hazards such as asbestos from the insulation or other building materials, heavy metals and dispersed household or industrial chemicals. Canadian public health agencies have guidelines for the public outlining the need to have homes professionally inspected post flood before starting renovations (Trotter 2013).

Canada’s susceptibility to negative flood-related health outcomes

Drawing conclusions from international comparisons of flood morbidity data is challenging since both post-flood morbidity and mortality can be hard to identify and attribute, and reporting is sporadic. However, as noted by Tapsell et al. (2002), some influences on morbidity are known. These are factors associated with flood itself (depth, area and speed), and the adaptive capacity of the affected population. The adaptive capacity is enhanced by the support the population receives before, during and post flood, and their capacity to anticipate, cope with and prevent negative flood-related outcomes. Adaptive capacity can be improved through education and increased risk perception (Shaw et al.,2015). Increased risk perception can manifest itself as the creation of family emergency plans, the development of community disaster engagement, the development of a high-functioning and prepared emergency response health care system, and the development of long-term government relief and reconstruction efforts. Social resilience balances inherent social vulnerability and is a product of education and risk perception. In a study by Akerlof et al. (2010) it was shown that although Canadians are aware of health impacts from climate change and flooding, most people in their survey did not acknowledge a clear association between the two without being prompted.

Discussion

This paper reviews human health impacts of floods in a Canadian context. Several large public databases including PubMed, Web of Science and Canadian Newsstand were searched with appropriate terms and subjected to inclusion criteria as outlined in the Methods. Such an approach may miss important literature (Greenhalgh and Peacock 2005). However, this paper does provide useful information for preparing Canada for future flood risk. For example, we describe how sub-populations in Canada who are particularly vulnerable to the ill effects of flooding include non-English/French speaking persons, lower socio-economic-level individuals, the very young or very old, single mothers, pregnant women, ethnic minorities, under/non-insured persons, and individuals with pre-flood morbidities (Phifer et al. 1988; Tapsell at al. 2002; Dutta et al. 2005; Ostry et al. 2010; Lowe et al. 2013; Dancause et al. 2015). Phifer et al. (1988) found individuals aged 55–64 years with low occupational status, such as unskilled or semiskilled laborers, were at a significantly higher risk for post-flood psychotic symptoms. Tapsell et al. (2002) created a social flood vulnerability index (SFVI) from post-flood focus groups, which identified several influential variables: age over 75 years, lone parents, pre-existing health problems and financial deprivation. Determining the vulnerability of a population can be a major challenge, as Canada has a diverse population in the domains of these identified variables (Statistics Canada 2011). A nationwide vulnerability assessment is an unmet, yet high, priority that would identify vulnerable communities prior to flood events and enable prioritization of the limited resources available in order to better mitigate the effects of flooding (Tapsell et al. 2002; Ostry et al. 2010).

Resilience in Canada’s health care system

Resilience to future challenges to public health is highly varied across Canada. Austin et al. (2015) studied how various provinces have been adapting to climate change. They point out that municipalities are adapting autonomously for the most part, although Quebec as a whole has initiated many more health-related climate change-flood adaptations than other Canadian provinces have. Austin et al. (2015) found adaptations at the federal level focus mainly on capacity building, and gathering information on infectious diseases and general health risks. Two examples of adaptation at a federal level are the Canadian Assessment of Vulnerabilities and Adaptive Capacity, and the pending Public Health and Waterborne Illness Research Tool (Food Secure Canada [FSC] 2015; Public Health Agency of Canada [PHAC] 2015). The Canadian Assessment of Vulnerabilities and Adaptive Capacity was designed to inform Canadians of the impacts of climate change and how vulnerable they may be to these impacts. The Public Health and Waterborne Illness Research Tool under development is an instrument used to evaluate the association between extreme weather events and public health impacts to control, prevent and respond to water–borne diseases in Canada (PHAC 2015).

Although more advanced than that of some countries, Canada’s health care system does not appear particularly resilient to natural disasters. Seeman et al. (2008) note that more than one quarter of Ontario hospitals examined did not have a process to identify, manage or minimize risks to the hospital’s business-level sustainability in the event of a disaster. Seeman et al. (2008) found a positive association between hospital boards that had approved risk preparation strategies and hospital boards having adapted six key governance practices. Therefore, even if it is made mandatory for hospital boards to have emergency management plans in case of a disaster, it still may be beneficial to also mandate particular governance practices which would support such plans.

All of the six specific governance practices examined by Seeman et al. (2008) are designed to increase accountability between hospital board members and their stakeholders. Hajat et al. (2005) showed that a comprehensive, risk-based emergency management program is necessary, and that it has great potential to reduce the adverse health effects from flooding. Keim (2008) states that floods and other climate change-related disasters create significant public health vulnerabilities that can easily exceed the capacity of a community. Prioritizing health service needs, securing health systems support and coordinating provision of care in an emergency are contingent upon the knowledge of baseline health disparities of the affected population (Runkle et al. 2010). Susceptibility to the health impacts of disasters can be reduced largely by preparation and contingency planning for emergencies (Keim 2008). For example, state service buildings, such as police stations, fire stations and hospitals, should be inspected for suitability before being used as gathering places post flood. Demonstration that the essential services associated with these public facilities are resilient and can be up and running quickly in an emergency is critical.

Axelrod et al. (1994) reviewed the American Midwest flood disaster of the Mississippi, Ohio and Missouri drainages of 1993, and found that between the increase in the number of patients and the decreased ability of the health care system due to damage and personnel loss, the health care system often did not have the “surge capacity” to service those in need. Axelrod et al. (1994) noted that ancillary health personnel in addition to primary-care physicians are needed in post-flood management. Post-flood response often requires an increase in essential primary care health services ranging from chronic illness care to maternal care, child health care, family planning, dental care, mental health care and immunization. Runkle et al. (2012) recognized a post-flood recipe for disaster, identifying a “perfect storm” of diminished health care system resilience coupled with the increase in demand for services. Runkle et al. (2012) also note the possibility of post-flood inverse care – “the people most in need of medical care are often the least likely to receive the care they need in the weeks and months following a disaster” (24). These authors conclude that in order for recovery efforts to be successful, response personnel must account for the health care needs of vulnerable populations at the following stages: (1) baseline, (2) phase I: acute response, and (3) phase II: long-term recovery. This team predicted six possible post-flood health system demands and pointed out that an effective emergency management plan should cover any of these possible demands, the most significant being an increase in acute care needs followed by an equal increase in post-primary care needs and a reduced long-term surge capacity (Runkle et al. 2012).

Vulnerability to flood impacts is also affected by the existing policy framework. The majority of flood management policy, such as land use and accompanying mapping, is administered at the provincial level or the local level in Canada. From 1975 to 1996, the federal Flood Damage Reduction Program (FDRP) was run in collaboration with the provincial and territorial governments. The objective of the FDRP was to reduce flood damage and prevent loss of life by discouraging development in areas vulnerable to flooding. This was done primarily through mapping the flood zones around hundreds of Canadian communities; however, essentially none of these maps account for climate change projections and must be updated with these additional stressors in mind (Shrubsole 2014).

There is no national flood impacts mitigation program in Canada. The new National Disaster Mitigation Program (NDMP) was launched in April 2015. The NDMP provides CAD $200 million over five years with the goals of: (1) focusing investments on significant, recurring flood risk and costs; and (2) advancing work to facilitate private residential insurance for overland flooding (Public Safety Canada 2015). The federal government also provides funds for structural flood controls and post-disaster relief.

Post-disaster relief is administered through Disaster Financial Assistance Arrangements (DFAA), which is a federal program that provides disaster recovery assistance to provinces and territories that suffer disaster damages exceeding CAD $1 per capita. The program distributes funding on a cost-sharing basis with provinces and territories, and as recovery costs increase, the federal government will absorb a proportionately larger share of the costs. Through the DFAA, assistance is first paid out to a province or territory, which is subsequently responsible for the design, development and delivery of financial assistance to those affected.

Provinces have taken different approaches to flood management, and therefore have different levels of preparedness. In BC, floodplain mapping is the role of local governments, which has led to considerable variety in approaches to flood protection. For example, many communities in BC do not have up-to-date floodplain maps or bylaws in place to regulate construction in flood-prone areas. In Alberta, Alberta Environment supervises the production of floodplain maps for communities as part of its Flood Hazard Identification Program. In addition, DFAA is administered in such a way that individuals who receive assistance once need to take precautionary measures before they are eligible for future assistance. Overall, Canadian flood policy is quite decentralized, with a variety of actions taken across provinces. This patchwork of public disaster assistance and partial private insurance has led to large data gaps on flood risk exposure that have contributed to the unavailability of overland flood insurance for some residential buildings in Canada (Sandink et al. 2015).

Adaptation in disaster response is especially needed at the community level. Therefore, local public health agencies also have great potential to build human resilience to climate-related disasters in conjunction with primary care services. Ebi et al. (2006) designed a framework to assess current climate change public health strategies. An example of a recent public health strategy for climate change that should be considered and tested locally is the creation of vulnerability maps with elevation, flood threat and land-use change data, to create usable flood risk information (Owrangi et al. 2014) for planning purposes. It is important that, while managing health system surge capacity and primary care services, public health strategies also be assessed to optimize them for post-flood recovery and future extreme conditions. Risk perception can be changed, and cities and communities can be redesigned to avoid disaster. Health care systems can build more capacity to respond to disasters. These actions would decrease morbidity from flooding.

Finally, strategies that address climate change itself need ongoing research, assessment and – where appropriate – implementation. Without this planning, the hazards of climate change will have an increasing impact on society, with the sustainability of the health system further stressed. A lack of preparation and action in this area is likely to create new, as yet unperceived, health effects for floods and other natural disasters.

Recommendations

Canada has been quite successful in reducing the number of deaths from flooding; however, the impact of indirect and post-flood effects has not been adequately addressed by all stakeholders (government bodies, hospital boards, communities, urban planners and academics). Based on the literature reviewed, the following evidence-based actions are recommended:

•	Flood education and capacity building should be increased for both the general public and the health care system. Education should be provided on the potential health impacts of natural disasters, as well as on the need for an emergency-ready health care system. Public education should focus on personal risk perception in order to produce a grassroots movement, with a push to improve the capacity of the health care system to cope with emergencies (Fitzpatrick-Lewis et al. 2010). Increasing public education on personal risk perception relating to natural disasters could also lower the vulnerability of the population and increase their capacity to anticipate and prepare for flooding and other disaster events. Cardwell and Elliott (2013) suggest that reframing climate change into a public health issue instead of an environmental issue will encourage the general public to push for more action regarding climate-related health threats. The resource guide developed by Paterson et al. (2014) is a useful resource for individuals, and could be included in a flood education plan.
•	Hospital boards should have mandatory emergency disaster plans. The six key governance practices researched by Seeman et al. (2008) should be implemented in all hospital boards, as they show association with a higher level, long-term planning. In order to create an emergency disaster plan, resources such as Paterson et al.’s (2014) toolkit can be used to help health care facility officials identify gaps in flooding preparedness, direct the allocation of resources and inform strategic planning to increase resiliency to climate change. The climate change health care toolkit includes an assessment checklist and a facilitator’s guide, along with the resource guide previously mentioned. The assessment checklist is designed to increase awareness of current resources (or lack thereof), and the facilitator’s guide is designed to facilitate discussion.
•	Current public health flood and environmental disaster strategies should be rigorously assessed (Ebi et al. 2006). Ebi et al. (2006) devised an approach to assessing public health strategies. H. Brown and Spickett (2014) also devised a health impact assessment to assess and compare climate change-related health effects. These should be reviewed for relevancy to applicable flood-related public health strategies.
•	Vulnerability is enhanced by social inequity. In order to protect the most vulnerable populations, flood-related public health strategies must address problems of social inequity that are likely to increase in the future. These populations include First Nations, young children, the elderly and others with compromised immune function or other social and environmental disadvantages.
•	Further research is needed to quantify the risks, costs and benefits of new flood- and environmental disaster-related public health strategies, such as those concerning water-borne diseases, vector-borne diseases, rodent-borne diseases, and land-use planning (J. J. Cheng and Berry 2013). To enhance the current evidence base, there is a need for attention to design, evaluation, reporting and adaptation of interventions, standardized health impact projection reporting, and knowledge translation (Hall et al. 2012). This critical research is needed for healthcare financial planners, practitioners and policy makers.
•	Precautionary principle. Until there is better evidence on how to balance economic growth and the health-related impacts of climate change, the precautionary principle should be deployed. The precautionary principle states that when an activity raises threats to human health or the environment, precautionary measures should be taken until cause and effect relationships are fully established scientifically (Foster et al. 2000). The precautionary principle can be applied to land-use change decisions, as well as other areas where risks are currently unknown.

Conclusion

While Canada has been quite successful in reducing the number of direct deaths from flooding, it seems to have paid less attention to the indirect and post-event health-related effects. This review of recent studies examining Canada’s susceptibility and the various negative health effects related to flooding suggests there is a need to increase the impetus to better assess Canada’s vulnerability to post-flood health impacts.

Adaptive capacity can be improved through social resilience, social disaster engagement, and a highly functional and prepared health care and government relief system. Tapsell et al. (2002) agreed that prior to experiencing a major flood, some communities have little awareness or understanding of the health risks of flooding at the community level. The majority of citizens in Tapsell’s study had not been in a previous flood, and were not prepared to cope with the flood or its aftermath. If people were more educated about flooding, associated health risks, adaptive capacity and the impacts of global warming, and were more aware of their own personal health risks (both physical and mental health) from disaster-related outcomes, they might be able to better avoid disastrous repercussions. In addition to education, the health care system must create emergency plans and assess those plans prior to a flooding event. If not, the health care system will be quickly overwhelmed and unable to cope with post-flood needs and flood-related morbidity. All primary and public health prevention strategies should be assessed as they are developed to ensure they are coordinated with non-government, academia and private post-flood programs. Vulnerable populations such as First Nations, the very young and the very old need particular attention.

Research is needed to develop new public health and climate change response strategies. As Hajat et al. (2005) described, there is currently inadequate evidence on the effectiveness of public health interventions to reduce flood-related morbidity. Until there is a more robust evidence base, the precautionary principle should be applied when designing flood-related public health strategies. That principle accompanied by these recommendations can add to community-level disaster resilience. Such action will be required to reduce flood-related destruction and morbidity in the climatic chaos predicted in the future.

Acknowledgements

We would like to thank Dr Paul Whitfield for his encouragement, editing and other guidance for this review, and Agnes Black for editing assistance.