The built environment is an important risk factor for unintentional injuries for all age groups.1^–6 Regulatory frameworks, notably building codes, are very influential in defining features of the built environment. Thus, improving building codes is a critical population health strategy to reduce injuries.78 Traditionally, the primary proponents of changes to the codes have been those in the building and construction industries, and those with disciplinary backgrounds in civil engineering and architecture.9 A fundamental shift in the orientation of these codes is underway, with a focus away from prescriptive codes and towards performance-based codes. This change opens up two important avenues for health professionals to provide input on the regulatory framework for the built environment so as to reduce the risk of injuries. The first involves advocating for evidence-based changes to building codes, an issue that has been discussed elsewhere.10^–12 A second avenue, and the focus of this article, concerns the use of indicators to monitor the performance of new codes. More specifically, this article describes the need for indicators that bridge the perspectives of the health and building sectors, for which predominant monitoring concerns are injury prevention and building performance, respectively.

I begin with some background on performance-based codes. The building-centric orientation of indicators in the building sector is then compared with the person-centric orientation of indicators in the health sector. Finally, injury prevention indicators that may help to bridge these sectors and implications for surveillance systems are considered.

PERFORMANCE-BASED CODES

For more than a decade, there has been an international movement to replace more prescriptive codes with performance-based building regulations.13^–15 This is shifting the content of codes from a description of how a building is to be constructed to what a building is expected to do, given its intended use.16 Performance-based codes are “legal instruments intended to ensure that buildings, when constructed in accordance with the regulations, provide socially acceptable levels of health, safety, welfare and amenity for building occupants and for the community in which the building is located” (p 91).17

The implications of performance-based building codes have received much attention. Global organizations such as the International Council for Research and Innovation in Building and Construction (http://www.cibworld.nl/website/pres.php) have commissioned numerous documents in support of this change and many international conferences have addressed this issue.

The primary benefits of performance-based codes, from the perspective of the building industry include: (a) providing a better understanding of and means to communicate user requirements; (b) allowing builders more flexibility, innovation, and cost-optimization in reaching design solutions; and (c) facilitating international trade.18 The uptake of these codes has led to a re-examination of accountability frameworks used in the housing industry. Indicators are an important component of these frameworks, providing a means to document the actual performance of buildings against targets and providing direction for the enforcement of new building regulations.171920

COMPARISON OF INDICATORS IN BUILDING AND HEALTH SECTORS

Several recurring themes are prominent in discussions of performance indicators in the building sector. Firstly, the underlying goals of performance-based indicators have a business orientation.20 For example, in a discussion of indicators for the objectives and goals of codes, Szigeti and Davis20 indicated that a notable change with the introduction of performance-based codes has been “the growing emphasis on the links to business objectives and goals, to…business processes, and to business results” (p 33). Secondly, there are frequent references to the life cycle of the building, as illustrated by discussions of sustainability, levels of degradation, and the continued performance of a building at required functional level(s).1720 Thirdly, levels of building obsolescence and utilization patterns are identified as important performance-based indicators to monitor asset management.20 The latter indicators are also reflected in more sophisticated models for monitoring assets that focus on the early identification of deteriorating infrastructure.21 Although some authors have described both technical (eg, materials science, mechanics, engineering, building science) and sociological (eg, ergonomics, human behavior, psychology) domains for performance indicators,22 a discussion of indicators that bridge these domains is largely absent from the literature.

Data to monitor the implementation of performance-based codes require an information infrastructure. Thus, with the identification of requirements for new performance indicators have come calls for shared, interactive databases that will provide housing and construction stakeholders with the information required to manage the life cycle of a building.20 Missing from these calls are considerations of how building databases might intersect with those in the health sector.

In the health literature, some preliminary efforts to integrate health and building databases have been described. For example, Newcombe and colleagues6 linked injury surveillance data from emergency room departments in Wales with a property database that included variables on housing type (eg, detached, flat conversions, terraced housing), and floor area. However, a discussion of performance-based building codes has not appeared. Rather, standard epidemiological indicators such as incidence rates, cumulative incidence rates, and attributable risk remain prominent in the field of injury prevention. These indicators are typically used for three purposes: to summarize trends in injury rates, to identify populations at risk, and to determine the impact of injury prevention programs. The person-centric orientation of these indicators is consistent with health sector goals. This orientation also reflects the dominance of behavioral solutions rather than built environment strategies to prevent unintentional injuries. The abundance of effectiveness studies and systematic reviews on injury prevention with this behavioral focus is illustrative of this point.452324

If the health sector is going to contribute to the evidence requirements of performance-based codes, a common taxonomy to identify features of the built environment that contribute to injuries is required. International classification systems have now been developed to standardize the documentation of built environment features implicated in injury events. The most recent iteration of the external classification system represents an important effort to further our understanding of environmental factors and injuries.2526 While this multi-axial and hierarchical classification system is promising, the place module captures general information about building design features (eg, place of occurrence— stairs, bathroom, toilet, kitchen, garden).26 Thus, this external classification system arguably suffers from the same lack of specificity regarding the built environment that is notable in many epidemiological studies of environmental risk factors for falls.

This lack of specificity is a critical concern. As an example, in the current round of revisions to the Canadian building code standards, specific code recommendations such as the height of risers, length of treads, and diameter and shape of handrails on stairs; also the slope, height of handrails, and turning circles on ramps are under review. The ideal evidence to support this review would link these very specific measures of the environment to health outcomes. Ideally, this evidence would be retrievable from both research studies and analyses of existing surveillance systems. But these sources have key limitations.

With respect to research, there is solid laboratory evidence providing the level of precision on environmental characteristics that is required to guide changes to the codes.27^–30 However, laboratory-based studies report on the functionality of these environmental features but not their morbidity or mortality outcomes. On the other hand, community-based injury studies, including those that use the new external classification system for injuries, often report injury morbidity and mortality outcomes, but lack the detailed descriptions of building features that are required to inform changes to performance-based codes. This evidence gap weakens the case for building code changes, as demonstrated causal links between specific features of the built environment and morbidity and mortality are largely absent from the research literature.

With respect to surveillance systems, data gaps and linkages between existing datasets in health and building sectors are problematic. As an illustration, following on from a community study that documented the positive association between access to universally installed grab bars and grab bar use,31 my colleagues and I explored the potential to link surveillance data on injuries resulting from bathtub falls with data from housing documenting the presence or absence of universally installed or owner-installed bathtub grab bars in the province of Ontario, Canada. We obtained information about the type of data available on building characteristics from housing authorities responsible for low-income seniors’ housing, as well as the type of data available from paramedic reports and emergency room admissions that were included in the Canadian Hospitals Injury Reporting Program. There were critical gaps which did not allow us to examine the relationship between the presence or absence of bathtub grab bars and injurious falls. The paramedic and emergency room data indicated the general location of a fall that led to an injury but rarely noted the presence or absence of a bathtub grab bar or its use or non-use at the time of a fall. Public housing authorities had data on universally installed grab bars, but had no information about grab bars that had been installed by individual tenants. We were unable to identify a database that contained information about bathroom grab bars in private homes, as these are not currently required under Ontario building code legislation (unless a residence is being built to meet standards for the disabled). Therefore these data are not captured as part of building inspections. Thus, there were important gaps in the information provided within each of these datasets. Even if linkages among the datasets were possible through common identifiers for individuals and buildings, they would not have provided the type of information required for this particular building code change. Notably, it is through the queries one might ask of integrated health and building data that these gaps in metrics and in routinely collected data become apparent.

From the health sector standpoint, it is my contention that one of the reasons for these types of gaps is the primary focus in the epidemiological literature on person-centric measures. For example, many epidemiological researchers have used refined measures of personal attributes that are known risk factors for falls. These include measures of balance,32 gait,33 and falls efficacy.3435 However, for the most part, it appears that epidemiologists have not used findings from laboratory-based studies in the fields of ergonomics and biomechanical engineering to inform the selection of study measures that would more precisely identify potent environmental hazards.

Thus, there is a fundamental difference in the orientation of standard indicators in the health and building sectors. Those in the health sector are person-centric, while those in the building sector are building-centric. Table 1 provides illustrative examples.

As shown in table 1, standard measures of injuries in the health sector do not take into account building longevity. This reveals another problem; buildings are normally constructed so as to provide many decades of use. In the case of hazardous building features (eg, stairs with inadequate tread length, guard rails that are too low), the cumulative contribution of these hazardous features over years of projected use is not captured by conventional indicators of injury rates. However, an integrated metric (shown in the fourth column) would provide a cumulative estimate of injuries over the lifecycle of the building. This estimate would probably be conservative, as it does not take into account maintenance failures. Nevertheless, cumulative projections of the number of injuries over the expected lifecycle of the building could then be compared for differences in design features (eg, stairs with or without handrails, bathtubs with and without grab bars). This would highlight priority decisions regarding changes to building codes.

A final issue concerns how we might narrow the range of comparisons for changes to building codes. Here, data from ergonomic studies would be helpful. Under laboratory conditions, it is often possible to test small differences in the performance of built environment features. Through such studies, the parameters that need to be assessed in community studies and via surveillance systems can be determined. For instance, ergonomic studies indicate that the average person clears about 1 cm in height with each walking step. This laboratory-derived measure provides a useful cut-off point to examine the uniformity of stairs and heights of thresholds for doorways. Similarly, pertinent hazard ranges for stair dimensions have been pinpointed by laboratory studies in the UK.36 These ergonomic studies have shown the range of stair dimensions that yield a significant increase in trips and thus provide important cut-off points for safety that can be used when examining stair dimensions and falls in community-based studies.

In summary then, much of the incidence data on injuries from the health literature lack specificity vis-à-vis the particular building elements that are involved, and fail to take into account varying usage patterns of building components. This seriously limits the utility of our standard injury indicators to inform new codes and to monitor the safety of specific performance-based codes. Furthermore, there are very few examples of injury studies in the health literature that identify the age of the building and the building code regulations that were in place at the time of construction. This prevents critical comparisons across studies to ascertain whether or not specific building code standards have been effective in reducing the risk of injuries.

TOWARD A NEW SET OF INDICATORS

The preceding discussion indicates that the predominant focus of the building industry concerns the performance of buildings. In contrast, the primary focus of those in the health sector is the safety of individuals who use those buildings. In this section, I present several ideas for metrics that bridge these sectors and reflect the interaction between individuals and building features. These suggestions are primarily aimed at those working in the health sector.

As discussed above, it is imperative that we use findings from laboratory-based studies when identifying measures of the built environment in epidemiological studies. This would yield much-needed evidence linking precise and potent features of the built environment with injuries, evidence that is critical to inform improvements in building codes.

With regard to cumulative incidence rates for injuries that involve human-built environment interactions, the denominator normally consists of the number of people exposed to a risk factor over a specific time period. However, this is a very crude measure of exposure time that may hide important relationships between the risk of injuries and the built environment. For example, location-specific rates are often used to estimate the proportion of injuries that occur in certain parts of the home. However, these rates do not take into account the much longer periods of time one spends in a bedroom relative to time spent using stairs or navigating thresholds. Various authors3637 have estimated that, when exposure is taken into account, stairs are among the most hazardous of environmental features, yet stairs are identified as the location for only about 10–15% of all falls.38 The cumulative rate, after adjustment for time spent in different locations of the home, would be a superior indicator. However, it would be onerous to try to estimate the amount of time individuals spend in various locations of a home for each epidemiological study. The development of an average time exposure factor to apply to cumulative injury incidence rates across population subgroups would assist in pinpointing environmental features that require priority attention. In essence, this would allow the estimation of cumulative building-element injury incidence rates.

The potential longevity of buildings flags the need for another injury indicator, one that would take into account the projected lifespan of building infrastructure and its expected lifetime contribution to injuries. This would involve calculating the projected cumulative incidence rate of injuries given the presence or absence of a particular environmental feature. In essence, this indicator would describe the cumulative incidence of injuries resulting from a specific building element (eg, presence of handrails with a particular circumference and shape on stairs) using building-years rather than person-years as the denominator. This would provide a means to compare estimated rates of cumulative injuries that are projected with building codes that permit more hazardous versus safer building features, as defined in ergonomic studies. It would also remind us that an inherent perpetrator of falls and injuries is the built environment. These new metrics would help to drive home the message that injuries may result from particular features of building design.

A final area that warrants attention is surveillance systems. Within the housing sector and building industry, there is much discussion about the need for integrated surveillance systems to support management processes over the life cycle of a building.19 Similarly, in the health literature there is discussion of the need for integrated surveillance systems but with a health focus.3940 Key documents guiding the development of injury prevention surveillance systems in the health field do not generally mention the housing and building industries nor the organizations responsible for developing building codes as key stakeholders to be involved in the process of developing injury surveillance systems.4142 The time is ripe for discussions of linked and complementary surveillance systems that would meet the needs of both sectors and inform improvements to regulations that can make the building infrastructure safer for the populations we all serve.

CONCLUSION

These indicators have been put forward with the aim of stimulating dialogue about new measures that might better inform policies influencing the design and construction of safer buildings. It is essential therefore that these new indicators resonate with those in both the health sector and building industries. It is time to re-examine our commonly used epidemiological indicators in the field of injury prevention to determine their utility to address the accountability requirements of performance-based codes. It is also critical that we initiate joint efforts to develop integrated surveillance systems in health and building sectors. With the introduction of performance-based codes, a re-examination of the metrics we use in the field of injury prevention is timely.