Space Syntax Analysis: Tools for Augmenting the Precision of Healthcare Facility Spatial Analysis

Transcription

1 Theory Space Syntax Analysis: Tools for Augmenting the Precision of Healthcare Facility Spatial Analysis Health Environments Research &DesignJournal 2016, Vol. 10(1) ª The Author(s) 2016 Reprints and permission: sagepub.com/journalspermissions.nav DOI: / herd.sagepub.com Ahmed Hassem Sadek, M.Arch 1,3, and Mardelle McCuskey Shepley, FAIA, EDAC, FACHA, LEED AP BDþC 2 Abstract Aim: This article reviews and addresses various spatial measures that provide and facilitate accurate descriptions of different layout typologies with particular focus on healthcare facilities. Background: Evidence-based design is a field of study that emphasizes the importance of using credible data in order to influence the design process. It addresses whether/how the design and operation of buildings support positive health outcomes through a growing collection of solutions informed by research and practical knowledge. In order to acquire this knowledge, we must consider the environment in a very precise and measurable way, which can then be investigated quantitatively with regard to various human behaviors and cognitive processes. Topical Headings: Space Syntax theory and methods are concerned with understanding how spaces are arranged and connected to one another. It provides quantitative measures of individual spaces and of the entire layout, thereby achieving great accuracy. Space Syntax researchers have used graph theory to define spatial measures and have produced computer tools to quantify the relational properties of spaces and to provide numerical and display values for each. Moreover, they have developed more precise tools that can be embedded within basic Space Syntax concepts in order to augment the credibility of the acquired data. Conclusion: The study elaborates on the effectiveness of Space Syntax and its newly developed measures for the field of healthcare facility design and research. We present a comprehensive model that brings together these measures with their correlated behavioral and perceptual consequences, thereby providing a point of departure for further investigation and exploration. Keywords evidence-based design, Space Syntax, healthcare, behavior, spatial analysis A growing body of research demonstrates the significant impact of the healthcare physical environment on a variety of health-related outcomes such as length of stay, pain, medication, stress, mood, and environmental appraisal (Huisman, Morales, van Hoof, & Kort, 2012; Ulrich, Berry, Quan, & Parish, 2010; Ulrich, Zimring, Quan, Joseph, & Choudhary, 2004). This places the built environment as a key factor in defining physical and psychological health for patients and 1 Department of Architecture, Faculty of Engineering, Assiut University, Assiut, Egypt 2 Department of Design and Environmental Analysis, Cornell University, Ithaca, NY, USA 3 Faculty of Architecture, Building and Planning, Melbourne School of Design, University of Melbourne, Victoria, Australia Corresponding Author: Ahmed Hassem Sadek, MArch, Department of Architecture, Faculty of Engineering, Assiut University, Assiut, Egypt. s: ahmed-sadek@hotmail.com; sadeka@student.unimelb. edu.au

2 Sadek and Shepley 115 their families as well as for healthcare staff. Evidence-based design (EBD) is a field of study that addresses whether/how the design and operation of buildings positively support health outcomes through a growing collection of solutions informed by research and practical knowledge. The approach has become popular in healthcare architectural design in an effort to improve patient and staff well-being, aid the patient healing process, reduce stress, and contribute to safety (Fouts & Gabay, 2008; Shepley, Harris, & White, 2008). The built environment has to be considered precisely to accurately predict its influence on people s mental-health, behavioral interaction, and engagement with the environment. Consequently, researchers have adopted Space Syntax, along with other tools and methods, to facilitate fulfilling this aim. The concept of Space Syntax, which was originally introduced by Hillier, concentrates on revealing the underlying social logic of spaces by developing strategies to describe their configurations and their effects on various social and cultural attributes (Bafna, 2003). It provides rich and diverse quantitative descriptions of the built environment configurations, particularly buildings and urban street networks, with a main focus on their arrangement and interconnections. This rigorous description allows for potential explanations of a variety of physical and psychological responses such as user movement, experiences, and cognitive knowledge of place (Montello, 2007). The built environment has to be considered precisely to accurately predict its influence on people s mental-health, behavioral interaction, and engagement with the environment. Accordingly, this article focuses on basic and newly refined Space Syntax tools used in healthcare facility research, including those that were developed with the aim of augmenting the precision of spatial analysis. We provide a literature review of the studies that currently involve, or potentially involve, implications for healthcare environments. Little research has focused on addressing the application of Space Syntax and its newly developed tools for healthcare research. Haq and Luo (2012) summarized studies that applied Space Syntax measures in healthcare settings until 2011; however, the focus was not on recently developed tools. Turner (2007a) addressed the technical description of a number of the programs written to perform Space Syntax analysis but did not focus on the newly developed measurements that underlie the precision of spatial analysis. According to the authors knowledge, no previous studies have thoroughly demonstrated and reviewed the existing and developed tools of Space Syntax in order to provide a comprehensive overview for further investigation and exploration within the field of healthcare design. We note that the spatial tools and constructs mentioned in this article are not exclusive applications to hospital layouts and fit the analysis of several building typologies. However, they have been addressed here in a way that serves the contemporary trends in the field of healthcare facility design and research. We also note that the new developments included in this study are not stand-alone methods that contradict the basic concept of Space Syntax. However, they are continuous and sequential improvements that have been introduced over time through ongoing research with the purpose of refining the basic analytical methods of Space Syntax. Method For the purpose of this study, electronic databases including PubMed, EBSCO, and SAGE and the websites of the Space Syntax biannual symposia were examined in order to define studies that carry robust implications and contribution to the research in healthcare environments between the years 1990 and Previous reviews in healthcare research addressing Space Syntax and developments in Space Syntax analysis were identified and their reference lists were inspected. The search was expanded by using a combination of the following terms: EBD, hospital design, healthcare design, quantitative representation of the space, spatial analysis, spatial configuration,

3 116 Health Environments Research & Design Journal 10(1) and behavior. In total, 112 papers were selected. The content of 29 of them was further reviewed and analyzed. The inclusion criteria were based on studies that are relevant to healthcare facilities and introduced refined syntactical analysis methods or empirical examinations. The following sections of this article briefly introduce the development of Space Syntax basic measures and constructs, followed by studies that have applied them to healthcare contexts. The article goes further by reviewing the attempts to introduce newly refined syntactical representations and measures that have the potential to increase the accuracy of spatial description. The developed tools were divided into discrete and combined measures with careful elaboration on the usefulness of each tool in addressing specific issues in healthcare design. The authors conclude by providing a comprehensive model that brings together the basic and newly developed spatial analysis tools in conjunction with their correlated behavioral and perceptual consequences with the intention of building a platform that allows for future research. Basic Space Syntax Measures Convex and axial maps are the most popular techniques for describing spatial configuration in Space Syntax. Convex maps, which are primarily used to analyze arrangement of programmatic spaces, are generated by partitioning a given spatial setting into a set of fewest and fattest convex spaces; each convex space then is defined with a node and its possible connections are defined with an edge, generating a justified graph that represents the spatial relations between adjacent spaces (Hillier & Hanson, 1984). When the focus is to capture behavioral characteristics such as movement, axial maps are more appropriate measures. They are defined as the longest sets of lines of sight that pass through all the open spaces in a study area (Hillier & Hanson, 1984, p. 91). The number of direct connections to other lines or spaces is called connectivity. The level of connection, in term of the number of turns along a path, is defined as the step depth, which describes the mutual relationship between all lines or spaces. Through this process, integration values can be assigned to each place in the layout. This indicates how well one space is connected to all other spaces in a system. For instance, a corridor with high integration means that it is closely connected (with fewer topological steps) to all other corridors in a given layout and vice versa. When integration is calculated for all spaces in a system, it is named global integration or integration-n. We can also calibrate integration based on connections at two-step depths which are named local integration or integration-3. The correlation between connectivity and integration values is used to define the intelligibility of the entire system. It identifies how far a person, in a local position, is able to build up a picture of the global structure based on its parts (the number of local connections). For more exploration of its complex nature and mathematical expressions, please refer to Space is the machine by Hillier (1996) and Hillier, Burdett, Peponis, and Penn (1987). Convex and axial maps are the most popular techniques for describing spatial configuration in Space Syntax. In order to increase resolution, Iida (2001) implemented a segmentation algorithm into a software program named Segmen (Turner, 2007b). The analysis was based on breaking continuous axial lines into segments according to their intersection with other lines. This approach facilitated finer scales of spatial analysis and helped differentiate systems that are characterized by uniform structures from those with more organic geometrical patterns (Al-Sayed, Turner, Hillier, Iida, & Penn, 2014). Segmented analysis appeared to have significant drawbacks when basing the analysis on topological configurations (fewest turns). The problem occurs as previously well-integrated axial line turns into more segregated shorter lines due to the calculated turns/steps associated with the adjacent segments on the same line. This, in fact, contradicts the logical structure of a space, as it treats steps to straight segments similarly to those crossing orthogonal lines (Turner, 2004). Additionally, the concept of topological step depth does not distinguish between different turn angles as it treats a slight shift of 15 as equal

4 Sadek and Shepley 117 to a right angle of 90 even though it is not a substantive turn (Turner, 2001a). Therefore, Turner (2001a) introduced a new syntactic model, named angular segment analysis (ASA) as a refined method to overcome these problems. The refined method still utilizes the concept of segmenting axial lines but bases the measures on the amount of angular change a long a path, by assigning angular weights to each connection in the system ranging from 0 when passing in a straight line to 2 when turns reach +180 (Turner, 2004). The concept of ASA shifted the segment representation to be a fundamental construct when performing spatial analysis. Its analytical power has been established in many studies within the Space Syntax research community including the seminal study by Hillier and Iida (2005). They investigated the strength of three types of analysis: topological (fewest turns), geometrical (least angle), and metric (shortest distance) in four urban areas in London. Movement patterns were most highly correlated with geometrical followed by topological configuration of spaces. Metric distance was far less correlated with both vehicular and pedestrian movements within the studied areas. The reason for that was attributed to the nature of metric distance analysis which does not take into consideration route complexities. Hillier, Turner, Yang, and Park (2007) proposed a different way to benefit from metric analysis rather than assigning direct metric weightings to Space Syntax measures. They argued that the best way to implement metric analysis is by using radius distance for these measures. This radius identifies how far calculations are allowed to reach within a certain distance. By selecting a subset of the system and processing the values accordingly, Hillier et al. (2007) were able to capture different local spatial phenomenon including formal and functional properties of the network (spatial differentiation of areas at the neighboring scale). ASA with a defined metric distance turned out to be one of the most recommended spatial representations within Space Syntax software (Al-Sayed et al., 2014). Isovist is another fundamental type of spatial representations that deals with the visual properties of the built environment. It is defined as the set of all points visible from a given vantage point in space (Benedikt, 1979, p. 49). The isovist area is the area of the visible field, usually assumed as having a 360 field of vision. Turner (2001b) developed a software program named UCL Depthmap basically to perform isovist analysis similar to what had been previously applied to axial lines and convex spaces. The utilized unit of analysis in this approach is the tile, which is a set of square tiles with a convenient dimension that can be laid on any floor plan. Consequently, the connectivity and integration values for each tile can be calculated. Two structures of the layout can be constructed, the first is the visibility structure (Visibility Graph Analysis [VGA]) which is developed when the tiles are placed at eye-level and only walls break up their interrelationships. However, when the tiles are placed at knee-level with the presence of furniture, then the accessibility structure can be constructed (Turner, Doxa, O Sullivan, & Penn, 2001). Space Syntax researchers have combined these measures and produced several computer software programs that provide tables with numerical syntax values in addition to colorcoded diagrams to display the distribution of those values on the plan layouts (Figure 1). With these software programs and constructs of Space Syntax, architectural plans and individual spaces within them can be quantitatively measured and described in rigorous ways, allowing them to be used as independent variables in various research projects and their consequent statistical analyses. Applications in Healthcare Facility Research Despite being relatively new in the field of healthcare facility research, a growing body of research has applied Space Syntax analysis in healthcare settings. Basic Space Syntax measures have been used to study wayfinding, privacy preferences, perceived quality of care, user cost, nurses movement, and evacuation patterns in hospital buildings (Haq & Luo, 2012). Wayfinding, which has received reasonable attention in the application of Space Syntax, was also investigated by Khan (2013b) to understand patients satisfaction and travel experience. Different associations were reported between satisfaction, travel behavior, and layout syntactic properties

5 118 Health Environments Research & Design Journal 10(1) Figure 1. Example of visual graph (left) and axial map (right) analysis (Cai & Zimring, 2013). according to gender, age, frequency, and number of visits. Basic Space Syntax measures have been used to study wayfinding, privacy preferences, perceived quality of care, user cost, nurses movement, and evacuation patterns in hospital buildings (Haq & Luo, 2012). In addition to the studies represented in Haq and Luo (2012), Zadeh, Shepley, and Waggener (2012) addressed the role of design in promoting efficiency and reducing redundancies in nursing units. Justified permeability diagram method was applied to quantitatively and qualitatively analyze the spatial ordering and space flow in five acute care settings and determine their potential effects on care quality. Setola, Borgianni, Martinez, and Tobari (2013) studied patterns of interaction among patient and medical staff. They analyzed the integration core (the top 10% of integrated lines) of both public and staff spaces in three hospitals. They demonstrated that interaction was occurring mostly in overlapping spaces when the two cores overlapped, while it was occurring in transition spaces between the two cores when they were not overlapping. Pachilova and Sailer (2013) and Sailer et al. (2013) also studied interaction between patients and caregivers by comparing spatial layouts of two different outpatient clinics using the analysis of axial step depth. They suggested that shared facilities for caregivers along with clear spatial separation of staff and patient areas may help in facilitating good and frequent communication among them. Cai and Zimring (2013) employed Space Syntax analysis to understand cultural differences between Chinese and U.S. designs of nursing units. The study revealed interesting culture differences regarding encounter, communication, territoriality, hierarchy, and privacy. From a similar perspective, Khan (2013a) investigated hospital design in Bangladesh. The study represented a reflection of the designer s cultural values and norms on the spatial organization of Bangladeshi hospitals. Space Syntax analysis was further combined with discrete event simulation in the process of designing a new emergency department in order to analyze the potential influence of spatial configuration on the effectiveness of visual surveillance, movement, and communication (Morgareidge, Cai, & Jia, 2014). The study demonstrated the importance of these tools for facilitating decision making, improving organizational performance, and reducing capital and operational costs. These studies, accordingly, represent promising evidence of the significance of Space Syntax theory in the field of healthcare facility design and research.

6 Sadek and Shepley 119 Figure 2. General visual connectivity (left) and target visual connectivity (right; Lu & Zimring, 2012). The pictures show the ability of target visibility analysis to better and more accurately discriminate between layout locations that have higher target visibility of patient beds compared to general visibility of all parts of the layout. Newly Refined Spatial Representations and Measures Scholars have also sought, in complementary ways, to increase the precision of spatial configuration measures. They have developed and improved additional measures and techniques in order to increase the accuracy of representing spatial characteristics. These attempts include the following: Target Visibility Analysis Previous research has demonstrated the importance of visibility for patient and staff outcomes. Better observation and faster response to patients, which contributes to reducing patient falls and mortality rates, are among the benefits of improved visibility. Additionally, it is associated with less travel time for nurses and increased communication (Lu & Zimring, 2012). Despite this importance, former attempts to develop methods depicting and analyzing visibility patterns in buildings focus mainly on generic visibility of all open spaces and omit foci on critical targets. Accordingly, Lu, Peponis, and Zimring (2009) developed a computational tool named target visibility analysis to quantify and graphically represent visibility to patients locations. The tool works by superposing a grid of tiles with even spacing on the original layout and defining targeted objects in order to calculate the mutual visibility (lines of sight) between them, thus allowing for the targeted visual connectivity value to be calculated (Figure 2). The proposed tool is a refinement to the standard analysis of visual fields as it enables the visibility measurement of a specific number of targets from definable observation locations. This allows the researcher to assign particular locations with more weight than others. In the calculation process, partially or fully visible targets will be considered visible. However, there are number of ways where targets can be redefined if specific parts of them are more important. Additionally, a distance radius can be applied in the analysis according to the user s interest....former attempts to develop methods depicting and analyzing visibility patterns in buildings focus mainly on generic visibility of all open spaces and omit foci on critical targets. The authors introduced the Targeted Visibility index (TVi) which is defined as the ratio of the average targeted visibility value of all locations in the setting, and number of pre-selected targets (Lu, 2010, p. 55). This provides a way to compare the degree of targeted visibility between various nursing layouts that have different numbers of patient beds. In order to demonstrate the importance of Target Visibility analysis as a tool that has capabilities

7 120 Health Environments Research & Design Journal 10(1) in assessing healthcare design proposals, Lu (2010) reanalyzed three typologies of nursing units radial, double, and single corridor layouts which were investigated in a previous study by Trites, Galbraith, Leckwart, and Sturdavant (1969). The results of the TVi corresponded completely with the Trites et al. s original empirical findings. Additionally, Lu and Zimring (2012) investigated distribution and behavior of nurses and physician in an intensive care unit (ICU) based on the visibility pattern of the setting. Staff positions were highly correlated with the target visibility measures compared to general visibility, which suggests targeted visibility as a particularly appropriate tool when investigating patterns of spatial use and behavioral consequences of medical staff in nursing units. In a later study, Lu, Ossmann, Leaf, and Factor (2014) replicated a study of Leaf, Homel, and Factor (2010) in an ICU with the aim of enhancing the methodology and complementing the original findings. They used target visibility analysis to calculate various visibility measures of patient rooms including patient head visibility and patient room visibility in addition to measuring field of view and the shortest walking distance from the nurse station to patient rooms. High ICU mortality rates were significantly associated with severely ill patients assigned to rooms with low visibility. This was aligned with the original outcomes of Leaf and colleagues. Spatial Positioning Tool (SPOT) Similar to the idea of target visibility is the concept of spatial positioning which was adopted by Markhede and Carranza (2007) to analyze face-to-face interaction in buildings that are strongly influenced by organizational and cultural features. They proposed the implementation of this tool to examine the relationships between occupied spaces as a supplement to regular Space Syntax analysis that create graphs and measures of all occupied spaces, ignoring the special status of specific occupied parts. Unlike the target visibility analysis, which bases its calculation on the tile as the unit of analysis, the SPOT adopts isovists to represent and analyze spaces. The position of the isovist can be determined according to subjects/objects spatial distribution, allowing for the visual situation to be illustrated on the generated graphs. SPOT produces two kinds of graphs: intervisibility, which shows overlapping isovist fields that become darker when many overlapping layers are superimposed, and network graphs which show the relative asymmetry integration (Figure 3). Unlike the target visibility analysis, which bases its calculation on the tile as the unit of analysis, the SPOT adopts isovists to represent and analyze spaces. Markhede, Miranda, and Koch (2010) demonstrated in their studies the power of SPOT by analyzing 600 workstations distributed on three floor plans in an office building located at Posten Headquarters, Sweden. In addition to intervisibility graph analysis, snapshot observations and a series of surveys asking questions about face-toface interactions were carried out. Face-to-face interactions were 50% higher at the workstations located in the more intervisible areas compared to those placed in the less intervisible places. This is also supported by other researchers in the field of Space Syntax such as Sailer (2007), who believes that organizational relations is needed as complement to the regular Space Syntax analysis. The concept of SPOT is implemented via an independent software written in Java which allows it to run in Windows, Mac, and Unix machines (Turner, 2007a). Furthermore, Carranza, Koch, and Izaki (2013) proposed a further development for SPOT by integrating it as a design tool in addition to its primary analytical role. The researchers developed algorithms to calculate isovists in real time and apply them to a new hospital ward in order to refine the design and visualize its effects simultaneously on issues significant to healthcare effectiveness like visibility, accessibility, and contact between caregivers and patients. The use of spatial positioning analysis maintains the incorporation of organizational aspects into the analysis of layouts, which allows the focus to shift to investigating interaction and knowledge sharing. Furthermore, spatial positioning has the potential to be applied in the field of healthcare design and research. For instance,

8 Sadek and Shepley 121 Figure 3. Intervisibility graph (top) and network integration graph (bottom) (Markhede et al., 2010). Darker areas indicate the zones that are visible from the largest number of the pre-identified positions (locations of interest). patients occupation of place can be tested in relation to different functions such as ICUs and infusion suites, where visibility along with privacy and communication are crucial for safety and other positive health outcomes. Additionally, the impact of different nurse station typologies, such as centralized and decentralized nurse stations, can be investigated. This might open the way to create representations for social relationships such as interactions, awareness of peers work, and knowledge exchange, allowing for testing and evaluating different combinations and solutions for existing and new layouts. Weighted Isovist Area With regard to critical functional areas in hospitals, where proximity to an observer is crucial for patient safety, Kim and Jung (2014) proposed a mathematical model to control the existing limitation in the classical measures of the isovist. As mentioned before, the isovist field represents the range and the area of the space that can be seen from a vantage point. However, the conventional measure does not discriminate between concentrated and dispersed spaces. Therefore, the authors developed the concept of weighted isovist area. The associated tool uses distance as a weighting factor to develop the isovist area, thus discriminating between different shapes and sizes. Distance, as a factor, can represent the information at the investigated points regarding quality of surveillance, privacy, and security. The model enables researchers to define the specific weighting functions according to their context. The weighted isovist is implemented by substituting the distance between the isovist origin and a point on the isovist boundary. It reveals areas that constitute large visible regions based on the needed information (weighting distance) from the observation point (Figure 4). The developed model is also implemented into Syntax2D, one of the spatial analysis software programs, to provide a ready-to-use tool that considers the differences between nearer and farther regions. Distance, as a factor, can represent the information at the investigated points regarding quality of surveillance, privacy, and security. The software Syntax2D also provides pathbased measures that generate isovists from a

9 122 Health Environments Research & Design Journal 10(1) Figure 4. General isovist area (left) and distance-weighted area (right) (Kim & Jung, 2014). series of observation points. The cumulative isovist of the path can be shown and the data from each of these points can be analyzed according to several measures. For instance, colored circles along the path are used to display the area value of each isovist (Turner, 2007a). The importance of the available information through a specific path is emphasized and illustrated by Lu and Seo (2012). They suggest quantifying route properties rather than the overall unit shape in order to predict nurse behavior and movement patterns. Multi-Layered Network Visibility may also play a significant role in conditions such as wayfinding and intelligibility of spaces, where the presence of a visibility link can reduce the degree of segregation of a desired destination. However, Space Syntax, when constructing the axial representation of a building, does not take this issue into consideration, causing what is called the visibility accessibility problem which occurs when spaces have accessibility barriers that do not hinder visibility. In order to overcome this problem, Dalton and Dalton (2009) developed the multi-layered tool, where elements of the building that can be seen, but not directly accessed, are taken into consideration through a formula implemented into the software WebmapAthome (version ). This process assigns a degree to these particular situations. Thus, a new type of integration is created, which incorporates visibility and accessibility into the syntactical representation. Beck and Turkienicz (2009) also address the influence of visual information on space intelligibility, where the layout of Pinacoteca Art Museum was analyzed in term of the intelligibility of three defined routes taking into consideration the visibility and permeability relations. The analysis depicted how visual information and connections may add more structure and levels of prediction to the visitors route, which is crucial for wayfinding situations and spatial cognition. In a recent study dealt with the same issue, Varoudis and Psarra (2014) developed and applied a new approach that takes into consideration 3- D visual relationships, extending the original VGA to a 3-D VGA model. The tool managed to precisely represent and discriminate locations that have highly visibility links and are, at the same time, not directly accessible (e.g., atrium and crossing bridges). The concept of multi-layered network can be applied to another issue when describing the space the signage system (Dalton & Dalton, 2009). The presence of signage can raise the intelligibility of a building as it aids in enhancing clarity of the route in the built environment. The concept can represent the presence of signage when analyzing a floor plan, thus altering the description of spaces and allowing for an integrated understanding of the built environment. By taking signage and visual information into account, a wide range of conditions can be represented. This can ultimately enhance the prediction of wayfinding difficulties in complex healthcare settings, which is a key factor in

10 Sadek and Shepley 123 frustrating patients and families, causing stress, and diverting staff time to giving directions (Ulrich et al., 2010). By taking signage and visual information into account, a wide range of conditions can be represented. This can ultimately enhance the prediction of wayfinding difficulties in complex healthcare settings. Place Syntax Place attractiveness is one of the determinants of accessibility research, where place determines a geographically specific space or location that has a particular content. On the other hand, Space Syntax deals only with spaces that have no specified content and therefore no measurable attractions. Noting these differences between space and place, Jiang and Claramunt (2002) have distinguished between geographic and geometric accessibility. They define geographic accessibility as the proximity of places, while geometric accessibility is defined as the proximity of spaces. In order to overcome this issue and to deal with the full concept of accessibility, Ståhle, Marcus, and Karlström (2005) developed a tool named Place Syntax which brings the Space Syntax description of the cognitive environment into a combined accessibility analysis model with the conventional descriptions of attraction. This suggests the possibility of loading geographical data and qualitative factors for improved predictions of pedestrian movement within Space Syntax. Space Syntax deals only with spaces that have no specified content and therefore no measurable attractions. Despite being developed mainly for urban planning and design, Place Syntax may lead to better understanding and prediction of patient movement and crowding within hospital layouts characterized by uneven distributions content attractions that might contradict the configurative descriptions of the space. Team-Base and Peer Distances Traditional classification of nurse station typologies such as centralized, decentralized, and hybrid layout doesn t provide an accurate reading for the fairly different spatial relationships that might exist in real practice where infinite combinations of those components exist (Cai & Zimring, 2012). Based on the concept of step depth, this study proposed a more systematic approach to quantify and categorize the spatial configuration of nursing units. They proposed team-base distance, which represents the visual proximity to team work, and peer distance (PD), which represents the visual proximity to peer alcoves, as fine-grained spatial metrics to evaluate behavioral differences between nurses such as frequency of interaction and awareness of peers work. Behavior mapping and co-awareness network analyses were also conducted beside the spatial analysis as a multi-layered approach to test the validity of the proposed spatial metrics. The results of the two ICU wings demonstrated a strong correlation between the proposed spatial metrics and nurse behavior and co-awareness. In particular, nurses assigned to lower PD alcoves had significant higher interaction ratios. Additionally, their awareness of other patients conditions was significantly higher than nurses in other alcoves. Overall, this research provided in-depth descriptions of nurse station typologies and their relation to nurses communication and learning. Combined Methods Alongside the efforts to develop advanced tools and construct measures for spatial analysis, researchers proposed and integrated new methods using Space Syntax in order to enhance the depiction of spatial properties and facilitate the investigation of specific situations. Nurse Assignment In a time and motion study conducted by Hendrich et al. (2009), the analysis of the collected data discerned that no statistically significant relationships existed between unit layout type (racetrack, radial, corridor, etc.) and time nurses

11 124 Health Environments Research & Design Journal 10(1) spent in patient rooms. However, noticeable differences between individual nurses in the same unit were observed. Therefore, the authors proposed investigating a sub-area of the unit layout within which individual nurses are assigned in order to understand the influence of spatial content on nurses movement. This was named nurse s assignment. Axial integration measures were calculated for the total number of patient rooms in each assignment allowing spatial properties to be assigned for each nurse. Parameters of movement patterns, such as number of entries to patient rooms and nurse stations, were considered as the dependent variables in the study. This generalized linear model demonstrated the ability to predict the impact of spatial properties of a nurse s assignment on the frequency of trips made to patient rooms. For instance, higher integration values for nurse assignment were associated with higher number of entries to patient rooms and the nurse station. Average time per visit was negatively associated with number of entries to patient rooms, however, the total number of entries was positively associated with the total time spent in patient rooms....higher integration values for nurse assignment were associated with higher number of entries to patient rooms and the nurse station. The syntactical spatial analysis was further developed by Choudhary, Bafna, Heo, Hendrich, and Chow (2010) to include visual connectivity of the corridors outside the rooms and visual step depth of the rooms from the nearest nurse station. The final developed model represented a significant tool that can be utilized to inform the assignment process in existing hospitals or investigate the potential influence of different assignment sets. Medication Administration In order to reveal additional rigorous information about the effect of hospital layout on nurse behavior and walking distances, Seo, Choi, and Zimring (2011) conducted a study in two different ICUs, one with a radial and smaller layout and the other with a larger racetrack layout. They determined that the overall shape and type of the unit, as studied in previous research, might be insufficient to accurately explain and predict nurses walking behavior. Instead the route characteristics which connect different unit components, such as patient room and medication area, contribute more to nurse behavior and walking distances. Therefore, they introduced a new method to observe nurses while performing a specific work process, namely, medication administration. Observing one task helps minimize the influence of other work processes on nurse behavior. Additionally, it allows for a more precise investigation of the effects of the physical spatial properties on the observed behavior. Visibility analysis for the shortest medication path was calculated in order to determine the percentage of the visible area of the unit that nurses could see while performing their tasks. Moreover, general target visibility analysis was performed to define differences in visibility of patient beds from both nurse stations. They found that walking distances were not less in the smaller unit as was expected, rather it was higher. Furthermore, the calculated number of extra stops was associated significantly with path visibility, which may play a role in increasing walking distance and task performance time. The study plays a role as well in demonstrating the applicability of the developed tool target visibility analysis. The findings of this study were supported by Lu and Seo (2012), where nurses did walk more in the smaller unit compared with the larger one. This result reinforced the assumption that specific spatial characteristics in a unit have greater effects on nurse behavior rather than whole spatial properties. Additionally, they hypothesized two measurements of route properties. The first one is the closeness value, which focuses on assessing whether the routes pass through the central nurse station and the second variable, named the path visibility index, is linked with quantifying the visibility from the path to the nurses work areas including the central nurse station and alcoves. This allows for better investigation of spatial properties and their contribution to behavioral differences.

12 Sadek and Shepley 125 Conclusion The effect of hospital layouts on patient, visitor, and staff behavior has been the focus of many rigorous research studies during the last decade using Space Syntax. The development of refined tools that precisely measure layout configurations allows researchers to better investigate occupant behavioral and perceptual consequences related to the built environment, thus, enhancing the credibility and quality of EBD. Moreover, these tools allow designers and clients to consider alternatives and assess proposals before complex buildings are constructed. This article adds to the existing knowledge of the use of Space Syntax in the field of healthcare facility research. It enhances the understanding of the effects of the built environment on occupants through the illustration of Space Syntax measures that depict these impacts. A summary of basic and newly developed measures and their correlated behavioral and social relationships is provided in Figure 5. This comprehensive model includes major research and design themes accompanied by related analytical tools and measures, in addition to the prospective outcomes of these applications. These rubrics are broken into a variety of categories in three clusters based on basic, newly, and combined Space Syntax methods. Direct links are provided to represent the relationship among the tools and their measurable consequences, creating a platform for future studies that aim to enhance outcomes for both patients and caregivers. It provides a design and research road map to facilitate researchers and designers selection of the appropriate measures. This review illuminates several attempts to refine and overcome some of the shortcomings in Space Syntax theory and its related measures as identified by Ratti (2004), Montello (2007), and others. Segmenting axial lines, incorporating metric information in the form of radius analysis, combining information on land use, and implementing multi-layered network allowed researchers to overcome several of these limitations. Respectively these attempts helped highlighting phenomenon related to discriminating different types of spatial configuration, edge effect and local phenomenon, place attractiveness, and representing visibility accessibility issues and elements of the built environments such as signage. Additionally, target visibility analysis, SPOT, and weighted isovist area allowed the analysis to shift from general visibility to critical targets and representing information availability. However, several issues remain to be explored including the impact of the vertical dimension, such as ceiling height and its proper representation, along with non-spatial characteristics of the environment such as surface colors, textures, and patterns. It is also beneficial accounting for individual group differences in responds to layouts. One of the contemporary research applications in hospitals is the performance analysis based on agent simulation. It was also introduced earlier into the Depthmap 7 software (Turner, 2007b), where the agent decision is based on the visibility graph. Integrating other parameters from the queuing theory such as capacity, waiting times, queue lengths, and cycle time into the agent decisionmaking process constitutes a promising direction for future investigations as well. The continuous improvements in rigorous measures for quantifying the spatial properties of the built environment contribute to the effectiveness of the theory of Space Syntax. These developments are imperative to maintaining Space Syntax as a strong and unique discipline and enhancing its contribution to the field of environmental psychology as a comprehensive theory. Implications for Practice Space Syntax analysis, including traditional and newly developed measures, ameliorates the description of spaces. Using this method, we can assess and validate healthcare facilities by comparing layouts in term of: supporting wayfinding and intelligibility, facilitating supervision and security, enhancing workflow and reducing redundancies, minimizing staff and clients walking distances, and promoting privacy within treatment areas. Specific design attributes can be also provided based on the analysis of tools such as target

13 126 Health Environments Research & Design Journal 10(1) Examinable research & design themes Design level Behavioral and perceptual level Spatial analysis tools and measures Outcomes highlight issues related to: Spatial ordering Accessibility Territoriality & hierarchy Flow Movement pattern Communication Spatial knowledge Convex maps & justified permeability diagram (JPD) Axial & segment analysis through: - Step depth - Least angle - Metric distance Efficiency Quality of care delivery Orientation quality Wayfinding Cultural influences Social relationship Privacy Visual surveillance Pattern of spatial use Face to face interaction Isovist & visual graph Target Visibility analysis Spatial Positioning Tool Observation Response time Falls rates Nurses travel time Knowledge sharing Mortality rates Available information Weighted Isovist Area Security Spatial cognition Frequency of interaction Awareness of peers work Multi-Layered Network Place Syntax Team-base distance & Peer distance Frustration & stress Staff time to give directions Learning Knowledge exchange Critical support Frequency of trips to patient rooms & movement patterns Walking distances & tasks performing time Nurse assignment Medication administration Basic Space Syntax measures Time spent in patient rooms Assignment process Fatigue Job satisfaction Figure 5. Summary of basic and newly developed spatial measures and their correlated behavioral and social relationships. visibility or spatial positioning, which provides predictions of patterns of spatial use. In the field of healthcare facility research, Space Syntax can be adopted to compare design trends, like nurse station typologies, and investigate their effects on social and organizational relationships such as: interactions, knowledge exchange, awareness of peers work, time spent with patient, and medication administration (walking distances and task performing time). Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The author(s) disclosed receipt of the following financial support for the research, authorship,

14 Sadek and Shepley 127 and/or publication of this article: This article was developed while the first author was a visiting scholar at the Center for Health Systems and Design at Texas A&M University. The author would like to express his sincere appreciation to the Egyptian Ministry of Higher Education that funded his stay there. References Al-Sayed, K., Turner, A., Hillier, B., Iida, S., & Penn, A. (2014). Space syntax methodology (4th ed.). London, England: Bartlett School of Architecture, UCL. Bafna, S. (2003). Space syntax: A brief introduction to its logic and analytical techniques. Environment and Behaviour, 35, doi: / Beck, M. P., & Turkienicz, B. (2009). Visibility and permeability: Complementary syntactical attributes of wayfinding. In D. Koch, L. Marcus, & J. Steen (Eds.), Proceedings of the Seventh International Space Syntax Symposium (Vol. 009, pp. 1 7). Stockholm, Sweden: Royal Institute of Technology. Benedikt, M. L. (1979). To take hold of space: Isovists and isovist fields. Environment and Planning B, 6, doi: /b Cai, H., & Zimring, C. (2012). Out of sight, out of reach: Correlating spatial metrics of nurse station typology with nurses communication and coawareness in an intensive care unit. In M. Greene, J. Reyes, & A. Castro (Eds.), Proceedings of the Eighth International Space Syntax Symposium (Vol. 8039, pp. 1 16). Santiago, Chile: PUC. Cai, H., & Zimring, C. (2013). Understanding cultural differences in nursing unit design with the support of space syntax analysis: Are Chinese nursing units designs different from their U.S. counterparts? In Y. O. Kim, H. T. Park, & K. W. Seo (Eds.), Proceedings of the Ninth International Space Syntax Symposium (Vol. 014, pp. 1 24). Seoul, South Korea: Sejong University. Carranza, P. M., Koch, D., & Izaki, A. (2013). SPOT with paths, and interactive diagram with a low complexity isovist algorithm. In Y. O. Kim, H. T. Park, & K. W. Seo (Eds.), Proceedings of the Ninth International Space Syntax Symposium (Vol. 062, pp. 1 13). Seoul, South Korea: Sejong University. Choudhary, R., Bafna, S., Heo, Y., Hendrich, A., & Chow, M. (2010). A predictive model for computing the influence of space layouts on nurses movement in hospital units. Journal of Building Performance Simulation, 3, doi: / Dalton, S., & Dalton, R. (2009). Solutions for visibility, accessibility and signage problems via layered graphs. In D. Koch, L. Marcus, & J. Steen (Eds.), Proceedings of the Seventh International Space Syntax Symposium (Vol. 023, pp. 1 8). Stockholm, Sweden: Royal Institute of Technology. Fouts, M., & Gabay, D. (2008). Healing through evidence-based design. Oncology Issues, Retrieved from issues/articles/mayjune08/fouts.pdf Haq, S., & Luo, Y. (2012). Space syntax in healthcare facilities research: A review. Health Environments Research and Design Journal, 5, Hendrich, A., Chow, M. P., Bafna, S., Choudhary, R., Heo, Y., & Skierczynski, B. A. (2009). Unit related factors that affect nursing time with patients: Spatial analysis of time and motion study. Health Environments Research and Design Journal, 2, Hillier, B. (1996). Space is the machine. Cambridge, England: Cambridge University Press. Hillier, B., & Hanson, J. (1984). The social logic of space. Cambridge, England: Cambridge University Press. Hillier, B., & Iida, S. (2005). Network and psychological effects in urban movement. In A. G. Cohn & D. M. Mark (Eds.), Spatial information theory (lecture notes in computer science) (pp ). Berlin, Germany: Springer. Hillier, B., Burdett, R., Peponis, J., & Penn, A. (1987). Creating life: Or, does architecture determine anything? Architecture & Comportment/ Architecture & Behaviour, 3, Hillier, B., Turner, A., Yang, T., & Park, H. T. (2007). Metric and topo-geometric properties of urban street networks: Some convergences, divergences, and new results. In A. S. Kubat, O. Ertekin, Y. I. Güney, & E. Eyübolou (Eds.), Proceedings of the Sixth International Space Syntax Symposium (Vol. 001, pp ). Istanbul, Turkey: ITU Faculty of Architecture. Huisman, E. R. C. M., Morales, E., van Hoof, J., & Kort, H. S. M. (2012). Healing environment: A review of the impact of physical environmental