A Novel Approach to Locating Community Clinics to Promote Health Care Accessibility and Reduce Health Disparities in Baltimore, Maryland

Abstract

Baltimore, Maryland’s entrenched racial residential segregation renders the city’s world-class medical facilities and services inaccessible to many Black residents living in its most divested neighborhoods. Arguing the need for post-pandemic health care facilities to address health inequities as a practice of care-giving, this article describes a project funded by the National Institutes of Health (NIH) to define a novel, transdisciplinary methodology for identifying ideal vacant sites for conversion into community clinics in Baltimore’s most vulnerable neighborhoods. Positioning architecture as a social determinant of health, this paper suggests ethical and methodological reorientations toward a compassionate approach to clinic design and placement.

• What do we already know about this topic?

• Community clinics have been demonstrated to help reduce disparities in underserved communities, and the recent COVID-19 pandemic further underscored the need for greater attention to temporary and “warm-site” clinics to promote greater health equity and reduce disparities during and beyond times of crisis.

• How does your research contribute to the field?

• We created a layered methodological approach that seeks to go beyond mere service provision and consider as well broader social determinants of health and physical and relational dimensions of health that can affect the efficacy and accessibility of clinics and provided services, both of which can work toward promoting equitable access to care in underserved communities.

• What are your research’s implications toward theory, practice, or policy?

• This article outlines important implications for more context-sensitive and -engaged architectural practice, amplifies a compassionate approach to interventions like community clinics which demand consideration of broader social determinants of health beyond curative medical service provision, and ultimately argues the need for a more nuanced approach to community clinic placement and design that includes attention not only to the health needs of underserved populations but also to historical, social, and other contextual factors that can materially affect access to and provision of needed services and resources.

Introduction

The COVID-19 pandemic is the most recent crisis to make the deep inequities in community health visible to society at large. In drawing attention to these inequities, the pandemic has offered an urgent and immediate example of the link between place, health, and stratified communities, and therefore an impetus for intervention. Baltimore, Maryland, in the Northeastern United States, is an instructive case study of community health and health (in)equity in a post-industrial city.

Using the National Institutes of Health (NIH)-funded project as our foundation, in the following section, we outline a 2-pronged call for meaningful change. The first is to recognize and position architectural intervention as a practice of compassionate caregiving. The second is to make the role of space and place, including the architecture therein, clearly visible as a social determinant of community health and healing. This may be achieved through reexamining and reframing the common practice of site analysis as a method to inform meaningful design and placement of community-serving clinical facilities.

Although we explore the context-specific realities of Baltimore, the implications of this work reverberate beyond the geographical boundaries of the city.

Baltimore: A City of Two Tales

We begin with a map of the population distribution of Baltimore, Maryland, the 30th most populous city in the United States (Figure 1). ¹ Although portions of this map show a mix of colors that suggest a diverse commingling of residents, particularly in the outer regions of the city, the central, urgent story it tells is of the extreme segregation of Black and White residents. Brown ² has described this phenomenon of the “two Baltimores” as the “White L” running through central and southeast Baltimore and the “Black Butterfly” spreading across the east and west parts of the city. These spatial patterns convey the fraught truth of a long history of racial inequity, economic vulnerability, and social precarity layered onto and exacerbated by Baltimore’s extreme residential segregation. ³ Brown explains that this pattern:

Figure 1.

Map of the population distribution of residents by race in Baltimore, Maryland. Source: City of Baltimore, Department of Planning, 2020.

“denotes not only where Black Baltimoreans are geographically clustered but also where capital is denied and structural disadvantages have accumulated due to the lack of capital access. Hence, the Black Butterfly is more than a demographic description; it’s a political, an economic, and a sociocultural description.” ⁴

Brown’s observations underscore an urgent need to look beyond simple abstractions of data to uncover the complicated stories of the social, historical, and economic contexts that define the sites in which we work, observe, and intervene. This deeper, more critical engagement provokes new insights and imperatives for intervention, especially regarding the role of architecture as an agent of health.

First, and most obviously, the map in Figure 1 reminds us that the city’s spatial and architectural contours have been shaped by urban interventions made in both the past and the present. We see, for example, an echo of segregationist policies of the early 1900s and of so-called “redlining” maps used in the 1930s to delineate economically “safe” areas for residential investment, as well as the more recent boundaries of economic investment into residential real estate, which still skew favorably toward low-poverty, largely white areas.^3,5 These regions also suggest distinct architectural realities. A stark example of this is the prevalence of abandoned homes in Baltimore: an estimated 16 000 of which—totaling 8% of all homes—populate the city (Figure 2). However, the neighborhoods of West Baltimore, these percentages are much larger, accounting for 32.5% of homes in the Sandtown-Winchester-Harlem Park neighborhood and 28.5% of homes in Upton/Druid Heights, according to 2019 data.^6-8 In this way, Brown’s “Black Butterfly” describes the outcome of entrenched, long-lasting racist interventions into Baltimore’s built, social, and economic fabric. ⁹

Figure 2.

Abandoned homes in Baltimore number at least 16,000 across the city, with concentrations higher in the neighborhoods of focus in this paper.

Source: Carol M. Highsmith, Public domain, via Wikimedia Commons.

Health Impact of Racial Residential Segregation on Minoritized Communities

We might also see the “Black Butterfly” as an embodied reality. Social epidemiological research has shown that racial residential segregation, which is all too visible in the map, and its accompanying economic and social factors (eg, poverty, racism) are fundamental causes of health disparities. ¹⁰ These social determinants of health disproportionately burden minoritized communities. For Black Baltimoreans, these disparate health outcomes are most visible in the spatial patterning of life expectancy across the city’s various neighborhoods (Figure 3). ¹¹ To take one stark example, in 2018, the average life expectancy in Baltimore was 72.7 years, yet in a predominantly white neighborhood, Cross-Country/Cheswolde (67% white), life expectancy was 84.7 years, and in a predominantly Black neighborhood, Upton/Druid Heights (91.9% Black), life expectancy was just 68.5 years.^12,13 This is a striking example of how, in the United States, a person’s zip code—a neighborhood-specific code used for postal service—is an indicator of health.^13-19

Figure 3.

Life expectancy in Baltimore (data from 2013). Note how Lawrence T. Brown’s “Black butterfly” and “White L” are visible in these patterns.

Source. Baltimore City Health Department.

Many of Baltimore’s neighborhoods are similarly divided by access to health care.^20,21 Recent initiatives have attempted to mitigate these discrepancies. For example, the launch of so-called Health Enterprise Zones (HEZs) in 2013 across the state of Maryland, including in the underserved neighborhoods of West Baltimore, aimed to:

address health and health care disparities among residents who are members of minority groups or have low socioeconomic status living in medically underserved areas by improving their access to care and providing services that improve their health behaviors. ²⁰

Although the HEZ program has shown some positive impacts, ultimately it has not proved sufficient, as investing only in medical services in Baltimore’s underserved communities leaves fundamental drivers of health and inequity largely untouched. This is an example of what public health scholars Rebekah Rollson and Sandro Galea describe as the “mismatch between spending on health care and poor health outcomes [which] can be explained principally by a disinvestment in the underlying forces that shape health” such as “supportive housing, education, early childhood care, public safety, the environment, and public health practice itself.” ¹¹ In other words, any action to intervene in health inequities and to dismantle underlying “spatial racism” must consider broader social determinants of health. ⁴

The Amplification of Architectural Determinants of Community Healthcare During a Pandemic

The outbreak of the SARS-CoV-2 virus in late 2019 and its rapid spread to a global pandemic in 2020 serves as a prism through which to view refracted layers of social and infrastructural vulnerability, as evidenced in the uneven patterning of illness and mortality across urban geographies. In Baltimore, Black residents were disproportionately affected by COVID-19 infections (1.4 times the incidence rate compared to white Baltimoreans) and associated mortality (1.6 times that of white Baltimoreans). ²² This crisis reminds us that breakdowns in health, whether due to chronic or infectious illnesses—either of which can become epidemic—exploit and amplify pre-existing social precarities.

Considering the devastating and discrepant impacts of COVID-19, Sandro Galea, Dean of the School of Public Health at Boston University, called for compassion to be the “animating force behind our thinking about health, and our thinking about how we go about informing the decisions we make” in the future and especially in response to such crises. ²³ Compassion is a deliberate word choice that denotes something deeper and more active than empathy:

Compassion motivates action because the phenomena we observe are unjust, not worthy of the world we would like to live in. . . . Compassion pushes us to understand how we have structured the world, and to ask how we can structure it better. ²³

Importantly, Galea adds that adopting a compassionate practice would require us to “embrace health as a public good.” ²³ To work toward a true, post-pandemic project of city-building, therefore, means to go beyond a “crisis paradigm” and to deliberately intervene in the social determinants of health as a means of “correcting historical injustices in the built environment that [have] contributed to health inequities,” as exposed and amplified by COVID-19.^24,25

Indeed, the systems that facilitate the spread of infectious illness among vulnerable populations go far beyond the capacity of a single pathogen and include economic, social, and, as we argue, architectural determinants of health. All architecture impacts human health in some way, whether directly or indirectly, intentionally or not. In a recent volume, Care and the City, Angelika Gabauer et al., observe how COVID-19 revealed crises in what they term “uncaring systems,” yet the pandemic also signaled the potential for innovation and new practices of community-building and care-giving. ²⁵ The call to action to which our research responds is about recognizing and rebuilding through both looking forward and backward.

Community Clinics as a Means to Intervene in Health Inequities

One historical precedent for intervening in inequitable systems and structures of care is the provision of community-based clinics for underserved communities. Community health centers are spaces that respond to the need for maximally accessible healthcare facilities by providing meaningful physical and relational access to health and social services to medically underserved and disenfranchised populations. These centers came into prominence in the context of the civil rights movement of the 1960s.^26,27 As architect and historian Helen Bronston has noted, the locations of these centers are important, as are their architectural presence and forms, all of which contribute to a practice of care. ²⁷ Research has linked community clinics to a range of positive outcomes, including improved healthcare, reduced discrimination, and empowerment within the served communities.^26,28 These centers have also been instrumental in providing essential services in emergency situations and disaster response. ²⁶ Despite these demonstrable, multi-pronged benefits, community health clinics are underfunded. This has resulted in the decline of these facilities and an accompanying increase in health disparities in urban areas. ²⁶

Elements of this community-based care and the precedent it set have come into renewed focus in the wake of COVID-19. The renowned architectural scholar Thomas Fisher has observed that a number of paradigmatic shifts were prompted by the pandemic, including a move back toward spaces of care that are integrated into local communities, rather than the monolithic teaching hospital institutions that are common in many large cities.^29,30 Community clinics have a history of adapting domestic or commercial spaces, and grocery stores, parking lots, and other vacant, underused, and unused spaces are used as places of care and resource provision. ²⁷ Regarding the latter point, underused and vacant spaces have long been used as “warm sites” (spaces that are unused but ready potential for medical pandemic response) and temporary or periodic healthcare sites.^31,32

As the above examples show, architects have a distinct and important role to play in creating spaces like community clinics that can compensate for inequities in health and healthcare. ³³ In a city like Baltimore, which boasts world-class hospitals yet is plagued by extreme inequities in the provision of care, we observe that this need for community clinics is especially acute. Baltimore is not the only city to experience such a phenomenon, yet it offers compelling evidence that the current healthcare system is not working. We posit that community clinics might be better mobilized to begin addressing these challenges.

Siting Community Clinics in Baltimore

In the beginning of 2021, and in the wake of COVID-19, an interdisciplinary team of researchers set out to investigate and define an innovative process for rapidly renovating vacant buildings located in Baltimore’s most socially vulnerable communities into temporary clinics. This research focused on the HEZ area and the area represented by zip code 21215 in order to identify methods of providing residents with maximally accessible community healthcare facilities and resources, especially during times of public health crises (Figure 4). ³⁴ Sponsored by the NIH, this project represented a methodological challenge for our team, as we worked without precedent and across disciplinary and spatial boundaries. The novelty of our work is expressed in our invention of a 3-pronged methodology to identify and assess possible sites and spaces suitable for clinic placement that includes a transdisciplinary literature review and data mining, site mapping, and site visits. A primary outcome of this work is a process that can be used to identify vacant sites that can be effectively adapted into spaces of community health care and health promotion more broadly, based on an index of neighborhood social vulnerability, various measures of physical and relational proximity, and architectural affordances for a range of care levels. Ultimately, this model can be used to address national, regional, or global crises at a local scale in a way that explicitly intervenes in health inequities in underserved and under-resourced communities.

Figure 4.

Socially vulnerable areas, defined by census tract and identified by our team’s calculation of social vulnerability (on right) within the study area (on left). Image by Andrew Bui, informed by the work of Salman Mohagheghi.

Defining a novel approach to the siting and design of community clinic spaces demanded that our team innovate new, nuanced methods for site selection and analysis. By first recognizing the importance of both the social and the spatial contexts in which community clinics operate, we were able to consider both the geographical as well as the economic and social access to the proposed sites. Although conventional measures of accessibility, including transit lines and physical distance, play important roles in considerations of resource provision, we had to recognize that these measures—like the maps above—obscure important realities in how people navigate cities and urban resources.^35-38 Thus, we contend that the effective siting of community clinics is a critical component of compassionate care-giving.

As a critical part of this work, we confronted the histories and spaces of neglect, discrimination, and disinvestment that have shaped these sites in Baltimore. ⁴ For each site we engaged with in this project, we posed questions about accessibility, use, and health(care) implications, for example:

• What urban qualities affect the accessibility of the site? What might attract someone to this site (vs another)? What might make someone unwilling or unable to access this site? What additional considerations might we account for in the specific context of Baltimore?

• What urban amenities near (within eyesight of or along major transportation paths) the selected site might draw people to a possible clinic at the site? Where are community members most likely to be during the day?

• What architectural qualities make the site suitable for care provision? How might the size, location, and design of the existing building affect service provision? Could the site accommodate needed services? What elements would be beneficial to retain? What might need to be changed?

In asking questions like these, we did not take a single site for granted; rather, as demonstrated below, our methodology demanded that we seek a range of insights in order to argue for or against each site’s suitability as a space for care. Ultimately, these parameters helped the team evaluate potential benefits and drawback of having a proposed health facility in proposed sites.

Methods of Informed Compassion

Selecting suitable sites and vacant buildings for conversion into temporary community health clinics demands a nuanced approach to clinic siting and design and resulted in situating architectural practice within a broader context of healthcare provision in underserved and vulnerable communities. As a result, we see our project as a model for intervening in a system of inequalities and recognizing architecture’s role in the past, present, and future of these dynamics.

To truly contend with architecture’s role as a social determinant of health and to take the field’s commitment to social justice and equity seriously, we argue that all architectural interventions must engage with health and/or broader social inequities in some way, and that even in spaces in which these conversations are already the norm, there is great value in building on current models and practices as a means to intervene in social determinants of health and health inequity. We note 2 methods in particular that can be particularly meaningful to that end, as evidenced by their value to our project in Baltimore, Maryland: mining (locating and collating publicly-accessible city- and tract-level data) and mapping data (plotting all meaningful datapoints on a single, layered map), and visiting sites.

Literature Review and Data Mining to Identify Site Selection Criteria

To identify optimal locations for potential temporary community clinics, we researched and defined an evidence-based, comprehensive list of neighborhood- and block-level urban elements and qualities that could affect the accessibility of temporary community health clinics in the most vulnerable and underserved Baltimore neighborhoods.

We began with a comprehensive literature review that sought to identify: general sources of health promotion; barriers to health and healthcare among residents in urban environments; and resources and other factors that aid or inhibit the accessibility of health-promoting resources and services. To this end, we sought general insights as well as examples that spoke to the specific context and needs of Black Baltimoreans, which we used to contextualize general insights about health outcomes and healthcare access within the specific community neighborhoods on which we focused during our study.

One important contribution of this work was our commitment to a relational view of place, which helped us to see the community we studied in terms of its social, temporal, and geographic context. In other words, our study was not just based on objective points on a map, but on the subjective, intentional beings that inhabit and navigate these places. Indeed, essential in defining and understanding place—and its relationship to health and behavior—is considering the meaning people ascribe to a place, and how they experience and navigate it. These experiential and navigational dimensions of place often differ from commonplace definitions of neighborhoods by expanding the geography that circumscribes residents’ daily activites.^36,37 Thus, critical to this relational perspective was our effort to calculate connections and separations by both physical distance and socio-relational distance. ³⁶ As anthropologist Shannon Mattern has noted, the “particular affective dimensions of urbanites’ spatial practices,” such as the “paths people carve through the city that might not align with official transit routes and pedestrian conduits,” can raise significant questions about definitions of what accessible may mean. ³⁸ We thus considered a range of factors that could make a chosen community clinic site feel more or less socially and spatially proximate to a person’s daily life. This required a novel, epistemological orientation to the concepts of a site and place, as well as greater sensitivity to the criteria and plots that gave life to our mapped sites.

Ultimately, we developed a rank-ordered list of 35 criteria specifying salient community elements across dimensions including physical proximity, relational proximity, safety, esthetics, and sanitation, all of which we saw as vital to success in providing health care to urban African-American communities in Baltimore. Importantly, though the neighborhoods we focused on for this study were circumscribed by official, pre-defined census tracts and the aforementioned Health Enterprise Zones (HEZs), our criteria recognized and valued as well the experiential qualities that might expand or nuance these conceptions of place as residents navigate to and from work, school, and social and cultural spaces that are meaningful to their daily lives.

Three primary themes that came from this list of criteria for assessing the suitability of sites to locate a clinic included:

1. Both physical and relational proximity matter for healthcare access. In terms of physical proximity, a site needs to be physically accessible or connected to surrounding residents. For example, the site needs to be located within 1/8 mile from multiple public transportation stops; preference should also be given to locations near dense housing and educational facilities. ³⁹ Though measures of physical distance and conventional measures of neighborhood zones are instructive, they are necessarily incomplete: in terms of relational proximity, a site needs to feel socially accessible as well and account for flows into surrounding spaces by people who may not live nearby but who may engage with local services or amenities as part of their daily lives.^36,37 For example, for many Black Baltimoreans, the church is an important institution.^40,41 Thus, the presence of or proximity to a church may allow a clinic to be more visible to the community and/or make it feel more accessible due to familiarity with a surrounding area, even if it is not the most physically proximate site to a person’s home or worksite.^36,37,42 In this way, this dimension amplifies the urban amenities that might bring people to or through the areas of focus in our study. Proximity to major transportation lines, for example, would ensure that a site is physically accessible; yet also, locations near major hubs and social institutions like churches or barber shops can help take into consideration what might be drawing people to a given place from not only the surrounding neighborhood but also from other neighboring areas further afield.

2. The physical qualities of a site have aesthetic and health implications. Criteria in this dimension look at the urban fabric at a smaller scale than those in the first dimension: namely, they explore the physical qualities of the site and immediate surrounding area (eg, sidewalks, neighboring buildings and lots) that could impact the use and experience of the site itself, as well as its appeal for renovation and use as a clinic. In other words, given numerous possible sites within a specific neighborhood area, which is the one that is most likely to promote meaningful use? For example, elements that have esthetic value may attract use. In contrast, the presence of excess refuse or inaccessible sidewalks may represent dangers in accessing the site and reduce appeal. Ultimately, we contend that a site should be attractive, but should mitigate any risk of harm.

3. Both perceived and actual safety of a site can affect the likelihood of access. Our model included elements that affect how people tend to assess the safety of a site and/or which may be perceived to be safer, including visibility and lighting. ⁴³ One measure of actual safety was constituted by data from the City of Baltimore. ⁴⁴ This dimension both elaborates the above category and also explores a specific lens that is unique to the areas of focus, namely the prevalence and perception of crime. Though these two elements can differ, both can impact the use and perception of space at a range of scales, from site to neighborhood, as well as the journey to a site or through a certain area. Consistent with our nuanced approach, we wanted to consider yet also go beyond crime data; thus, we kept an eye to urban elements that could shape the ways in which possible users might positively or negatively perceive a site, and if the latter, what might detract from a site’s perceived safety, including lighting, evidence of illegal dumping of refuse, or seclusion from populated areas.

Importantly, in each of these broad categories, we strove to go beyond deterministic, deficit-oriented thinking in order to focus on contextually-specific resources and elements within these neighborhoods that were health-promoting and which recognized residents’ agency in navigating urban systems. In reflecting upon this aspect of our work, one lesson comes in the form of the value added by a local perspective. Of the 6 team members who undertook this research, 4 were residents in Baltimore, and 3 had lived and worked in the areas we studied. This invested our work with a sensitivity to local nuances and ensured the additional resonance of our work with the lived experience of the communities included in the study.

Mapping Health Resources to Identify Potential Care Opportunities

We also undertook a data collection and mapping effort in dialog with the literature review, in which we plotted existing health centers and resources in and around our defined area to understand what resources and services were available. As shown in Figure 5, we began by creating a map via ArcGIS—a cloud based, collaborative map making analysis software used to visualize and analyze data overlaid on maps—of the tracts identified in and earlier phase of this project (Figure 3). Figure 6 shows how we added layers of data to the map to indicate as many health-related facilities and resources as could be identified by our team through multiple data streams. These mapped resources included existing healthcare facilities such as hospitals, clinics and pharmacies, and secondary health promotion facilities such as parks and schools. Where possible, we used publicly-available data to map additional community resources, qualities, and elements that were identified through the literature review as important for healthcare access. These resources and elements included housing, churches, barber shops, and transit lines. Figure 6 also shows all mapped data points relevant to our study, including vacant lots and vacant buildings to map potential sites for intervention; resources that could promote accessibility, such as public housing locations, bus stops and routes, day care facilities, schools, religious institutions, barber shops, beauty salons, pharmacies; and additional information such as Baltimore blue police light locations.^44,45

Figure 5.

Socially vulnerable tract areas from Figure 3 translated into Google Maps. By Andrew Bui.

Figure 6.

Layered data from public sources (Google, City of Baltimore). By Andrew Bui.

Figure 7 shows how we created a connectivity matrix to identify potential block-level sites based on proximities to ideal and nonideal conditions, with our literature review and measures of proximity guiding criteria for selection in these categories.

Figure 7.

Sample map of idea and unideal locations, in tracts 1504 and 1303. Darker blue areas indicate desirable locations; large red circles are undesirable locations. Small dots indicate available sites (abandoned buildings). Image by Andrew Bui.

This 2-pronged approach, which includes both identifying and mapping health-related data in the areas of study, provided immense value to our project. First, this approach allowed us to see factors that can impact residents’ access to and perception of a site with greater nuance and across multiple scales. The layering of data points we employed also helped us to pose broader questions about what factors might impact access and use, as well as why that might be the case. In this way, we were able to offer more tangible translations of general insights from the literature into a local context, in order to ensure maximal accessibility for the specific interventions we envision in the form of community clinics.

Still, there were and are limits to this approach. Firstly, though we aimed to include a range of health-promoting resources and services borne out from our literature review, there are a few notable gaps, including our lack of consideration of LGBTQIA+ services (which are undoubtedly essential health-promoting resources for queer, intersex, and transgender Baltimoreans in these served communities); nor did we include prior investment of community-based initiatives like our own. In future studies, we advocate the inclusion of these factors, as well as any additional health-promoting resources or services relevant to the targeted community. Additionally, though our layered approach was helpful in identifying resource-rich areas that likely already draw in prospective clinic users, such an approach could obscure the quality or meaningfulness of mapped services or resources and/or also preference already overserved areas. These factors are critical to weigh in site selection.

Order-ranking, Studying, and Documenting Shortlisted Sites

Once we determined block-level data to inform site selection for potential clinics, we engaged two processes in order to select possible sites at the building scale. These sites included abandoned homes or buildings such as stores and warehouses that were in or adjacent to residential areas. First, we developed an evidence-based, rank-ordered list of architectural and spatial features that were ideally suited for temporary community health clinics for medically underserved communities in Baltimore. Criteria for selecting these sites included: suitability (or modifiability) of properties’ exterior/interior conditions; site qualities pertaining to accessibility, safety and security, infrastructure and utilities, usability, esthetics, and environmental and natural features; and the condition of, impact on, and connection to immediate surroundings (eg, adjoining uses). These criteria were used, with the creation of a rubric (see Table 1), to help identify specific sites that would be ideally suited for conversion into clinics, which could function at either a temporary or longer-term scale, or, alternatively, as “warm sites” that could be maintained for periodic activation in response to crises or other urgent community needs.

Table 1.

Criteria for Assessing Adaptability of Site for Community Care Provision. By Naomi Hemme.

Area	Theme	Supporting literature	Parameter	Desirable condition	Evaluation criteria	Documentation	Qualitative/quantitative measurement	Preliminary Assessment		Detail Assessment (scale 1-5)					Weighed Rating
										1 minimally qualified, 2 somewhat qualified, 3 qualified, 4 highly qualified, 5 exceed expectations
								Qualifier	Weighting	1	2	3	4	5
Accessibility	Built and social environments	1 (p. 38), 7 (p. 68, 133)	Are there any public transit access points?	Accessible by buses and subway with drop off points within 1/2 mile radius	Number of public transit access points within 0.5 mile radius	Mapping and photo documentation	Quantitative								0
			How accessible is the site for disabled patients? (type of terrain, ADA access)	Condition of typo allows convenient access to site by users.	Sidewalk terrain and topographic features, terrain from public transit access points to site; review ADA guidelines	Mapping and photo documentation	Quantitative								0
			How many residents can it reach?	Ability to reach a minimum of 80% of intended population w/in 5 mile radius	How many residents can it reach within a 1 mile radius? Within 30 min travel time by car? By public transportation?	Publicly available data	Quantitative								0
			Proximity to major streets	Accessible to major throughfares and/or highway with 5 min of vehicular travel.	Define and list major streets within proximity to the site	Mapping and photo documentation									0
			Walkability/bikeability	Dedicated bike lanes and safe, ADA accessible sidewalk	Any bike lanes in the area? Road and sidewalk conditions	Mapping and photo documentation	Both								0
			Visibility of entrances/any separate points of entry for patient/visitor/staff/vender	Signage and entry visible from the street within 100 ft	Number of entrances and how hard/easy it is to access each of these entrances	Photo documentation	Both								0
Safety/Security	Social environment	5 (p. 2-190)	How visibility is the building at street level?	Signage and entry visible from the street within 100 ft	Visibility from street	Photo documentation	Quantitative								0
			Is are around the building well-lit? Any dark corners/areas?	Presence of operational street lights, reasonable safety measures can be implemented, eg, outdoor lights can be installed in otherwise dark areas.	Presence of operating street lights, visual inspection	Photo documentation	Both								0
			Are there any controlled access points	Multiple controlled access points available	Number of controlled access points and the means of controlling access	Photo documentation	Both								0
Support services	Social environment	9	Proximity to hospitals, social services, homeless shelters	10-min travel time to area hospital with ER service; accessible to social services/homeless shelter by public transit or within walking distance.	Number and type of support service within 1 mile radius	Mapping	Quantitative								0

Note. Table 1 Selected criteria for assessing the adaptability of site for community care provision, as excerpted from our rubric. Red cells indicate essential qualities, yellow are desired. Table by Naomi Hemme.

Based on this list of criteria and an initial site visit by one team member, we identified 13 sites that were suitable for a formal, more robust site visit conducted by 5 team members. Figure 8 outlines the formal group site visit itinerary. Figure 9 shows the results of the visits to each site in September of 2021, during which the team took notes and photographs, and documented observations about urban elements, site conditions, and architectural affordances pertaining to the accessibility and safety of the site. Importantly, though only one team member was responsibility for conducting the aforementioned initial site visit, the proposed sites were finalized and discussed with the full team to achieve consensus. Additionally, when the larger, full-team site visit occurred, each of these sites were not seen in isolation: rather, walking to, and around each site allowed us to view additional possible sites within the broader locations we decided on as a team.

Figure 8.

Itinerary for site visit, mapped onto study area. Image by Andrew Bui and Naomi Hemme.

Figure 9.

Graphic representation of site visit observations. Image by Andrew Bui.

Site Visit Revelations and Limitations

These site visits were an especially rich part of our study. Visiting the sites, many of which we walked to on foot, others of which we drove to in a car, allowed us to experience each space in its social as well as its architectural context. We could imagine possible flows to and from the site and its use, envision therapeutic possibilities for architectural features like porches, and assess the community presence at the site. In one example, we noted the presence of a neighborhood organization and met the owner of the identified site, who was the leader of a local community group. Although the consistent reminders of historical neglect were evident at each site, walking through these neighborhoods also allowed us to observe local strengths, and signs of resilience and resourcefulness in the visible presence of maintenance work, investment, and community activity.

The site visits also offered us new, first-hand insights we had not considered through our mapping and literature review methods alone, and they allowed us to test assumptions we had made on paper. For example, Sites 1 and 2 met many of our criteria in terms of accessibility via car and public transit and also boasted close proximity to Druid Hill Park, a local community health resource. However, visiting and observing these sites in person helped us realize that the many steps required to access the building, located on a steep hill above the sidewalk, would be a barrier at the human scale, which was not something easily remedied through architectural intervention. Additionally, we saw through our visits that conventional measures of walkability, often concerned primarily with walking time or distance to resources, would not be sufficient for our work as they did not account for the significant barriers and challenges to accessibility that we observed, such as eroded sidewalks.^46,47 In future work, we would visit multiple times, including at night, and standardize our documentation process to focus on what to notice, what to document and how to record this information, and how to increase reliability and consistency. ⁴⁸

Visiting the sites caused a profound awareness of the scale of urban abandonment in these Baltimore neighborhoods. Guided by both a compassionate approach and committed to a relational view of place, in addition to seeing and acknowledging this fact, we also sought to see beyond the breakdown and deficit to acknowledge the people and potential that comprised these communities. As Sandro Galea explains:

Compassion pushes us to understand how we have structured the world, and to ask how we can structure it better, not because we may suffer but because others are suffering and that is not how the world should be. ²³

In this way, we see a compassionate approach as one that requires a commitment to hope and possibility as much as it does to humanity and justice, action and intervention. Regarding clinic site identification and placement, a compassionate approach considers not just the investment of public health services and infrastructures in underserved communities, but also applies a social equity lens to the consideration of everyday activities, navigational agency, and subjective experience to recognize and advocate maximal physical and relational accessibility of these services. ¹⁸ Recognizing a holistic view and experience of health and by placing clinics in close proximity to valued and used community institutions—churches, schools, barber shops, and others—is one means to recognize and amplify strengths in the community.

Results

Of the 13 sites we visited, the team agreed that 3 were not suitable for conversion into a community clinic due to observed safety concerns, inaccessible entryways that could not be easily remediated through architectural intervention, and unavailability (not abandoned as originally thought). Of the remaining sites, the team came to a consensus on 3 that were most compelling for possible conversion into a community clinic. Importantly, these sites are diverse in size—and thus possible scope of care—and each are defined by unique built and contextual affordances, which would need to be carefully considered by community and caregiving stakeholders in finalizing site selection.

The first site, a 3-story townhouse on a well-kept, 2-lane street connecting Druid Hill Park and North Avenue, was located between an alleyway and a small parking lot, an ideal location due to proximity to housing, a popular café, a major park, a local community organizing space, and because of its ability to expand services into the adjacent open spaces should flex space, distancing, or increased ventilation be necessary in planned and emergent scopes of care. Because of its spatial flexibility (proximity to open space) and closeness to major thoroughfares, this site could potentially serve as a temporary satellite health clinic to support local healthcare infrastructure in the event of a natural disaster or a public health emergency (the adjacent open space could accommodate capacity surges and isolation needs). Its relative proximity to green space and nature allows for added program elements to address post-disaster/traumatic event mental and physical health and wellness. The connection to the local community could allot this site to provide a holistic range of wellness-oriented services, including mental health services and health promotion activities, perhaps even integrating activities in the nearby park. Alternatively, structure integrity and utility infrastructure notwithstanding, the building could serve as a warm site for community-based care which could be activated, expanded, or shifted depending on need.

The second site, a 2-story row home located on a busy, 4-lane road a block from the parking lot of Mondawmin Mall, was an ideal location due to its accessibility for both car and bus travel (a major bus stop is nearby, adjacent to the parking lot), proximity to a neighborhood church, and small front yard space. While visiting that site, our research team noted a vacant storefront directly across the street, which could also be a possible site for all of the reasons noted above, as well as the accessibility of a stairless, ground-floor entry and large commercial space. Similar to the first site, because of the availability of nearby open space (parking lot and an indoor mall), its closeness to public transportation, and its proximity to a well-known landmark (the mall), this site could serve as a temporary health clinic in emergencies where having readily available indoor space is preferable (assuming arrangements were made with the property owner). This site could be considered as a semi-permanent health clinic that serves the community’s seasonal or recurring healthcare needs, for either ambulatory care or potentially for longer stays (though space of the site itself is rather limited and could likely only accommodate a handful of inpatient beds).

Finally, the third site, located in the Penn-North neighborhood, just off of North Avenue, was accessible by car, subway, and bus and located near housing and public services. This site was an abandoned commercial space, likely a warehouse, and boasted not just ample interior space, but also a large parking lot which, like in the first site, could translate to expandable or flexible service provision. The amount of potential indoor open space on this site makes it a good candidate for a temporary health clinic that could service large numbers of visitors while also shielding them from the elements (snow, rain, wind, heat, etc.) with opportunities to expand its service area into the parking lot. For example, this site could be used for service provision, vaccinations, testing, or other ambulatory or outpatient care. That said, in contrast to sites 1 and 2 whose domestic architectural form could easily be adapted to accommodate longer stays on the part of residents served, in this third site, provisions for longer or overnight stays would be difficult or require significant architectural intervention.

Ultimately, as suggested above, the choice of a site—for example, between these 3 locations—would ultimately depend on desired scope of service (these sites are all different sizes) and would need further professional assessment to calculate the cost and scope of architectural intervention for renovation into a clinic. Still, each of these locations are beneficial according to the context-specific physical and relational characteristics defined in our study.

Conclusion

Two clear implications have emerged from our findings. First, we must recognize the impact of place on health and therefore take responsibility for the health consequences—intended or unintended—of architectural interventions. Second, and in service of this ideal, we need what Gabauer et al describe as “more nuanced and spatially sensitive strategies that would pay attention to the particular needs and risks of specific groups.” ²⁵ In other words, the critical and complex engagement with the social, economic, and ecological context of a given site must become a central architectural concern.

The research project upon which this paper is based is a step toward a compassion-based, future-oriented approach to architectural practice in the service of responses to the current COVID-19 pandemic public health crisis. The methodology we propose—in particular its attention to the need to consider the location for temporary clinics placement in addition to more standard decisions about the scope of care—can be instructive for practitioners and policy-makers responding to both urgent health needs and seeking to intervene in broader and more persistent health inequities in the landscape of care in urban environments.

Future directions for this research are twofold. First, it would be valuable to triangulate findings from this research with interviews or focus groups with people who live near and/or access services in close proximity to the identified sites to verify that our own assumptions about and accessibility are accurate. Such focus group interviews could also provide a more robust understanding of community-defined and -perceived health that could potentially add more nuance and texture to the more quantitative measures obtained through local public health records. Second, once sites are confirmed and validated, partnerships with local agencies and institutions could be forged to proceed with the site renovation and clinic operations and, ultimately, post-occupancy evaluations exploring both architectural and care-oriented outcomes. Possible agencies include those involved in the HEZ initiative, local organizations incentivizing and funding the conversion of vacant homes into residences or other community-serving resources, and/or public health agencies planning for the next pandemic (or other public health crisis). Ultimately, it is our hope that this work, its implications and suggested future directions, help to inspire additional research and interventions that take a holistic, equity-focused approach to community-based care in these and other underserved neighborhoods.