INTRODUCTION
The physical environment of healthcare facilities is increasingly recognized as a critical component of care quality, influencing patient safety, psychological well-being, and staff performance [
1]. In mental health settings, where patients may experience prolonged stays and heightened sensitivity to environmental stressors, spatial conditions such as crowding, privacy, and access to daylight may have particularly profound effects on care experiences and recovery processes [
2].
Previous studies suggest that psychiatric ward design, including spatial density, room configuration, sensory stimulation, and privacy-related features, may affect aggression, emotional well-being, perceived control, and dignity among inpatients [
3-
6]. However, existing studies have largely relied on guideline reviews, qualitative evaluations, or stakeholder perceptions, and relatively few have used objective spatial metrics derived from architectural plans [
7-
9].
Architectural floor plans provide a unique opportunity to examine structural characteristics of care environments. By offering measurable indicators of ward density, spatial allocation, and functional organization, plan-based analyses can support systematic assessment of institutional priorities and constraints embedded within mental health care systems [
8].
Korea provides a particularly relevant context for such an investigation. The country has historically relied on long-term inpatient psychiatric care, with a substantial proportion of services delivered by private psychiatric hospitals. Although recent policy initiatives emphasize deinstitutionalization, patient rights, and recovery-oriented care, the extent to which existing inpatient environments align with these goals remains unclear.
This study aimed to examine the spatial characteristics of inpatient psychiatric wards in Korea using nationwide architectural plan data, focusing on ward density, room composition, seclusion room conditions, and ward typologies as structural components of the mental health care system.
METHODS
Study design
This study conducted a cross-sectional analysis of architectural plans collected through a nationwide survey to examine the spatial characteristics of inpatient psychiatric wards in Korea. Architectural plans have been used in prior healthcare research as objective sources for evaluating environmental conditions and organizational characteristics of care settings. Spatial indicators derived from floor plans can serve as proxy measures for institutional priorities and constraints embedded within healthcare systems [
10].
The analysis focused on ward-level physical configurations as a structural component of the mental health care system, rather than on individual patient outcomes. Architectural floor plans were used as objective data sources to assess ward density, spatial allocation per patient, room composition, seclusion room characteristics, and ward typologies across different types of psychiatric medical institutions.
Because this study used only architectural floor plans and did not include human participants, medical records, identifiable personal information, or other human data, IRB review and approval were not required.
Data collection and sample selection
According to the Health Insurance Review and Assessment Service registry, 408 psychiatric medical institutions with inpatient psychiatric wards were operating in Korea as of June 2024. As part of a nationwide data collection initiative, architectural floor plans were sought from all eligible institutions through local public health authorities. Of the 408 institutions, 172 (42.2%) submitted one or more ward-level floor plans. Because completely comparable information was not available for the 236 nonparticipating institutions, potential differences between participating and nonparticipating institutions could not be fully assessed.
The submitted plans covered all institutional categories legally permitted to provide inpatient psychiatric care in Korea, including general hospitals with psychiatric wards, including tertiary hospitals, psychiatric hospitals, hospitals, Korean medicine hospitals, and psychiatric clinics. The submitted materials were screened in two sequential stages. First, the 172 institutional submissions were assessed for minimum suitability for spatial identification according to the following criteria: 1) adequate resolution and legibility; 2) clear delineation of inpatient ward boundaries; and 3) identifiable functional spaces, including patient rooms, seclusion rooms, nursing stations, patient activity areas, shared bathrooms, and circulation spaces. A total of 132 institutions met these criteria.
Second, the submissions from these 132 institutions were further assessed for sufficient dimensional information and clearly defined spatial relationships to permit quantitative analysis. Of these, 111 institutions met the second-stage criteria and were included in the final analytic sample. The final sample comprised 383 inpatient psychiatric wards and 20,608 beds, representing 64.5% of the institutions that submitted plans and 27.2% of all eligible institutions nationwide.
Exclusions at both screening stages were based on prespecified technical criteria related to the quality and completeness of the submitted documentation. Although the final sample was not a probability sample and cannot be considered nationally representative, its inclusion of all provider categories and its substantial number of wards and beds provided broad typological coverage for descriptive analyses and comparisons across hospital types.
Variables and spatial indicators
Spatial variables were derived from architectural drawings using standardized measurement and classification procedures. The following key indicators were analyzed:
Ward size and bed capacity
• Gross floor area of each inpatient ward (m²) and number of beds per ward.
Floor area per bed
Floor area per bed (m²/bed) was calculated by dividing the total ward floor area by the number of beds within the ward. This indicator was used as a proxy measure for spatial density and patient-level space allocation, which are closely related to privacy, safety, and quality of inpatient environments.
Room type composition
Patient rooms were classified by bed count into the following categories:
• Single-bed rooms, two-bed rooms, three-bed rooms, four-bed rooms, five-bed rooms, and rooms with six or more beds.
For each occupancy category, we calculated the number of beds located in that room type. This approach estimated the proportion of inpatient bed capacity exposed to high-occupancy shared-room environments.
Seclusion room characteristics
Seclusion rooms were analyzed with respect to:
• Net floor area (m²), proximity to nursing stations, adjacency to public circulation spaces, presence of windows, and presence of in-room toilets.
Variation in seclusion room size was examined to assess the consistency of spatial standards across institution types.
Ward typology
Ward layouts were classified according to corridor configuration, which has been shown to influence patient behavior and environmental quality in long-term care settings. The typology categories included:
• Double-loaded corridor, central hall type, double-loaded hall type, dual double-loaded corridor, central circulation type, single-loaded corridor, and mixed corridor type.
Classification criteria
Institutions were stratified for comparative analysis using the following criteria:
• For descriptive tables, institutions were reported using their original legal categories. For inferential analyses requiring adequate cell sizes, hospitals and traditional Korean medicine hospitals were combined with psychiatric hospitals, resulting in three analytic categories: general hospitals, psychiatric hospital-level institutions, and clinics.
This stratification allowed examination of structural disparities related to institutional role and care-delivery context.
Data analysis
All measurements were derived directly from architectural floor plans using consistent scaling methods. Descriptive statistical analyses were conducted to summarize spatial characteristics across wards and institutions.
Comparative analyses were performed to identify differences in spatial indicators by hospital type. Results are presented using means, medians, ranges, and proportional distributions. Differences in floor area per bed across hospital types were examined using one-way analysis of variance, followed by Tukey’s honestly significant difference (HSD) test for post-hoc pairwise comparisons. This analysis was restricted to wards with complete data on ward floor area and bed capacity (n=383 ward units).
RESULTS
A total of 391 ward units from 111 psychiatric institutions were initially reviewed. After excluding cases with missing or unidentifiable spatial data (e.g., “unidentifiable” or “no dimensions”), 383 units were included in the final analysis. For three-group comparisons, the analytic categories comprised general-hospital wards (n=25), psychiatric hospital-level wards (n=339), and clinic wards (n=19).
Ward density and floor area per bed
Analysis of ward-level floor plans revealed substantial variation in ward density across institution types. Wards with similar gross floor areas often accommodated markedly different numbers of beds, indicating divergent spatial allocation strategies. Across the 383 analyzed wards, the mean floor area per bed was 14.0±6.2 m²/bed, with significant differences observed across hospital types (F=41.25, p<0.001). General hospital wards showed the largest floor area per bed (28.63±8.41 m²/bed, n=25), followed by psychiatric hospitals (13.08±3.25 m²/bed, n=338) and clinics (11.75±2.12 m²/bed, n=19). Post-hoc Tukey HSD tests showed that general hospital wards had significantly larger floor area per bed than psychiatric hospitals and clinics (p<0.001), while the latter two groups did not differ significantly. Psychiatric hospitals therefore operated with less than half the per-bed floor area observed in general hospitals (
Figure 1).
In wards with comparable floor areas of approximately 900-1,100 m², bed capacity ranged from 20-30 beds in general hospitals to more than 100 beds in psychiatric hospitals. This resulted in pronounced differences in floor area per bed (
Supplementary Figure 1 for representative ward layouts). These findings indicate that inpatient psychiatric wards in Korea are generally characterized by high bed density and limited patient-level spatial allocation, particularly in psychiatric hospitals and clinics.
To examine whether spatial conditions differed by institutional year of establishment, we conducted an additional descriptive analysis of floor area per bed by hospital type. Institutions for which the year of establishment could not be ascertained because institutional identity was not disclosed were excluded from this analysis, yielding an analytic sample of 91 institutions. As shown in
Figure 2, general hospitals with psychiatric wards maintained relatively high floor area per bed across establishment periods, ranging from 27.93 to 31.17 m²/bed. In contrast, psychiatric hospitals showed lower floor area per bed overall, with mean values decreasing from 14.68 m²/bed among institutions established before 1990 to 11.80 m²/bed among those established in 2010-2019. These descriptive patterns indicate that floor area per bed varied by both hospital type and establishment period.
Room composition and multi-bed room dominance
Across all hospital types, inpatient bed capacity was heavily concentrated in shared rooms. As shown in
Table 1 and
Figure 3, beds in rooms accommodating four or more patients accounted for 89.8% of measured inpatient bed capacity overall, whereas beds in single-bed rooms accounted for only 0.7%. This pattern was observed across all institution types: beds in rooms accommodating four or more patients accounted for 87.0% of inpatient bed capacity in general hospitals, 90.0% in psychiatric hospitals, and 84.5% in clinics. In psychiatric hospitals, where this pattern was particularly marked, beds in five-bed and six-or-more-bed rooms together accounted for more than 60% of measured inpatient bed capacity, with beds in six-or-more-bed rooms alone accounting for 45.3%. These findings indicate that inpatient psychiatric care in Korea is predominantly delivered in high-occupancy shared-room environments with limited availability of single-bed rooms.
Seclusion room size and environmental characteristics
Across the final analytic sample, 794 seclusion rooms were identified among 20,608 inpatient beds, corresponding to approximately one seclusion room per 26.0 beds, or 3.9 seclusion rooms per 100 beds. This overall figure may obscure variation across individual institutions and should not be interpreted as indicating compliance by every institution.
Area measurements were available for 597 seclusion rooms in general hospitals, psychiatric hospitals, and clinics. Seclusion room size varied widely, ranging from 3.0 to 22.95 m², representing more than a sevenfold difference (
Table 2).
Mean seclusion room size differed substantially by institution type. General hospitals had the largest average seclusion room area (12.27 m²), followed by psychiatric hospitals (7.61 m²) and clinics (4.66 m²). Seclusion rooms in psychiatric hospitals averaged approximately 62% of the size observed in general hospitals. Seclusion room size varied substantially across hospital types (
Figure 4).
Location and feature data were available for 704 seclusion rooms. Overall, 75.9% were located adjacent to nursing stations, facilitating staff observation. However, 24.1% were accessed directly from public circulation spaces, potentially limiting privacy.
Regarding interior conditions, 55.4% of seclusion rooms had windows, meaning that 44.6% lacked windows; only 13.1% included an in-room toilet. These features were not consistently associated with institution type, indicating the absence of standardized environmental criteria for seclusion rooms.
Table 3 summarizes key spatial characteristics of seclusion rooms, including proximity to nursing stations, access routes, and internal environmental features.
Ward typology and corridor configuration
Table 4 summarizes the distribution of ward corridor typologies across hospital types, highlighting the predominance of double-loaded corridor layouts. A total of 383 ward layouts were classified according to corridor configuration. The double-loaded corridor typology was overwhelmingly dominant, accounting for 83.6% of all analyzed wards.
Alternative typologies were relatively rare. Single-loaded corridor layouts, central circulation types, and mixed configurations collectively accounted for less than 5% of all wards. Central hall and hybrid corridor-hall types were observed in a limited number of facilities.
The dominance of double-loaded corridor layouts was consistent across hospital types, reflecting the structural persistence of ward configurations that may limit access to daylight, external views, and environmental variation within circulation areas. Representative floor plans illustrating the six corridor typology categories identified in the analysis are shown in
Figure 5.
DISCUSSION
Principal findings
This nationwide architectural plan analysis provides empirical evidence of structural constraints embedded in inpatient psychiatric care in Korea. The findings indicate that psychiatric wards are characterized by high bed density, limited floor area per patient, and a predominance of large shared-room environments, alongside substantial variability and inconsistency in seclusion room standards. In addition, the overwhelming reliance on double-loaded corridor layouts suggests limited access to therapeutic environmental features such as daylight, external views, and spatial differentiation. Taken together, these results suggest that the physical environments of inpatient psychiatric care in Korea remain largely oriented toward capacity maximization and operational efficiency, rather than patient-centered, rights-based, or recovery-oriented care models.
Ward density and spatial allocation as system-level constraints
One of the most striking findings of this study is the pronounced disparity in floor area per bed between general hospitals and psychiatric hospitals. Psychiatric hospitals operated with less than half the per-bed floor area observed in general hospitals, despite serving a patient population that may experience longer lengths of stay and greater sensitivity to environmental stressors.
This disparity should not be interpreted solely as an architectural shortcoming. Rather, it reflects system-level constraints shaped by reimbursement mechanisms, staffing structures, and the historical role of psychiatric hospitals within Korea’s mental health care system. High bed density may function as a compensatory strategy in settings where financial margins are narrow and staffing costs are relatively fixed.
Consequently, spatial crowding in psychiatric wards is not merely a design issue but a structural outcome of how inpatient psychiatric services are financed and organized. Any effort to improve spatial standards without addressing these underlying systemic factors may therefore face significant implementation barriers.
The Medical Aid reimbursement system provides an important context for interpreting these high-density ward environments. In Korea, many psychiatric inpatients are Medical Aid beneficiaries, and inpatient psychiatric care for this group has historically been reimbursed through a relatively low fixed per-diem payment system. This payment structure may limit the capacity of psychiatric institutions, particularly those serving a high proportion of Medical Aid patients, to recover the costs of environmental improvements such as reducing room occupancy, expanding therapeutic communal spaces, and improving ward-based staffing and therapeutic activities. Therefore, the high-density ward configurations observed in this study should not be interpreted solely as architectural design choices, but also in relation to broader financing structures and socioeconomic inequality.
The analysis by institutional year of establishment provides further support for this interpretation. General hospitals with psychiatric wards maintained relatively high floor area per bed regardless of establishment period, whereas psychiatric hospitals did not show comparable improvement among more recently established institutions. Rather, psychiatric hospitals established in later periods showed lower mean floor area per bed than those established before 1990. This pattern should be interpreted cautiously, because institutional year of establishment does not necessarily correspond to the construction or renovation year of each psychiatric ward. Nevertheless, the findings suggest that improvements in general hospital facility standards may not have been accompanied by equivalent spatial improvements in psychiatric hospitals. This discrepancy may reflect the weaker regulatory and financial incentives for psychiatric hospitals to expand floor area, reduce bed density, and redesign wards around therapeutic environmental quality.
Multi-bedroom dominance and implications for dignity and privacy
International research on hospital design has consistently highlighted the limitations of large multi-bed rooms, noting their associations with reduced privacy, heightened sensory stimulation, and challenges in infection control and staff workflow [
11]. Nevertheless, the present findings indicate that large shared rooms remain the dominant configuration across inpatient psychiatric facilities in Korea.
The analysis showed that beds in rooms accommodating five or more patients accounted for more than 60% of measured inpatient bed capacity, whereas beds in single-bed rooms were exceedingly rare. This configuration contrasts with contemporary international design guidance for psychiatric inpatient units, which generally favors single en-suite bedrooms, smaller bed clusters, and the elimination or minimization of shared sleeping accommodation [
3,
12].
Large shared rooms may facilitate observation and reduce staffing demands, but they also limit personal space, compromise privacy, and restrict patients’ ability to regulate sensory input and social interaction. In long-term inpatient contexts, such conditions may contribute to heightened stress, reduced autonomy, and diminished dignity.
Although this study did not assess clinical outcomes directly, the spatial patterns observed suggest a misalignment between current ward configurations and contemporary mental health care principles that emphasize recovery, autonomy, and human rights. Importantly, the persistence of high-occupancy shared rooms across institution types indicates that this pattern is systemic rather than exceptional.
Seclusion rooms: variability, ambiguity, and lack of standardization
Seclusion rooms represent one of the most ethically sensitive spaces within inpatient psychiatric care. Previous studies and practice guidelines emphasize that seclusion should be minimized and used only as a last-resort intervention after de-escalation has failed, rather than as a routine management strategy [
13]. From this perspective, the physical characteristics of seclusion rooms are not neutral; they may actively shape patient experiences during periods of acute distress.
International human rights frameworks further emphasize that the physical conditions of mental health facilities are integral to protecting dignity and preventing degrading treatment. The World Health Organization’s QualityRights initiative underscores the need for environments that support safety, privacy, and recovery while discouraging reliance on coercive practices [
14].
The present analysis identified extreme variability in seclusion room size and environmental conditions, even among institutions serving similar functions. In the analyzed sample, psychiatric hospitals accounted for most identified seclusion rooms and provided substantially smaller seclusion spaces than general hospitals. In addition, 44.6% of seclusion rooms lacked windows, and only a small minority included in-room toilets. Such variability suggests the absence of clear, enforceable standards regarding the minimum spatial and environmental conditions required for safe and dignified seclusion.
These findings raise important questions about how seclusion is conceptualized within the mental health system. Without consistent spatial criteria, seclusion rooms risk functioning primarily as containment spaces rather than as environments designed to support de-escalation and patient safety. Establishing clearer spatial and environmental benchmarks may therefore be a necessary component of broader efforts to reduce coercive practices.
Persistence of corridor-dominated ward typologies
Previous research in long-term care settings has demonstrated that corridor configuration is closely associated with resident behavior and environmental experience. In a study of dementia care facilities, Choi and Yang [
4] reported that single-loaded corridor layouts and corridors with visual connections to outdoor spaces were associated with higher frequencies of positive resident behaviors, whereas double-loaded corridor layouts were linked to reduced activity and environmental engagement.
Although conducted in dementia care facilities, these findings suggest that corridor typology may also be relevant to psychiatric inpatient settings, where long lengths of stay and heightened sensitivity to environmental stimuli are common. The overwhelming dominance of double-loaded corridor layouts observed in the present study may therefore represent a structural constraint on therapeutic environmental quality.
The dominance of double-loaded corridor layouts across nearly all analyzed wards highlights the structural inertia of inpatient psychiatric environments. Such layouts are efficient in terms of circulation and supervision but may limit access to daylight, external views, and spatial variation within corridors and shared circulation areas.
Although alternative typologies, such as single-loaded corridors or courtyard-based layouts, were observed, they accounted for only a small fraction of wards. This suggests that existing building stock and incremental renovation practices constrain the feasibility of more therapeutic spatial configurations.
The findings imply that meaningful environmental improvements may require not only renovation at the ward level but also long-term strategies for building replacement, ward downsizing, or service reconfiguration aligned with broader mental health system reforms.
Policy implications: aligning facility standards with workforce and reimbursement reform
The findings of this study have important policy implications. Because this analysis was based on architectural floor plans, it directly demonstrates the structural conditions under which inpatient psychiatric care is delivered in Korea. However, the policy response should not be limited to architectural regulation alone. The observed high bed density, limited personal space, predominance of large multi-bed rooms, and inconsistent seclusion room environments should be understood as outcomes of the interaction between facility standards, staffing requirements, reimbursement mechanisms, and the historical dependence of the Korean mental health system on long-term inpatient care.
First, strengthening facility standards should be accompanied by workforce reform. In high-density psychiatric wards, even well-designed spaces may fail to function therapeutically if staffing levels are insufficient to support continuous engagement, individualized de-escalation, and recovery-oriented psychosocial care. Conversely, overcrowded wards managed with limited staff may increase tension between patients and staff and may reinforce a custodial model of care in which observation and behavioral control replace therapeutic interaction. Therefore, human rights-oriented environmental reform should not be framed solely as a matter of floor area, room occupancy, or architectural fixtures. It should be implemented as part of an integrated policy package that includes improved nursing and mental health professional staffing standards.
Second, facility improvement requires a financially sustainable reimbursement framework. Converting large multi-bedrooms into single- or double-bed rooms, expanding therapeutic common areas, installing anti-ligature fixtures, and improving seclusion room environments all require capital investment and may increase operating costs by reducing the number of beds accommodated within a given floor area. This issue is particularly relevant to psychiatric institutions serving a high proportion of Medical Aid beneficiaries, for whom inpatient psychiatric care has historically been reimbursed through a relatively low fixed per-diem payment system. Such payment structures may unintentionally encourage institutions to maintain high bed density, minimize staffing costs, and underinvest in psychosocial treatment programs. Accordingly, raising environmental and human rights standards without corresponding reimbursement reform may lead to uneven implementation, financial pressure on private psychiatric hospitals, or regulatory avoidance rather than meaningful improvement in patient care.
Finally, reform should be gradual and system-aware. Immediate uniform enforcement of high spatial standards may be difficult for older psychiatric hospitals and could risk reducing necessary inpatient capacity before community-based alternatives are sufficiently established. A staged approach may therefore be more feasible: minimum standards for safety and dignity should be established first, followed by progressive targets for room occupancy, floor area per bed, seclusion room conditions, and therapeutic communal space. Such an approach would transform facility standards from a simple regulatory requirement into a policy lever for improving the quality, safety, and human rights orientation of inpatient psychiatric care [
15].
Taken together, the present findings suggest that improving psychiatric inpatient environments in Korea requires more than renovation of individual wards. Spatial reform must be integrated with reimbursement reform, workforce development, and broader service-delivery redesign. Only when facility standards, staffing capacity, and financial incentives are aligned can psychiatric inpatient care move from a high-density custodial model toward a rights-based and recovery-oriented model of care [
16].
Limitations
This study has several limitations. First, the final analytic sample was limited by institutional participation and the quality of submitted documentation. Although floor plans were sought from all 408 eligible psychiatric medical institutions with inpatient psychiatric wards nationwide, only 172 institutions submitted floor plans, and 111 institutions provided documentation of sufficient quality for quantitative analysis. Nonparticipating institutions could not be fully characterized, and institutions with inadequate or incomplete floor plans may have differed from included institutions in ways not captured by the available data. Therefore, selection bias cannot be excluded, and the findings should not be interpreted as statistically representative of all psychiatric inpatient facilities in Korea. Second, the analysis was based on architectural floor plans and did not capture real-time occupancy levels, staffing ratios, or patient experiences. Third, clinical outcomes such as symptom improvement, length of stay, use of restraint, or use of seclusion were not assessed. Forth, while ward typologies were classified systematically, qualitative aspects of environmental experience could not be fully evaluated. Fifth, psychiatric safety features such as ligature risk, acoustic conditions, observation blind spots, door hardware, material durability, and staff-patient interaction patterns could not be assessed from floor plans alone. Despite these limitations, the use of nationwide architectural plan data provides rare objective insight into the structural conditions of inpatient psychiatric care and offers a foundation for future research integrating spatial, operational, patient-experience, and clinical dimensions.
Conclusion
Inpatient psychiatric wards in Korea operate under structural conditions characterized by high density, limited personal space, and inconsistent seclusion room standards. These conditions reflect systemic features of the mental health care system rather than isolated architectural choices. Incremental, system-aware reforms that integrate spatial standards with financing and service delivery are needed to support rights-based and recovery-oriented inpatient care.