The Role of the Built Environment as a Therapeutic Intervention in Mental Health Facilities: A Systematic Literature Review

Abstract

Objectives:

This systematic literature review synthesizes and assesses empirical research concerning the use of the built environment as a therapeutic intervention in adult mental health inpatient facilities. The review explores the impact of facility design on patient outcomes.

Background:

There is a growing recognition that the built environment in mental health facilities must strike a balance between ensuring safety and providing a therapeutic atmosphere. A review addressing how facility design contributes to this therapeutic environment is warranted.

Methods:

Database searches were conducted in CINAHL, Embase, PsychInfo, PubMed, and Web of Science from inception up to March 10, 2022. The Scottish Intercollegiate Guidelines Network (SIGN50) critical appraisal checklists were used to assess the quality of included studies.

Results:

Of the 44 peer-reviewed studies identified from nine countries, several factors emerged as vital for the therapeutic environment in mental health inpatient facilities. These included personal spaces prioritizing privacy and control of the environment, daylight-optimized spaces, versatile communal areas promoting activities and interaction, designated areas for visits and spiritual/contemplative reflection, homelike environments, the inclusion of artwork in units, open nursing stations, and dedicated female-only areas. Yet, there is a need for research yielding stronger evidence-based designs harmonizing with therapeutic needs.

Conclusion:

This review offers initial guidance on designing mental health facilities that foster a therapeutic environment, while highlighting that the influence of facility design on mental health inpatients is considerably under-researched.

Introduction

The significance of the physical environment in healthcare settings cannot be overstated. Over the years, its role in enhancing the quality of healthcare delivery has been researched and is now better understood. Several literature reviews have shown that well-designed physical healthcare environments can positively impact patient health, well-being, and safety (Dijkstra et al., 2006; Huisman et al., 2012; Jamshidi et al., 2020; Ulrich et al., 2008). Among such contributions, Ulrich et al. (2008) identified several design factors that play a pivotal role in enhancing patient outcomes, including single-bedded rooms, access to daylight, ergonomic designs, and appropriate acoustic environment and ventilation systems.

However, when considering specialized areas like mental health settings, the design conversation takes on different dimensions. Historically, the design of these facilities has been primarily safety-oriented, leading to the practice of “safe wards”—a combination of physical structures and operational protocols intended to minimize the risk of self-harm and suicide. However, tThere is a growing recognition that the design of mental health facilities should balance safety features with therapeutic environments to promote well-being and recovery (Chrysikou, 2019; Lundin, 2021; Oostermeijer et al., 2021; Ulrich et al., 2018).

There is a growing recognition that the design of mental health facilities should balance safety features with therapeutic environments to promote well-being and recovery.

To contribute to this evolving discourse, this systematic literature review aimed to identify and synthesize empirical research on the use of the built environment as a therapeutic intervention in mental health inpatient facilities. Thus, the question guiding this review was as follows: What design factors of adult mental health inpatient facilities influence patient outcomes and satisfaction? This review covered a wide range of patient outcomes, such as individual experiences (e.g., emotional states, feelings of safety, and overall well-being), behavioral responses (e.g., social interaction), and direct clinical interventions (e.g., medication intake and length of stay). Additionally, the review considered safety outcomes that can indirectly be influenced by the therapeutic environment, including aggression, seclusion, self-harm, and suicide. Outcomes directly related to the physical safety of patients (i.e., falls, infections) were excluded.

Several literature reviews have made notable contributions on exploring the impact of mental health inpatient facilities on social interaction (Jovanović et al., 2020) and on seclusion and restraint (Oostermeijer et al., 2021). A literature review by Weber et al. (2022) synthesized research on relationships between physical and socio-physical environmental characteristics of inpatient mental health facilities and their effect on patient mental health, well-being, and recovery outcomes. Our review aimed to expand on this knowledge by adopting a different perspective. Unlike Weber et al. (2022), who concentrated primarily on outcomes, our review synthesized evidence based on design factors. Additionally, we included forensic facilities and Intensive Psychiatric Care Units (IPCU) within our scope. In doing so, we aimed to encompass a broader range of design influences, deepening our understanding of their effects on patient outcomes. Another distinguishing aspect of our review laid in its approach to recognizing the complex nature of research that examines the impacts of facility renovation and relocation. In such scenarios, patients experience multiple design changes, making it challenging to associate specific design interventions to specific patient outcomes due to the number of alterations introduced. This approach diverges from Weber’s methodology, which established connections between outcomes and specific design factors that were part of a multifactorial intervention involving multiple design changes. Our review took a novel approach by dedicating a section to examining these scenarios, rather than isolating specific design factors of a multifactorial intervention and associating them to specific outcomes. In the discussion and implications for practice, we identified the discernible trends and shared elements across these multifactorial interventional studies. This approach aligns with the methodology employed by Dijkstra et al. (2006) in their earlier review and highlights the continuity in research direction.

Through our review, we aimed to enhance the discourse on the role of design in mental health inpatient facilities, offering the evidence base to inform future design practices. Furthermore, this review will support the development of a quality assessment tool being produced by the Scottish Government and NHSScotland Assure, aimed at assessing the safety and therapeutic quality of mental health estates.

Method

This review was commissioned by the Scottish Government and NHSScotland Assure, who provided the research team with the review question and inclusion criteria. The review was conducted by a multidisciplinary team of two healthcare scientists, a research manager, and an architect, all with a background in the healthcare-built environment and research.

Search Strategy

Searches were conducted in five electronic databases, including CINAHL (Ebsco), Embase (Ovid), PsycINFO (Ovid), PubMed, and Web of Science from their inception up to March 10, 2022. The reference lists of all included studies were hand searched for potentially relevant articles. In addition, extensive handpicking was informed by the reference lists from studies that were identified but did not meet the inclusion criteria (e.g., systematic literature reviews and critical reviews). Search strategies were developed by the lead author (L.R.L.) and adapted to each database using a comprehensive list of search terms related to the setting, population, and intervention or phenomenon of interest. The search strategy was guided by the principal architect (S.G.) and reviewed by a librarian for accuracy (Supplementary File 1).

Eligibility Criteria

For studies to be included in the review, they had to meet the following criteria:

• Investigate the effect of the facility design (architecture, interior design, and ambient features) on patient outcomes. This may include clinical, behavioral, emotional/well-being, or safety outcomes that can indirectly be influenced by the therapeutic environment. Studies that report on patient experiences or perceptions of the facility design, or staff insights on the impact of the design on patient outcomes.

• The study is conducted in adult mental health inpatient facilities including acute and long-stay services, psychiatric intensive care units, and forensic units.

• The study consists of empirical research that has been peer-reviewed and is written in English.

No restriction was placed on the type of study design and country. Studies were excluded if they focused on child, adolescent, geriatric, or emergency departments or were conducted exclusively with populations diagnosed with organic mental illnesses. This is because the balance of benefits, constraints, and challenges associated with these populations might differ from the target population of this review. However, if these populations and settings were part of a more extensive sample that met our inclusion criteria, they were considered for inclusion. Studies that exclusively focused on the design of smoking areas, sensory rooms, or general inpatient settings were also excluded. Moreover, studies were ruled out if they were based on expert opinion, literature reviews, commentaries, audits, and dissertations or if they were presented as posters or abstracts at conferences.

Study Selection

A first screening (reading the title and abstract) of all the studies and a second screening (assessing the full text of eligible studies) was conducted independently by two reviewers (L.R.L. and J.K.) based on the inclusion and exclusion criteria. Disagreement between the reviewers occurred 0.05% in the first screening (n = 32 studies) and 29.09% in the second screening (n = 15 studies). A consensus meeting between reviewers resolved disagreements. A third reviewer (G.O.) made the final decision when a consensus could not be reached. For this literature review, we used the Scottish Intercollegiate Guidelines Network’s SIGN50 checklists to assess the methodological quality of the studies and adopted their evidence grading system (SIGN, 2019). However, in applying this methodology, we noticed its lack of specific criteria for qualitative and controlled pre- and post- intervention studies. Recognizing the indispensable value of these study designs in our context, we adapted the SIGN50 hierarchy to integrate qualitative research as Level 3 evidence and controlled pre- and post-intervention studies as Level 2 evidence. Adjustments to the evidence hierarchy are detailed in Supplementary File 2.

Data Extraction, Synthesis, and Integration

Data from the included studies were extracted by the lead author and included, where available, the authors, year, country, study design, sample population, number and characteristics of participants, type and characteristics of the inpatient facility, phenomenon of interest or intervention, data collection method(s), outcome measures, results, limitations identified by the authors, sources of funding, and design factor investigated or reported in the study.

Following the Joanna Briggs Institute methodology for mixed-methods review, the extracted quantitative data were transformed into qualitized data to facilitate integration with data extracted from qualitative studies (Lizarondo et al., 2020). Data were categorized based on design factors, followed by similarities in the outcomes, to produce a set of integrated findings. The criteria for incorporating findings into the synthesis included either the presence of multiple studies or a single in-depth study that offered comprehensive data. The confidence in the body of evidence that formed each finding was assessed based on the quantity of the studies, the SIGN50 level of evidence assigned to each study, the study limitations, the consistency across the studies, and the generalization of the finding. The certainty in the body of evidence for each finding was graded as either very low, low, moderate, or high.

Results

The final selection consists of 44 studies (see Figure 1), which include 19 quasi-experimental studies—of which only five involved a control group—, two randomized control trials, three retrospective cohort studies, three cross-sectional studies, two post-evaluations, and 15 qualitative studies. An overview of the characteristics of the included studies can be found in Table 1. An expanded version of the study characteristics can be found in Supplementary File 3.

Figure 1.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart for selection process of studies in the systematic review.

Table 1.

Characteristics of the Included Studies.

Study/Country	Study Design	Participants and Patient Diagnoses	Methods	Design Feature	Outcome	Brief Findings	Evidence Grade
Baldwin, 1985 (United Kingdom)	Mixed methods Controlled pre–post intervention	Patients (number unreported) Staff (number unreported) Individuals with severe intellectual disabilities, psychogeriatric patients, and those with various psychiatric disorders	Medication rates Seclusion rates Casual incidents Points earned daily for good behavior Staff and residents' perception of the ward	Furniture	Seclusion Aggression Casualty incidents Social interaction Medication intake	Rearranging furniture into a group setting reduced seclusion hours, improved social interactions as perceived by nurses, and resulted in fewer casualty incidents; there were no changes in medication rates.	2+
Beauchemin & Hays,1996 (Canada)	Retrospective cohort study	Patients (n = 174) Major depressive d. single/recurrent, bipolar, depression Not Otherwise Specified (N.O.S.)	Admissions records (2 years)	Daylight	Length of stay	Sunny rooms reduced the length of stay by 15%.	2−
Benedetti et al., 2001 (Italy)	Retrospective cohort study	Nonpsychotic depression patients consecutively admitted over a period of 3 years (Patients stating in eastern rooms, n = 290 unipolar, n = 113 bipolar; Patients stating in western rooms n = 125 unipolar, n = 74 bipolar)	Records (All admissions for nonpsychotic depression over a 3-year period; length of stay)	Daylight	Length of stay	Morning-exposed eastern rooms reduced bipolar patients’ hospital stay by 3.67 days. No effect in unipolar patients.	2−
Borckardt et al., 2011 (United Sates)	Quantitative Pre–post intervention	Patients (n = 446) Staff (n = 340) Diagnoses unreported	Records of seclusion and restrain Quality of care forms Observations	Interior design	Seclusion and restraint Satisfaction	Warm color in walls, rugs, and furniture rearrangement, significantly increased patient ratings of the environment and reduced seclusion and restraint use.	3
Brown et al., 2020 (United Kingdom)	Qualitative Phenomenological	Staff and patients (n = 40) Diagnosis unreported	Photoproduction methods	Door Single room	Well-being	The clinking noise of traditional door mechanisms contributed to patients’ perception of being in a prison-like environment. Private rooms allow for a place to relax.	3
Butler et al., 2020 (United Kingdom )	Mixed methods Pre–post intervention	Patients (n = 17) Diagnoses unreported	Patient Experience Data Intelligence Center (PEDIC) reports Focus groups	Artwork	Satisfaction	Artwork contributed positively to the perceived comfort of the ward and its increased recommendation among friends.	3
Christenfeld et al., 1989 (United Sates)	Mixed methods Controlled pre–post renovation	Patients (n = 81; n = 37 ward intervention and n = 44 control wards) Staff (number unclear) Psychotic and schizophrenic	Questionnaires Semi-structure interviews	Architecture and interior design	Violent behavior Nonfunctional behavior Satisfaction	The new bedrooms, shower rooms, and dining rooms, did not improve satisfaction, affect depression or nonfunctional behavior. The new dayroom increased satisfaction and self-image. Patient violence was reduced by nearly 50%.	2+
Corey et al., 1986 (United Sates)	Quantitative Pre–post renovation	Patients (pretest, n = 66; posttest n = 56) Staff (pretest, n = 65; posttest, n = 56) Diagnoses unreported	The Ward Atmosphere Scale (WAS) Questionnaires	Interior design	Social interaction Anger and aggression Seclusion and restraint	Homelike features did not influence perceived social support, anger, aggression, or staff control.	3
Csipke et al., 2016 (United Kingdom)	Qualitative Participatory research	Interviews: patients (n = 10) and staff (n = 10) Questionnaire: patients (n = 36), staff (n = 61) Psychosis, bipolar, other	Semi-structured interviews Photography Qualitative component of a questionnaire	Artwork Control of the environment	Well-being and satisfaction	Artwork enhanced well-being, but unadjustable fixtures induced feelings of disempowerment in patients.	3
Curtis et al., 2007 (United Kingdom)	Qualitative Case study	Ex-patients of the new hospital (n = 6), users of local psychiatric services (n = 7), nursing staff (n = 10), and consultants (n = 3) Diagnoses unreported	Unstructured interviews	Single room Spaces for prayer	Well-being Sense of safety	Single-bedded rooms were linked to improved well-being and a sense of safety. Suitable spaces for prayer and faith-related practices were also associated with well-being.	3
Cutler et al., 2021 (Australia)	Qualitative Phenomenological	Ex-patients with multiple admissions to an acute mental health unit (n = 15) Self-reported diagnosis: bipolar disorder, depression, schizophrenia, psychosis, posttraumatic stress disorder, and anxiety	Semi-structure interviews	Lockable doors	Sense of safety	A lack of secure bathroom locks created feelings of unsafety due to potential accidental or intentional intrusions.	3
Donald et al., 2015 (Australia)	Qualitative Phenomenological	Patients (n = 20) Diagnosis unreported	Semi-structured interviews Focus groups	Treatment rooms Day space	Well-being	A lack of amenities and distractions contributed to boredom and a depressing atmosphere. Floor-to-ceiling glass in treatment rooms created confusion due to perceived privacy loss.	3
Edwards & Hults, 1970 (United Sates)	Mixed methods Pre–post renovation	Patients (number unreported) Staff (pretest n = 11; posttest n = 12) Diagnoses unreported	Number of staff–patient interaction Questionnaires (patients and staff) In-depth interviews (patients) Clinical observations	Nursing station	Well-being Staff-to-patient interaction	The removal of the window glass from the nursing station showed a shift to direct verbal communication, with staff spending more time interacting with patients in the dayroom, fostering quicker responses and less feelings of intrusion or fear.	3
Gallop, 1999 (Canada)	Qualitative Phenomenological	Patients (n = 18) Female patients with sexual and/or physical abuse history, primary diagnosis of depression, bipolar disorder, dissociate identity disorder, and schizophrenia	Semi-structure interviews	Artificial light Female only ward/area	Satisfaction Sense of safety	Preferences leaned toward soft lighting in bedrooms for increased comfort. Concerns arose over mixed-gender wards, with a call for segregated or all-female units for enhanced safety and control.	3
Gbyl et al., 2017 (Denmark)	Retrospective cohort study	Patients (n = 29) Bipolar depression, unipolar single-episode depression, and recurrent unipolar depressive disorder	Report files Administrative notes	Daylight	Length of stay Depression severity	Length of stay was significantly lower in south-east (SE) rooms than in north-west (NW) rooms, and while not statistically significant, a greater reduction in depression severity was observed in SE rooms.	2−
Holahan & Saegert, 1973 (United Sates)	Quantitative Controlled post-intervention	Patients (n = 50) Schizophrenic, other psychotic, nonpsychotic, and deferred	Interviews Observations	Interior design	Social behavior	Homelike features promoted more socializing and positive attitudes toward the physical environment.	2−
Ittelson et al., 1970 (United Sates)	Quantitative Observational comparative	Patients (number/diagnoses unreported)	Observations Behavioral mapping	Single/double room	Social behavior and interaction	Single and twin rooms fostered increased social activity, whereas dormitory rooms led to more isolated passive behavior.	3
Jenkins et al., 2015 (United Kingdom)	Quantitative Pre–post relocation	Patients (number unclear) Schizophrenia, schizotypal and delusional disorders, and mood disorders	Records (Total duration of all seclusions; number of seclusion incidents; number of secluded patients; and reporting verbal and physical aggression)	Architecture and interior design	Seclusion Aggression	Compared to the older mixed-gender unit, en suite bedrooms, gender-specific areas, and enhanced visibility and control features showed a significant decrease in seclusion episodes and aggressive incidents.	3
Jovanovic et al., 2020 (United Kingdom and Italy)	Cross-sectional Descriptive	Patients (n = 2,130) Primary diagnosis of ICD-10 psychotic, affective, and anxiety disorder	Questionnaire Client Assessment of Treatment (CAT)	Spaces for visiting Female only ward/area	Well-being Satisfaction	Hospitalization on mixed-sex wards and having suitable spaces for visitors were associated with higher patient satisfaction, particularly among female patients.	3
Kalantari & Snell, 2017 (Canada)	Mixed methods Post-occupancy evaluation	Staff (n = 98) Diagnoses unreported	Interviews Surveys	Nursing station	Satisfaction Rehabilitation Recovery Incidents Stress Staff-to patient interaction	Open nursing stations had no significant effect on interactions or incidents. Vibrant colors improved patient rehabilitation, but their impact on stress and satisfaction was inconclusive.	3
Kulkarni et al., 2014 (Australia)	Mixed methods Controlled post-intervention	Patients (n = 65) Staff (n = 20) Psychotic, mood disorder, postpartum psychosis/depression, anxiety disorder, eating disorder, and personality disorder	Safety incidents reports (for the 6 months postrenovation period) Questionnaire (patients and staff) Interviews with staff	Female only ward/area	Perception of safety Satisfaction	Women in wards with female-only areas felt safer, were more satisfied with their stay, and perceived the ward environment as more friendly, welcoming, and therapeutic.	2−
Molin et al., 2021 (Sweden)	Qualitative Phenomenological	Patients (n = 11) Staff (n = 13) Bipolar syndrome, schizophrenia, depression, borderline personality disorder, attention deficit/hyperactivity disorder, and delusional syndrome	Semi-structured interviews with patients Focus group with staff	Single room Lockable doors Day space Artificial light Spaces for visiting Control of the environment	Well-being Sleep quality Staff-to-patient interaction	Single rooms, personal storage, lockable doors, and control over lighting enhanced feelings of safety and well-being. A dedicated activity room enhanced interaction and well-being, and the lack of a visitation room outside the ward was seen as negative.	3
Nanda et al., 2011 (United Sates)	Mixed methods Pre–post multiple interventions	Patients (number unreported) Staff (n = 22) Diagnosis unreported	Rate of pro re nata (PRN) medication Nurse feedback based on observations on patient behavior Focus groups	Artwork	Agitation and anxiety	Nature art significantly reduced patient agitation and anxiety levels, enhancing the calming effect of the lounge area, unlike abstract art.	2−
Nicholls et al., 2015 (Australia)	Quantitative Pre–post relocation	Patients (n = 100) Staff (n = 92) Diagnosis unreported	Moos WAS	Day space Space for prayer Spaces for visiting	Aggressive behavior Social interaction Seclusion and restraint	Moving to a new mental health facility with enhanced amenities did not change aggression, social interaction, seclusion, or restraint levels.	3
Novotna et al., 2011 (Canada)	Qualitative Case study	Staff (n = 40) Substance misuse patients, mood and anxiety disorder	Focus groups with staff Behavioral mapping	Architecture and interior design	Recovery Staff-to-patient interaction	The new low-rise, homelike design improved patients’ privacy and control but increased safety concerns and observation difficulties for staff, despite encouraging more staff–patient communication in shared areas.	3
Olausson et al., 2021 (Sweden)	Qualitative Phenomenological	Patients (n = 19) Schizophrenia, schizotypal, and delusional, mood and mental and behavioral disorder	Photovoice	Lockable doors Day space Control of the environment	Well-being Social interaction	Lockable doors and personalization of rooms enhanced well-being. Spaces that promote activities such as training, creative workshops, and gardening increased social contact between inpatients and a sense of being a person.	3
Olausson et al., 2019 (Sweden)	Qualitative Phenomenological	Patients (n = 11) Schizophrenia, schizotypal, delusional d., mood d., neurotic, stress-related, and somatoform disorders	Photovoice	Single room Day space Control of the environment	Well-being Satisfaction Social interaction	Personalized, comfortable spaces with adequate storage and activity areas enhanced satisfaction, alleviated loneliness, and fostered social contact and personal identity.	3
Olver et al., 2009 (Australia)	Quantitative Pre–post relocation	Patients (n = 15, same pre–post relocation) Chronic schizophrenia, schizoaffective disorder, and intellectual disability	Records Brief Psychiatric Rating Scale	Architecture and interior design	Seclusion Aggression	Transitioning from an old to a new facility resulted in significant reductions in seclusion events and duration However, the decrease in aggressive incidents reported was not statistically significant.	3
Pyrke et al., 2017 (Canada)	Mixed methods Pre–post renovation	Patients (n = 47 preassessment/n = 47 postassessment) Depression and anxiety disorders, bipolar or schizophrenia and related disorders and without acquired brain injury, developmental delay, neurological/neurodegenerative disease, or sleep disorder	Questionnaire Sleep diary Actigraphy	Single room Control of the environment	Sleep quality	Moving from a dorm-style facility to a modern, single en suite room facility significantly improved measurable sleep quality.	3
Reavey et al., 2019 (United Kingdom)	Qualitative Phenomenological	Patients (n = 20) Diagnoses unreported	Photovoice	Single room	Sense of safety Well-being	Single rooms enhanced feelings of safety and provided greater control over personal psychological boundaries.	3
Rogerson et al., 2021 (United Kingdom)	Cross-sectional Descriptive	Staff (n = 191) Comorbid psychiatric disorders	Questionnaire	Exterior view	Aggression	Wards with views of both green space and man-made structures had significantly higher rates of physical incidents compared to those with only green space views.	3
Sachs et al., 2020 (United States)	Mixed methods Post-evaluation	Staff (n = 29) Patients (n = 33) Diagnoses unreported	Feedback forms followed by three listening sessions	Single room Control of the environment	Well-being	En suite private rooms, offering personalization, environmental control, and adequate storage, fostered a sense of empowerment, disconnection, and homeliness.	3
M. Shattell et al., 2015 (United Sates)	Qualitative Phenomenological	Patients (n = 13) Staff (n = 16) Self-reported diagnoses: depression, bipolar disorder, substance abuse, and schizophrenia; one participant reported issues with “cutting,” one identified as having “suicidal ideation,” and another identified “suicide attempt”	Interviews	Nursing station	Staff-to-patient interaction Well-being	Glass enclosures in nursing stations can create an impersonal barrier to communication, inciting feelings of imprisonment and emotional separation.	3
M. M. Shattell et al., 2008 (United Sates)	Qualitative Phenomenological	Patients (n = 10) Staff (n = 9) Depression, substance abuse, bipolar disorder, posttraumatic stress, anxiety, and borderline personality disorder	Interviews	Nursing station	Staff-to-patient interaction	Glassed-in nurses’ stations hinder direct visibility and communication with patients.	3
Southard et al., 2012 (United Sates)	Quantitative Pre–post renovation	Pretest: patients (n = 41); staff (n = 12) Posttest: patients (n = 40); staff (n = 13) Depression, substance abuse, and bipolar disorder	WAS	Nursing station	Anger and aggression Seclusion and restraint	Transitioning from an enclosed to an open nursing station had no significant effect on anger, aggression, or control measures.	3
Tyson et al., 2002 (Australia)	Mixed method Pre–post relocation	Staff (n = 80) Diagnosis unreported	Observations	Architecture and interior design	Staff-to-patient interaction	In the new acute ward that included more day space and a mix of single-, double-, and four-bedded rooms, staff–patient interactions increased and became more positive. In the long-term wards, with less day space and single rooms, interactions decreased but improved in quality.	3
Ulrich et al., 2018 (Sweden)	Quantitative Controlled pre–post relocation	Patients (n = 42) Schizophrenia or other psychosis, bipolar disorder, personality disorder, and suicide risk	Records	Architecture and interior design	Compulsory injections Physical restraint	Implementing stress-reducing environmental design significantly reduced physical restraints and injections in a new hospital.	2−
Vaaler et al., 2005 (Norway)	Randomized control trial	Patients (n = 56) Schizophrenia and mood disorders	The Positive and Negative Syndrome Scale for schizophrenia Global assessment scale split version Brøset violence checklist or the global assessment Records of violent incidents Questionnaire for patient satisfaction	Interior design	Length of stay Violence Satisfaction Symptoms of psychopathology	Redesigning seclusion areas showed nonsignificant trends toward shorter stays, reduced symptoms, and no instances of vandalism.	1−
Van der Shaaf et al., 2013 (Netherlands)	Cross-sectional Descriptive	Patients (n = 23,868 admissions of 14,834 patients) Schizophrenia, schizotypal, delusional, mood (affective), behavioral and personality disorder, mental and behavioral disorders due to psychoactive substances	Rate of seclusion Number of seclusions per patient Time secluded	Private spaces	Seclusion	A greater private space per patient is linked to a reduced likelihood of seclusion.	3
Vethe et al., 2021 (Norway)	Randomized cross-over trial	Patients (n = 12) Diagnosis unreported	Melatonin levels Sleep (spent in each sleep stage, sleep onset latency, time awake after sleep onset, rapid eye movement sleep onset Latency, and sleep efficiency) Neurocognitive arousal Subjective sleepiness Potential side effects	Artificial light	Sleep quality	Blue-depleted light exposure improved sleep quality with longer sleep durations, less melatonin suppression, and reduced arousal, without affecting subjective sleepiness or side effects.	1−
Whitehead et al., 1984 (United Sates)	Quantitative Pre–post renovation	Patients (number unreported) Acute and chronic illness; diagnosis DSM-III criteria [DSM-III: Diagnostic and Statistical Manual of Mental Disorders, third edition]	Behavioral environment assessment technique and attitudinal measures—the Sepulveda psycho environmental assessment record	Nursing station Single/double room Spaces for visiting	Social interaction Well-being	Redesign led to more staff in the day room, likely due to the open nursing station However, patient frequency near the station remained unchanged. Shared rooms with only one other person, private visitor areas, and the open station contributed to an improved sense of well-being for patients.	3
V. J. Wood et al., 2013 (United Kingdom)	Qualitative Case study	Caregivers (n = 9) Diagnoses unreported	Interviews Group discussions	Spaces for visiting Spaces for prayer	Well-being	Spaces for visitors and spiritual areas were important for strengthening family ties and well-being.	3
D. Wood and Pistrang, 2004 (United Kingdom)	Qualitative Phenomenological	Patients (n = 9) Bipolar affective disorder, depression, schizophrenia, borderline personality disorder Staff (n = 7)	Interviews	Single room Female only ward/area	Sense of safety	Shared rooms were associated with feelings of unsafety. Female patients perceived mixed wards as reducing their safety.	3
Yakov et al., 2018 (United Sates)	Quantitative Pre–post intervention	Patients (number unreported) Schizophrenia and psychotic spectrum disorders, substance use, and depressive and bipolar disorders	Assault rates Percentage of patient restraint hours per 1,000 patient hours between 4:00 p.m. and 7:00 p.m.	Artificial light Acoustics	Assaults Restraints	Low lighting using only safety light and natural light decreased restraints and assault.	3

The selected studies were published between 1968 and March 2022, deriving from nine countries, including the United States (n = 13), United Kingdom (n = 10), Canada (n = 5), Australia (n = 6), Sweden (n = 4), Norway (n = 2), Netherlands (n = 1), Denmark (n = 1), Italy (n = 1), and one study conducted jointly in Italy and England (n = 1). The largest study included 14,834 participants (cross-sectional study), and the smallest study included nine participants. Of the ward-level studies, the largest study included 199 wards, and the smallest included one. Studies were conducted at acute mental health inpatient facilities (n = 13); IPCUs (n = 2); forensic inpatient settings, including high, medium, and low security (n = 8); and multiple studies involved more than one type of mental health inpatient setting (n = 4). In 21 studies, the type of adult mental health inpatient setting was unspecified. Participant diagnoses were not disclosed in 16 studies and data sample sizes were unspecified in six studies. Several studies also included data from geriatric (Baldwin, 1985; Borckardt et al., 2011), organic mental illness (Baldwin, 1985), and child and adolescent services (Borckardt et al., 2011). In one mixed-methods study, only the qualitative study was relevant for this literature review (Csipke, 2016).

Patient outcomes identified included medication intake, length of stay, quality of sleep, seclusion and restraint (physical and pharmaceutical restraint), aggression (verbal and physical aggression), violence and vandalism, social interaction and activity (staff–patient, patient–patient, and patient–visitor interaction), satisfaction with the environment, stress, agitation and anxiety, sense of safety, and well-being (e.g., feelings of being bothersome and threatening to the staff, hope, power and control, sense of integrity and self-respect, negative self-image, and prison-like and homeliness feeling).

Of the 44 studies identified, 34 directly explored the influence of design factors on patient outcomes, reported on patients’ experiences or presented staff and carers perspectives on how such design factors impact patients. The synthesis of these studies is presented first and systematically organized by design factors—ranging from patient bedrooms and communal areas to clinical spaces, interior design interventions, and ambient factors. This is followed by the findings from the remaining ten studies that focus on the processes of either transitioning patients to a new facility or refurbished one.

Patient Bedroom

When considering patient bedrooms, privacy stood out as a recurrent theme in many studies, highlighting its role in providing patients with a sense of autonomy and control over their surroundings. One particular study by Van der Schaaf et al. (2013) found that when patients have more private space, they are less likely to be secluded.

Single rooms

Looking into the specifics of bedroom configurations, a total of ten studies reported the positive impact of having single or twin rooms on patient outcomes. Ittelson et al. (1970) found increased social activity in patients staying in single and twin rooms, while dormitory rooms were characterized by more isolated passive behavior (e.g., lying on a bed from two thirds to three fourths of the time spent in their rooms). In the study by Whitehead et al. (1984), patients described that sharing a room with only one other person made them feel better compared to dormitory style. Pyrke et al. (2017) identified an improvement in sleep quality after the renovation of a facility that transitioned from dormitory-style rooms to single rooms. This renovation also incorporated a climate control system (ventilation, lights, and acoustics) as well as soundproofed walls and doors. Similarly, qualitative data indicated that single rooms contributed positively to patient well-being (Brown et al., 2020; Curtis et al., 2007; Molin et al., 2021; Olausson et al., 2019; Reavey et al., 2019; Sachs et al., 2020) and feelings of safety (Curtis et al., 2007; Molin et al., 2021; Reavey et al., 2019; D. Wood and Pistrang, 2004). In these studies, participants perceived single rooms as places for quietness and safety where they could withdraw from staff, other patients, and the overall soundscape environment.

Lockable doors

In one study, staff highlighted the importance of lockable doors in patient rooms for enhancing patients’ sense of safety (Molin et al., 2021). Similarly, patients emphasized that these doors imparted feelings of comfort and normality (Olausson et al., 2021). In addition, ex-patients in the Cutler et al. (2021) study expressed concerns about feeling unsafe in bathrooms without locks or those that could be easily opened from the outside, citing potential inadvertent or deliberate intrusions by other patients.

Control of the environment

Several studies emphasized the importance for patients to have control of the environment in their rooms. Having personally controllable room appliances and fixtures, such as windows, showers, heating, light and acoustics, was associated with a sense of empowerment (Csipke et al., 2016), well-being (Molin et al., 2021; Sachs et al., 2020), and improved sleep quality (Pyrke et al., 2017). In addition, a sense of homeliness was associated with patients having personalization over rooms (e.g., photographs, artwork, and gifts) (Olausson et al., 2019, 2021; Sachs et al., 2020), comfortable beds, and the use of personal bedding (Olausson et al., 2019, 2021). Having storage space in patient rooms was also evidenced in two studies as contributing positively to patient satisfaction (Molin et al., 2021; Sachs et al., 2020). Furthermore, lockable cabinets and private wardrobes for personal belongings were highlighted as important for patients’ perceptions of safety and control (Molin et al., 2021).

Communal Spaces

Spaces for activities, visitors, and spiritual/contemplative reflection

Several studies reported that, based on patients experiences of the physical environment, day spaces designed to promote activities increased their interaction with staff (Olausson et al., 2021; Molin et al., 2021) and reduced feelings of boredom (Donald et al., 2015; Olausson et al., 2019). Further outlining the importance of specific spaces, several studies highlighted the value of designated areas, both communal and private, for visitor meetings. These areas were associated with strengthening family ties and enhancing well-being (Jovanović et al., 2020; Molin et al., 2021; Wood et al., 2013), as well as lifting patient spirits (Whitehead et al., 1984). The provision of spiritual or contemplative areas also resonated positively with patient well-being in two studies (Curtis et al., 2007; V. J. Wood et al., 2013). By contrast, Nicholls et al. (2015) identified, based on patients’ and staff perceptions of the new “atmosphere,” no change in levels of aggression, social interaction, seclusion, and restraint after moving to a new facility that included space for prayer/reflection, recreational courtyards, multiple therapeutic spaces, and family-friendly space. However, the authors acknowledged that the results could have been influenced by the nature of the atmosphere which involves more than just the building or environment; it also requires social occupation and interactions.

Female only ward/areas

Several studies found an association between female-only areas and an enhanced sense of safety. First, a study comparing a mixed-sex ward to a similar ward comprising a female-only area identified that the latter had fewer incidents and female patients had a better experience of care, sense of safety, and satisfaction with the environment (Kulkarni et al., 2014). Similarly, in two qualitative studies, female patients perceived mixed-sex wards as reducing their safety and control (Gallop et al., 1999; D. Wood & Pistrang, 2004). They were concerned about male patients and wished for an all-female unit or a unit with segregated communal areas (Gallop et al., 1999). In addition, female patients from mixed-sex wards reported feeling vulnerable when sexually harassed or intimidated by male patients (D. Wood & Pistrang, 2004). However, due to the previous experiences of sexual and physical abuse, and/or safety trauma of the studies population sample, the generalization of these results should be treated with caution. In contrast, a recent cross-sectional study involving 2,130 patients from 18 acute mental health hospitals with mixed-sex and single-sex wards found that hospitalization in a mixed-sex ward was associated with higher patient treatment satisfaction. Interestingly, compared to male patients, for whom the effect of the type of wards was marginal, females reported higher satisfaction with mixed-sex wards (Jovanović et al., 2020).

Clinical Spaces

Nursing station

Changes in the configuration of nursing stations, transitioning from traditionally glass-enclosed nursing stations to more open designs, have been associated with improved patient–staff interactions and communication (Edwards & Hults, 1970; M. Shattell et al., 2015; M. M. Shattell et al., 2008), as well as well-being (Whitehead et al., 1984). Two studies observed that staff spent more time in the dayroom when the nursing station glass was removed (Edwards & Hults, 1970; Whitehead et al., 1984). Additionally, patients reported feeling less bothersome and threatening to the staff. They also expressed a greater sense of power and control in their interactions with nursing station staff, coupled with feelings of freedom as opposed to sensations of imprisonment and confinement (Edwards & Hults, 1970; M. Shattell et al., 2015).

In contrast, a post-occupancy evaluation by Kalantari and Snell (2017) that included only staff, found that open nursing stations did not enhance patient satisfaction, rehabilitation, or staff–patient interactions without compromising safety or reducing the number of incidents between patients. However, the results from this study should be interpreted with caution due to potential staff bias in their perception of the safety offered. Southard et al. (2012) found no statistically significant difference in patient therapeutic milieu when comparing the nursing station before and after removing the glass enclosure. They also recorded no perceived increase in patient anger or aggression toward staff.

Treatment rooms

As well as privacy in bedrooms and bathrooms, ensuring privacy in treatment rooms was identified as contributing positively to patient well-being. In the study by Donald et al. (2015), treatment rooms with floor-to-ceiling glass, observable from the rest of the unit, were perceived by patients as confusing. Essentially, a space they considered to be “private” became visible to others.

Interior Design Interventions

Furniture/seating arrangements

Only one study purposely investigated the effects of changes in furniture and seating arrangements as a single intervention. Baldwin (1985) found that rearranging ward furniture from non-social patterns into group seating arrangements in the dayroom—three or more chairs facing within three feet radius of the table—contributed to reducing the number of hours of seclusion and the number of casualty incidents. However, no changes in medication rate were observed.

Artwork and color

This review included three studies that identified the positive effect of artwork on patients. Nanda et al. (2011) found that the presence of realistic nature, abstract-representational, and abstract art led to a reduction in the average daily medication required, compared to days without it. Notably, this effect was more pronounced for realistic nature art than for abstract-representational or abstract art. Nurses also expressed that nature art had a calming effect on patients compared to the abstract art, which was described as irritating, or even headache-inducing. In Butler et al. (2020), participants reported increased satisfaction with the physical environment when nature and abstract-related wall paintings—created by professional artists with input from patients—were displayed in the IPCU. Csipke et al. (2016) reported that patients described the artwork as “brightening up” the ward and enhancing their sense of well-being.

Regarding color, an interventional study by Kalantari and Snell (2017) used different color tones across the unit—vibrant and saturated in treatment areas and warmer and more soothing colors in communal areas and bedrooms. While no association was found between the color and patient satisfaction or stress—as perceived by staff—the study concluded that the color scheme improved patient recovery and rehabilitation.

Ambient Features

Acoustic environment

Two studies reported the effects of various acoustic environmental features on patient outcomes. Brown et al. (2020) found that certain door design features, such as the use of traditional mechanical cylinders—evident from the clinking of keys carried by staff—were perceived by patients as evoking feelings of being in prison. These feelings were further exacerbated by the spatial layout, characterized by high ceilings and large open spaces. On the other hand, Yakov et al. (2018) assessed the impact of minimizing noise from staff and patients, combined with soft music in lieu of television, and introducing live therapeutic music on rates of assaults and restraint. Although there was a reduction in those outcomes, the study could not conclusively attribute these results to the noise reduction interventions as they were implemented concurrently with a low-lighting intervention. Nonetheless, when considering both interventions together, the study observed a significant 72% reduction in assaults and restrains 11 months post-implementation.

Artificial light

A longitudinal, pre- and post-intervention study found that low lighting, using only safety and natural lighting in the day room, was positively associated with a reduction in assault and restraint rates during the evening hours compared to the baseline period (Yakov et al., 2018). Similarly, Gallop et al. (1999) found that patients favored soft lights in their bedrooms for increased comfort, compared to the staff’s standard use of flashlights for routine checks and nighttime visibility. In another study, the lighting—termed “healing lights”—in bedrooms of a new facility were perceived by nurses to enhance patient sleep quality and relaxation (Molin et al., 2021).

Lastly, a randomized crossover trial involving 12 patients showed that a blue-depleted light environment, applied in the evening from 06:30 pm to 06:50 am in bedrooms, bathrooms, hallways, and common areas, improved sleep quality— the total sleep time was 8.1 minutes longer compared to the control. However, the sample might not resemble all patients as some left the ward from 08:00 to 05:00 pm and could have been exposed to higher lights levels during the day (Vethe et al., 2021).

Daylight

Three retrospective cohort studies showed the positive effect of bright and sunny rooms on the length of stay of patients experiencing a depressive phase of bipolar disorder. However, this impact was unclear for unipolar depressive patients. Beauchemin and Hays (1996) identified that patients with depression who stayed in brighter, sunnier rooms (reaching up to 50,000 lux on a bright day) had a 2.6 day shorter length of stay than those in dull rooms (with a maximum of 3,000 lux). More recently, Gbyl et al. (2017) identified that patients with a depressive phase, either as part of a unipolar or bipolar disorder, had a shorter average length of stay of 29.6 days when accommodated in rooms receiving more direct natural light (facing southeast with up to 60,000 lux) compared to a control group in rooms facing northwest (maximum of 3,000 lux). Similarly, the study by Benedetti et al. (2001) involving 602 unipolar and bipolar patients reported a 3.67-day shorter length of stay for depressed patients with bipolar affective disorder staying in eastern rooms (receiving direct sunlight in the morning) compared with those in western rooms (direct sunlight in the evening). This study, however, did not note a sunlight effect on the length of stay for unipolar depression, a finding that contrasts with Gbyl et al. (2017). Given their design, these studies have limitations that might imply a lack of causality for the observed difference in the length of stay. For instance, factors such as the time patients spent in their rooms, how often curtains were opened or closed, variations in pharmacological regimes, or illness severity, were not reported in any of these studies.

Outside view

This review included only one study examining the impact of the content of the view from the room window on patients. Rogerson et al. (2021) conducted a cross-sectional study across 101 wards and identified higher aggressive behavior in wards that offered a view of both nature and man-made structures compared to a view exclusively of nature. Nonetheless, the findings of this study should be treated with caution, given that survey responses were unevenly distributed across all wards. Additionally, the study did not account for certain variables that might influence aggression, such as the addition of extra staff to facilitate observations.

Multifactorial Interventions

Interior design

In four studies, modifications to the interior design that provide a more home-like environment, such as warm wall colors, a wall and door painted with bright solid colors, placement of decorative throw rugs, plants, and modern furniture, and seating arrangements for small group interactions, were found to have an impact on patient outcomes. For example, Borckardt et al. (2011) found, when compared to the previous facility design, a reduction in the use of seclusion and restraint along with an increase in patients’ satisfaction with the new physical environment. Similarly, Holahan and Saegert (1973) identified enhanced social interaction, patient satisfaction, and less isolated passive behavior in a renovated ward compared to a nonrenovated one. Vaaler et al. (2005), when comparing a traditional design seclusion area with another that was identically structured but redecorated to mimic a home-like environment, identified that the latter reduced vandalism and increased patient satisfaction, especially in women. There was no substantial effect on symptoms of psychopathology or length of stay. In contrast, Corey et al. (1986), using the Ward Atmosphere Scale (WAS) to assess how staff and patients perceived the inclusion of home-like features before and after the redesign of a mental health facility, did not find an impact on perceived social support (functional measure of social interaction), anger and aggressive behavior, or staff control (functional measure of seclusion and restraint).

Interior design and architectural features

This review included six studies that explored the impact of the renovation of an existing facility or the relocation to a new purpose-built unit or hospital involving simultaneous alterations in both the design and architecture. In the study by Jenkins et al. (2015), the relocation to a new IPCU that included en-suite bedrooms, gender-specific areas, a layout that enabled higher visibility, control of the environment, spaces for activities, and lockable toilets, led to decreased aggressive incidents and the number and total duration of seclusion episodes, compared to the old ward. Similarly, in a study by Ulrich et al. (2008), a comparison between the old and control facility and a new facility incorporating enhanced stress-reducing environmental features (e.g. privacy, control, visibility, natural light, art and communal areas) revealed a decrease of 6.8% in patients receiving injections (pharmaceutical restraints) after the transition. However, there were no differences for patients who required physical restraints. Although, the number of physical restraints for patients who received at least one type of restraint decreased by 50%.

Christenfeld et al. (1989), involving pre- and post-assessments of four control and two experimental wards, found that extensive design changes in the day room, and the transition to an open nursing station, led to a 50% reduction in violent episodes. The study did not find a difference in nonfunctional behavior between renovated and control wards. In addition, patients reported greatest satisfaction and an improvement in self-image with the renovated dayroom. Similarly, the findings from Tyson et al. (2002) implied that increased day space enhanced staff-patient interaction, yet due to other changes accompanying the relocation, these results should be interpreted with caution. Olver et al. (2009) reported that the relocation to a new purpose-built secure unit that included more indoor and outdoor recreational areas, single rooms and access to more daylight, led to a reduction in the number of seclusion events and seclusion episodes exceeding 4 hr. However, the study did not find a reduction of aggressive incidents as perceived by staff. Novotná et al. (2011) identified that the new low-rise, homelike design not only improved privacy and control for the patients, but also increased safety concerns and observation difficulties for staff. In addition, behavior mapping activities on the units revealed increased staff–patient communication, occurring primarily in the patients lounges, adjacent kitchens, and lobby areas.

Discussion

This review included 44 studies where the impact of different factors of the facility design on patient outcomes were investigated or reported based on patient, staff, or carer experiences of the physical environment. This number differs from a similar literature review conducted by Weber et al. (2022), which included 26 studies. The difference in the quantity of included studies could be attributed to our broader inclusion criteria. For example, unlike Weber’s literature review, we included research from forensic units and studies using the WAS. We also included five studies—excluding those from forensic settings—published between the end date of Weber’s search (September 2020) and ours (March 2022). In contrast, our review excluded studies focusing solely on organizational factors like open/locked wards. Although they play a significant role in mental health inpatient settings, we chose not to include these in the review based on our consideration that they are more related to organizational procedures than to the physical design itself, and hence would require a separate comprehensive literature review. Despite the larger quantity of studies included, our review echoes Weber’s in acknowledging the difficulties and complexities of conducting research in these environments.

Delving into our results, several findings emerged from the literature review that can guide future design efforts to contribute to the therapeutic environment and inform best practice. To provide a concise overview, we compiled a summary table to outline the findings, along with their confidence levels and recommendations for practice (see Table 2). First, our review found evidence suggesting that incorporating a more home-like environment into mental health wards could minimize the institutional feel and enhance comfort and satisfaction. Specific changes might include incorporating plants, selecting comforting colors for walls, or modern furniture that make the space feel more like a home than a clinical setting. The provision of private spaces, specifically single-bedded rooms, and enabling control of the environment, such as ventilation, lighting, acoustics, and personalization of rooms, emerged as other crucial design factors. They were found to enhance feelings of safety, well-being, and sleep quality, pointing to designs that focus on patient privacy and environmental control. Such insights align with Paopulas et al. (2014), whose review synthesis identified the provision of home-like designs and private spaces to be associated with decreased vandalism and violence, and increased social interaction.

Table 2.

Confidence Level and Recommendations for Practice

Design Factor	Finding	Studies	SIGN50 Evidence Levels and Key Points (Consistency, Limitations, and Generalizability)	Confidence in the Finding	Recommendations for Practice
Patient bedroom—single rooms	Positive impact of single rooms on increased social activity, enhanced sleep quality, and improved feelings of safety and well-being.	Brown et al. (2020), Curtis et al. (2007), Ittelson et al. (1970), Molin et al. (2021), Olausson et al. (2019), Pyrke et al. (2017), Reavey et al. (2019), Sachs et al. (2020), Whitehead et al. (1984), and D. Wood and Pistrang, (2004)	Supported by 10 qualitative studies categorized as Level 3 evidence. There is consistency in the findings related to feelings of safety and well-being across studies. However, findings on social activity and sleep quality are based on single studies, implying a lack of robust evidence across these outcomes. There was increased potential for confounding variables in the study reporting improvements in sleep quality, as single rooms were also equipped with climate control and soundproofed walls and doors. Two studies were conducted in forensic units.	High	Prioritize the use of single rooms when designing or renovating mental health wards, especially to enhance feelings of safety and well-being.
Control of the room environment	Control of the room environment, such as windows, showers, heating, light, and acoustics, was associated with increased empowerment, well-being, and sleep quality.	Csipke et al. (2016), Molin et al. (2021), Pyrke et al. (2017), and Sachs et al. (2020)	Supported by four studies categorized as Level 3 evidence. There is consistency in the finding related to well-being in two of the studies. A single study supports sleep quality and empowerment.	Moderate	Should include, where reasonably practicable, adjustable environmental room controls (e.g., light, temperature) to accommodate individual preferences.
	A sense of homeliness was associated with room personalization (e.g., photographs, artwork, gifts) and having comfortable beds and personal bedding.	Sachs et al. (2020), Olausson et al. (2021), and Olausson et al. (2019)	Supported by three studies categorized as Level 3 evidence. Two studies were authored by the same individual and conducted in forensic units.	Low	Consider the use of inpatient room personalization, bearing in mind the security and safety considerations of each setting, especially in forensic units. Based on current evidence, emphasizing comfort in bedding choices may also be considered.
	Lockable doors in patient rooms and bathrooms enhanced patients’ sense of safety and well-being.	Cutler et al. (2021), Molin et al. (2021), and Olausson et al. (2021)	Supported by three qualitative studies categorized as Level 3 evidence. Two studies focus on bedroom doors and one on bathroom doors. One study was conducted in forensic units.	Low	Consider the use of lockable doors in patient rooms to enhance patients’ sense of safety and well-being, keeping in mind the necessary safety mechanisms for emergencies. More research is still needed to identify mitigation of possible risks.
	Availability of storage space in patient rooms was linked to increased patient satisfaction. Lockable cabinets and private wardrobes were associated with patients’ perceptions of safety and control.	Molin et al. (2021) and Sachs et al. (2020)	Supported by two studies categorized as Level 3 evidence. There is consistency in the findings related to storage space enhancing satisfaction across studies. A single study emphasized the importance of these storage spaces being lockable.	Very low	Consider piloting storage space, lockable cabinets and private wardrobes in patient rooms to increase patients’ satisfaction and perceptions of safety and control.
Spaces that promote participation in activities	Day spaces that promoted participation in activities enhanced staff-patient interaction and reduced feelings of boredom. However, the mere presence of these spaces does not guarantee improved outcomes.	Donald et al. (2015), Molin et al. (2021), Nicholls et al. (2015), Olausson et al. (2019), and Olausson et al. (2021)	Supported by four studies categorized as Level 3 evidence. There is consistency in the findings related to both improved interaction and reduced boredom. Two studies were authored by the same individual and conducted in forensic units. However, a further study (Level 3 evidence) by Nicholls et al. (2015) found no impacts on levels of aggression, seclusion, or restraint, emphasizing that factors beyond the physical environment, such as social occupancy and interactions, could influence outcomes.	Moderate	Should include, where reasonably practicable, day spaces that promote participation in activities to reduce feelings of boredom.
Designated areas for visitors	Designated spaces for visitors reinforced family relationships and boosted patient well-being.	Jovanović (2020), Nicholls et al. (2015), Molin et al. (2021), Whitehead et al. (1984), and V. J. Wood et al. (2013)	Supported by four studies categorized as Level 3 evidence. However, a further study (Level 3 evidence) by Nicholls et al. (2015) found no impacts on levels of aggression, social interaction, seclusion, and restraint, emphasizing that factors beyond the physical environment, such as social occupancy and interactions, can influence outcomes.	Moderate	Should include, where reasonably practicable, designated visitor areas to foster positive interactions between patients and visitors and improve patient well-being.
Spiritual/contemplative areas	Provision of spiritual/contemplative spaces contributed to patient well-being.	Curtis et al. (2007), Nicholls et al. (2015), and V. J. Wood et al. (2013)	Supported by two studies categorized as Level 3 evidence. However, a further study (Level 3 evidence) by Nicholls et al. (2015) noted no change in specific behavioral outcomes and highlighted the broader complexities of the atmosphere.	Low	Consider the use of spiritual or contemplative spaces to support patient well-being.
Female only ward/communal areas	Female-only areas were associated with enhanced sense of safety for women with sexual and physical abuse or safety traumas.	Gallop et al. (1999), Jovanic et al. (2020), Kulkarni et al. (2014), and D. Wood & Pistrang (2004)	Supported by three studies categorized as Levels 2 and 3 evidence. Two studies (Level 3 evidence) indicate a preference of female patients for female-only areas, highlighting concerns about safety in mixed-sex wards. A study (Level 2 evidence) by Kulkarni et al. (2014) identified better experiences in wards with female-only areas, including fewer incidents. Generalizability is limited due to the backgrounds of certain study populations, especially those with histories of trauma. However, contradicting these findings is a study (Level 3 evidence) by Jovanović et al. (2020) identifying a higher treatment satisfaction for females in mixed-sex wards.	Moderate	Should include, where reasonably practicable, female-only areas or segregated communal spaces to enhance the feeling of safety and control, especially for female patients with histories of trauma or sexual abuse. However, individual preferences can vary, and some women might feel more satisfied using mixed-sex areas of wards.
Nursing station configuration	Transitioning from enclosed to open nursing stations was shown to positively impact patient–staff interactions, communication, and patient well-being.	Edwards & Hults (1970), Kalantari and Snell (2017), M. M. Shattell et al. (2008), M. Shattell et al. (2015), Whitehead et al. (1984), Southard et al. (2012), and Whitehead et al. (1984)	Supported by seven studies categorized as Level 3 evidence. There is consistency in the findings related to the benefits of open nursing stations across five studies. However, Kalantari and Snell (2017) did not find open stations enhanced outcomes, possibly due to staff bias. Southard et al. (2012) did not note increased aggression when the glass was removed. Two studies were authored by the same individual.	Moderate	Should include, where reasonably practicable, an open configuration to enhance patient–staff interactions, communication, and patient well-being. However, considerations should be given to staff safety. More research is needed to ascertain the best configuration for differing environments and patient populations.
Treatment room	Ensuring privacy in treatment rooms contributed positively to patient well-being. Rooms with observable floor-to-ceiling glass were perceived as confusing due to the lack of privacy.	Donald et al. (2015)	Supported by one study categorized as Level 3 evidence.	Very low	Consider piloting patient privacy in treatment room designs to improve patient well-being.
Furniture/seating arrangements	Rearranging ward furniture into social seating arrangements reduced the number of seclusion hours and casualty incidents. No change in medication rate was observed.	Baldwin (1985)	Supported by one study categorized as Level 2+ evidence. The study investigated the effects of changes in furniture and seating as a single intervention and observed reductions in seclusion hours and casualty incidents. However, medication rates remained unchanged. The study was conducted in maximum security forensic units.	Low	Consider optimizing seating arrangements to foster social interaction and reduce seclusion and adverse incidents.
Artwork	The presence of artwork, particularly realistic nature art, reduced the need for medication, had a calming effect on patients, increased satisfaction with the physical environment, and enhanced overall patient well-being.	Butler et al. (2020), Csipke et al. (2016), and Nanda et al. (2011)	Supported by two studies categorized as Level 3 evidence and one as Level 2. There is consistency in the findings related to the positive impact of artwork, across studies. There was supporting evidence in relation to realistic nature art but abstract art, on the other hand, received mixed feedback. One study was conducted in an IPCU setting.	Moderate	Should include, were reasonably practicable, artwork across the facility. Prioritize realistic nature pieces, given their observed calming effects and potential to reduce medication needs. Caution is advised when incorporating abstract pieces due to some negative feedback, but further research would solidify these findings.
Color schemes	Different color tones across mental health wards may influence patient recovery and rehabilitation but not significantly impacted patient satisfaction or perceived stress.	Kalantari and Snell (2017)	Supported by one study categorized as Level 3 evidence. While the study did not find a direct link between color schemes and patient satisfaction or stress levels, there is a potentially positive effect on patient recovery and rehabilitation.	Very low	Consider piloting varied tones of color across the facility to support patient recovery and rehabilitation.
Acoustic environment	Door design features, particularly those with traditional mechanical cylinders evoked feelings of being imprisoned.	Brown et al. (2020)	Supported by one study categorized as Level 3 evidence. Conducted in a medium-secure forensic unit.	Very low	To guide subsequent decision making, mental health facilities should consider piloting studies on the benefits of nontraditional locks.
Artificial lighting	Low or soft lighting in the dayroom was associated with reduced evening assault and restraint rates. In bedrooms, it enhanced comfort and improved sleep quality and relaxation. Evening blue-depleted lighting in common areas and bedrooms further improved sleep quality, increasing total sleep time.	Gallop et al. (1999), Molin et al. (2021), Vethe et al. (2021), and Yakov et al. (2018)	Supported by three studies categorized as Level 3 evidence and one study Level 1 evidence. There is consistency in the findings related to positive effects of tailored lighting across studies. The study (Level 1 evidence) on evening-blue depleted lighting was informed by the results from Vethe et al. (2021). This study had a small sample, with potential daytime exposure affecting results.	Moderate	Should include, where reasonably practicable, low, soft, or evening blue-depleted lighting in dayrooms and bedrooms, while adhering to safety standards to foster a calming atmosphere and improve sleep quality.
Room brightness and orientation	Bright, sunny rooms, especially those receiving direct morning sunlight, were associated with a shorter length of stay for patients in a depressive phase of bipolar disorder. The impact on unipolar depressive patients was inconclusive.	Beauchemin and Hays (1996), Benedetti et al. (2001), and Gbyl et al. (2017)	Supported by three retrospective cohort studies categorized as Level 2 evidence. There is consistency in the findings related to positive effect of brighter rooms on bipolar depressive patients across studies. These studies have potential uncontrolled confounders, such as the time patients spent in their rooms, curtain usage, variations in drug treatments, or severity of illness. However, inconsistent findings were observed for unipolar depression. Gbyl et al. (2017) indicated a benefit, while Benedetti et al. (2001) did not.	Moderate	Should include, where reasonably practicable, patient room designs which maximize the potential for direct sunlight to shorten recovery duration. The orientation should be a key consideration, maximizing exposure to morning sunlight.
Room window—view content	Rooms with a mixed view of nature and man-made structures were associated with higher aggressive behavior compared to rooms with a view exclusively of nature.	Rogerson et al. (2021)	Supported by one cross-sectional study categorized as Level 3 evidence. Uneven survey response distribution across wards and nonconsideration of potential confounders like extra staffing could have affected the findings’ reliability. The study includes forensic and no-forensic inpatient units.	Very low	To guide subsequent decision making, mental health facilities should consider piloting studies on the benefits of nature views.
Interior design—home-like environment	Incorporating a home-like environment in mental health improved satisfaction and behavior.	Borckardt et al. (2011), Holahan & Saegert (1973), Vaaler et al. (2005), and Corey et al. (1986)	Supported by three studies categorized as Levels 1, 2, and 3 evidence. There is consistency in the findings related to positive effects of a home-like environment on patient satisfaction. Improved behavior was consistent in two studies. However, a further study (Level 3 evidence) by Corey et al. (1986) did not find an impact on social support, anger, aggressive behavior, or staff control. It is important to note that this finding come from studies observing patients transitioning between facilities, rather than direct measurements of the impact of home-like environments.	Moderate	Should include, where reasonably practicable, home-like features to potentially enhance patient satisfaction and social interactions.

Delving into our results, several findings emerged from the literature review that can guide future design efforts to contribute to the therapeutic environment and inform best practice.

Communal spaces designed to promote interaction and activity was also a common theme, emphasizing their role in reducing feelings of isolation and boredom, and contributing to improved overall well-being. Additionally, the need for dedicated family visitation areas and spaces for spiritual/contemplative practices came to the fore, associated with improved well-being, indicating the importance of incorporating these spaces into facility designs. The provision of spaces for females, especially those who have experienced past trauma, was deemed valuable, suggesting a need for gender-focused areas or wards to enhance their feeling of safety and well-being. Our findings revealed a preference among patients for open nursing stations, as they tend to foster improved staff–patient interactions and better communication, leading to a more collaborative care atmosphere. However, while open configurations have their merits, staff safety and specific environmental factors must be taken into account. More research is essential to determine the best nursing station configuration, considering varying environments and patient populations.

Daylight emerged as a pivotal design consideration, particularly for patients during depressive episodes of bipolar disorder. Room designs that maximize daylight intake was found to positively influence their recovery. Lastly, the integration of artwork, especially those depicting nature, emerged as a positive design strategy throughout our review. Such aesthetic integrations were correlated with reduced stress levels and heightened patient satisfaction. However, it is worth noting the nuances in selecting artwork, in particular, abstract artwork. This was found by Salonen et al. (2013) to have potential negative implications for patients in general healthcare facilities. In line with these insights, our recommendations urge practitioners to exercise caution when opting for abstract art pieces. Yet, it is essential to acknowledge that further research is required to provide a more definitive stance on this observation.

Contrary to our initial expectations, the available body of evidence did not offer an exhaustive exploration of the topic. For certain design factors, we identified only one single study highlighting its importance. While this does not undermine the potential significance of these elements, it does prevent us in forming robust, evidence-based recommendations. Yet, these factors have been recommended in evidence synthesis for general facilities. For instance, Ulrich et al. (2008) presented compelling evidence that views of nature can alleviate patient pain and stress, and that access to daylight and proper lighting can mitigate depression. Furthermore, while certain factors have not been studied or reported for mental health facilities, there is growing evidence from general healthcare facilities. For example, evidence synthesis has shown the positive effect of an appropriate thermal environment and exposure to natural elements on well-being (Iyendo et al., 2016; Salonen et al., 2013). While we have provided preliminary insights and recommendations based on the current literature, it is evident that more research is essential to fully comprehend and underscore all design factors that can contribute to the therapeutic environment of mental health facilities. Furthermore, it is vital to note that the recommendations provided should be adapted to the specific constraints and requirements of individual settings and patient populations for the most effective implementation.

While we have provided preliminary insights and recommendations based on the current literature, it is evident that more research is essential to fully comprehend and underscore all design factors that can contribute to the therapeutic environment of mental health facilities.

Limitations and Future Research

We acknowledge that the review findings are rooted in studies that presented some methodological limitations. For example, most of the pre- and post-intervention studies lacked control groups or did not match the intervention wards, neither did they included controls for possible confounding factors such as changes in organizational procedures, management protocols, treatment methods, or changes in the staff's mood level. Furthermore, several studies used small convenience samples that limited the generalization of findings, and often participants were different pre- and post-intervention. However, we recognize that the transient nature of these populations means that it is almost inevitable that participants will change over time. Other limitations are related to the methodological quality of qualitative studies such as lack of triangulation and reflexibility. Additionally, the quality of the responses in some studies could have been influenced by factors such as the passage of time, the benefit of hindsight, or possibly by several social and cultural factors. Future research should address these limitations.

Moreover, the limitations extend to the methods of our review. We solely focused on empirical, peer-reviewed studies in English, thus excluding valuable insights from gray literature and non-English research. Additionally, we noticed that some studies used a general label, “adult inpatient psychiatric settings,” which might include other types of settings not specifically part of our research criteria. However, we believe that the information we obtained from these studies is significant and contributed positively to the review. For the future, we encourage researchers to specify the type of mental health setting they are studying. This extra detail will greatly help future reviews to categorize and understand the research more effectively.

During our review, we encountered difficulties in the effective application of the SIGN50 methodology (SIGN50, 2019). This primarily arose because the study designs frequently observed in this specific research domain, such as controlled pre- and post- intervention studies or qualitative studies, are not fully addressed by SIGN50. This highlights the need for alternative methodologies better tailor to assess the methodological quality inherent to the typical study designs in this area.

While our systematic review aimed to highlight the impact of individual design factors on patient outcomes, it is crucial to recognize an inherent limitation in isolating these factors in interventional studies that involved multiple design changes. Future research might consider conducting controlled studies that manipulate one design feature at a time to provide a clearer understanding of their individual impacts.

Conclusions

This systematic literature review examined the existing body of published literature on the topic and shed light on the gaps in knowledge. Although the studies included in this review have limitations, initial findings suggest various design factors that should be considered when designing or renovating a facility to enhance its contribution to the therapeutic environment. These encompass (1) patients’ personal space designed to emphasize privacy and control of the environment; (2) spaces designed to optimize daylight exposure; (3) flexible communal areas promoting activities, socialization, and designated areas for visits and spiritual reflection; (4) environments that evoke a home-like feeling, minimizing the institutional aura; (5) the inclusion of artwork across the unit; (6) open designs for nursing stations; and (7) dedicated female-only areas or wards. Future research is still needed to provide more robust evidence-based designs that align with the therapeutic and clinical environment.

Implications for Practice

• The current body of literature provides an indication of how the built environment of mental health inpatient settings can be used as a therapeutic intervention for patients.

• Based on research conducted in inpatient mental health settings, practitioners can identify and implement design factors that contribute positively to patient outcomes to produce evidence-based designs, tools, guidance, and policies.

• Several gaps in the evidence and limitations are identified, providing avenues for further investigation in this area.