A randomized controlled trial of Golden Ratio, Feng Shui, and evidence based design in healthcare

Background In a global effort to design better hospital buildings for people and organizations, some design principles are still surrounded by great mystery. The aim of this online study was to compare anxiety in an existing single-bed inpatient hospital room with three redesigns of this room in accordance with the principles of Golden Ratio, Feng Shui, and Evidence-Based Design. Methods In this online multi-arm parallel-group randomized trial participants were randomly assigned (1:1:1:1) to one of four conditions, namely Golden Ratio condition, Feng Shui condition, Evidence-Based Design condition, or the control condition. The primary outcomes were anxiety, sense of control, social support, positive distraction, and pleasantness of the room. Findings Between June 24, 2022, and August 22, 2022, 558 individuals were randomly assigned to one of the four conditions, 137 participants to the control condition, 138 participants to the Golden Ratio condition, 140 participants to the Feng Shui condition, and 143 participants to the Evidence-Based Design condition. Compared with baseline, participants assigned to the Evidence-Based Design condition experienced less anxiety (mean difference -1.35, 95% CI -2.15 to -0.55, Cohen’s d = 0.40, p < 0.001). Results also showed a significant indirect effect of the Feng Shui condition on anxiety through the pleasantness of the room (B = -0.85, CI = -1.29 to -0.45) and social support (B = -0.33, CI = -0.56 to -0.13). Pleasantness of the room and social support were mediators of change in anxiety in the Evidence-Based Design and Feng Shui conditions. In contrast, application of the design principle Golden Ratio showed no effect on anxiety and remains a myth. Interpretation To our knowledge, this is the first randomized controlled trial linking design principles directly to anxiety in hospital rooms. The findings of our study suggest that Feng Shui and Evidence-Based Design hospital rooms can mitigate anxiety by creating a pleasant looking hospital room that fosters access to social support. Clinical trial registration The trial is registered with ISRCTN, ISRCTN10480033.


ABSTRACT • Background
Many patients experience stress during hospitalization (Andrade & Devlin, 2015;Zijlstra, 2021).The theory of supportive design conceptualizes that the physical environment improves the sense of control, access to social support and positive distraction which positively effects patients' well-being (Ulrich, 1991).

• Main research question
What is the effect of a single-bed inpatient room design on patients' stress?• Design (including population, confounders/outcomes) The study is an online randomized controlled trial with a between-subjects design.Participants are assigned to one of four conditions, namely the current patient room, a patient room with architectural changes, a patient room with interior changes, a patient room with architectural and interior changes.Participants are included when they have been hospitalized in the last 5 years in a Dutch hospital.Perceived stress, sense of control, social support, positive distraction, pleasantness of the room, and environmental sensitivity are measured using a questionnaire.

• Expected results
It is expected that participants exposed to changed design features anticipated less stress than participants that are exposed to the current patient room (control condition).It is expected that this positive effect is mediated by sense of control, positive distraction, social support, and perceived pleasantness.It is also expected that participants who score higher in environmental sensitivity perceive more sense of control, positive distraction, social support and perceived pleasantness in the new design and, therefore, perceive less stress.
Patients are often concerned about their disease and recovery in inpatient settings.In these hospital settings, the impact of the design of inpatient rooms on patients is still not well understood.Knowledge regarding the influence of this environment on patients is essential for facilitating the quality of health care.But how to design patient rooms that actually improve patients' well-being?Therefore, is important to understand the holistic experience of patients in a patient room.Patient room design is receiving growing attention to lower hospital-acquired infection rates during hospitalization (Taylor et al., 2018).For that reason, hospitals are nowadays designed with an increased number of single rooms (Larsen et al., 2014).Although the building costs for Dutch hospitals are high and approximately around €3.000,-per square meter, building decisions are still mainly based on experience and intuition but not on scientific evidence (Becker & Parsons, 2007).The theory of supportive design conceptualizes that the physical environment improves the sense of control, access to social support and positive distraction which positively effects patients' well-being (Ulrich, 1991).Several studies have shown the advantages of single patient room elements.These studies showed that, for example, plants, view on nature, daylight, and opportunity for social support can positively affect patients physical and psychological well-being (Dijkstra, 2009;Suess & Mody, 2018;Ulrich et al., 2004).However, until now there have been no empirical tests of the effects of a holistic single inpatient room design on patients' well-being.
The aim of this study is to test whether design principles of a holistic inpatient room has stress reducing effects, because they improve the perceptions of control, social support, positive distraction, and perceived pleasantness of the room.

• Research question (Bank or FAIR data)
What is the effect of a single-bed inpatient room design on patients' stress?

Description study design
The study is an online randomized controlled trial with a between-subjects design.Participants are assigned to one of four conditions, namely (1) the control condition which is the current UMCG patient room, (2) a patient room with architectural changes, (3) a patient room with interior changes, (4) a patient room with architectural and interior changes.
It is expected that participants exposed to a new design anticipated less stress than participants that are exposed to the current patient room.It is expected that the positive effect of the design is mediated by sense of control, positive distraction, social support, and perceived pleasantness.
Participants perceive more sense of control, positive distraction, social support, and pleasantness of the room, and, therefore, perceive less stress.It is also expected that participants who score higher in environmental sensitivity perceive more sense of control, positive distraction, social support and pleasantness in a new design and, therefore, perceive less stress compared to the control condition.
To answer the main question participants will be asked to complete a questionnaire that will take approximately 20 minutes to complete (see Appendices).In collaboration with the Dutch Patient Federation participants will be selected and approached to ask if they wish to participate in this study.When participants are willing to participate in this study they will receive an informed consent and online questionnaire.This questionnaire contains photographs and a video of the patient room and contains general questions about the person, questions about their perception of the room and questions about the psychological well-being.• <text on screening procedure, including informed consent procedure or give reasons if one or more participants will not be asked informed consent and underpin these with strong arguments> o Total number of participants who will not be asked informed consent for screening: o Total number of UMCG participants who will not be asked informed consent for screening: • <text on recruitment, including informed consent procedure or give reasons if one or more participants will not be asked informed consent and underpin these with strong arguments> o Total number of participants who will not be asked informed consent: o Total number of UMCG participants who will not be asked informed consent: ☐ Data/biomaterials will be obtained from an already existing internal or external (UMCG/non-UMCG) bio-or databank (see Section 1. Study organization).☐ Data/biomaterials will be obtained from a previous study ('FAIR data'internal/external; see Section 1. Study organization). in case one or more participants will not be asked informed consent, the objection registry will be checked for these participants and the data from those who objected will be excluded from the analyses.NA 5.4.4Informed consent (IC): access to identifiable participant data in case one or more study team members will have access to direct/indirect identifiable participant data, informed consent will be/has been obtained for this access.NA 5.4.5 IC: Collaboration with commercial parties In case of collaboration with commercial/profit organizations, informed consent will be/has been obtained for this type of collaboration NA 5.4.6 IC: Linking with other registries In case the data will be linked with other registries, informed consent will be/has been obtained for this linkage(s) NA 5.4.7 IC: Incidental findings In case there is a risk of incidental findings, informed consent will be/has been obtained to return findings to the participant NA 5.4.8IC: FAIR Data In case data collected for the present study will be shared for future studies, informed consent will be obtained for this NA 5.4.9IC: other aspects NA or <other relevant aspects of the study for which informed consent will be/has been obtained> 5.4.10Withdrawal • Can participants withdraw informed consent before publication and will all data/ biomaterials of that participant be destroyed • Does the participant information letter contain information on how to withdraw yes yes

Research Data Management Plan (RDMP)
In this study the data will be collected, processed, and archived in accordance with the General Data Protection Regulation (GDPR) and the FAIR (Findable, Accessible, Interoperable, Reusable) principles under the responsibility of the Principal Investigator.A research data management plan (RDMP) <has been/will be> drawn up to describe the further operational details and procedures.<For details on the RDMP please find the RDMP enclosed in the Appendix> ⎕ the RDMP section below is completed ☒ a separate RDMP document will be attached to this protocol -See appendix I 5.5.1 Data collection • Only essential baseline characteristics and data required to answer the research question(s) will be collected.

yes
• Tooling (eg.software and procedures) used for collecting, processing, analysing, and storing data will be compliant with the UMCG policy and Standard Operating Procedures in the UMCG Research Toolbox.

NA
Not a UMCG research 5.5.2Anonymization and pseudonymization • Data will be anonymised during data collection (i.e.data cannot be linked back to the participant) Yes skip section 5.5.2 No <if no, explain why it is not possible to anonymise the data during collection> • Data will be pseudonymized by use of <a code list/an encryption key> during data collection.

NA
• Indirect and direct identifiable information collected will be minimized and only collected for the purpose of this study NA • Direct identifiable information (e.g.contact details, code list/encryption key/subject identification log) will be stored separately from pseudonymized data <in the electronic patient files (EPD), in an electronic file, on paper in the study file, other> NA 5.5.3Data access (during the study)

NA
• Review procedure, conditions and agreements for re-use of data and access to data by other researchers will be drawn up.

NA
• For this study a discipline specific metadata standard will be chosen (i.e. to increase interoperability and re-use). NA

Management of biomaterials
Will biomaterials be collected, processed, analyzed and/or stored for the purpose of this study Yes No skip section 5.6 5.6.1 Retrospective study (see sections 1, 5.2.2, and 5.4.1)If biomaterials will be used from a secondary/further use biobank that has not been approved by the Board of Directors of the UMCG, how will be prohibited that biomaterials necessary for future diagnostic/treatment purposes will be used in the present study.

Biomaterials collection
• Only biomaterials required to answer the research question(s) will be collected NA • What biomaterials will be collected <text> 5.6.3Pseudonymization and access to biomaterials • Does the storage unit of the biomaterials comprise information that the participant (in)directly identifies, other than the participant's number and / or the sample number.

NA
• Biomaterials can only be accessed by the Principal Investigator and study delegates after authorization by the Principal Investigator NA 5.6.4Sharing of biomaterials (during and after study completion) In case biomaterials (and data) will leave the UMCG, will you contact the loket Contract Research to arrange the proper contracts?(Loket_Contract_Research@umcg.nl)NA 5.6.5 Biomaterials storage (during and after study completion) • Where and how will the biomaterials be stored • Biomaterials will be stored 15 years after the study is completed NA • What will be done with the remaining biomaterials after study completion (eg.destroyed, returned to biobank/previous study, stored) 5.6.6 Biomaterials re-use and access after completion of the present study NA ☒ skip section 5.6.6 • Biomaterials will become available and shared for re-use and participants will be asked informed consent for this ('FAIR data')

NA
• Biomaterials will be made findable by including the description of the study (and type of biomaterials in the UMCG FAIR data catalogue and other discipline specific catalogue(s).

NA
• Review procedure, conditions and agreements for re-use of biomaterials and access to biomaterials by other researchers will be drawn up. NA

Burden, Risks & Benefits (Prospective studies only)
• If participants are patients: Can be deviated from the standard care / diagnostic procedures (e.g. can medical treatment be postponed or limited)

NA
• Burden Filling in the questionnaire takes approximately 20 minutes time of participants

Data analysis
• Justification of sample size (e.g.power analysis) With an effect size of d = 0.4 this study requires a minimum of 834 participants to have a mean effect that has practical relevance (d = 0.4) with a power of 0.90 and an alpha of 0.05 (Brysbaert, 2019).This makes the probability 61% to establish the expected difference.

Participant information after the study
Will participants be informed about the study results yes/no

Research revenue
In case the study will result in revenues (e.g. as a result of the use of data/biomaterials or successful licensing of intellectual property or manufactured products), will you contact the loket Contract Research to arrange the proper contracts?

NA/yes/no
Describe what will be done with the revenues.

Algemene informatie
De Hanzehogeschool Groningen en het Universitair Medisch Centrum Groningen doen onderzoek naar het ontwerp van ziekenhuiskamers in Nederland.
In deze vragenlijst krijgt u foto's en een filmpje te zien van een ziekenhuiskamer.Er worden vragen gesteld over uzelf en over uw ervaringen en mening over een ziekenhuiskamer.

Als u niet wilt deelnemen of wilt stoppen met uw deelname aan het onderzoek
Het is aan u om te beslissen of u al dan niet aan het onderzoek deelneemt.Deelname is vrijwillig.Als u aan het onderzoek deelneemt, kunt u altijd van gedachten veranderen en besluiten te stoppen, op elk moment tijdens de vragenlijst.U kunt dan de vragenlijst afsluiten.De gegevens die tot dat moment zijn verzameld, zullen nog steeds voor het onderzoek worden gebruikt.
Om uw privacy te beschermen krijgen uw gegevens een code.Uw naam en andere gegevens die u direct kunnen identificeren worden daarbij weggelaten.In rapporten en publicaties over het onderzoek zijn de gegevens niet tot u te herleiden.
Ja Nee
Hoe lang is het geleden dat u opgenomen bent geweest in het ziekenhuis voor minimaal 1 nacht?

Nauwelijks of helemaal niet
Een  to a precision of one decimal place more than the individual measurements; standard deviation will be reported to a precision of two decimal places more than the individual measurements; and min/max will have the same precision as the individual measurements.For qualitative parameters (categorical or ordered), frequency counts and percentages of each category will be calculated by intervention.Percentages will be reported up to 1 decimal place.Baseline characteristics will be summarized overall, and by treatment when applicable.

General considerations
There will be no missing values on any of the self-reported items, because the online questionnaire procedure requires an answer to all questions to complete the questionnaire.
Missing values are only possible when participants stop halfway through the questionnaire.

Primary outcomes
Analysis of covariance (ANCOVA) will be conducted to test the main effect of the interventions on patients' anxiety and stress as well as the interaction effect of intervention and environmental sensitivity on patients' anxiety and stress.It is expected that participants who score higher on environmental sensitivity will react stronger to the intervention.Therefore, we will test whether environmental sensitivity moderates the effect of the intervention on each dependent measure (perceived anxiety, sense of control, positive distraction, social support, atmospherics of the patient room, and pleasantness of the room).Hayes Process Macro (Hayes, 2013) will be used to test the moderation effect.

Secondary outcomes
Separate linear regression analyses will be conducted to test the association of the intervention on sense of control, positive distraction, social support, atmospherics of the patient room, and pleasantness of the room.

Template non-WMO Study Protocol, version 7; Page 44 of 48
It is expected that patients perceive less anxiety when they perceive more sense of control, positive distraction, social support, atmospherics of the patient room, and pleasantness of the room.Therefore, we will test for the indirect effect of the intervention on anxiety through the five mediators.This mediation analysis includes intervention as independent variable, and sense of control, positive distraction, social support, atmospherics of the patient room, and pleasantness of the room as mediator, and perceived anxiety as outcome.Hayes Process Macro (Hayes, 2013) will be used to test the mediation effect.

Statistical methods used for efficacy
For quantitative parameters, descriptive statistics will be mean, standard deviation (or SEM), or median and minimum-maximum, and number of valid observations by intervention group, depending on normality of data.Mean or median will be reported to a precision of one decimal place more than the individual measurements; standard deviation will be reported to a precision of two decimal places more than the individual measurements; and minimummaximum will have the same precision as the individual measurements.
For qualitative parameters (categorical or ordered), frequency counts and percentages of each category will be calculated by intervention group.Percentages will be reported up to 1 decimal place.Where applicable, 95% confidence intervals around percentages will be calculated.
A 2-tailed P-value less than 0.05 indicates statistical significance.All analyses will be performed using SPSS.
Template non-WMO Study Protocol, version 7; Page 45 of 48

Presentation of results
Statistical output will be generated in accordance to this Statistical Analysis Plan and results will be presented in accordance with international standards for reporting clinical studies, as a manuscript to be submitted to a medical journal (CONSORT) In addition, results of all analyses will be presented in more detail using predefined tables of baseline data, efficacy results, as specified in the Tables and Figure below.

(See Appendix III Informed consent and questionnaire)
• Direct identifiable information can only be accessed by the Principal Investigator and study delegates after authorization by the Principal Investigator.storagedat the Hanze University of Applied Sciences with strict Hanze University security and back-up policy.•Paper source data and study files will be archived within the UMCG facilities.NA • Source data, study files and digital data will be stored 15 years after the study is completed.
• Data will become available and shared for re-use and participants will be asked informed consent for this ('FAIR data') NA • Data will be made findable by including the description of the study (and type of data (i.e.metadata) in the UMCG FAIR data catalogue and other discipline specific catalogue(s).

•
Will the participants risk any injuries and/or other discomfort when they participate in the proposed study Procedure to assess if a finding should be returned to the participant, or not o Procedure to inform the participant

•
Statistical analysis: Describe in detail the statistical analysis used to test the research question.See document 80D1.01TPL Statistical Analysis Plan, on DocPortal for guides lines.

Table 4
Baseline data -self reported psychophysical well-being

Table 5
Descriptive statistics

Table 6
Results Mediation Analysis on patients' anxiety and sense of control

Table 7
Results mediation analysis on patients' anxiety and positive distraction

Table 9
Results mediation analysis on patients' anxiety and perceived pleasantness Figure 2 Presentation interaction effect of intervention and environmental sensitivity on patients' anxiety