Field Observation in Mental Health Inpatient Settings: An Integrative Review and Reporting Guideline

Abstract

Using field observation in mental health settings is challenging and necessitates careful planning and transparent data collection and analysis. Lack of requirements for reporting observation methods increases the potential for researchers to omit important details, leading to ambiguity in interpretation, challenges in replicating the study, and misinformed conclusions. Despite the availability of reporting guidelines for qualitative methods such as interviews and focus groups, guidelines or frameworks for reporting field observation methods are lacking. There is a need to address this gap by starting with an examination of current research that employs and reports on observation methods. As a result, we conducted an integrative review to analyze and synthesize research studies that implemented qualitative observation in mental health inpatient settings. We navigated the detailed reporting of contextual information, bias, rigor, participants, observer information, observation procedures, ethical considerations, data analysis, and limitations. Informed by the strengths and limitations of the observation methods used and described in the literature, we developed a guideline for reporting field observation research methods in mental health inpatient settings. The guideline outlines the key characteristics and information to include when reporting observation procedures. The guideline has five categories covering: i) context, ii) access and ethics, iii) observer-related factors, iv) observation procedures, and v) data collection and analysis. It is designed for reporting field observation methods, but may also help researchers plan and conduct field observation.

Plain Language Summary

Research using direct observation of patients, health care providers (HCPs) and their daily activities on mental health inpatient units can provide important information. Observation can help understand the situations and challenges that patients and staff experience while receiving or providing mental health care. However, observing and collecting data in mental health settings can be difficult and requires careful planning. Researchers may not have clear guidelines on how to report their observation methods and important details can be missed. This makes it hard to understand the research, repeat the study, and reach accurate conclusions. While there are good reporting guidelines for interviews and focus groups, reporting guidelines are lacking for field observation. To fill this gap, we reviewed existing research studies that used observation methods in mental health inpatient settings. We focused on relevant aspects of each study, such as, the purpose and context, the people involved, ethical issues, the procedures used, potential biases and how the data were analyzed. Based on our review findings, we created a guideline to help researchers report their observation methods more clearly. The guideline includes five categories: study context, ethical considerations, factors related to the observer, the observation procedures used, and how data are collected and analyzed. The guideline not only helps with reporting but can also guide researchers in planning their observation studies.

Mental health services are provided in a variety of community and institutional settings for individuals affected by mental illnesses. Supports and services range from self-help and peer support, virtual and digital resources, to inpatient admissions that may include involuntary detainment and treatment (Mental Health Commission of Canada, 2024). Individuals receiving mental health services frequently struggle with substance use disorders in addition to other physical comorbidities (Ngune, 2024; Sartorious, 2013). Risks such as self-harm, suicidal intent, and aggressive behaviors may also be prevalent for those living with mental illness (Sadath et al., 2023; Scocco et al., 2019).

Conducting research that focuses on mental health issues is challenging as the issues are sensitive, subjective and more difficult to measure than many other health conditions (Malla et al., 2015). The cognitive and decision-making abilities of individuals with mental illness may also be impaired due to the nature of the mental illness and the side effects of psychotropic medication. This adds complexity to different aspects of the research process, including the informed consent process, participant recruitment, data gathering, and outcome evaluation (Dalal, 2020). Furthermore, from the individual’s perspective, feeling labelled (stigmatized) or feeling that their health concerns are not taken seriously (epistemic injustice) may foster reluctance to engage in research (Knaak et al., 2017). Researchers must think profoundly and deliberately about ethical issues and problems, ensure the research encounter protects the dignity and safety of all participants, and use rigorous research methods to produce high-quality results.

Observation research, one type of qualitative research, offers researchers a different lens from which to explore the various dimensions of a mental health setting (Johnston et al., 2024), For inpatient settings, it can illustrate what is enacted within the day to day routines, activities and person to person interactions among individuals receiving care, their families, and healthcare professionals (Palinkas, 2014). The observation method is valuable for providing detailed data in a naturalistic way (Mulhall, 2003). This data can foster insights not normally accessible through other research approaches, such as structured interviews or quantitative methods (Carnevale et al., 2008). Observation findings have strength on their own, but when used alongside other qualitative methods, can add visibility to what is said or claimed by having captured what is actually done.

Observation research originated in the social sciences and has been slow to be adopted in health research (Weston et al., 2021). This resulted, in part, from the past reluctance of health research ethics boards to accept observation as a viable method in human health research (Oeye et al., 2007). This position has changed and observation is now emerging as a valuable and unique method in health research (Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, and Social Sciences and Humanities Research Council of Canada [CIHR, NSERCC, & SSHRCC], 2022). However, as health researchers turn to observation to provide an additional perspective on clinical issues, it is essential that effective resources and guidance are available to support rigorous observation research.

Fortunately, guidance on the ‘doing’ of observation research is readily available and well established in the scholarly literature (Howell, 2018; De Munck & Sobo, 1998; Festinger et al., 1969). From the well-known seminal works of Gold (1958) and Spradley (1980) to Hammersley and Atkinson (2007) and individual and team accounts of how observation methods were used in specific research studies (Johnston et al., 2025), there are many approaches to guide researchers in conducting observation research (Fetters & Rubinstein, 2019). Ethnography, with a long history of observation research, is the methodology that is used by many health researchers when conducting observation research in health settings (Horrill et al., 2024; Molloy et al., 2019; Waldemar et al., 2019). The guidance in this area is robust.

What is noted from our review of the literature, however, is the apparent lack of any reporting guidelines for describing the observation method and how it was implemented in a study. The core features of the method have not yet been delineated for the purpose of reporting on the quality and strength of the method. The aim of this integrative review is to initiate a process that will lead to the development of standardized reporting guidelines when using observation methods in mental health settings. It will begin with a systematic, integrative review of qualitative research that involves the use of observation as a data collection method in mental health inpatient environments. The findings from this review will contribute to the guideline development.

Background

Observation research conducted on mental health inpatient units has been used to examined a variety of phenomena and identify a range of issues that have clinical implications (Hagen et al., 2020; Sercu et al., 2015). Observation as a method involves the systematic description of events, artifacts, documents, and records of interactions, behaviors, experiences, cultures, or phenomena (Marshall & Rossman, 2014). It allows researchers to understand nonverbal cues, interactions, time spent in events, and diverse modes of communication.

Observation as a method can be implemented in several different ways. The observer can be an active participant in the setting, termed participant observation, or maintain a quiet, unobtrusive presence in the setting, termed nonparticipant observation (Spradley, 1980). The observation can be structured or unstructured and used alone or with other research methods such as interviews, focus groups, and document analysis (Fix et al., 2022; Palinkas, 2014). In a structured (a priori) observation, the observer is focused on specific events or phenomena; in an unstructured observation, the observer freely captures what is happening at the moment (Fetters & Rubinstein, 2019). In addition, the observer’s membership (engagement and relationship with participants) can be varied and needs to be described (Gold, 1958; Spradley, 1980). Attention to detail, such as careful planning, observer training, minimizing observer bias, a well-designed sampling strategy, consistency of the observation process, ethical consideration, and transparent data analysis, is necessary to ensure the observation method’s trustworthiness (Fix et al., 2022).

Respecting the rights and welfare of individuals with mental illnesses are crucial considerations when conducting research (Harris et al., 2021). Due to the sensitive nature of gathering observation data in secure mental health inpatient units, conducting the observation method should be justified, well-planned, and reported in detail (CIHR, NSERCC, SSHRCC, 2022, sect.10.2,). Strategies to safeguard patients and healthcare personnel should also be carefully considered (Harris et al., 2021). Privacy and confidentiality are paramount. Consequently, the transparency and trustworthiness of the researchers about their observation techniques play an important role in participant engagement (Wang et al., 2015). Researchers must be forthright about the choice of site selection, the target population, the duration and frequency of observation, how data will be collected and analyzed, and the research context.

As noted, specific guidelines or frameworks for reporting observation methods are lacking, which forces researchers to use general recommendations/guidelines to direct their reporting. For example, the Standards for Reporting Qualitative Research (SRQR) provides general guidelines for reporting qualitative research (O'Brien et al., 2014), the Consolidated Criteria for Reporting Qualitative Research (COREQ) focuses on interviews and focus groups (Tong et al., 2007), and the Journal Article Reporting Standards for Qualitative Research (JARS-Qual) recommend what should be reported in qualitative research (Levitt et al., 2018). Researchers may also use established frameworks for other forms of qualitative research, such as ethnography or case study research, to guide their observation research. Specific research methodologies, textbooks, and other publications may offer instructions on applying field observation. Nevertheless, without guidelines or frameworks that provide recommendations for reporting observation methods, researchers may omit important details, leading to ambiguity in interpretation, challenges in replicating the study, and potentially misinformed conclusions.

Considering the importance of reporting observation methods in general and the uniqueness of mental health settings, we sought to review, analyze, and synthesize research articles that use qualitative observation in mental health inpatient settings. We examined the detailed reporting of contextual information, bias, rigor, participants, observer information, observation procedures, data analysis, and ethical considerations. Using the strengths and limitations of the observation methods identified in the review and additional literature, we have developed a reporting guideline for applying and reporting future research using field observation methods in mental health settings.

Aim and Review Questions

This integrative review aims to analyze the use of field observation research in inpatient mental health settings. We addressed the following research questions:

• How have field observation research methods been used in mental health settings?

• Based on the reviewed literature, what recommendations can be made for reporting and applying field observation research in mental health settings?

Methods

The studies that employed the observation research method in mental health inpatient settings were identified, analyzed, synthesized, and themed using an integrative literature review methodology. An integrative review effectively combines the data from numerous studies and, for this review, allows a more thorough understanding of the procedures used during observations, including findings from various research designs. The review process can shed light on the methodological strengths and limitations of the method and help guide future research design and reporting. The integrative review also guides the creation of recommendations, frameworks, policies, and guidelines (Whittemore & Knafl, 2005). In this study, we sought to better understand how the observation method is used in mental health inpatient environments to inform the development of a reporting guideline for field observation research methods. To ensure the strength of the review process and the credibility of the conclusions, we followed the Whittemore and Knafl (2005) framework to guide this review paper: identifying the problem, searching the literature, evaluating data, analyzing data, and presenting the synthesized findings (Whittemore & Knafl, 2005).

Searching the Literature

We systematically searched the following databases: MEDLINE (via Ovid), CINAHL Plus (via EBSCOhost), Embase (via Embase.com), and APA PsycINFO (via EBSCOhost) for studies published between January 01, 2012, and June 10, 2022. Ten years was purposefully chosen to ensure a manageable number of articles for analysis. The four databases were searched because together they provide good coverage of psychological and psychiatric literature and qualitative research. We used appropriate combinations of keywords and subject terms related to observation, ethnography, researcher as participant, video recording, mental disorders, and mental health patients and facilities. The search strategies were written by a health sciences librarian (KR) in consultation with other team members. Detailed search strategies are available in Appendix I.

Inclusion and Exclusion Criteria

In this review, we considered original, qualitative articles only. We did not consider quantitative articles, literature reviews, editorials, letters, commentaries, opinion papers, case reports, nor grey literature. We included studies that met the following criteria: i) primary qualitative research articles that ii) used qualitative observation research methods, iii) were published in the English language, and iv) were conducted in adult inpatient mental health settings. We excluded studies that i) did not provide a description (or that provided only a short description) of the observation methods, ii) were conducted in forensic settings, nursing homes, or community settings, iii) had participants that were all individuals with dementia, iv) used quantitative analysis, or v) had no full text available. We excluded studies conducted in forensic settings due to the different contexts, goals, populations, and treatment approaches used, as compared to adult inpatient settings. We excluded studies conducted in individuals with dementia due to their specific cognitive function nature, which did not represent the population admitted to acute inpatient services.

Evaluating Data

All citations identified by the searches were loaded into EndNote 20 (Clarivate, PA, USA), and duplicates were removed. The remaining citations were loaded into Covidence (Veritas Health Innovation, Melbourne, Australia). Two independent reviewers (AD and WJM) assessed the titles and abstracts against the eligibility criteria for this review, the conflicts were solved by consensus, and non-eligible articles were removed. A full-text review was conducted for the remaining articles and assessed against the eligibility criteria by the same independent reviewers; conflicts were solved by consensus, and non-eligible articles were removed. Thirty articles met the eligibility criteria and were considered for analysis. Two independent researchers (CI) and (KR) randomly checked the extraction tables to confirm and validate the findings with their sources.

Data Extraction

We collected and reported information on the following 30 data items: study title, authors, year of publication, country in which the study conducted, aim of the study, study setting, study design, conceptual/theoretical framework, type of observation (participant or non-participant), observer position (external or internal), nature of observation (overt or covert), observation agenda (structured, semi-structured, or unstructured), length per observation, total length of observation, time period of the observation, description of observation, reference for observation method used, number of observer(s), observer role (complete observer, observer as participant, participant as observer, or complete participants), use of other data collection methods (e.g., interview, focus group), population description, sampling methods, total number of participants, reported ethical concerns, informed consent process, type of informed consent (individual or organization), data analysis methods, if a separate analysis of the observation data was completed, that is, not merged with other data collection methods, main findings, and reported limitations regarding the observation.

Methodological rigor was assessed using the JBI Quality Assessment Tool for qualitative studies (Lockwood et al., 2015) (Table 1). The studies were categorized as weak, moderate, or high quality according to a weighted approach that gives more weight to questions related to the concepts of dependability in qualitative research (Questions 2, 3, 4, 6, and 7) (JBI, 2016).

Table 1.

JBI Quality Assessment

Study authors	1*	2	3	4	5	6	7	8	9	10	Rating
1 Allard and Bleakley (2016)	Yes	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	Yes	Yes	High
2 Berlin and Carlström (2015)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	No	Yes	High
3 Blacker et al. (2017)	Yes	Yes	Yes	Yes	Yes	No	Yes	Unclear	Yes	Yes	High
4 Deacon and Cleary (2013)	Yes	Yes	Yes	Yes	Yes	Unclear	No	Unclear	Yes	Yes	Moderate
5 Ellegaard et al. (2020)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
6 Grant et al. (2014)	Yes	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	Yes	Yes	High
7 Gurtner et al. (2022)	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	High
8 Hagen et al. (2020)	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	High
9 Hannah (2018)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
10 Holyoake (2013)	Yes	Yes	Yes	Yes	Yes	Unclear	No	Yes	Yes	Yes	Moderate
11 Hung et al. (2014)	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	High
12 James et al. (2017)	Unclear	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
13 Kallakorpi et al. (2019)	Unclear	Yes	Yes	Yes	Yes	No	Yes	No	Yes	Yes	High
14 Kessing (2020)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
15 Lindgren et al. (2015)	Unclear	Yes	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	Moderate
16 Maddock (2015)	Unclear	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
17 McAllister et al. (2021)	Unclear	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	High
18 McKeown et al. (2020)	Unclear	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	High
19 Molloy et al. (2019)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
20 Nagayama and Hasegawa (2014)	Yes	Yes	Yes	Yes	Yes	No	No	Yes	Yes	Yes	Moderate
21 Nossek et al. (2021)	Unclear	Yes	Yes	Unclear	Yes	No	Yes	Yes	Yes	Yes	Moderate
22 Rio et al. (2017)	Yes	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	Yes	Yes	High
23 Rise et al. (2014)	Unclear	Yes	Yes	Yes	Yes	No	Yes	Unclear	Yes	Yes	High
24 Salzmann-Erikson (2015)	Unclear	Yes	Yes	Yes	Yes	Unclear	Yes	Unclear	Yes	Yes	High
25 Sercu et al. (2015)	Unclear	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	High
26 Terkelsen and Larsen (2013)	Unclear	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	Yes	Yes	High
27 Ulfseth et al. (2015)	Unclear	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	High
28 Unhjem et al. (2018)	Yes	Yes	Yes	Yes	Yes	Unclear	Yes	Yes	Yes	Yes	High
29 Voogt et al. (2014)	Unclear	Yes	Yes	Yes	Yes	No	Yes	NA	Yes	Yes	High
30 Waldemar et al. (2019)	Unclear	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	High

Assessment Questions: 1) Is there congruity between the stated philosophical perspective and the research methodology? 2) Is there congruity between the research methodology and the research question or objectives? 3) Is there congruity between the research methodology and the methods used to collect data? 4) Is there congruity between the research methodology and the representation and analysis of data? 5) Is there congruity between the research methodology and the interpretation of results? 6) Is there a statement locating the researcher culturally or theoretically? 7) Is the influence of the researcher on the research, and vice- versa, addressed? 8) Are participants, and their voices, adequately represented? 9) Is the research ethical according to current criteria or, for recent studies, and is there evidence of ethical approval by an appropriate body? 10) Do the conclusions drawn in the research report flow from the analysis, or interpretation, of the data?

Data Analysis

We reported study characteristics using frequencies and key percentages and grouped the extracted findings according to the various procedures used when conducting observation as a method. Synthesis was used to determine how observation procedures were alike and different across the 30 studies. We identified key items from the data describing the observation and organized them into a guideline containing recommendations for reporting. In addition, to better understand the use of observation as a method in mental inpatient settings, we analysed and synthesized data related to i) the study participants and the focus of the inquiry, ii) the findings that were generated, and iii) the limitations that researchers identified when conducting observation research.

Results

Study Selection

Our search in the selected databases yielded a total of 1,736 articles. After removing duplicates and excluding 1,008 articles deemed irrelevant based on their titles and abstracts, we fully screened 115 articles. Of these, 85 were excluded for not meeting the eligibility criteria. The primary reasons for exclusion were:

• Lack of description of the observation method (18 articles)

• Not using observation as a research method (13 articles)

• Studies conducted in forensic settings (10 articles)

• Studies not conducted in inpatient psychiatric health settings (9 articles)

Consequently, 30 articles were included in this review. Detailed information is presented in the PRISMA flowchart (Figure 1) (Page et al., 2021)

Figure 1.

PRISMA Flowchart

Methodological Rigor

Of the studies reviewed, 19 were deemed high quality, while 11 were rated moderate. The primary concerns regarding rigor in some articles were the failure to address congruities between the stated philosophical perspectives and the research methodologies, ambiguity concerning the researcher’s cultural or theoretical stance, unaddressed influences of the researcher on the research (and vice versa), and the inadequate representation of participants and their perspectives (Table 1).

Study Characteristics

Studies were carried out in the following countries: the United Kingdom (n = 7, 23%), Norway (n = 5), Sweden (n = 3), Denmark (n = 3), Canada (n = 2), Australia (n = 2), the United States (n = 1), Finland (n = 1), Japan (n = 1), Belgium (n = 1), Germany (n = 1), Ireland (n = 1), the Netherlands (n = 1), and Switzerland (n = 1). Most studies were published between 2013 and 2015 (n = 14, 47%); ten studies were published between 2019 and 2022, and six were published between 2016 and 2018. As for study designs, 19 (63%) were ethnographic, three were case studies, three utilized grounded theory, and the remaining had varied designs described as qualitative process evaluation, participatory action research, qualitative descriptive, and qualitative-empirical research. The study design was unclear in one study.

The Scope of Observation as a Method

Studies in this review were categorized by participant type as a first step in the synthesis process. This set of categories highlights the source of the evidence and establishes the focus of the observation. Of the 30 studies, seven (23%) had only participants that were individuals admitted to inpatient mental health wards (Ellegaard et al., 2020; Hagen et al., 2020; Hung et al., 2014; Kessing, 2020; Lindgren et al., 2015; McAllister et al., 2021; Ulfseth et al., 2015). Observation of these participants addressed some aspects of their current hospitalization, such as, relationships that develop among people receiving care, how patients experience suicidality, the features of daily life on the ward, and the perspective of those who can self-admit when needed.

Ten studies had a one-health-discipline-only sample, mostly nurses (Deacon & Cleary, 2013; Holyoake, 2013; Kallakorpi et al., 2019; Molloy et al., 2019; Nagayama & Hasegawa, 2014; Sercu et al., 2015; Unhjem et al., 2018; Voogt et al., 2014), except for one social work sample (Nossek et al., 2021) and one mixed health care professional sample (Rio et al., 2021). Studies of nurses working in mental health wards primarily reflect what nurses do, such as, nurses’ occupational activities, how different populations are cared for by mental health nurses, nursing interventions delivered by nurses, how nurses included the views of patients in care planning meetings, and how nurses engage with distressed individuals while maintaining selfhood. The remaining 13 studies (43%) had mixed samples, which usually comprised patients and health providers (Allard & Bleakley, 2016; Berlin & Carlström, 2015; Blacker et al., 2017; Grant et al., 2014; Gurtner et al., 2022; Hannah, 2018; James et al., 2017; Maddock, 2015; McKeown et al., 2020; Rise et al., 2014; Salzmann-Erikson, 2015; Terkelsen & Larsen, 2013; Waldemar et al., 2019). These studies, with both patients and HCPs as participants, explored the experiences of coercion on locked wards, ward culture regarding restraint use, the actions of patients and staff in the context of involuntary confinement, multidisciplinary teamwork and recovery focused practices in interactions between patients and nurses (Table 2).

Table 2.

Review Study Characteristics: Study Aim, Design, Participants and Conceptual Framework

Authors and (study date)	Aim(s) of study	Study design	Participant description and setting	Number of participants	Conceptual theoretical framework
1. Allard and Bleakley (2016)	To determine how clinical teams will make meaning of top-down policy directives perceived as obstructive to bottom-up, tailored, patient-centred care in the face of throughput pressure	Ethnography	Multidisciplinary healthcare practitioners in one emergency department (ED) and one mental health ward (MHW) in two distinct NHS England healthcare settings. A large teaching hospital (ED) and a medium-sized acute admissions MHW serve a predominantly rural population in the United Kingdom	64 practitioners were recruited for observation and video recording	Cultural-historical activity theory (CHAT)
2. Berlin and Carlström (2015)	To find out how artifacts are camouflaged and to determine the significance of this camouflaging	Ethnography	Professionals such as physicians, psychiatric nurses, medical secretaries, psychologists, occupational therapists, physiotherapists, treatment assistants, consultants, attendants and social welfare officers were involved in five psychiatric wards at a university hospital in Sweden	60 staff were interviewed, and 66 observations conducted	None
3. Blacker et al. (2017)	To understand the complex healthcare cultures in the National health service (NHS), focusing on an acute adult in-patient psychiatric ward	Unclear	Three groups of participants: Members of the supervisory group, staff on the ward at which the observations took place, & patients in an acute psychiatric ward within a mental health unit located in a rural location in the Midlands, United Kingdom	Not reported for observation	None
4. Deacon and Cleary (2013)	To explore the occupational activities of mental health nurses in an acute inpatient mental health ward	Ethnography	Nurses and nursing assistants and whomever they came into contact with during the periods of participant observation in an inner-city mental health unit in England, for example, patients and psychiatrists	18 registered nurses and 16 nursing assistants were observed during the study	None
5. Ellegard et al. (2020)	To develop a grounded theory of patients’ experiences with patient-controlled admission	Grounded theory	Mental health patients. The study was part of a 3-year Danish multicenter project. The multicenter project took place in 11 mental health units, and 10 of those participated in this study. Each unit was at a public mental health hospital in one of the five Danish regions	26 patients	None
6. Grant et al. (2014)	To examine the implementation of a smoke-free grounds policy through the lens of culture within psychiatric settings in Northern British Columbia	Ethnography	Tobacco-related interactions within and between health care professionals and patients in two hospital-based adults (one urban and one rural), inpatient psychiatric units in Northern BC, Canada	Participant observation with 11 participants & 19 interviews with HCPs & 2 staff members at two hospitals	None
7. Gurtner et al. (2022)	To examine the nature of interactions and communication between health care professionals and patients during regular interprofessional psychiatric consultations	Ethnography	Patients, psychiatrists, junior physicians, nurses, psychologists, social workers, and therapists in two different acute psychiatric inpatient wards in a large representative Swiss psychiatric hospital	71 consultations involving (ward A) 37 patients, 3-5 junior doctors, 1–3 nurses & 0-1 consultants. Ward B: 34 patients, 3-4 junior physicians, 2 nurses, & 2–5 allied health professionals	None
8. Hagen et al. (2020)	To explore how patients experience their suicidality and how they experience being in a psychiatric hospital	Ethnography	Women who were voluntarily admitted to two general psychiatric acute wards at a psychiatric hospital in Norway & were struggling with suicidality	Eleven women	The theory of liminality
9. Hannah (2018)	To examine how clinicians attach notions of culture to the social categories they use in their work	Ethnography	Clinicians, staff, & management acute psychiatric inpatient unit in a private psychiatric hospital in Massachusetts, United States	Not reported for observation	None
10. Holyoake (2013)	To access and produce ethnographies about the cultural meaning associated with inpatient mental health nursing observation	Ethnography	Nurses from six psychiatric inpatient units in the United Kingdom	Not reported for observation	The Goffman framework of inmate (service user) and staff world (nurses)
11. Hung et al. (2014)	To explore how the physical environment in a geriatric psychiatry unit may play a role in either supporting or obstructing patient and family care needs	Focused ethnography	Geriatric patients with depression, dementia, or both with responsive behaviours and four family members in a locked 16-bed geriatric psychiatry unit in a 60-year-old urban community hospital in Canada	Seven participants: Patients & family members	None
12. James et al. (2017)	To explore implementation of a complex intervention (safewards) on mental health wards during a cluster RCT	Qualitative process evaluation	Staff members from 16 acute psychiatric wards for adults of any gender in England	564 unit staff members	None
13. Kallakorpi et al. (2019)	To describe nurses’ experiences caring for immigrant patients in psychiatric units	Focused ethnography	Nurses working at adult psychiatric units in a university hospital district in Finland cared for the immigrant patients who participated in the study	Five nurses and 9 patients	None
14. Kessing (2020)	To explore the unfolding social relations between inpatients in psychiatric wards (how inpatients supported one another and used these relations to challenge the mental health professionals)	Ethnography	Individuals who were admitted to two psychiatric wards in Denmark. An acute ward and open ward in the secondary sector	Fourteen patients	A relational sociological approach
15. Lindgren et al. (2015)	To describe the features of everyday life in psychiatric inpatient care as experienced by women who self-harm	Ethnography	Women with self-harming behaviour in two locked acute psychiatric wards in Sweden	Six patients	None
16. Maddock (2015)	To explore multi-disciplinary teamwork in mental health settings, particularly what aids & hinders the process of multidisciplinary team-working and the social work contribution in such teams	Case study	Multidisciplinary inpatient mental health team meeting in Irish mental health service team: the clinical lead, social worker, community psychiatric nurses, psychologist and occupational therapist	Five multidisciplinary team members from one mental health team were observed during four team meetings	Effective collaborative practice
17. McAllister et al. (2021)	To understand i) how engagement is experienced in acute wards & ii) needs of service users, carers, & HCPs to inform future collaborations	Participatory action research	Nurses’ patient interaction in community and acute mental health ward in inner London, United Kingdom	No information regarding observation	None
18. McKeown et al. (2020)	To explore ward cultures about the use of restraint & the impact of the REsTRAIN yourself (RYS) intervention	Rapid ethnography	Staff and service users in 14 acute mental health wards within seven NHS trusts across the North west of England	Unclear	None
19. Molloy et al. (2019)	To explore the culture of mental health nursing practice re: The care of aboriginal and Torres strait islander mental health service users	Ethnography	Professional mental health nurses practise in two regional health services in Queensland and New South Wales, Australia	Unclear	The paradigm of multi-sited ethnography
20. Nagayama and Hasegawa (2014)	To elucidate the process of nursing care that nurses provided for patients in long-term seclusion to guide them towards release from seclusion	Grounded theory	Nurses in 3 long-term psychiatric wards with seclusion rooms at two institutions in Japan	Eighteen nurses	Symbolic interaction
21. Nossek et al. (2021)	To examine the integration of peer support workers in adult psychiatric hospitals in Germany. To clarify how PSWs perceive their roles and how the latter evolve over time	Qualitative-empirical research	Multidisciplinary team members & peer support workers at their regional meetings in 3 adult psychiatric hospitals in Germany	Unknown number of HCPs who attended 25 interprofessional team meetings & 5 transregional peer support worker meetings	None
22. Rio et al. (2021)	To explore how mental health nurses make accountability decisions about the involvement of consumers in care planning	Focused ethnography	Twelve nurses and additional 23 multidisciplinary HCPs agreed to be observed during care planning practices on one publicly funded inpatient unit in an Australian capital city	35 MH professionals Inc. Nurses who participated in care planning mtgs on one unit over six months	Recovery philosophy
23. Rise et al. (2014)	To investigate the experiences of service users & HCPs who took part in implementation of a comprehensive development plan to enhance user involvement in a mental health hospital	Case study	Health care professionals in team meetings with 3 to 15 members at each in mental health community hospital in central Norway that provided inpatient treatment, outpatient treatment, and ambulatory services	Between 3 and 15 participants at each of the 10 development meetings	None
24. Salzmann-Erikson (2015)	To describe how the limitation of patients is being practiced in psychiatric intensive care units	Ethnography	Staff (nurse assistants and registered nurses) work daily in three psychiatric intensive care units in Sweden, located in different large cities	There were 15 staff members on each unit. One nurse and 3 nurse assistants were allocated to each shift	None
25. Sercu et al. (2015)	To explore how stigma may give meaning to mental health nursing identities	Case study + ethnography	Nurses in 4 wards in two psychiatric hospitals in the region of Ghent, Belgium, each with distinct approaches to categorizing and treating mental health needs	Thirty-three nurses	Stigma for mental health service users and mental health nursing
26. Terkelsen and Larsen (2013)	To explore how patients and staff act in the context of involuntary commitment, how interactions are described and how they might be interpreted	Ethnography	Staff and people who are committed involuntarily for mental health and substance abuse problems or dual diagnosis (DD) in locked psychiatric ward in Norway	Sixteen patients, 2 psychiatrists and 20 milieu-therapists gave consent to participate	None
27. Ulfseth et al. (2015)	To explore how social relations emerge & are facilitated among people with mental illness in the context of everyday life during a psychiatric treatment stay	Ethnography	Inpatients with persistent mental illness in a psychiatric centre in a town in Norway	Ten patients	None
28. Unhjem et al. (2018)	To describe what nurses self-disclose to patients in mental healthcare and what reasons they have for self-disclosure	Qualitative descriptive	Registered nurses from four units (both open and closed units) in three districts located in a small town and rural villages in mid-Norway that attended to adult patients with diverse diagnoses	Sixteen nurses	None
29. Voogt et al. (2014)	To observe the actions of psychiatric nurses when providing structure and identify results in order to better understand providing structure as a complex nursing intervention	Grounded theory	Nurse and patient interactions in dual diagnosis unit and crisis intervention ward in a mental health care facility (intensive care wards) in The Netherlands	Participant numbers not reported. Total patients on both units was 48; total number of nurses was 30. Fifty-two ‘providing structure’ events were analysed	None
30. Waldemar et al. (2019)	To explore how recovery-oriented practice is reflected in the interactions between patients and health professionals around treatment in two mental health inpatient wards in Denmark	Ethnography	Patients staying in two mental health inpatient wards in the capital region of Denmark (an open ward and a locked intensive care ward) and the health professionals employed there, along with various other people, such as relatives visiting, health professionals from other wards, and cleaning staff	Participants: All staff patients, visitors present during observation sessions on 2 units. Approx. 36 patients and 55 HCPs had potential to be observed	None

A wide range of phenomena are examined through observation methods on mental health inpatient units. These phenomena highlight unique aspects of mental health care and treatment.

Observation Procedures

Reporting observation procedures varied among studies. We excluded many studies due to the lack of detailed descriptions of their observation procedures. The included articles provided information such as type of observation, nature of observation, observation structure, length per observation, total length of observation, number of observations, period of observation, observation description, and use of observation methods reference.

Regarding the type of observation method, 22 (73%) studies employed participant observation, while the remaining eight conducted non-participant observation. All observations were overt, with no study using covert methods. Most adopted a semi-structured approach (15 studies, 50%) when collecting observation data; seven studies used a structured approach, three opted for a non-structured method, and it was unclear in five studies. Fifteen studies did not report the duration of each observation. Among the studies that did, the average observation length was 4.5 h, with durations ranging from 15 min to 8 h. Spradley (1980) was the most frequently cited source for observation procedures or observer roles, referenced in four studies. This was followed by Miles and Huberman (1994), Gold (1958), and Hammersley and Atkinson (2007), cited in two studies. Nine studies did not provide any references for their observation methodology.

Additional details regarding observations were reported, including a description of how field notes were taken and the information included, in addition to the methods and timing of their transcription or processing. Interactions, engagements between the observer and participants, and management strategies employed during these interactions were reported. Ethical considerations were reported, particularly about participant consent and data confidentiality and how it was handled during observations. No ethical dilemmas were identified. Procedures regarding how observational data were anonymized, stored, and managed to assure participant confidentiality and data integrity were also outlined. Additionally, observer Information, such as, their background, positionality, and reflections on their role and impact on the observation, was provided in some studies. Moreover, the reports delineate how reflections and analytical thoughts were recorded and distinguished from observational notes (Table 3).

Table 3.

Study Characteristics: Observations, Additional Data Collection Methods, Analysis and Ethics

Study ID	Nature, type of obs & observer position	Description of observation	Other data collection methods	Data analysis methods	Separate obs analysis	Reported ethical concerns	Informed consent form
1. Allard and Bleakley (2016)	Non-participant observation with an external observer in observer as participant role	One observer collected semi-structured data on interactions between interdisciplinary staff & artifacts focused on patient quality & safety. Working locations in 2 settings, a psychiatric ward & an emergency department, were used to conduct observations. Data were collected over 12 months, six months in each location. A video record (to validate field notes) was used. Total observation hours = 320. Total video hours = 21	Interviews	Data treated as narratives, analyzed thematically and theorized using cultural-historical activity theory (CHAT)	No	A period of researcher familiarization was supported by ethics committee	Sixty-four practitioners provided individual consent
			Informal conversation			This allowed the researcher to meet informally with many practitioners before initiation of the study	Patients have not consented
2. Berlin and Carlström (2015)	Non-participant observation with an external observer in observer as participant role	One observer collected unstructured data on the actions/events of health staff focused on the concealment of select artifacts. Sixty-four observation sessions, totalling 87 h, were conducted over 48 months (2008 – 11). Each observation was b/t 1.0 and 2.5 h	Interviews	Inductive analysis inspired by grounded theory	No	None	Not reported
			Artifacts review
3. Blacker et al. (2017)	Non-participant observation with an external observer in the observer as participant role	The observer conducted non-structured observations to explore multidisciplinary teamwork in mental health settings & the social work contribution in teams. Following the 1-h session (total 6 h), detailed process notes were written for discussion with their supervisory committee	Supervisory discussions	Thematic analysis	No	None	Not stated
4. Deacon and Cleary (2013)	Participant observation with an internal observer in participant as observer role	Field notes & informal talks were recorded - reflexive & analytic comments made by the observer. Data consisted of the materials & resources used by nurses during their work. Focus was the physical fabric of the setting, the people within it, significant events & ward routines. Data were collected over two years	None	Used constant comparative method & deviant cases for validity. Themes emerged & were developed during data collection	a Yes	None	Ethics approval included organizational consent meaning consent from individuals during observation periods was not required
				Themes were then returned to the field notes as an analytic framework for their coding indexing technique			However, informed consent was not assumed. The researcher sought consent from every individual encountered while data collecting. Private spaces such as bedrooms were never entered without invitation
5. Ellegaard et al. (2020)	Participant observation with an external observer in the role of the participant as observer	Two days (16 h) of semi-structured field observations conducted on two units by one observer to capture patient concerns, attitudes & strategies with the patient-controlled admission. Focus was on verbal & nonverbal interaction between patients and the HCPs, & observation of reflexive sessions as a research team to identify biases	Interviews surveys	As required for classic grounded theory studies. The study used the constant comparative method with iterative data collection and analysis	No	None	Participants received oral and written information about the study before data collection and again on the day of data collection. All participants provided written consent
			& a personal & social performance scale
6. Grant et al. (2014)	Participant observation with external observers in the role of observer as participant	Participant observations focused on tobacco-related interactions within and between HCPs and patients for which participant consent was obtained. 69.5 h of observation. Observations took place at various times during the day and all days of the week (2009-2010)	Interviews	Inductive thematic analysis	No	None	Researchers obtained individual consent for participants. Informal talk was collected as part of the observation
			Document review
			Fieldwork, informal conversation
7. Gurtner et al. (2022)	Participant observation with an external observer in the role of observer as participant	One observer was on the ward for seven 4-h periods & attended the consultations b/t Dec 2018 - March 2019. Structured observations using a focused guide - OPTION - were developed to capture HCP’s skills in engaging individuals in decisions & extent to which pts were involved in decision-making. Informal conversations before or after the consultations were also collected using unstructured field notes. A qualitative ethnographic design was chosen	Spontaneous conversations with the participants	Thematic analysis	a Yes	None	Managers of wards a & B distributed written study information to the HCPs & obtained informed consent for observation
							Patients in ward a & B were contacted by the manager or a nurse from the ward & were informed about this study and, if willing, gave their oral and written individual consent
8. Hagen et al. (2020)	Participant observation with an external observer in observer as participant role	One observer (senior mental health nurse) conducted 160 daytime hours of observation during three months in 2018. Observer was present (usually silent) in one or two of the conversations nine participants had with their health team member. Observer wrote field notes continuously & extensive reflections at the end of the day. Field notes contained informal conversations with patients and observations in the ward and reflections on both	Interviews	Inductive systematic text condensation, a form of thematic analysis, was followed by a deductive process where the data and preliminary findings were interpreted in light of liminality. Theory	No	None	The participants signed an individual informed consent form to allow the researcher to conduct participatory observation in the ward after consultation with the therapists
			Conversations
9. Hannah (2018)	Participant observation with an external observer in the participant as observer role	One observer on 1 ward collected semi-structured data on how clinicians think of racial & ethnic identity & how they account for it when treating patients. Observer attended group sessions that involved most individuals on the psychiatric ward, arts and crafts groups, and team meetings. Observer interacted with all staff & those receiving care. Observations occurred 6 h daily, 2-3 days per week over nine months	Interviews	Analysis of interview transcripts and observational fieldnotes proceeded inductively, involving immersion in the entire dataset to extract core themes	No	None	Individual informed consent was obtained from all individual participants included in the study
10. Holyoake (2013)	Participant observation with an external observer in the participant as observer role	Structured observations were conducted for 500 h over 14 months at a number of inpatient sites by one observer. The observer explored the cultural meaning associated with inpatient mental health nursing observation	Interviews	The ethnographic interview and observational data were analyzed using a semiotic and symbolic frame of reference	No	None	Individual consent for interview. No individual patient consent for observation
			Focus groups
11. Hung et al. (2014)	Participant observation with external observers in the observer as participant role	Five observers conducted 20 hours of semi-structured observations with patients and their families on days, evenings, and nights. Obs occurred in 15 to 60-min intervals over three months in the spring	Interviews	Thematic analysis	No	None	A process consent approach was used to reinform and recheck consents and assents with all participants throughout data collection
12. James et al. (2017)	Participant observation with external observers in the observer as participant role	11 research assistants (RAs) collected structured data, using a structured collection sheet, on the quality of intervention delivery in an RCT over 16 weeks. 565 observations were collected & a typology was developed highlighting intervention implementation. RAs observed for about 2.2 h per week per ward. They recorded field notes after each visit. RAs had a ½-day training session & received written guidance on observation and how to record good-quality notes	Focus groups	Inductive thematic analysis	Yes	None	Written individual informed consent from the 564 staff participating in this study
13. Kallakorpi et al. (2019)	Participant observation with internal observer assuming observer as participant role	One observer focused on a nurse caring for individuals who immigrated between May 2008 & December 2009. Semi-structured observations took place on psychiatric unit. Observer reflected on field notes	Interviews	Inductive content analysis	No	None	Individual informed consent was obtained from all participants
			Documentation of the nursing care provided to the immigrant participants
14. Kessing (2020)	Participant observation with an external observer in the role of the participant as observer	Semi-structured observation data were collected by 1 observer through involvement in ward activities. Observations occurred for 20 h per week for five months. There was a mutual awareness of the relationship b/t the researcher & participants as informed consent was renegotiated throughout the 5-month study. Participants could opt out for different observations. Field notes were written in common areas, and elaborated upon after the session	Interviews	Patient interactions in field notes were identified and coded by comparing incidents, which helped identify subtle patterns and significant events. The theoretical framework was then identified and applied to develop further analytical depth	No	None	The researcher introduced herself as a researcher when meeting patients & asked for their approval. Patients could opt out of observation
							A poster describing the study was hung in the common area of the two wards
							All participants provided oral or written individual consent
15. Lindgren et al. (2015)	Participant observation with an external observer in the role of the participant as observer	External mental health nurse collected non-structured data for 150 h over six months in 2010. Observations were conducted during weekdays between 6 am & 10 pm and lasted 1.5–5 hours. Six women were observed. Nurses and the involved patients were informed about the study. Obs took place in public areas, staff areas and the women’s rooms. Field notes were taken, dictated into a tape recorder and transcribed as running text	Informal interviews	Qualitative content analysis	a Yes	None	Two nurses on each ward informed the participants (men/women) about the purpose of the study and requesting their participation. The information was given both orally and in writing
							The participants asked to give their individual written informed consent to participate
16. Maddock (2015)	Non-participant observation with an external observer in the role of the observer as participant	1 observer collected structured observations by observing four weekly multidisciplinary team meetings (1 per week for four weeks). The professionals’ involvement in patient cases & interactions between professionals were noted on a pre-prepared obs schedule	Interviews	A content analysis	Yes	None	To attain individual informed consent, an information sheet about the purpose of the study was provided for each participant to sign
17. McAllister et al. (2021)	Non-participant observation with an external observer in the observer as participant role	Eighty hours of structured observations were conducted by one observer over seven months (May-November 2019). Observation occurred between Mon & sun, 0730–2130 h & was performed at 15-min intervals starting with the 1st nurse met & continued until all nursing staff were observed. Field notes were recorded on a tablet and focused on nurse-patient interactions—with minimal participation by the observer	Interviews	Thematic analysis	Yes	None	Written consent was obtained before the interviews
							Permission for observations was obtained at a unit level before the study started
							Participation in observations was on an opt-out basis, although nobody declined to be observed
18. McKeown et al. (2020)	Participant observation with the external observers in a observer as participant role	Two observers (supervised by two researchers) collected semi-structured data for 1260 h over 20 months (July 2014-February 2016). Observers were present on the wards at for approximately a week, mirroring entire staff shifts and observing key practices such as staff team handover, medication rounds & lunchtime activity. Structured field notes, were written away from the ward at break times and the end of each shift day	Interviews	Thematic analysis	No	None	Not stated
19. Molloy et al. (2019)	Non-participant observation with the external observer in the role of observer as participant	One observer collected semi-structured observations at two health services between Feb 2014 and Oct 2016. Field notes contained descriptions of interactions with MH nurses, observed nursing practices, physical environment & organizational structures	Interviews	Inductive grounded theory approach	No	None	No information about the observation
			Document review
20. Nagayama and Hasegawa (2014)	Participant observation with the external observer in the role of observer as participant	More than one observer conducted 97 h of semi-structured observations from Sep 2011 to Nov 2012. The observers watched the interactions b/t nurses and patients and recorded their reactions to each other’s behaviour, facial expressions, tone of voice, etc.	Interviews	Modified grounded theory approach	No	Researchers discontinued participant observation when exacerbation of a patient’s symptoms was anticipated	Individual consent from eighteen nurses who consented to participate, and verbal consent from patients
21. Nossek et al. (2021)	Non-participant observation with the external observers in the observer as participant role	Using an observation protocol, two observers collected structured data in multidisciplinary team meetings and transregional social work meetings to examine psychiatric social workers’ integration and role-finding process in adult psychiatric hospitals. Obs were recorded on handwritten notes and, following the observation session, were documented on a computer. No patient data were recorded or used. Observation took place over six months, from April to October 2016	There were other data collection methods but only observation reported in this paper	Qualitative content analysis	a Yes	None	Individual written informed consent was obtained from all participants included in the study
22. Rio et al. (2017)	Non-participant observation with an internal observer in the observer as participant role	Nonparticipant observations of care planning practices focused on clinical handovers and multidisciplinary meetings. Nurses were observed for how they included the views of care recipients who were not present. Observations were documented in field notes and a reflective journal from Aug 2015 to Feb 2016	Interviews	Systematic qualitative analysis	No	None	Informed consent for observation of team meetings was not required. Staff had knowledge of the research aims & were aware their actions were observed & field notes kept. Staff received written study information
23. Rise et al. (2014)	Participant observation with external observer in an observer as participant role	1 observer collected semi-structured data during team meetings regarding the redevelopment of the hospital over seven months (autumn 2009-spring 2010). Ten meetings were observed that included between 3 and 15 participants. The observer sat with meeting participants & made notes through the meeting. Reflective summaries were written after the session	Interviews	Grounded theory analysis	No	None	Not reported
24 Salzmann-Erikson (2015)	Participant observation with an external observer in the observer as participant role	One observer collected semi-structured data for approximately 200 hours over 18 months in three psychiatric ICUs (25, 8-h observational shifts). The observer focused on watching staff in their daily work about ethnographic standpoints such as spaces, activities and goals. Observations were terminated when no new information was forthcoming from the newly sampled units	Interviews	The data collection and analysis were related to ethnographic approaches	No	Involuntary patients not permitted to participate	Information meetings were held with the staff at each unit to inform them about the study and to obtain informed consent from each individual
						Researchers emphasize nursing’s ethical duty/moral obligation regarding bioethics and relational ethics	All staff agreed to participate and signed a consent form
25. Sercu et al. (2015)	Participant observation with the external observer in the participant as observer role	One observer collected semi-structured data focused on the intricate relationship between stigma and mental health nursing identity. Observations occurred during team meetings, therapeutic activities & meal times. During idle periods, time was used to talk informally with staff, cautiously examining the organization of care, stigma, and identity conceptions. When field notes became repetitive after about 200 observation hours, the data collection was stopped. The observation took place between 2011 and 2012	Interviews	Constructivist grounded theory approach	No	None	All were personally informed about the goals and methods of the research and received a prospect. The observation was only undertaken once all service users and professionals agreed on the researcher’s presence
26. Terkelsen and Larsen (2013)	Participant observation with external observer in the participant as observer role	One observer visited the ward 48 times over 4 months in 2010 and conducted ethnographic fieldwork usually three times per week for 2 to 7 h divided by daytime, evenings and two nights. Field notes were written while observing or on the same day. Observer focused on how and where people were interacting & moved from common areas to rooms reserved for staff. The observer was able to observe and participate in different situations. Informal conversations with patients and staff were recorded	Interviews	The study aimed at detecting patterns, themes, repetitions, contrasts and paradoxes in the material	No	Although the participants consented, it was problematic to ask people who were involuntarily detained to participate in a study. They did not participate	Sixteen patients, 2 psychiatrists, and 20 milieu therapists gave individual consent to participate
27. Ulfseth et al. (2015)	Participant observation with an external observer in the participant as observer role	One observer collected semi-structured data by participating in daily activities and general conversations with patients. Over the period of 1 year, the observer had frequent observation of 2–6 h, most at the beginning of the study and once every two weeks at the end of data collection. The observer was actively passive. Detailed field notes and the narratives were written down soon after the observation. Facts, descriptions and reflections structured the field notes	Narrative	An interpretative analytic approach in accordance with the ethnographic method	a Yes	None	Patients & staff were informed verbally & in writing about the study, & they knew why & when the researcher was there
							Written information was placed in the living room & the staff room
							The researcher only followed pts. Who wanted to participate
28. Unhjem et al. (2018)	Participant observation with an internal observer in the observer as participant role	One observer conducted semi-structured observations at each nurse’s place of employment and focused on the personal information the nurse shared. Sixteen nurses were observed. Observations were about 4 h long on day and evening shifts. Field notes were written. The observation took place between July 2013 and 2014	Interviews	Systematic text condensation served as an analytical framework	b Yes	None	Participating nurses signed an individual consent form
			Focus groups
29. Voogt et al. (2014)	Participant observation with an internal observer in the observer as participant role	One observer collected semi-structured data on two mental health wards over six months (Aug 2009-Jan 2010). Data were collected at different times that optimized nurse-patient interactions, e.g., wake-up, mealtimes, and patient meetings. 52 events were observed across the two units. Observer took a non-obtrusive stance so as not to disturb the routines. Notes were taken by the observer of the nurse-patient interactions. Following the event, these notes were elaborated upon to record as much detail as possible. Data were checked after 25 recorded obs to ensure the events met inclusion criteria. No changes were made. Detailed structure & process info were collected about each event	Informal conversations included with observation data	Grounded theory analysis	a Yes	None	Patients/nurses informed individually & in group meetings about the study. Information also provided in written format
							Patients’ & nurses’ consents were recorded in the minutes of group meetings
							Before each observation, observer asked nurses if the patients were stable enough to be observed & if any patients did not want to be observed. When this proved to be the case, those patients were indeed excluded from the observation
30. Waldemar et al. (2019)	Participant observation with an external observer in the observer as participant role	One observer collected 84 h of structure observation over 21 days (84 h) within three months (November 2014 to January 2015). The observer role shifted from nonparticipant to observer as participant when the observer helped patients and staff with practical matters. Observer was primarily an observer only. Observations took place on day and evening shifts, weekends and weekdays. They involved attending daily conferences, meetings between patients and their nurses, team consultations, & informal conversations with those on the unit	Informal ethnographic interviews consisting of informal conversation	Field notes were analysed deductively and inductively using qualitative content analysis	a Yes	Researcher did not see any situations where a patient was subjected to coercion through physical restraint or compulsory treatment	Written consent obtained from administration of the mental health hospital
						Prior to the observations it was decided that researcher would inform local managers of any unethical practice. This was not necessary	Verbal consent obtained from head nurse & psychiatrist of two wards
							Participants were informed of the study & their rights via flyers & posters on the wards
							Verbal consent obtained for informal interviews & one on one meetings between patients and HCPs

^aYes, means that field observation was the only data collection method used in the study.

^bYes, means there was a separate observation analysis for one research question but data was merged for other questions.

Observer Information

Information about the observer was typically related to the observer’s relationship with the research setting, the number of observers, and the role of the observer. Twenty-five studies (83%) employed observers who had no association with the research setting, known as external observers. Five used internal observers, which identifies an observer who has had, or currently has, an association with the setting. Most studies had a single observer (23 studies, 77%), with the number of observers ranging from one to eleven. The number of observers was unclear in three studies. The most common role was “observer as participant”, employed in 21 studies (70%). The “participant as observer” role was employed in nine studies.

Population and Setting

Most studies employed convenience sampling to recruit participants, as evidenced in 21 studies (70%), followed by purposive samples (n = 8 studies) and snowball (one study). Participant numbers ranged from 5 to 564. Notably, nine studies did not include clear details about participant numbers. Regarding the setting, 20 studies were conducted in one environment only, while 10 took place in multi-centre locations. Most research was carried out across multiple units, as seen in 20 studies (Table 3).

Analysis of Observation Data

The predominant data analysis method was thematic analysis, which was used in ten studies. Grounded theory and content analysis were applied in nine and five studies, respectively. Additional methods included systematic text condensation, systematic qualitative analysis, and interpretative ethnographic analysis.

As a data collection method, observation was commonly paired with other methods. Only seven studies utilized observation as the sole method of data collection. Interviews with observation were the most frequently combined data collection method. Additionally, focus groups, document analysis, and artifact analysis were employed in conjunction with observation. Of these, 11 studies presented separate analyses for the observations and other data collection methods. Nineteen studies did not separate observation analysis from other qualitative methods (Table 3).

Findings From the Studies in the Review

The qualitative findings derived from the 30 studies were broadly categorized into four groups using the constant comparison method (Thorne, 2016). This analysis addressed aspects of the first research question - how the observation method is used in mental health inpatient settings. An examination and synthesis of the categorized findings pinpointed what was produced and developed from the observation data. Four growing bodies of evidence about mental health inpatient care have been identified, and observation methods are contributing to the evidence in a meaningful way. Study findings clustered around four meta-findings: patient-centred care, diversity and culture, teams, and professional, ethical dilemmas.

Patient-Centredness

Patient-centred care (PCC) involves ensuring personhood and individuality in care, particularly concerning patients’ unique experiences and needs, while developing and maintaining therapeutic and social relationships with patients (Barry & Edgman-Levitan, 2012; Byrne et al., 2020). Seventeen studies in the review addressed some aspect of PCC, including nurse-patient interactions and the cultivation of relationships through self-disclosure (Blacker et al., 2017; Unhjem et al., 2018), ensuring patient autonomy and safety (Ellegaard et al., 2020), physical and mental wellbeing (Nagayama & Hasegawa, 2014), and fostering a compelling emotional connection between nurses and patients (Lindgren et al., 2015), and others.

Diversity and Culture

Navigating cultural, ethnic, and social differences offers an opportunity to explore HCP experiences in providing care while embracing diversity within mental health inpatient settings. Kallakorpi et al. (2019) delineated how nurses articulate the reasons for their immigrant patients’ illnesses and described the nursing experiences and impact of cultural factors when caring for this population. Molloy et al. (2019) described nurses’ mental health practices towards Indigenous Australians as individualized which stemmed from the nurses’ beliefs about Indigenous Australians and their experiences with indigenous communities. Hannah (2018), however, noted how clinicians tended to minimize the importance of race and ethnicity, and instead focused their attention on issues that more directly impact their daily work.

Teams

Team dynamics refer to the team members’ functions, integration, interaction, teamwork, and challenges. Five studies addressed the challenges that persist in maintaining professional identity within teams (Blacker et al., 2017; Maddock, 2015), integrating into multi-professional teams (Deacon & Cleary, 2013; Nossek et al., 2021), sustaining effective teamwork in mental health settings (Maddock, 2015), changing practice patterns (Grant et al., 2014) and how these challenges influence decision-making and negotiation patterns (Blacker et al., 2017; Deacon & Cleary, 2013). Teamwork is an important dynamic in mental health, partly because of its strong focus on safety. HCPs need to know they can rely on each other.

Professional, Ethical Dilemmas

Six studies in the review developed findings related to stigmatization and/or mental health practice issues that posed challenges to HCPs (Allard & Bleakley, 2016; Berlin & Carlström, 2015; Holyoake, 2013; McKeown et al., 2020; Salzmann-Erikson, 2015; Sercu et al., 2015). Finding legitimacy in mental health practice and navigating ethical situations were common occurrences for HCPs. This element benefits from good team functioning.

Observation Limitations

Like any data collection method, the observation process and its interpretations present limitations and challenges. The duration, depth, and number of observations were highlighted as factors that can impact the interpretation of results (Berlin & Carlström, 2015; Nossek et al., 2021). A limited number of observations can result in several issues including translational challenges. Extending the duration or increasing the number of observations are two solutions to achieving a more comprehensive understanding of the findings (Berlin & Carlström, 2015).

The observation process can also be influenced by inherent biases or be constrained by its contextual relevance (Hagen et al., 2020; James et al., 2017; McAllister et al., 2021; McKeown et al., 2020; Nossek et al., 2021; Rio et al., 2021; Rise et al., 2014; Terkelsen & Larsen, 2013; Unhjem et al., 2018). The behaviour being observed might itself impact the observation. Additionally, the observer’s perception can skew results, emphasizing the need for in-depth reflection and discussion with the research team (Blacker et al., 2017; Gurtner et al., 2022; Rio et al., 2021).

Several other factors present challenges. Relying on a single observer can introduce bias but this can be mitigated by having multiple researchers involved in data analysis (Deacon & Cleary, 2013). This solution is used with caution, however, as it may lengthen the analysis process and foster interpretational contestability. Additionally, being an insider researcher has advantages, such as easier access, but it can also blind the observer to certain aspects of the research phenomenon, again demonstrating the need for reflection and discussion (Unhjem et al., 2018). When observations are restricted and only focus on a specific profession, or a few activities (e.g., meetings), or they exclude some participants (e.g., patients), the limited scope raises caution about the transferability of the study findings and weakens the quality of the study (Blacker et al., 2017; Grant et al., 2014; Hung et al., 2014; Lindgren et al., 2015; Maddock, 2015; McAllister et al., 2021; Nagayama & Hasegawa, 2014; Rise et al., 2014; Salzmann-Erikson, 2015; Sercu et al., 2015; Voogt et al., 2014).

Accessibility challenges for both participants and researchers can also impede comprehensive data collection (Rio et al., 2021). Waldemar et al. (2019) noted difficulties in accessing spontaneous private conversations between patients and health professionals. Hagen et al. (2020) indicated that some patients were uninformed about the study due to oversight by the nurses, which led to lost research opportunities. Factors like patients’ need for privacy and rest can further complicate the observation process (Ellegaard et al., 2020). Observers must strive to be unobtrusive to avoid disturbing patients and regular ward routines, which can sometimes result in an incomplete record of events (Voogt et al., 2014).

Guideline for Reporting Field Observation

Upon reviewing what was reported about field observation in the included studies, and supported by other related literature, we first developed a framework with over 55 items to be considered when reporting the observation method and procedures in a qualitative study. The items are spread across five topics: context, access and ethics, observer-related factors, observation procedure, and data collection and analysis. Rationales are included to support each topic (see Appendix 2).

While providing a valuable starting place, the framework is not suitable as a reporting tool. It is cumbersome, too long and not formatted for useability. Taking guidance from the Equator Network website (2025) and other reporting guidelines, we removed some items, synthesized others and formatted the tool to make it more useable for reporting on the observation method. The five main topics are retained with 34 information requests/items across the five topics (see Appendix 3). We hope this guideline will assist researchers in planning how to use field observation in their studies, thus, supporting the study’s methodological rigor and strengthening its conclusions. Although the guideline items were developed primarily using the research literature on mental health inpatient units, it has applicability for reporting observations in other mental health settings. Whether the study is situated in a community mental health centre or a residential setting for substance use, the majority of methodological and ethical considerations in the guideline will need to be accounted for in all observation research.

Discussion

This field observation integrative review was conducted in mental health inpatient settings and comprised studies that explored a range of phenomena. It illustrated many challenges presented by this method, all of which require serious thought and planning (Palinkas, 2014). The importance of trained observers and the implementation of transparent data-collecting processes was evident (James et al., 2017; Tuval-Mashiach, 2017). Confidence in a study’s observation findings will be enhanced by fully reporting the observation method, including contextual information, people, procedures and rigorousness. Developing a comprehensive description and report of the observation method will add to the quality of the study overall and improve the study’s methods in future research.

By conducting an integrative review, we developed a guideline with information for applying and reporting field observation research methods. Using mental health inpatient settings as the environment for the review adds extra value to the guideline, as it represents a challenging healthcare setting with unique features. Considering a context that likely has a higher level of complexity provides potential for the reporting guideline to inform and benefit a wider range of different health settings. We anticipate that this reporting guideline for field observation can be used as a reference or guidance for researchers conducting field observation in any setting. The new guideline is not intended to replace other qualitative reporting guidelines, such as SRQR, COREQ or JARS-Qual. The guideline is designed to cover comprehensive aspects of the field observation method and not all research processes (e.g., title, purpose, introduction). However, it is intended to supplement current qualitative reporting guidelines that fail to address field observation methods.

Context for Guideline Development

As context for the development of the reporting guideline we examined three aspects of the 30 studies in our integrative review that did not directly address the guideline development. First, we explored the types of study participants and the focus of each research study to gain an understanding of the knowledge that was being developed and the nature of the research that was conducted using an observational lens. The research had a professional stance with more studies involving health professional participants than participants who were receiving inpatient care. Clinical practice issues were the focus of the majority of studies.

Second, the findings of the 30 integrative review studies were also analysed and synthesized in an attempt to assess the impact of the findings and identify what they reveal about the mental health inpatient setting. Based on study findings, four main areas of inquiry were determined. These are i) patient-centredness, ii) diversity and culture, iii) teams, and iv) professional, ethical dilemmas. Each of these areas have significance for the clinical area and was illustrated by research that advanced patient- centred care, was open to learning and celebrating other cultures and differences, valued effective team functioning and resolved ethical situations regularly.

Finally, synthesized findings resulting from the limitations described by the authors of each study were examined to ensure that common methodological challenges were understood and represented as appropriate in the guidelines. By exploring the studies from the perspectives of participants, purpose, findings and limitations, we were able to understand the observation method as much more than simply a technique. It manifested as an approach that influences the study from beginning to end.

Guideline Creation

One of the challenges faced in this review is determining the observer’s role or typology (level of participation and engagement with participants). Much literature has tried to map the varied roles of the observers during field observation. The studies in this review used classical sources to refer to the observer role, such as Spradley’s classification of involvement (1980) (non-participant, passive participation, moderate participation, active participation, and complete participation) and Gold’s typology (1958) (complete observer, observer as participant, participant as observer, and complete participant). As reviewers, we chose to classify the role according to Gold (1958) as it comprised fewer categories with clear distinctions between them. However, the complete observer and complete participant roles were excluded as they require that the participants be unaware of the observer or observation. These two roles can raise ethical concerns because of the lack of information given to participants and the waiver of informed consent to being observed. This is problematic in contexts such as healthcare or mental health settings. We differentiated between the role of (observer as participant) and (participant as observer) by the level of engagement. Passive or minimal engagement was assigned to the observer as participant role. This includes observers’ acknowledging their role and informal conversation between the participants and observers. Being a complete observer without any engagement is not realistic in closed healthcare units such as inpatient mental health units. Even non-participant observation may include minimal or passive participation. Therefore, for this review, we assigned the role of observer as participant to observers who conducted non-participant observation. The active involvement and engagement of the observer was assigned participant as observer role. Most of the literature on this issue was social research; these roles must be redetermined for research conducted in healthcare settings.

The study’s findings should be considered, including its strengths and weaknesses. The study provides a novel framework (guideline) for reporting field observations in mental health settings. The study takes advantage of reviewing research articles using field observation methods in mental health settings to develop this framework supported by other related literature. A few studies have contributed to developing guidelines or templates for researchers in field observation in healthcare settings. For example, Fetters and Rubinstein (2019) developed a “3 Cs” template of content, context, and concepts as a practical approach to collecting and recording unstructured field observation data. The template was developed to help researchers use data collection tools for field observation as the ethical committees may require them to review and be used as a training resource for observers. Fix et al. (2022) developed a practical guide for health researchers on when and how to conduct direct observation methods. To our knowledge, this is the first guideline for reporting field observation research methods in mental health settings. The limitation of this framework is that it depends on reviewing articles conducted in inpatient mental health units and excludes mental health community settings and other health settings, which may limit its use in settings other than inpatient mental health settings. However, inpatient mental health settings represent a more complex setting that requires extra attention to detail than other healthcare settings. Therefore, this guideline may also fit a broader range of other healthcare settings. Another challenge we faced conducting this study was represented in extracted data; many authors did not explicitly report all necessary information. This challenge further supports the need for a reporting framework.

We believe this review’s findings will interest researchers, healthcare providers, managers, ethics committees, and policymakers interested in using or assessing observation research methods in mental health settings. The study provided an overview of how studies conducted and reported field observation research in mental health settings. It described the reported observation process, synthesizing its outcome challenges, limitations, and recommendations. The guideline of reporting field observation methods has important implications for researchers, reviewers, and readers of studies that used the observation methods in mental health settings. Future research is recommended to enhance the current qualitative guidelines by incorporating the recommendations of this framework or developing a new guideline that considers the recommendations of the reporting field observation method.

Conclusion

This integrative review involved the extraction, description, and synthesis of data from qualitative studies that used field observation research methods in mental health inpatient settings. It is the initial phase addressing the creation and refinement of a reporting guideline for observation methods in mental health services. Thirty studies in which observation as a method was carried out on mental health inpatient units were the primary source of data for the guideline. Using guidance from the Equator Network, the reporting guideline will be further developed and then tested. Although field observation research is growing in mental health settings, our review demonstrated that the lack of description of field observation methods was common. A guideline for reporting field observation was developed to help researchers plan, apply, and report field observation research methods.

Supplemental Material

Supplemental Material - Field Observation in Mental Health Inpatient Settings: An Integrative Review and Reporting Guideline

Supplemental Material for Field Observation in Mental Health Inpatient Settings: An Integrative Review and Reporting Guideline by Joy Maddigan, Ahmad M. Deeb, Kristen Romme, and Chantille Isler in International Journal of Qualitative Methods