Individuals with severe symptoms of a mental health condition, which pose a danger to their safety and health and that of others, require treatment in in-patient settings (IPPN & NHS, 2015). Around 23,600 persons were admitted into mental health units in 2014 with a median duration of stay of 23 days (HSIC, 2014). In-patient suicide is a serious adverse event and is defined as intentional self-harm leading to death (DOH, 2014) in the hospital following admission on a voluntary basis or involuntary basis sanctioned by the Mental Health Act (IPPN & NHS, 2015). Despite a significant decline in deaths compared to previous years, 153 deaths continue to occur each year in mental health wards (NCIHS, 2015).
Critically appraising evidence ensures that interventions are effective and applicable (Aveyard & Sharp, 2013). Using the PICO framework (Bettany-Saltikov, 2012), the purpose of this paper is to review prior research on formal observation in order to identify best practices and analyse their implementation in practice. It aims to answer the question “Is formal observation effective in preventing adult mental health in-patient suicide?” A search of 3 electronic databases (CINAHL, Medline, and PsycINFO) was done using the keywords “effectiveness, formal, observation, prevention, in-patient, suicide” to locate the 5 primary, full-text, English language research articles employed in this review.
Review of the Literature
Six studies published between 1996 and 2011 were retrieved and used in this review. Four used quantitative descriptive methods and two were qualitative. The findings are inconsistent in regard to the effectiveness of formal observation in preventing in-patient suicide. Study limitations also precluded strong conclusions regarding effectiveness. Steward, Bowers and Ross (2011) conducted a retrospective cross-section case study using patient data from 84 mental health units in 31 London hospitals. The purpose was to investigate the incidences of conflict and containment specifically during the first two weeks following admission (Steward, Bowers & Ross, 2011). The use of an instrument with good reliability and the large sample is a positive aspect of the study. Patients were initially selected randomly but there was a high rate of refusal to participate (Steward, Bowers & Ross, 2011) which may have affected the representativeness of the sample which has implication on generalisability (Aveyard & Sharp, 2013). There may be bias from patient notes as a data source if staff did not document their care accurately or thoroughly.
Steward, Bowers and Ross (2011) noted that patients at high risk of suicide based on assessments were prescribed formal observation. Whilst there were incidences of attempted suicides, these occurred after the patients were discharged home and were no longer observed (Steward, Bowers & Ross, 2011). As such, the authors concluded that risk assessment and formal observation may have a protective effect (Steward, Bowers & Ross, 2011). However, the research design precludes analysing for a cause-and-effect relationship between the two variables that would strengthen confidence in this conclusion (Boswell & Cannon, 2014).
Two articles were qualitative studies. Cardell and Pitula (1999) elucidated what a therapeutic environment is in the context of formal observation. To generate the qualitative findings, independent thematic analysis conducted by each of the researchers and an independent researcher, exhaustive interviews to create thick descriptions, and the use of quotes to best represent the themes. These aspects contribute to the credibility of the findings (Streubert & Carpenter, 2011). From the point-of-view of patients, nurses who observe were most therapeutic when they demonstrated optimism which fostered in patients hope, positivity, and the willingness to being solving problems (Cardell & Pitula, 1999). Nurses who acknowledged patients by greeting and talking to them, validating their emotions, and conveying their interest were also therapeutic as they alleviated loneliness and anxiety. Moreover, nurses who observe but also initiated activities such as crafts or sports that distract the patient from thinking about suicide were therapeutic by helping reduce negative thoughts and anxiety (Cardell & Pitula, 1999).
Emotionally supportive observing nurses, by virtue of their compassion, affirmation, and encouragement, were also therapeutic. The participants felt that formal observation was protective and generated feelings of relief as the nurse’s physical presence made it more difficult to carry out suicide plans (Cardell & Pitula, 1999). Formal observation is thereby meaningful to patients and effective in preventing suicide when it is implemented in a therapeutic way. However, participants in the study by Cardell and Pitula (1999) similarly acknowledged the disadvantages of formal observation which are the lack of privacy, invasion of personal space, and a feeling of confinement among patients. This aspect of the intervention clearly impacts feasibility and appropriateness.
Mackay, Paterson and Cassells (2005) enable a two-way view of formal observation by exploring mental health nurses’ perspectives through a thematic analysis of qualitative interviews. The study employed several techniques such as member checking which ensure validity (Streubert & Carpenter, 2011). For nurses, formal observation was more complex than just looking at the patient. The process also involved preventive interventions, maintenance of patient safety, assessments, communication with patients and fellow staff, and providing therapy that meant spending time with the patient (Mackay, Paterson & Cassells, 2005). As such, nurses similarly viewed their role in formal observation as therapeutic. Nurses were of the view that a repertoire of skills was needed to effectively implement formal observation in a therapeutic way but that the level of skill significantly differs among staff. For this reason, raising the skill level of the staff would be important to ensuring that formal observations are therapeutic and also prevent patient suicide.
A more recent study conducted by Bowers et al. (2011) employed a different approach, namely a retrospective study of the incident reports pertaining to psychiatric patients who attempted suicide using highly lethal means but survived. The authors accessed data from the Department of Health’s National Patient Safety Agency. However, there may be bias arising from the data collected, namely when nurses failed to accurately describe the attempted suicide in their incident reports. Information from the reports was coded into categories including the patient’s observation status and what factors prevented the suicide (Bowers et al., 2011).
A sample of 244 cases was used and findings show that in 80% of these cases, suicide was prevented because the patients were discovered by ward staff (Bowers et al., 2011). In 62% of 120 reports which provided sufficient detail, patients were under formal intermittent or constant observation or other types of observation and the discovery occurred in the course of checking up on the patient (Bowers et al., 2011).
In 27% of the 120 reports, nurses performed general observations of patients and exhibited behaviors that were deemed “caringly vigilant and inquisitive” which served to prevent patients from fatal self-harm (Bowers et al., 2011). These behaviors included following patients who showed signs of distress, carefully listening, and noticing the absence of patients, suspicious patient actions or the appearance of physical illness (Bowers et al., 2011). Other behaviors were noting how long patients were in the toilet and acting immediately to unusual noises in bedrooms. Thus, Bowers et al. (2011) demonstrated that the effectiveness of formal observation lies in the ability of nurses to be observant, vigilant, and caring all at the same time.
The other three studies found formal observation ineffective. In a descriptive case study, Dodds and Bowles (2001) shifted from formal observation to individualised patient activities and noted differences in nursing and patient outcomes including the rate of self-harm and suicide. As the study was limited to only one mental health ward, a major limitation lies in generalising feasibility and effectiveness in other settings. It is worthwhile to note that the formal observation previously implemented in the chosen mental health ward was one that was a routine task of going around patient rooms and looking through the doors (Dodds & Bowles, 2001). Patients were kept in their rooms and there was little social interaction between patients and nurses. The problem with this type of formal observation was the violation of patient privacy and lack of patient engagement. There were also a high number of suicide attempts in the ward.
It is also worthwhile to note the larger context of this type of formal observation. There was a hierarchical rather than team-based collaborative relationship between nurses and physicians which left nurses disempowered to provide individualised patient care and to allocate staff based on true patient needs (Dodds & Bowles, 2001). Formal observation exemplified this situation as it was ordered by physicians for long periods of time deemed unnecessary and was time and human resource consuming (Dodds & Bowles, 2001). Thus, formal observation was found to be meaningless, ineffective, inappropriate, and difficult to sustain. This type of formal observation was therefore reduced and eliminated and in its place was implemented a team-based approach wherein nurses had more control over their work. One-on-one activities with high-risk patients were planned which allowed nurses and patients time to interact whilst keeping the patients in sight (Dodds & Bowles, 2001). Patients were also included in the planning and weekly evaluation of their care.
After 18 months of implementation, the authors found a 67.1% reduction in the incidence of deliberate self-harm among patients compared to baseline (Dodds & Bowles, 2001). Patients also found the gift of time from nurses very valuable. The need for agency nurses to perform formal observation declined leading to significant savings within one year. Essentially, the change described by Dodds and Bowles (2001) is similar to the therapeutic observation strategies described by Cardell and Pitula (1999). Joint nurse-patient activities permitted patients to be within eyesight but were also therapeutic and engaging at the same time. However, Dodds and Bowles (2001) pointed out how this approach can only be implemented within the context of a collaborative and patient-centred approach to care.
Finally, Steward, Bowers and Warburton (2009), in a longitudinal analysis of 16 mental health wards, did not find a significant association between the incidence of self-harm and the formal observation of at-risk patients. Over a period of 2 years, data on attempted suicide incidence and the number of hours of formal observation were collected. The limitation of this study is that it did not establish the comparability of the types of formal observation employed by the selected wards given that differences exist in the definition as demonstrated by Cardell and Pitula (1999), Dodds and Bowles (2001), and Bowers et al. (2011). Also, it fails to explore the organisational factors that may contribute to poor patient outcomes in the way that Dodds and Bowles (2001) did. However, Steward, Bowers and Warburton (2009) did control for demographic, diagnostic, conflict history, and use of agency nurses as factors that can possibly influence patient outcomes. The number of hours of formal observation was not found to impact self-harm incidence.
The studies located reflect the need to conduct high-quality research to generate stronger and up-to-date evidence regarding the effectiveness of formal observation in preventing in-patient suicide. However, the studies did present the perspectives of both nurses (Bowers et al., 2011; Dodds & Bowles, 2001; Mackay, Paterson & Cassells, 2005) and patients (Cardell & Pitula, 1999; Dodds & Bowles, 2001) and the relationship between the quality of care and the care environment (Dodds & Bowles, 2001). The need for standardised definitions and practices was highlighted by 5 of the studies as well. Based on the evidence, formal observation coupled with therapeutic approaches and patient engagement is best practice in promoting patient safety and wellbeing (Dodds & Bowles, 2001; Cardell & Pitula, 1999; Mackay, Paterson & Cassells, 2005). This approach enables patient-centred or individualised care. Nurses must be trained and dedicated to provide vigilant care (Bowers et al., 2011; Mackay, Paterson & Cassells, 2005), employ therapeutic methods such as social interaction and structured activities, and engage patients (Dodds & Bowles, 2001; Cardell & Pitula, 1999; Mackay, Paterson & Cassells, 2005). However, best practice is possible within a collaborative interdisciplinary team framework (Dodds & Bowles, 2001) with clear policies (Steward, Bowers & Warburton, 2009; Steward, Bowers & Ross, 2011).
Strategies for Introducing Practice Change
Practice change is a prerequisite to improving the practice of formal observation so that it becomes truly effective, appropriate, meaningful, and feasible in preventing in-patient self-harm. There are factors that promote and inhibit the adoption of best practices. A strong promotive factor is external pressure arising from current national reforms aiming to improve the quality of patient care (Charlesworth, Smith & Thorlby, 2015). Inhibitory factors include shortages in nursing staff as change may add to work stress, lack of knowledge and skills in leading and translating evidence into practice, and staff resistance which signifies opposition to leaving one’s established comfort zone (McSherry & Warr, 2008; McCormack, Manley & Titchen, 2013; Melnyk & Fineout-Overholt, 2011; Sullivan & Decker, 2009).
Lewin’s change theory on unfreezing, moving, and refreezing is a useful framework in implementing practice change (McSherry & Warr, 2008). Unfreezing pertains to establishing the basis for change. In implementing improvements in the formal observation of patients at risk for self-harm, it is important to assist the staff in realising the ineffectiveness of current practice. This can be done by presenting data on the high number of suicides and attempted suicides in the unit and communicating a common vision to achieve an improved quality of care to reduce the incidence. The ethical and legal duty to protect patients must be reiterated (Varcarolis, 2013). The alternative practice, i.e. to foster therapeutic interventions and engagement, will be presented as an alternative.
In order to elicit buy-in, input from the staff must be elicited in a participatory way (Bach & Ellis, 2011) and taken into consideration in the planning of practice change that entails the creation of new guidelines and workflow. In this manner, one elicits staff engagement which fosters commitment to change (Bowers, 2011). In order to ensure that the patient-centred approach is acceptable to patients, service users and their families must be able to influence the change plan as well (Morrow et al., 2012). Input should be elicited directly from current and former patients and their families. Piloting the study can generate evidence that the new intervention is effective, feasible, appropriate, and meaningful and will help convince stakeholders about change (Aveyard & Sharp, 2013; Gerrish & Lacey, 2010).
During the moving stage, the enhanced formal observation practice will be implemented. It is important to enable the staff to perform the new practice. As such, education and training sessions must be given until the staff feel confident (Bach & Ellis, 2011; Bowers, 2011) that they can establish therapeutic relationships with patients at risk and possess the necessary communication skills which are important in nursing in general, and mental health nursing in particular (Varcarolis, 2013). Emotional and other forms of support, such as staffing, must be given to build competency in the new practice, reduce resistance, and enhance staff morale (Bach & Ellis, 2011; Bowers, 2011; Sullivan & Decker, 2009). The change must be tied directly to measureable outcomes (Melnyk & Fineout-Overholt, 2011), namely suicide and attempted suicide incidence, which are periodically measured in order for the staff to realise the direct impact of their adherence to improved practice (McCormack, Manley & Titchen, 2013). In the refreezing stage, the new practice becomes cemented as the new norm (McSherry & Warr, 2008) by being incorporated in policies and performance evaluations. Nurse leaders adept in change management achieve success through the above strategies which also address the inhibitors of change.
Conclusion
References
Aveyard, H., & Sharp, P. (2013). A beginner's guide to evidence-based practice in
health and social care (2nd ed.). Maidenhead, Berkshire: Open University Press.
Bach, S., & Ellis, P. (2011). Leadership, management and team working in nursing.
Exeter: Learning Matters.
Bettany-Saltikov, J. (2012). How to do a systematic literature review in nursing: A step-by-
step guide. Maidenhead, Berkshire: Open University Press.
Boswell, C., & Cannon, S. (2014). Introduction to nursing research: Incorporating
evidence-based practice (3rd ed.). Burlington, MA: Jones and Bartlett Learning.
Bowers, B. (2011). Managing change by empowering staff. Nursing Times, 107(32/33),19- 21. Retrieved from http://www.nursingtimes.net/download?ac=1236479
Bowers, L., Alexander, J., Bilgin, H., Botha, M., Dack, C., James, K., Stewart, D. (2014). Safewards: The empirical basis of the model and a critical appraisal. Journal of Psychiatric and Mental Health Nursing, 21, 354-364. doi: 10.1111/jpm.12085
Bowers, L., Dack, C., Gul, N., Thomas, B., & James, K. (2011). Learning from prevented suicide in psychiatric inpatient care: An analysis of data from the National Patient Safety Agency. International Journal of Nursing Studies, 48, 1459-1465. doi:10.1016/j.ijnurstu.2011.05.008
Bowers, L., Nijman, H., & Banda, T. (2008). Suicide inside: A literature review on inpatient suicide. Retrieved from http://www.kcl.ac.uk/ioppn/depts/hspr/research/ciemh/mhn/projects/litreview/LitRev Suicide.pdf
Cardell, R., & Pitula, C.R. (1999). Suicidal inpatients’ perceptions of therapeutic and nontherapeutic aspects of constant observation. Psychiatric Services, 50(8), 1066- 1070. doi: http://dx.doi.org/10.1176/ps.50.8.1066
Charlesworth, A., Smith, J., & Thorlby, R. (2015). The Coalition Government’s health and social care reforms: 2010-2015. Retrieved from http://www.nuffieldtrust.org.uk/our- work/projects/coalition-governments-health-and-social-care-reforms
Dodds, P., & Bowles, N. (2001). Dismantling formal observation and refocusing nursing activity in acute inpatient psychiatry: A case study. Journal of Psychiatric and Mental Health Nursing, 8, 173-188. doi: 10.1046/j.1365-2850.2001.0365d.x
Gerrish, K., & Lacey, A. (2010). The research process in nursing (6th ed.). Chichester,
West Sussex: John Wiley.
Health and Social Care Information Centre (HSCIC) (2014). Inpatients in mental health wards experiencing long stays in hospital. Retrieved from http://www.hscic.gov.uk/article/5222/Inpatients-in-mental-health-wards- experiencing-long-stays-in-hospital
Mackay, I., Paterson, B., & Cassells, C. (2005). Constant of special observations of inpatients presenting a risk of aggression or violence: Nurses’ perceptions of the rules of engagement. Journal of Psychiatric and Mental Health Nursing, 12, 464-471. doi: 10.1111/j.1365-2850.2005.00867.x
McCormack, B., Manley, K., & Titchen, A. (2013). Practice development in nursing and
healthcare (2nd ed.). Chichester, West Sussex: John Wiley.
McSherry, R., & Warr, J. (2008). Introducing practice development to facilitate
excellence in care. Maidenhead, Berkshire: McGraw-Hill Education.
Melnyk, B.M., & Fineout-Overholt, E. (2011). Evidence-based practice in nursing and
healthcare: A guide to best practice (2nd ed.). Philadelphia, PA: Lippincott Williams and Wilkins.
Morrow, E., Boaz, A., Brearley, S., &Ross, F. (2012). Handbook of service-user
involvement in nursing and healthcare research. Chichester, West Sussex: John Wiley.
National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (NCISH) (2015). In-patient suicide under observation. Retrieved from http://www.bbmh.manchester.ac.uk/cmhs/research/centreforsuicideprevention/nci/rep orts/ipobsreport.pdf
Nursing and Midwifery Council (NMC) (2015). The Code: Professional standards of practice and behavior for nurses and midwives. Retrieved from https://www.nmc.org.uk/globalassets/sitedocuments/nmc-publications/revised-new- nmc-code.pdf
Stewart, D., Bowers, L., & Ross, J. (2012). Managing risk and conflict behaviours in acute psychiatry: The dual role of constant special observation. Journal of Advanced Nursing, 68(6), 1340-1348. doi: 10.1111/j.1365-2648.2011.05844.x
Stewart, D., Bowers, L., & Warburton, F. (2009). Constant special observation and self-harm on acute psychiatric wards: A longitudinal analysis. General Hospital Psychiatry, 31, 523-530. doi:10.1016/j.genhosppsych.2009.05.008
Streubert, H.J., & Carpenter, D. (2011). Qualitative research in nursing: Advancing the
humanistic imperative (5th ed.). Philadelphia, PA: Wolters Kluwer Health/Lippincott Williams & Wilkins.
Sullivan, E.J., & Decker, P.J (2009). Effective leadership and management in nursing
(6th ed.). Upper Saddle River, N.J: Pearson Education.
Varcarolis, E.M. (2013). Essentials of psychiatric mental health nursing (2nd ed.). Missouri, MO: Elsevier.
