Who’s got the obs sheets? Can QI methods reduce violence and restrictive practices on inpatient mental health wards?

You are in the ward office, you hear the same question multiple times every day… “who’s got the observation sheets?”

Everyone’s busy, the office is full, but who’s “out on the ward” and available to the service users? You want to do more to provide therapeutic care, but you’re not sure if the observation process is therapeutic? It’s busy and you can’t find time to do more. Sound familiar?

Observations are a means to keep someone safe, often implemented by a head around the door and a brief ‘hello’, but arguably more like surveillance than therapeutic engagement and increasingly involving unproven or harmful technologies. This approach has been criticised for feeling impersonal and failing to build meaningful relationships (Cusack et al., 2018).

Observation practices are often linked with restrictive practices, which can cause psychological harm, damage staff and patient relationships, and even be retraumatising (Kontio et al., 2012; Cusack et al., 2018), with physical and psychological risks to both service users and staff (Soininen, et al. 2013). Alternatives to observations emphasise empathetic interaction, meaningful activities, and therapeutic community approaches (Kontio et al., 2012).

A recent QI programme across a large UK mental health Trust, tested three interventions that aimed to augment the observation process and improve therapeutic care on 55 different wards. These included wards for adults, adolescents, and older persons covering acute, intensive care, forensics and rehabilitation services.

Observations are a means to keep someone safe, often implemented by a head around the door and a brief ‘hello’.

Methods

This Quality Improvement (QI) project aimed to test interventions designed to improve observation completion and therapeutic engagement across 55 inpatient wards in East London NHS Foundation Trust using the Model for Improvement framework. Quality Improvement in healthcare uses systematic, data-driven methods in practice settings to enhance patient care, safety, and outcomes through iterative cycles of testing and learning.

Clinical teams identified 25 initial ideas to enhance observations and engagement; through a group consensus process, three interventions were selected for implementation and testing as follows:

1. Board Relay: a sheet for observation recording is always kept in the hands of staff and handed over between them. Staff then give a verbal handover of how each service user presented in the previous hour and any critical information that will help the next staff member to conduct the observations safely and therapeutically.
2. Zonal Observations: involves designating the ward into different zones where allocated staff observe and engage with patients individually and as a group over a set period.
3. Life Skills Recovery Workers (LSRW): are employed on inpatient wards to deliver coordinated therapeutic activities. Initially, they worked twilight shifts (2–10 pm), but shift patterns were locally adapted to meet patient needs in some areas, including mornings and weekends.

Each participating ward implemented a combination of these interventions that they believed were most suited to their context. Intervention effectiveness was measured through recording rates of observation completion, incidents of violence and aggression, restrictive practices, and staff sickness using baseline routine data as a comparator. Data were tracked over 18 months using Statistical Process Control (SPC) charts, used to gather data before during and after the change (NHS Institute for Innovation and Improvement, 2010).

Service users participated in early workshops, contributing to brainstorming the 25 initial ideas and voting on the three interventions selected for scale-up. Co-production was facilitated during monthly QI team meetings and the trust-wide learning sessions, where service users shared feedback and helped adapt interventions to local contexts. For example, replacing the clipboard with a paper copy of the form and a coloured bib in a child and adolescent ward.

Quality Improvement uses data-driven methods in practice settings to enhance outcomes through iterative cycles.

Results

There were ten measures used, to assess improvement – summarised below in Table 1. The results compare data collected during the implementation of the interventions between 2022 to 2023 with baseline data collected over an 18 month period between 2021 and 2022. Resulting data are presented as percentage completion of observations or as number of incidents per 1,000 Occupied Bed Days (OBD).

Table 1- Summary of Reported Effects

*OBD – Occupied Bed Days

Improvements in completion of general observations and intermittent observations were sustained following the implementation of Board Relay and Zonal Observations.

It’s interesting to note that improvements in observation recording occurred prior to the interventions being implemented. Authors suggested that this may have been explained by an increase in pre-intervention scrutiny on observation practices, this phenomenon has been observed in many areas of practice and is known as the Hawthorne effect (McCambridge, et al. 2014). There were further improvements in observation practices two to three months following implementation of the intervention.

A significant reduction in aggression was reported:

• The largest effect being reported on incidents of racial aggression which reduced by an impressive 60%
• Verbal aggression reduced by 38% and physical violence by 23%
• These results were achieved soon after implementation and did not reduce further over time.

The use of restrictive interventions declined:

• The largest reductions observed in the use of seclusion which reduced by 38%
• Prone restraint reduced by 35%
• Rapid tranquillisation reduced by 26%
• Physical restraint reduced by 16%.

Additionally, there was also a reduction in staff sickness by 16%.

These findings show that relatively simple interventions, developed from ideas generated by clinical teams and service users, may deliver significant and meaningful change. The organisational buy-in from executive level through to clinical teams and the provision of support from QI experts were critical to allow local adaptation and implementation, whilst maintaining the core principle of the intervention.

Co-production is reported, but not fully visible in this QI study

Conclusions

This large-scale QI project demonstrated the importance and effectiveness of staff and service user engagement in service improvement. The use of a Board Relay, Zonal Observations, and Life Skills activities enhanced therapeutic engagement and observation completion. There was also a demonstratable improvement in observation compliance, a reduction in violence and restrictive practices, and decreased staff sickness.

Whilst the authors acknowledged other factors may have influenced the effect on outcomes, and there are some gaps in understanding which of the selected interventions, or which combination of interventions had the greatest effect. Nonetheless these finding have indicated that a systematic approach to quality improvement can prompt meaningful change.

Strengths and limitations

This project has several significant strengths. Prioritising therapeutic engagement is vital because of the adverse physical and psychological effects restrictive interventions have on service users and staff, and negative impact on the therapeutic relationship (Butterworth, 2022; Chieze, et al. 2019; Haugom, et al. 2019; Muir-Cochrane & Baird, 2015). The project reflects global priorities, with the World Health Organisation advocating for the elimination of coercive practices (WHO, 2025).

The QI project was implemented on a large scale across 55 wards, which included a diverse range of clinical services. Data collection and monitoring using SPC charts is a robust QI method for evidencing improvement (NHS, 2010), with data being collected before, during implementation and following, to identify whether improved outcomes were sustained. The use of co-production with service users and clinical teams indicates a meaningful collaboration and shared ownership of the changes.

There are however several limitations to consider. The involvement of service users is reported but not fully outlined, including who participated, when, and how their input shaped design, delivery or outcomes, it would have been interesting to understand how this diversity was mirrored in the reported co-production.

The absence of a control group, as identified by the authors, means there could be other factors not directly attributable to the interventions which resulted in the reported improvements. Future research should include a control group.

Due to there being multiple different interventions implemented at the same time, there are challenges to identifying which, if any, intervention had a better outcome associated with it. There are important contextual factors that have not been accounted for in the analysis including ward acuity, local adaptations of interventions, staffing, skill mix, the impact of COVID-19, and seasonal variations.

Additionally, the report did not fully outline which combinations of interventions were implemented in each setting leaving gaps in understanding of how these interventions can be tailored to different contexts.

Using staff sickness as an outcome measure has not been fully justified and is difficult to associate with the intervention implementation.

The reported results perhaps overstate the impact on ‘therapeutic engagement’ which has not been evaluated or reported. There is an assumption that the reported reductions in restrictive practices provide a proxy for improved therapeutic engagement. Although not applicable to all areas, the use of targeted validated tools for example a therapeutic engagement questionnaire or qualitative evaluation methods could be used to further explore therapeutic engagement.

While this large-scale QI project showed promising reductions in restrictive practices, the absence of a control group limits causal conclusions. Future research using randomised controlled trials would provide stronger evidence by controlling for confounding factors and establishing whether these interventions directly cause improvements in patient care.

Implications for practice

This study offers a number of important findings, which have practical implications for inpatient mental health services.

Firstly, it emphasises the importance of therapeutic engagement within psychiatric observations. Observations are used as a means to manage risk and improve safety, but they also offer opportunities for meaningful, purposeful interactions with service users who may prefer to collaborate on safety planning. The use of Zonal Observations showed how relatively simple changes to a process can enhance relational security. Interventions by Life Skills Recovery Workers (LSRW) provided meaningful activities that can reduce boredom, increase service users’ life skills and prevent incidents.

The second success is the organisational engagement with clinical teams. This process allowed clinical teams, who invariably hold the answer to many solutions the organisation faces, to be able to explore and bring about real change. This is a real-world and practice-based approach to research and development. Its impact and levels of engagement can be contrasted with more traditional and extractive methods whose findings take a long time to influence practice.

Thirdly, the need for local adaptability and flexibility in implementation is highlighted. The QI interventions were tailored to ensure they were safe and effective in areas with differing ward layouts, staff skill mix, patient demographics and service user needs. For example, in older person wards, Zonal Observations were not appropriate due to falls risks, and LSRW shifts were adapted to meet local needs. Whilst there was adaption, the fidelity of the purpose of the interventions was retained, particularly the focus on staff visibility, communication and engagement.

This approach demonstrated the need for organisations, with large and diverse inpatient portfolios, to support local implementation strategies within defined parameters. This QI project demonstrated how having an overarching QI project at organisational level can provide the oversight required to sustain project delivery and intervention fidelity whilst also enabling localised tailoring to meet the needs of different settings.

There were also a number of significant organisational factors. The key leadership sponsorship, QI project support, and local learning systems enabled not only implementation, but for these to be adaptable and sustainable. The workforce existing QI knowledge, and access to this training during the project, as well as support from quality improvement advisors appears key to its success.

Whilst this project showed positive results in areas of violence, restrictive interventions and observations recording completion, it doesn’t capture therapeutic engagement, so more work is needed here.

Pre-intervention impacts also suggest that a simple focus on observation practices may have contributed to an increase in compliance, which is consistent with the Hawthorne effect (McCambridge, et al. 2014). Individuals may modify their behaviour as a result of knowing they or their practices are receiving increased scrutiny, with subsequent positive effects sometimes occurring as a result.

Overall, this project illustrates the impact achieved when staff work together at all levels of the organisation alongside service users to generate meaningful and positive changes that improve safety, experience and staff wellbeing.

Sometimes just shining a light on practice can spark improvement.

Edited by

Simon Bradstreet.

Links

Primary paper

Aurelio M, Singh S, Chitewe A, et al. Improving therapeutic engagement and observations on inpatient mental health wards in the English National Health Service: lessons from using quality improvement to scale up interventions. Int J Qual Health Care. 2025;37(3):mzaf070.

Other references

NHS Institute for Innovation and Improvement, 2010. The Handbook of Quality and Service Improvement Tools

Kontio R, Joffe G, Putkonen H., et al (2012) Seclusion and restraint in psychiatry: patients’ experiences and practical suggestions on how to improve practices and use alternatives. Perspectives in Psychiatric Care. 48(1), 16–24.

Soininen P, Valimaki M, Noda T, et al (2013) Secluded and restrained patients’ perceptions of their treatment. International Journal of Mental Health Nursing, 48(1), 16-24.

Cusack P, Cusack FP, McAndrew S, et al (2018) An integrative review exploring the physical and psychological harm inherent in using restraint in mental health inpatient settings. International Journal of Mental Health Nursing. 27(9), 1162-1176.

McCambridge J, Witton J. and Elbourne D.R. (2014) Systematic review of the Hawthorne effect: New concepts are needed to study research participation effects. Journal of Clinical Epidemiology, 67(3), pp. 267–277.

Photo credits