Trauma informed care on acute inpatient units: reducing self harm and restrictive practices


The NHS Long Term Plan (2019) is now 4 years old and within it there was laid out a 10-year plan for mental health services to be trauma-informed. However, as a Consultant working on an acute inpatient unit, what trauma informed care (TIC) might actually look like and how effective it may be on inpatient wards is still a relatively unknown quantity from a research perspective. I am pleased to blog about Nikopaschos et al’s service evaluation of the implementation of two trauma informed practices on their acute inpatient unit: Power Threat Meaning Formulation (PTMF) Team Formulation and Psychological Stabilisation intervention (Nikopaschos et al, 2023).

For some background information as to how important trauma informed care is in mental health care, head to these previous Woodland blogs to help you set the scene:

Aneta Zarska blogged about trauma informed care in mental health: why we need it and what it should look like. This was an experience-based co-design and the authors emphasised concepts related to human connectedness as vital and acknowledged the multi-dimensional factors that challenge these concepts in mental health settings.

Sarah Carr posted about the publication of an overview of the literature on trauma informed approaches. Adverse childhood experiences and their effects on health and mental health in later life.

The Kings Fund have also blogged about tackling poor health outcomes and the role of trauma informed care, highlighting the key message that a change in the way front-line staff understand the impact of trauma is required, which in turn can then influence their practice. Staff however do need support, education, and guidance to make this happen.

Four road cross over each other, and are surrounded by green trees

This study is an evaluation of the implementation of two trauma informed practices on their acute inpatient unit: Power Threat Meaning Formulation (PTMF) Team Formulation and Psychological Stabilisation intervention.


This is a retrospective service evaluation based on monthly quantitative incident data collected from an NHS adult acute inpatient mental health unit (consisting of two wards) at a North London Hospital. Data from the first year of the evaluation (July 2017–June 2018), and prior to the introduction of trauma informed care, was compared with data from the four years following the introduction of trauma informed care (July 2018–June 2022). The number of incidents of self-harm, seclusion and restraint were collected and analysed for statistical significance.

The 2 trauma informed care interventions implemented were:

  1. Power Threat Meaning Formulation (PTMF) Team Formulation:
    A model of trauma informed Team Formulation, informed by the PTMF, delivered to the inpatient multi-disciplinary team (MDT) in a one-hour weekly meeting (alternating fortnightly between the two wards). Facilitated by two trauma informed care Champions who were all senior MDT members.
  1. Psychological Stabilisation Interventions
    A Stabilisation Manual was developed by the authors (Nikopaschos et al., 2020) comprising one introductory session and ten intervention sessions. The manual was based on the Cwm Taf Morgannwg University Health Board Psychological Therapies Department (2017) Stabilisation Pack and draws on a range of established strategies and skills for first-stage trauma work (Linehan, 2014).


The study was based on 2,332 discrete episodes of inpatient care which involved over 1,625 different inpatients. Summary demographics include (to the nearest %):

  • 48% female; 52% male
  • Age range from 15 to over 75, with the majority age 24-49yrs
  • 44% were Black, Asian or from multiple ethnic groups
  • 29% White; and 27% other or not stated

Most people admitted to inpatient care presented with symptoms diagnosed as psychosis (41%) followed by depression/anxiety (11%), bipolar affective disorder (10%), personality disorder (8%), substance use (6%), acute stress reaction (4%), other (3%) and 17% had no diagnosis recorded. Most inpatient care episodes were under Mental Health Act detention (61%).

Implementation of Power Threat Meaning Framework Team Formulation

  • Weekly Team Formulation commenced in July 2018 and ran 147 times until the end of June 2022
  • 1,170 staff attendances at the meeting
  • Mean weekly attendance of 8 staff per meeting, not including the 2 facilitators

Implementation of Psychological Stabilisation training

  • Weekly Stabilisation training for the ward staff commenced in November 2018 and ran 119 times until the end of June 2022
  • 706 staff attendances at the training
  • Mean weekly attendance of 6 staff per meeting

Rates of reported incidents

Self harm Seclusion Restraint Staff Sickness days
Before TIC 93 123 193 340
Year 1 post TIC 80 80 134 207
Year 2 post TIC 43 72 139 459
Year 3 post TIC 31 105 194 1,173
Year 4 post TIC 10 94 152 666

Note: TIC = trauma informed care.


  • Before trauma informed care, the monthly mean number of self-harm incidents was 7.75, and this decreased to 3.42 in the four years post-trauma informed care.
  • There was an overall self-harm incident reduction of 55.9% (p=0.00), r=0.42 (p=0.00)


  • Before trauma informed care, the monthly mean number of seclusion incidents was 10.25, and this decreased to 7.31 in the four years post-trauma informed care.
  • There was an overall seclusion incident reduction of 28.7% (p=0.02), r=0.30 (p=0.02)


  • Before trauma informed care, the monthly mean number of restraint incidents was 16.1, and this decreased to 12.9 in the four years post-trauma informed care.
  • There was an overall restraint incident reduction of 19.8% (p=0.27), d=0.55 (p=0.05).
a white arrow pointing towards the camera, there's a person standing on the start of the arrow.

Self harm, seclusion and restraint incidents all decreased in the four years post-trauma informed care implementation. Self harm reduced by the greatest percentage, at 55.9% reduction.


The findings suggest that PTMF Team Formulation and Psychological Stabilisation training may contribute to significant reductions in self-harm and restrictive interventions (seclusion and restraint) on adult inpatient mental health wards, highlighting the benefits of trauma informed care for inpatients.

The authors have completed qualitative interviews with staff and service users from the unit (pending publication) which will support a better understanding of the mechanisms of this change beyond the often-cited theory of shifting the focus from ‘what’s wrong with you’ to ‘what happened to you?’.

Strengths and limitations

With the paper reporting on a service evaluation it doesn’t fit as neatly into critical appraisal frameworks as other pieces of research, however we can still use the key questions to structure our thinking:

  1. Does this study address a clearly focused question?
    • Yes – implementing a standardised model for trauma informed care (comprising two practices) that is easily replicated in other settings; the evaluation reviews whether this has an impact on patient safety incidents.
    • However, the aim of the study (“to assess the impact of introducing a model of Trauma-Informed Care (TIC), comprising weekly Power Threat Meaning Framework (PTMF) Team Formulation and weekly Psychological Stabilisation staff training, to a National Health Service (NHS) adult acute inpatient mental health unit over a four-year period”) is very broad, and a more specific question would be better.
  1. Did the study use valid methods to address this question?
    • Yes – there was a substantial follow up period of 4 years post-intervention within the naturalistic NHS environment enhancing ecological validity. The intervention was structured in accordance with a standard Protocol and Quality Measure, ensuring consistency and attendance to all key areas.
    • However, the use of incident reporting systems (Datix in this case) does have an inherent limitation in that, were all incidents reported onto the system? Or indeed was there over reporting? Different hospitals and trusts have different reporting cultures; did the implementation of the intervention create a potential conscious or subconscious reporting bias within staff?
    • A randomised controlled trial is the gold standard for testing the safety and efficacy of an intervention, so this is the ideal methodology to reliably measure the impact that these service models are having on patients.
  1. Are the valid results of this study important?
    • Yes, maybe – the significant reductions in self harm are important and are clear year on year. My hesitation is that two interventions were implemented at the same time, which is always going to create some doubt as to which is contributing towards the results. Would one of the interventions on their own create the same results for self harm reduction? Or perhaps something else entirely is leading to the reductions in self-harm? The lack of randomisation in this study means we cannot be sure that the self-harm outcomes are the result of the two interventions.
    • However, the impact on episodes of seclusion and restraint did not reduce in the same linear fashion as the incidents of self harm. There were spikes in the data which corresponded with large increases in staff sickness as a result of COVID-19. The authors report that this meant an increase in agency staff who hadn’t been trained in the intervention; does this imply that success may be contingent on a stable staffing group?
    • Averaging the monthly incidents over the total period of follow up time led to a statistically significant reduction however, given the marked variation in number of incidents year on year. I wonder if this gives an accurate reflection of the effect of the interventions.
 a multicoloured puzzle has been put together but there is one piece missing.

The authors used incident reporting system ‘Datix’, but this raises the question of whether all incidents were reported, and whether incidents were reported consistently.

Implications for practice

So, as a Consultant Psychiatrist on an acute inpatient unit are these valid, important results applicable to my patients or local population?

In short, yes – the acute inpatient units in the study are of a recommended size, however there is a variation to my own clinical practice in the diagnostic split of the patient group. In my Trust, our inpatient population has a much higher preponderance of personality disorder and trauma spectrum disorders. As such, would this make the interventions more or less effective?

To enable implementation, it would seem that the only local resources required are the trauma informed care Champions and allocated time. In the evaluation the champions included the Principal Psychologist, the Deputy Borough Director (who was an Occupational Therapist by background), the Lead Occupational Therapist and Matron. Two of these were providing an hour of Multi-Disciplinary Team formulation and two providing an hour of psychological stabilisation training per week. Ensuring consistent prioritisation of face-to-face delivery of the intervention by a similar senior staff group may be a challenge in some places; either due to work load or geographical constraints.

It is of note that based on the preliminary data of the evaluation, 13 other adult inpatient mental health wards in the study NHS Trust are now implementing the approach with support from the authors. If data from these boroughs is available in the future then a larger dataset could facilitate superior statistical modelling and control for confounding variables (it would also hopefully not have the influence of a global pandemic).

The authors may have missed a trick as the outcomes are focused solely on patient safety incidents (which are of course important), however the implementation of trauma informed care can have a wider impact on the mental health of patients and staff. It may be the qualitative information as yet to be published that clarifies this, although any future study may benefit from a wider range of outcomes.

Often studies recommend future research including randomised control trials to increase the generalisability and validity of findings. However, the authors rightly note that the ethical implications of an experimental design, where individuals in acute psychological distress are randomised to a control group without access to potentially beneficial practice, may outweigh the benefits of such a study.

Having robust research for a model of TIC on an acute inpatient unit which may reduce potentially re-traumatising interventions is a great start on the journey to transforming inpatient care. This is very positive news for the most unwell mental health patients, especially considering the media publicity about coercion on inpatient wards (e.g., Dispatches, 2022; Panorama, 2022) highlighting the urgency that is required to address this.

If locally you don’t have a structure for inpatient trauma informed care, the model described here suggests a (so far) successful approach, which could be a very good start indeed.

scrabble pieces spell out mental health next to a leafed plant

This study focused solely on patient safety incidents, however trauma-informed care can have a wider impacts on the mental health of patients and staff.

And finally: Understanding what people have been through can be key to helping them to find themselves – when they didn’t even know they were lost (Avicii 2013).

Statement of interests

I work on an acute inpatient unit and have an interest in trauma informed care and novel models of care and EMDR. I however have had no specific involvement with the researchers or the process of publication of this article.


Primary paper

Nikopaschos F, Burrell G, Clark J and Salgueiro A (2023) Trauma-Informed Care on mental health wards: the impact of Power Threat Meaning Framework Team Formulation and Psychological Stabilisation on self-harm and restrictive interventions. Front. Psychol.

Other references

Cwm Taf Morgannwg University Health Board Psychological Therapies Department (2017). Stabilisation Pack. Cwm Taf Morgannwg University Health Board. Available online at:

Dispatches (2022). Hospital Undercover Are They Safe? Channel 4. Available online at:

Linehan, M. (2014). DBT Skills Training Manual. New York: Guilford Publications.

NHS Long Term Plan (2019) nhs-long-term-plan-june-2019.pdf (

Nikopaschos, F., Burrell, G., Holmes, S., Rhodes, N., and Boado, C. (2020). Stabilisation Manual. London: CNWL. Available online at:

Panorama (2022). Undercover Hospital: Patients at Risk. BBC. Available online at:

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+