How many nursing staff should you have on a mental health ward to achieve the best outcomes for service users? It’s a very important question with a lot to unpack. What outcomes are we talking about? Reductions in negative outcomes like self-harm and aggression or improvements in positive outcomes like increasing wellbeing or fewer readmissions or both? Some of the negative experiences of mental health wards have been discussed here recently. Then, there is what we mean by nursing staff (registered nurses, nursing associates, healthcare assistants) and what mix of these is best?
Some countries and jurisdictions, including Scotland, Wales, Victoria (Australia), and California (USA), have legally required mandatory minimum staffing levels (e.g. Van den Heede et al., 2020). Some people have called for this in England too, but many of these policy decisions are based on research done on general and surgical wards (P. Griffiths et al., 2023). Is that research relevant to mental health? And what happens when hospitals do not reach the legal limit? They usually get a fine and fining a hospital that is already struggling with staffing numbers is unlikely to make the situation better. As we have previously discussed, we already have problems retaining staff (Karpusheff, 2023; Long et al., 2023) and while well-staffed wards would help, they are not the only issue.
Many English hospitals use staffing guidelines that are based on clinical judgement and can be changed based on the numbers of services users on the ward and how seriously ill they are (acuity), but even this has problems. For example, ‘clinical judgement’ will vary from one clinician to another.
A recently published study from Woodnut and colleagues (Woodnutt et al., 2025) has pulled together the research literature on staffing in mental health wards, to see what evidence we currently have and if recommendations can be more evidence-based.
Do we know the best staff mix for mental health wards?
Methods
Woodnutt et al’s study was a systematic review of the existing research on staffing levels and incidents of violence in acute mental health wards. The aim was only to look at this one outcome. It is therefore relatively narrow, but it is better defined and better represented in the literature than some other outcomes. The search criteria are clear and appear suited to the research questions and a large selection of research databases have been searched, suggesting that all relevant published papers will have been captured.
Due to the wide variation in literature, a narrative synthesis approach was adopted which prioritised papers based on their quality for inclusion in the analysis and findings. Overall, the authors have paid a lot of attention to the quality of the studies. This is important as much of the evidence was low quality and as such, confident conclusions are difficult to draw.
The relationship between staffing and incidents of violence and aggression on wards is complex and guidelines are unclear.
Results
35 studies were included in the systematic review, all of which were observational, that is they did not include randomisation of participants. There was considerable variation between the types of study. These included:
- International comparisons (2)
- Those which reviewed national data (14)
- Ward based review (17)
All included studies were based in high income countries. Most studies were of low to moderate quality with just one assessed as being high quality.
The main finding of this review is that the relationship between staffing levels and violence is complicated. For example, some studies found an association between increased staffing levels and reduced violence while others found the opposite effect (increased staffing levels were associated with increased incidents of violence).
Some studies showed that increasing staffing decreased incidents of violence, but again, others showed the exact opposite, while some showed no relationship at all.
There was more positive evidence around skill mix, studies either showed increasing the ratio of registered nurses (compared to assistant or associate staff) decreased violence or made no difference. Findings on the levels of registered nurses on wards were equally mixed with studies finding both positive and negative associations with levels of violent incidents.
This general level of inconsistency in the data may not come as shock to clinicians or former inpatients; however, it is still important to evaluate the state of the current literature so that the nuances can be addressed in future studies.
Results on the relationship between staffing and violence on wards were far from clear.
Conclusions
The authors conclude:
The current staffing crisis in mental health nursing underscores the urgent need for robust evidence to guide policies and improve patient outcomes. However, existing research on the relationship between staffing, skill- mix and conflict/containment outcomes is limited by inconsistent findings, poor study design and inadequate adjustment for confounders.
Strengths and limitations
This is a well conducted review; the search strategy is likely to have captured the most relevant and salient published research available. The quality assessment is thorough and candid about the flaws of the included studies. This is woven through the analysis.
As the authors note, this was undertaken as part of a PhD, with the majority of work done by the first author in isolation. This can lead to some bias being introduced, but the team did try to address this.
The review is unable to draw firm conclusions about staffing levels or make any recommendations. However, this is because the current evidence is low quality. This is an important finding in itself, and not a criticism of the review. Systematic reviews of this type can be important not only for summarising and synthesising knowledge across studies, but also for highlighting serious shortcomings in the current evidence base.
The authors are honest about these limitations including the lack of disambiguation (of both the types of incidents occurring and types of staff included in the studies) in the methods of the included papers. They have made suggestions for future work and research in this area. However, I would personally like to have seen more discussion of acuity and of how ‘violence’ was defined.
Systematic reviews play an important role in highlighting uncertainty and gaps in the evidence base.
Implications for practice
The antecedents of violence in mental health settings are varied and numerous. Staffing is likely to have some impact, due to the interpersonal nature of mental health care. However, this is far more complicated that just numbers of people on the ward. Further and better-quality research is urgently needed to take account of the nuance of this subject including acuity levels, staff experience, and a wider range of safety outcomes.
At this point safe staffing levels of mental health wards are not, and cannot be evidence-based, which is a disservice to both staff and service users. This review therefore serves an important role in articulating an inconsistent and weak evidence base in an extremely important area of mental health practice.
The main implication of this work is that the current research on mental health staffing levels is simply not good enough. This review cannot make recommendations on the best way to staff a ward to reduce violence because the evidence is just not there to do so. This is a symptom of the lack of regard and esteem that is given to mental health services. In 2015, NICE began to write safe staffing guidance for mental health services in England and stopped without publishing. Ten years later, we appear to be no closer to understanding the best way to staff our mental health units for the benefit of service users and staff. This has led to mental health units having some of the worse retention rates in the NHS and service users reporting repeated poor experiences at best, and actively being traumatised at worst.
Given the seriousness of the topic and the impacts of violent incidents on staff and patients, significant investment is needed into high quality research around mental health staffing. We urgently need:
- Increased funding for research into mental health staffing.
- High quality research into staffing levels and skill mix and the effect that has on safety in mental health settings. This should include inpatient and community settings.
This is a complex area, much more complex than general health care, and this is one of the reasons it has not received much attention. However, the interpersonal nature of mental health treatments mean it is crucial to achieving positive outcomes for service users. As the Government increasingly encourages new roles such as the nursing associate, the need for understanding staffing and skill mix becomes even more important.
Safe staffing levels on wards are not, and cannot be evidence-based, because we don’t have sufficient research. This is a disservice to everyone involved.
Links
Primary paper
Woodnutt, S., Hall, S., Libberton, P., Ball, J., Dall’Ora, C., & Griffiths, P. (2025). The Association Between Nurse Staffing and Conflict and Containment in Acute Mental Health Care: A Systematic Review. International Journal of Mental Health Nursing, 34(2), e70039.
Other references
Griffiths, P., Saville, C., Ball, J., Dall’Ora, C., Meredith, P., Turner, L., & Jones, J. (2023). Costs and cost-effectiveness of improved nurse staffing levels and skill mix in acute hospitals: A systematic review. International Journal of Nursing Studies, 147. https://doi.org/10.1016/j.ijnurstu.2023.104601
Karpusheff, J. (2023). What one research synthesis tells us is missing from the NHS workforce plan. The Health Foundation. https://www.health.org.uk/news-and-comment/blogs/what-one-research-synthesis-tells-us-is-missing-from-the-nhs-workforce-plan
Long, J., Ohlsen, S., Senek, M., Booth, A., Weich, S., & Wood, E. (2023). Realist synthesis of factors affecting retention of staff in UK adult mental health services. BMJ Open, 13(5), e070953. https://doi.org/10.1136/BMJOPEN-2022-070953
Van den Heede, K., Cornelis, J., Bouckaert, N., Bruyneel, L., Van de Voorde, C., & Sermeus, W. (2020). Safe nurse staffing policies for hospitals in England, Ireland, California, Victoria and Queensland: A discussion paper. Health Policy, 124(10), 1064–1073. https://doi.org/10.1016/J.HEALTHPOL.2020.08.003
