Community mental health care support the largest number of patients within England’s mental health services. Yet, when it comes to research on patient safety in these settings, there’s still a lot we don’t know (Averill et al., 2023). In a previous study on nursing staff in England, I found that, for the first time, more mental health nurses are now working in community settings than in hospitals (Woodnutt et al., 2024). Despite this shift, most of the existing research on patient safety in mental health focuses on inpatient care (see, for example, the recent coverage here of a systematic review of adverse experiences in mental health wards), and overall, there is a lack of evidence around patient safety (Thibaut et al., 2019).
To address this gap, (Averill, Sevdalis and Henderson, 2025) undertook a first-of-its kind study looking at the free-text information that is routinely reported by staff in incident reports. This approach is crucial in mental health settings, where most staff-reported observational data lack numeric scales for comparison and is instead entered as free-text. While the National Reporting and Learning System (NRLS), does compile quantitative information from the National Health Service (NHS) and make this freely available, most of these data are counts of incident types, rather than detailed notes written by staff.
These notes contain a rich source of information – and yet, they have not previously been studied for patterns or themes, and therefore the information is not widely available to disseminate outside of the NHS. It is important that dedicated researchers are involved in this process. They have skills and experience in mitigating potential biases that could arise when looking at the data – and help discern real or truthful relationships that ultimately improve the quality of care.
Unlocking insights from staff-reported incident notes could transform patient safety in community mental health care, where research remains limited.
Methods
Averill et al. (2025) conducted a mixed-methods study to identify themes in community-based incident reports and evaluate proposed safety solutions. They did this by analysing reports from 22 NHS Trusts, which represents a 10% sample of England’s 225 Trusts. Working with NHS England’s Patient Safety Data Team, they retrieved all incident reports from 1st January to 31st December 2019, filtering for adults (18-65) and excluding those related to institutional care or hospital settings. This resulted in the identification of 4,804 reports. The severity of harm in these included reports was: 3,020 described as no harm, 918 low harm, 465 moderate harm, 35 severe harm, and 366 deaths.
A subset of 1,443 reports underwent inductive content analysis, where researchers coded data without predefined themes to allow patterns to emerge. To ensure reliability, 10% of the sample was coded independently, with discrepancies resolved through team discussions, with a third researcher acting as an arbitrator if needed. The researchers then analysed the proportion of incident reports across different categories.
Results
The results can be split into three main categories:
1. The nature of reported incidents
Community-based patient safety incidents were commonly associated with issues in investigations, documentation, referrals, communication, administration, treatment, medication, and diagnosis. Delays, errors, and miscommunication frequently led to repeated tests, treatment disruptions, and, in severe cases, patient harm or death. Diagnosis and assessment incidents were the most common cause of moderate (35.9%) and severe (3.4%) harm, while medication errors (17.1%) and administration issues (13.0%) were frequently reported with no harm. Suicide and self-harm were the leading contributors to fatal outcomes, with suicide accounting for 52.1% of deaths. Other key concerns included delays in referrals, failures to act on symptoms, and communication breakdowns.
2. Factors that contributed to incidents
Key factors contributing to patient safety incidents included errors in investigations and documentation, communication failures, administrative and process issues, treatment delays, and medication errors. Misprocessed lab tests, incomplete records, and referral mistakes often stemmed from staff inexperience and system failures. Communication breakdowns between professionals, services, and patients led to missed referrals and critical miscommunications about care. Administrative errors, such as mishandled appointments and medication issues, were exacerbated by staffing shortages and unclear responsibilities. Treatment delays, including inadequate risk assessments and hospital admission issues, further compromised patient safety.
Harmful outcomes, including suicide, self-harm, overdoses, and violence, were frequently linked to intoxication and impulsive acts, service inaccessibility, and inadequate risk assessments.
3. Safety solutions
Proposed solutions for patient safety incidents focused on reinforcing policies, improving communication, streamlining administrative processes, and enhancing staff training and supervision. Key measures included clearer documentation, procedural changes, and better inter-team collaboration for investigations, referrals, and diagnoses.
Communication issues could be addressed through improved clinical handovers, role clarification, and inter-agency coordination. Administrative errors could be mitigated via policy reinforcement, staff training, and enhanced information-sharing. Treatment and procedural improvements promoted better coordination, and continuity of care for high-risk patients.
Medication safety strategies included structured handovers, patient counselling, staff training, and double-checking procedures. In cases where service influence on harm was unclear, interventions could focus on restricting access to means, harm reduction, and proactive monitoring.
This study highlights patient safety incidents related to communication, documentation, and treatment errors, with solutions proposed.
Conclusions
This study summarises rich data sources for patient safety from a large sample and provides invaluable insight into an area of research that is under explored.
The key findings include:
- Common incidents included documentation, communication, and medication, with unique incidents in community mental health, such as Mental Health Act legal errors and delayed (MHA) assessments.
- Harmful outcomes like self-harm and suicides were prevalent, but the role of mental health services in preventing these was not always clear.
- Proposed solutions often emphasised reinforcing existing policies and service user-directed approaches, with less focus on detailed explanations or systemic improvements.
- The findings highlight that mental health patient safety requires a unique approach, different from physical health or other care settings.
Mental health patient safety requires a unique approach, different from physical health or other care settings.
Strengths and limitations
The study’s strengths include a systematic, manual coding approach and a mixed-methods design that enriches routine data. However, limitations include the potential for sampling biases and underreporting of incidents (and therefore systematic bias within the data).
Whilst not a limitation of the study design, differences in how mental health services define and report incidents may affect the data. Services handling higher-risk patients might tolerate certain behaviours more than others, leading to inconsistencies. Additionally, staff may underreport incidents to avoid accountability, creating gaps in the data. Involving patients and carers in routine reporting is challenging due to time constraints, though it is more common for serious incidents.
Despite these challenges, using routine data in this way is often the best available source to explore relationships. Therefore, neither of these ‘limitations’ are likely to have significantly changed the conclusions – as the authors robustly managed the data that were available, and the size of the sample likely reduced some of the ‘noise’ in the data.
The reporting system does allow public reports, but this is underutilised, as no reports from the public were found in the research. Therefore, whilst efforts were made to identify and involve the voice of patients and their families/carers, these were not included. Future research could include deeper ethnographic approaches to increase involvement from patients and their families/carers in understanding adverse incidents.
Involving patients and carers in routine reporting is challenging due to time constraints.
Implications for practice
This study provides unique oversight of common incidents in community mental health services, which is, as yet, unknown in the academic literature. These findings could be used by clinical teams and patients to consider how to best approach care, identify potential hazards and risks, and seek to engineer solutions to minimise harms. In effect, the findings allow clinicians and researchers a framework to begin to understand how safety is conceptualised in community mental health services.
In a previous review of English data on incidents and staffing, I compared how incidents in mental health services tend to focus on patient factors such as self-injury or aggression as leading categories (Woodnutt et al., 2024). This conceptual model differs from physical health care settings where there is much greater focus on service-related factors (such as missed assessment) which lead to consequences for patients. Averill et al.’s (2025) new research promotes the view that community mental health services are in themselves a distinct population when it comes to monitoring and assessing the safety of patients. We can only hope that this study precipitates more research in this area given the ubiquity of community based supports in the UK.
Averill et al.’s analysis is encouraging as it highlights service-level incident categories, reflecting a shift toward monitoring service behaviour rather than just patient actions. This helps identify gaps and improve care, thus improving care quality and safety for patients.
Mental health services—and the patients they support—continue to face issues with legacy stigma and epistemic injustice, often attributing incidents to patient behaviour rather than systemic shortcomings. Whilst self-injury is the leading incident in England, framing it solely as an adverse event risks ignoring it as a part of mental illness and missing opportunities for preventive care.
What we don’t know a lot about in mental health care is when care is missed. In physical health hospital settings, a significant amount of research has been done to identify care that is left ‘undone’ at the end of shifts (Ball et al., 2014; Griffiths et al., 2018) – and this can be used to consider what an appropriate minimum number of staff is for a set group of patients. This new study from Averill et al. (2025) helps to light the way for mental health patient safety research and importantly broadens the focus to the places in the NHS where most people receive support.
The findings allow clinicians and researchers a framework to begin to understand how patient safety is conceptualised in community mental health services.
Links
Primary paper
Averill, P., Sevdalis, N. and Henderson, C. (2025) ‘Patient safety incidents within adult community-based mental health services in England: A mixed-methods examination of reported incidents, contributory factors, and proposed solutions’, Psychological Medicine, 55:e8. https://doi.org/10.1017/S0033291724003532
Other references
Averill, P. et al. (2023) ‘Conceptual and practical challenges associated with understanding patient safety within community‐based mental health services’, Health Expectations, 26(1), pp. 51-63. https://pubmed.ncbi.nlm.nih.gov/36370458/
Ball, J.E. et al. (2014) ‘‘Care left undone’ during nursing shifts: associations with workload and perceived quality of care’, BMJ quality & safety, 23(2), pp. 116-125. https://doi.org/10.1136/bmjqs-2012-001767
Griffiths, P. et al. (2018) ‘The association between nurse staffing and omissions in nursing care: A systematic review’, Journal of advanced nursing, 74(7), pp. 1474-1487. https://pubmed.ncbi.nlm.nih.gov/29517813/
Thibaut, B. et al. (2019) ‘Patient safety in inpatient mental health settings: a systematic review’, BMJ open, 9(12), p. e030230. https://pubmed.ncbi.nlm.nih.gov/29517813/
Woodnutt, S. et al. (2024) ‘Analysis of England’s incident and mental health nursing workforce data 2015–2022’, Journal of Psychiatric and Mental Health Nursing, 31(5), pp. 716-728. https://pubmed.ncbi.nlm.nih.gov/38258945/
