Systems-based approaches for learning from patient safety incidents: a recent discussion at the Patient Safety Management Network

Summary

This blog provides an overview of a discussion at a Patient Safety Management Network (PSMN) meeting on 26 August 2022. The discussion considered the use of two different system-based approaches for learning from patient safety incidents recommended by the NHS Patient Safety Incident Response Framework (PSIRF).

The PSMN is an informal voluntary network for patient safety managers. Created by and for patient safety managers, it provides a weekly drop-in session with guests to talk through issues of importance, offer peer support and create a safe space for discussion. You can find out more about the network here

Content

As part of implementing the NHS Patient Safety Strategy, there are currently a number of new initiatives being rolled out across the NHS which are intended to achieve its vision of continuously improving patient safety. This includes the development of the Learn from patient safety events (LFPSE) service, for recording and analysing patient safety incidents, a new framework for involving patients in patient safety and the Patient Safety Incident Response Framework (PSIRF).  

PSIRF sets out the NHS’s approach to developing and maintaining effective systems and processes for responding to patient safety incidents. It promotes systems-based approaches for learning from incidents, rather than methods that assume simple, linear identification of a single cause.

A key aim of the Patient Safety Management Network (PSMN) is to provide those working in patient safety with a shared space to discuss new policies that impact their work, and to share knowledge and resources with their peers. In this session, the Network considered how the systems-based approaches to incidents recommended by PSIRF can be implemented in practice. They focused on an example that used two of these tools in relation to a specific patient safety incident—an After Action Review (AAR) and an observational analysis,

Patient safety incident

The example discussed in this meeting was shared by a Patient Safety Manager who had applied two PSIRF tools to a specific patient safety incident that took place on a surgical ward, where an elderly, partially-sighted patient was due to be discharged. The original intention was to consider the implications of applying these tools to two separate incidents, but due to the level of discussion around the first incident there was not time to do this. However, Network members agreed on the value of having a future session focused on another example.

In this case, an Internationally Educated-Nurse (IEN) came to issue a patient with medication on discharge. On the ward they took out medication from a POD locker (‘Patients Own Drugs’ - a bedside cabinet designed to offer safe and secure medication storage) and, when distracted by another task in a busy ward, put this on a side along with medication issued for the patient by a pharmacy.

Subsequently, the patient took away both sets of medication, however it transpired that the medication in the POD locker belonged to another patient. The patient took the incorrect medication following discharge and was subsequently readmitted to hospital with an irregular heartbeat.

To analyse this incident, the Patient Safety Manager decided to apply the Human Factors and ergonomics tools being promoted through PSIRF rather than undertake a Root Cause Analysis recommended by the Serious Incident Framework.

After Action Review

In response to this incident, first an AAR was undertaken. This is a structured review based on four questions aimed at understanding what happened, why it happened and how it can be done better by those responsible and involved in the incident.

This review was undertaken by the Patient Safety Manager and a colleague in the Patient Safety team, and involved all staff involved in this incident including the Ward Manager and the IEN. In this case, there were specific reasons why it did not involve the patient, although AARs often can. The review concluded that:

• What was expected to happen? The patient should have been delivered the correct medication by the IEN, which should have been checked by another staff member as the IEN was still waiting their PIN (registration code from the Nursing and Midwifery Council).
• What actually occurred? The patient was sent home with the medication belonging to another patient which when taken, resulted in a hospital re-admission.
• Why was there a difference between what was expected and what actually happened? The IEN was not aware they required supervision discharging patients with medication (the ward coordinator was not aware they did not have their PIN yet), the POD locker was not emptied after the last patient and the ward was short staffed.
• What was the learning? Staff coordinating the ward need to be aware of IEN capabilities, IENs need to be aware of restrictions prior to receiving their PIN, POD lockers require checking on discharge of patients.

Observational analysis

The Patient Safety Manager felt that the AAR alone hadn’t necessarily provided the team with enough insights into the issues involved in the incident and decided to apply another recommended PSIRF tool, an observational analysis. The intention of this was to understand how the ward worked in relation to patients receiving medication from POD lockers on discharge, seeing ‘work-as-done’ as opposed to ‘work-as-imagined’ by staff in this area.

This observational analysis was done using a locally adapted version of the tools recommended by PSIRF. Findings of the observation included:

• Environment – The ward was busy, noisy, and hot, with lots of activity taking place in a small space. The POD lockers themselves were not easily visible as white boxes on white wall.
• Person – There was limited communication between porters and nurses and limited dialogue/handover/briefing before a patient transfer.
• Organisation of work – Workload was extremely high. The POD locker was not checked routinely, and it was unclear whose responsibility it was to check medications in the lockers.
• Tasks – The task of checking a POD locker once a patient has moved is a simple one, but needs to be performed by a trained nurse and faces the competing priorities of patient care and patient flow.
• Technology and tools – POD lockers are not all in the same places, not all nurses have keys to them and there are no visual cues to check the lockers when a patient is moved or when there are drugs in them.

Evaluating the findings of this observational analysis, a key issue not picked up as clearly by the AAR in this case concerned the POD lockers—namely the lack of operating procedures and routine around these, limited staff having access to them and there being no clear responsibility for checking and clearing them.

Following completion of the AAR and observational analysis, both documents were uploaded to the Trust’s incident reporting system and an outcome letter was shared with the patient’s family, detailing what issues had been found and what action would be taken to address these. The family were appreciative of the information and were reassured that learning was being applied that would prevent future harm to patients.

The action to address the issues identified in the observation included referral to the Trust-wide Medicines Management Committee for review of the need for improvements in the management of medication and POD lockers.

Network discussion

In the subsequent discussion of these approaches to analysing and learning from this patient safety incident, there were a range of reflections from Network members:

In relation to the specific patient safety incident:

• It was noted that in this case, involving the IEN in the AAR was positive as it provided immediate reassurance to the staff member that the aim of this review to learn rather than blame, as the IEN had concerns about the professional consequences of this error and the potential impact on their employment status.
• There was a discussion of whether it would make sense to do a short-term fix with regards to the POD lockers, such as painting them a distinct colour, and whether this would have a significant positive impact, or potentially unintended consequences if done in isolation of other quality improvement activities.
• An interesting outcome in this case is that the AAR review seemed much more centred on the individual involved in the incident, while the observational analysis drew our wider environment factors.

In relation to the application of PSIRF tools more broadly:

• There were questions about how information from AARs, observations and other new PSIRF tools would subsequently feed into organisational plans. It was posited that these could be reviewed at regular intervals (for example, every three months) by the patient safety team, and their insights used to feed into an organisation-wide quality improvement project, or a thematic review. PSIRF highlights the new approaches and tools to be adopted, but organisations need to consider how they respond to the outcome of new tools and how information is reported and acted on with quality improvement projects and organisational oversight.
• There was a question about whether the staff conducting the AAR and observational analysis got the right support. A question was posed as to whether there could be an opportunity for a constructive friend challenge by a Human Factors expert or discussion about how this was approached afterwards?
• There was an acknowledgement that sometimes it can be difficult to define what observations fit into which SEIPs categories—for example, something in the ‘Environment’ that may also fit under the ‘Technology and tools’ heading. Also, a question was asked as to whether this matters as long as the learning is recorded.
• It was noted that training for PSIRF tools is covered in Healthcare Safety Investigation Branch training, but that it would be helpful if there were also simple practical guides to help staff when undertaking these reviews.

In relation to the observational analysis:

There was also discussion about how to approach observations of this type. Many highlighted the issue that when staff know they are being observed, they potentially act differently. The question was raised as to how close you get to seeing an accurate reflection of ‘work-as-done’—is the presence of someone observing having a significant impact on how activities are being approached? Other points raised included:

• An observational analysis of this type can be easily done in a hospital, but how effective or simple would it be to perform it in a community setting, for instance if the issue occurred involving a nurse in a patient’s home?
• Would it potentially be better to do observations while also doing a shift on a ward, as opposed to joining simply to do an observation? Or would this add in unexpected bias into the process?
• Is there more to be done for staff to understand how to ‘observe well’? With training or guidance from Human Factors/ergonomics experts?
• If the aim is to create an open, learning culture, it is important that staff are aware they are being observed so that they do not feel they are being spied on.
• It is important to clearly communicate the aim of an observational analysis to staff, highlighting that it is fundamentally to understand their work and improve safety.

Concluding comments

At the session there were a number of positive reflections on the use of new PSIRF tools and their potential to improve learning from patient safety incidents. The discussions also underlined the importance of ensuring that staff have the appropriate support and training to help embed the use of the tools and develop how the outcomes of each tool inform improvement and organisational oversight.

How to get involved in the PSMN

Are you a patient safety manager interested in joining the Patient Safety Management Network? You can join by signing up to the hub today  When putting in your details, please tick ‘Patient Safety Management Network’ in the ‘Join a private group’ section. If you are already a member of the hub, please email claire@patientsafetylearning.org

Related Reading

Applying the After Action Review for the PSIRF – some real life examples (10 March 2022)
Observational tools, Human Factors and patient safety: a recent discussion at the Patient Safety Management Network (9 March 2022)
Patient Safety Management Network – the time is now (25 October 2021)