Problem

In 2014, the ‘triage ward’ system was introduced into three boroughs (Lewisham, Lambeth and Croydon) in South London and Maudsley (SLaM) NHS Foundation Trust. Similar to an acute medical admission unit, the triage ward would rapidly assess and treat all new admissions. The patients would either be discharged or admitted to a ‘locality ward’ for further treatment. Patients transferred from triage to locality wards were not handed over, although a few with pressing issues may be discussed informally through emails, phone calls or ad hoc face-to-face meetings. This leads to potential problems. In order to identify which tasks were outstanding, the receiving medical team needed to read through electronic records which was time consuming. Often it was unclear from these records which tasks were complete and which were pending; thus, tasks and referrals could be missed or duplicated. More than once, a referral to another specialty was duplicated due to inadequate handover. Physical health issues were not immediately apparent. For example, a patient with poorly controlled diabetes may not have optimal blood glucose monitoring as planned. Overall, the lack of handover impaired efficiency, patient safety and overall quality of care. In addition, patients were often transferred out of hours, and so, the necessary phone call or email to handover important information could be delayed or not made at all.

We aimed to improve efficiency, patient safety and overall quality of care by formalising the medical handover between triage and locality wards, by developing a system whereby a written succinct and pertinent handover was provided for every patient transferred, over a 4-month period. After piloting the project at the Lewisham site, a secondary aim was then to expand the project to the two other boroughs, Croydon and Lambeth, which were using the triage model.

Background

The project was based in SLaM NHS Foundation Trust which provides the widest range of NHS mental health services in the UK. Overall, the trust serves a local population of 1.3 million people in some of the most deprived areas of London.

For the past decade, the importance of safe handover has become ever more salient as the European Working Time Directive and resultant shift patterns increase the frequency of transfer of care between teams. The British Medical Association in collaboration with the National Patient Safety Agency, the General Medical Council, the Royal College of Physicians and the Royal College of Surgeons have all released expansive guidance on handover.1–4

Handover recommendations from these bodies and related Quality Improvement (QI) projects tend to focus on improving handover between shifts. 5 6 There has been less focus on handover of transferred patients between medical admission units (MAUs) and wards for instance, or from intensive therapy units to wards, although some QI projects have begun to address this.7 8 With mounting bed pressures, patients are being transferred between acute units and wards out of hours, often without the responsible doctor’s knowledge.7 8

In psychiatry, this has traditionally been less of a problem, as patients are usually admitted to a ward in which they remain until discharge. However, at SLaM NHS Foundation Trust, a triage system was introduced, in an attempt to improve efficiency of inpatient care.9 The Lewisham Triage system started in 2003, Lambeth in 2012 and Croydon in 2014. In this system, all acute admissions are directed to one triage ward that has a capped length of stay (7–10 days). Similar to an MAU, the aim is rapid stabilisation of the acute episode and either discharge from triage ward or transfer to a locality ward.9 For Lewisham and Lambeth, there are 900–1000 admissions a year, 50% are discharged within 7 days and the rest transferred for a longer stay.9 Again, like MAU, patients from these triage wards are often being transferred out of hours and to date, without any formal handover process between doctors in place.

Patients with serious mental illness often have complex physical health needs and an improper handover ‘can be a major contributory factor to subsequent error and harm to patients’.1 This project aimed to address the lack of medical handover, an unforeseen consequence of the introduction of the triage ward system in psychiatry.

Baseline measurement

We used a short survey to measure junior doctors’ opinions on the current state of handover before our intervention. We repeated the measure 4 months after implementing change. To avoid potential confounders, the same doctors were surveyed before and after.

The survey was anonymous, so it should be a valid reflection of doctors’ opinions. There were no major changes in the transfer system other than our intervention while the project was underway, so it would be very unlikely that changes in survey results from baseline could be due to much else except our intervention. The rationale for choosing this measure was that our intervention was solely for junior doctor use, so the perceived utility of our intervention for these stakeholders was important. While it was anticipated that patient outcomes would be improved if key medical information was handed over, it was difficult to find a way of capturing this and attributing any improvement directly to the handover system. Serious incident (SI) data were collected preintervention and postintervention, but much of this data was for incidents not related to the handover: for example, violence, absconding and staffing issues. This limitation is discussed below.

The survey questions were as follows:

How adequate do you think the current handover system is overall?

How important is it to have a handover system?

Currently, how easy is it to ascertain which jobs are outstanding for patients transferred from triage ward?

What form do you think the handover should take? Options: verbal, email, brief entry on electronic notes, brief discharge summary on electronic notes

Do you have any suggestions for how the handover process could be improved?

We asked doctors to answer the first three questions using 1–5 Likert scales, with one being the lowest rating.

The results of the baseline measurements for all three boroughs are displayed in the attachment.

We obtained SI data 6 months preintervention as a baseline measurement and compared this with data 6 months following the intervention. This was used to monitor changes in patient outcomes.

The baseline results showed that doctors on the locality wards found the current handover system inadequate. They found it difficult to find the information they needed and thought a brief entry on our electronic records (electronic patient journey, ePJS) would be the best format. The vast majority of doctors thought it was important to formalise handover. See online supplementary file: baseline result.

Design

During the pilot project in Lewisham, the baseline measurement survey formed our main consultation process, and using these results, we came to the final design for our intervention, a formal documented handover on ePJS, example below:

Triage Junior Doctor handover:

• Outstanding jobs: has refused bloods, physical, ECG, so these are still pending

• Referrals: neuropsychiatry referral complete 4/7/14—see correspondence, needs to be chased

• Physical health problems: type 2 diabetic on metformin, blood sugars  stable

• Other comments: nil

We concluded the above would be the most effective handover system for two reasons. First, it is the format that the majority of the doctors thought would be the most helpful (see Baseline measurements section). Second, it had a consistent format that would convey the important information succinctly and clearly. To ensure the workload of triage doctors was not significantly increased, we included only the most pertinent information.

The main problem we anticipated at this stage was that patients could be transferred without our knowledge (ie, not ‘captured’) and thus not receive a handover.

A simple strategy was devised to increase ‘capture rates’, whereby each evening, a doctor on triage ward would write down a list of the current inpatients on the ward. The next morning this list would be compared with the updated patient list on the ward whiteboard. The patients no longer on the ward had a handover duly written.

Given the only marginal increase in workload, we felt the intervention would likely be sustainable over the long term. To ensure continuity between 6 monthly rotations of doctors, we incorporated the handover system into the ‘Junior Doctor’s Guide’ for Lewisham triage ward, so that it became an expected part of the junior doctors routine work. We also involved the triage ward consultant (RR) who was similarly committed to sustaining the project across rotations.

Strategy

Lessons and limitations

The notable strength of the project was its simplicity and ease of implementation, which resulted in improvement in the measured outcomes across all three boroughs and also contributed to its sustainability in the pilot borough. We achieved sustainability in Lewisham by incorporating the handover into systems that were already in place, so that it became a natural part of the junior doctors’ daily work. We also used relationships with permanent staff to promote this. These strategies proved successful. Using the survey as both a consultation device and as a baseline measure meant that the stakeholders, the junior doctors themselves, had agreed on its intrinsic worth and utility prior to implementation; there was an appetite for this intervention. Thus, the receiving doctors expected and benefited from the handover, which motivated the doctors on triage to ensure it was completed.

The main challenge and limitations of the project came after the roll-out to other boroughs, and it showed how even a very simple intervention is difficult to replicate between sites. For example, at the time of the project in Croydon, there were far less transfers than other sites due to ward closures, so it did not give an accurate picture of how the implementation would work when the system was running as normal. This unforeseen factor gave a very small sample size, which is a limitation of the project. The ward subsequently closed prior to sustainability data being collected at 1 year.

In Lambeth, the project was not sustained after 1 year, where it was challenging to successfully implement the handover due to a number of human factors. The handover was only taken up by one of the doctors working on the ward, and therefore, data were only obtained over a 3-week period rather than 4 months. A possible reason for the low uptake among the junior doctors could be related to the additional work required in completing the handover entry. This is despite the general consensus that having a handover system would be helpful and that it was quick to complete. It has been observed in other quality improvement projects that changing the handover culture of departments can be challenging particularly when a handover culture is deeply embedded10; the changes are perceived as more time consuming than the current system in place11 or thought to involve additional work.12

One way to improve the implementation of the handover project would be to ensure that all the key team members were fully behind the project and finding ways to motivate them. If we were to do the project again, we would ensure all the doctors on Lambeth triage were directly involved in the surveying and consultation process, so that they would be more likely to realise the worth of the project and be accountable to the junior doctors expecting a handover. We would also attempt to involve the triage ward consultant which may help with encouraging the trainees to adopt the new handover system. The importance of engaging all the key team members, particularly senior staff, has been identified in other handover quality improvement projects.10 13

Another limitation was capturing improvement in patient outcomes as a result of the intervention. SI data offer a broad  snapshot preintervention and postintervention, but given the complex nature of ward life, it would seem impossible to ascertain how improvement in patient outcomes could be directly attributed to a better handover, which itself covers multiple aspects of patients’ care.

In terms of the limits of generalisability, we have shown that the project was applicable to other boroughs within the trust, but more work is needed to make it as successful as the Lewisham pilot. The handover template was tailored to the specific needs of a psychiatric junior doctor, receiving a patient from the SLaM triage ward system, and we worked within the capabilities of our electronic record system. It is therefore limited to this specific setting. However, features of the methodology of this project are generalisable, for example, working with systems already in place and consulting stakeholders before the intervention. This feature both raises awareness and gives stakeholders a sense of ownership in the project. This feature could create bias, in the sense that the same doctors were surveyed before and after implementation, and they thus had invested in the intervention to a small extent. It would, however, serve to reduce variability of surveying a different group of junior doctors.

The capture rates were low in Lambeth, and it would be useful to test the hypothesis above: that the motivation of the ward doctors on triage is key to capture rates and that motivation depends on level of involvement in the project which then leads to accountability.