Transforming triage

By Kate Horton, Nina Johns, Laura Goodwin and Sara Kenyon on 25 May 2018 Midwives Magazine

Triage based on clinical priority is standard procedure in A&E departments, and should be rolled out to maternity units too, argue Kate Horton, Nina Johns, Laura Goodwin, and Sara Kenyon.

Triage systems are designed to prioritise the provision of care according to clinical need. Prioritisation follows a standardised initial consultation and physiological assessment, and depends on the presenting condition, the severity of physical symptoms and measurement of vital signs. Triage is well established in emergency medicine, but no standardised triage system currently exists within maternity care.

The physiological changes in pregnancy – increased heart rate, lower blood pressure and increased respiratory rate – mean that general parameters of standard triage tools may not be applicable. The underlying good health of the maternity population may also mask the severity of maternal illness if a specific assessment is not undertaken. Existing triage tools also do not allow the assessment of the condition of the unborn baby.

Previously, all women attending maternity units, other than for scheduled antenatal clinic visits, were reviewed on labour wards. Confidential enquires (CMACE, 2011; CEMACH, 2007) have identified high numbers of women attending with problems that were pregnancy-related, but not labour, whose attendance at the delivery suite diverted midwifery staff and clinical care from women in active labour. National recommendations that women attending with unscheduled visits should be seen in areas away from the delivery suite are now accepted throughout the UK. However, maternity units were not designed for this and ‘triage’ departments have evolved in other physical spaces, such as day assessment units and labour induction bays.

These maternity triage departments have expanded without standardised pathways, and their workloads continue to grow without appropriate organisational and clinical systems.

Standardising the system

The need for a standardised maternity triage system has been recognised in recent UK confidential enquiries (CMACE, 2011; CEMACH, 2007) as well as many local reported clinical incidents. The need to develop guidelines for the triage of pregnant women has also been highlighted by the American College of Obstetricians and Gynecologists (2016), which advocates the use of tools such as the Birmingham system described below to improve quality and efficiency. The demand for such a system has been further emphasised by the NICE guideline for safe midwifery staffing (NICE, 2015), which defines a delay of 30 minutes or more between presentation and triage as a ‘red flag event’.

The Birmingham symptom-specific obstetric triage system (BSOTS) was co-produced in 2013 by midwives and obstetricians from Birmingham Women’s and Children’s NHS Foundation Trust and researchers at the University of Birmingham. The system is based on established triage systems in emergency medicine and uses a uniform assessment with clinical prioritisation of the common conditions that women present within maternity triage.

Initial assessment

An initial standardised assessment of each woman identifies her presenting condition, key clinical symptoms and physiological indicators. Symptom-specific prioritisation algorithms use this information to define the level of clinical urgency using a four-category scale. The guidelines for levels of clinical urgency and immediate subsequent care were developed using current available evidence, and consensus statements with the agreement of the local maternity team.

The initial triage assessment, lasting around five minutes, is completed within 15 minutes of the woman’s attendance, and includes the taking of a brief maternal history, completion of baseline maternal observations (temperature, pulse, respirations, blood pressure), assessment of pain levels, abdominal palpation and auscultation of fetal heart rate (if the woman is antenatal). This assessment is used to define a category of clinical urgency, which guides timing of subsequent assessment and immediate care using algorithms.

These standardised symptom-specific algorithms are specific to the primary reason for attendance – abdominal pain, antenatal bleeding, hypertension, suspected labour, ruptured membranes, reduced fetal movements, feeling unwell and postnatal concerns. The clinical priority allocated is colour-coded depending on urgency: red, amber, yellow and green. The algorithms also guide immediate care of the woman and direct further investigations.

The initial triage assessment is carried out in the ‘triage room’ – ideally a single identified room – although that room may change/rotate if women cannot be moved after assessment. Following their initial triage assessment, women will be seen in the order of their clinical need and are informed when this is likely to be. If designated as being ‘amber’ priority, they will stay in the triage unit for urgent assessment; if ‘yellow’ or ‘green’, they can return to the waiting area until further assessment. Women prioritised as ‘red’ are usually transferred immediately to the labour ward for emergency care or imminent birth.

Local midwifery and medical staffing will depend on how busy the triage department is, with the priority to undertake the initial triage within 15 minutes of arrival.

In large and busy departments, it is recommended that one midwife is responsible for initial triage, while another midwife undertakes subsequent care and investigations. In smaller departments, one midwife can undertake both roles but needs to remain responsive to new arrivals and interrupt the ongoing care of women to triage others.

Using a standardised triage system such as BSOTS improves the overall management of the department – in addition to the care of individual women – by maximising the effective use of available resources and allowing midwives and medical staff to easily view the clinical workload. BSOTS enables midwives to have an overview of the number of women in the department; those who have not yet had their initial triage assessment and, following initial assessment and triage, the level of clinical urgency for each woman.

BSOTS also creates a shared language between clinicians, enabling medical staff to prioritise which woman should be reviewed next, facilitating effective handover between staff, and prompting rapid escalation of clinical and staffing concerns.

The BSOTS triage system has been used by Birmingham Women’s Hospital since 2013 and is well established as the standard of care for women presenting with unscheduled pregnancy-related concerns. Following initial review of the system, three other West Midlands maternity units have implemented BSOTS in their triage departments and worked with the University of Birmingham to further evaluate its safety and effectiveness.

A mixed-methods design was selected for evaluating the impact of the introduction of BSOTS, the objective of which was to evaluate the recognised features of a robust triage system according to the following criteria:

1. Utility: System must be easy to understand and simple to apply by clinicians.
2. Validity: System should measure what it is designed to measure; it should measure clinical urgency as opposed to severity or complexity of illness.
3. Reliability: The application of the scale should be independent of the clinician performing the role and be consistent. ‘Inter-rater reliability’ is the term used for the statistical measure of agreement that is achieved by two or more raters using the same scale.
4. Safety: Triage decisions must correspond with objective clinical criteria and must optimise time to medical intervention.

Improving safety

An initial evaluation at Birmingham Women’s Hospital showed that BSOTS increased the number of women seen within 15 minutes of attendance to maternity triage from 38% (159/421) to 53% (209/391) – relative risk 1.4 (1.2, 1.7 (95% confidence interval (CI)) (Kenyon et al, 2017). The system also appeared to reduce the time spent waiting for assessment and the interval between attendance to medical review for those who required it.

Numbers of women who reattended, when they were next seen by maternity services and the investigation of predefined maternal and neonatal morbidities suggested validity and improved safety.

To evaluate the reliability of the BSOTS triage algorithms, 30 midwives who had undergone the training agreed to complete eight scenario cases. Excellent inter-rater reliability (intraclass correlation 0.961 and 95% CI 0.91-0.99) was demonstrated using the cases with the BSOTS algorithms. Scenarios with the highest and lowest clinical importance were most consistently assessed accurately, and all maternity units using BSOTS showed the excellent inter-rater reliability.

These results show that BSOTS is relatively easy to use – achieving the utility criterion – and the triage decisions appear to correspond with the clinical situation and improve safety. This was confirmed by findings from focus group interviews, which demonstrated that midwives felt the introduction of BSOTS helped them manage and organise the department. They reported that they felt the safety of women and their babies had improved and that the system, although standardised, afforded them opportunities to use their clinical judgement when appropriate.

To enable further review and planning, a national consensus meeting was held in May 2017, chaired by the director of the Birmingham Clinical Trials Unit and comprising representatives of the RCOG, the National Maternal and Neonatal Health Safety Collaborative (NMNHSC), service users, the PROMPT faculty and West Midlands clinical networks. At the meeting, it was agreed that no further evaluation was required as the system appeared to increase safety for mothers and babies, as well as clinicians. To that end, BSOTS has been included as a safety option by the NMNHSC, and local maternity clinical networks and the West Midlands Academic Health Science Network are supporting implementation. The RCM and the RCOG also support its implementation and the system is being rolled out nationally. 

Kate Horton is delivery suite matron at Birmingham Women’s Hospital; Nina Johns is an obstetrician at Birmingham Women’s Hospital; Laura Goodwin is a research fellow at the University of Birmingham; and Sara Kenyon is professor of evidence-based maternity care at the University of Birmingham

Interested in using BSOTS?

If you are interested in more information or implementing the system in your triage department, contact the BSOTS team via kate.horton@bwnft.nhs.uk