Task shifting: a key aspect to improving care for women at risk of preterm birth

Abstract

Local problem

Until April 2021, women presenting to maternity triage with symptoms of threatened preterm labour (TPTL) and/or preterm premature rupture of the membranes (PPROM) were triaged by a doctor. Depending on the acuity on the labour ward, women in triage often had a long wait for a doctor’s review. These delays create anxiety for women and impair the capacity of triage midwives to care for other women.

Methods

The Plan-Do-Study-Act method of quality improvement was used for this project. 3 months prior to the intervention, the baseline assessment was women’s wait time for medical review when presenting with TPTL and/or PPROM.

Intervention

Triage midwives were trained in performing speculum examination on preterm (<37 weeks’ gestation) women to allow quicker review. Waiting time for review by a midwife vs doctor was compared using data collected between January and December 2021.

Results

88 eligible women were identified. 44 cases (intervention group) had their initial assessment by the triage midwife, while 44 cases (control group) had their initial assessment by a doctor. The mean waiting time between arrival and performance of quantitative fetal fibronectin (qfFN) in the intervention group was 67 min (SD=42.7), compared with 127 min (SD=61.2) in the control group (p<0.001). However, there was no significant difference in the waiting time between arrival and discharge/admission.

Conclusion

Women presenting with symptoms of TPTL are reviewed on average twice as quickly by the triage midwife compared with a doctor, allowing a quick reassurance for those where TPTL/PPROM has been excluded. However, the overall waiting time in triage was similar, as women in our unit currently need a doctor’s review before discharge.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• While symptoms of threatened preterm labour (TPTL) are a common reason to present to triage, the vast majority of women can be reassured that they are not at imminent risk.

• Long wait times for women in pain and at risk for preterm birth being reviewed by a doctor contribute to anxiety. Finding ways of improving review pathways is therefore important.

WHAT THIS STUDY ADDS

• We discuss how a simple task-shifting intervention—teaching triage midwives how to perform speculum examination on women under 37 weeks’ gestation—can optimise patient care and the key features of this quality improvement project that contribute to its sustainability.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Implementation of midwifery triage for TPTL halved waiting times, allowing quicker reassurance and optimisation of triage area.

Problem description

In our unit and in most maternity units nationwide, speculum examination on preterm women and, consequently, any point-of-care tests are performed by doctors only (general practitioner (GP) trainee or obstetrics and gynaecology (O&G) trainee). This review pathway leads to long waiting times, as doctors covering maternity triage are often busy on the delivery suite and, subsequently, increases anxiety for women presenting in threatened preterm labour (TPTL).

Available knowledge

TPTL is a common reason for pregnant women to present to hospital.¹ Among those presenting, only a few women will be at risk of true imminent preterm birth and require life-saving time-critical treatments, such as antenatal corticosteroids (ANCS), magnesium sulphate and in-utero transfer.^{2 3} However, with the help of preterm birth risk assessment (eg, speculum and the use of predictive tests), the vast majority of women will be reassured and discharged home.⁴ TPTL is known to be associated with a high level of anxiety, whether or not preterm labour is confirmed or only a ‘false alarm’.⁵ The challenge is, therefore, not only to identify those in threatening preterm labour from those who need reassurance but also to do so in a timely manner.

With National Health Service (NHS) workforce pressures in the UK likely to continue to rise,⁶ strategies to expedite patients’ review are more important than ever. In light of recent Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries reports and the Ockenden⁷ and East Kent enquiries⁸—both of which exposed serious failings in UK maternity care and continue to shape national practice—it is clear that prompt recognition and escalation of high-risk pregnancies is key to improving neonatal outcomes.² One strategy to improve efficiency and resource allocation is task shifting: training of healthcare professionals, such as midwives, in developing advanced practice and specialist skills.^{9 10} Another strategy is to optimise current triage processes, as was achieved with the Birmingham Symptom-specific Obstetric Triage System, a triage system developed solely for maternity services with the aim to triage women within 15 min after arrival.¹¹ The goal of this quality improvement project (QIP) was to see whether the performance of speculum and point-of-care tests by triage midwives, rather than doctors, significantly improved the waiting time for women attending with symptoms of TPTL.

Rationale

Our midwife-led QIP emerged during the creation of the Quantitative Innovation in Predicting Preterm birth (QUiPP) app toolkit, in which our unit was a stakeholder.¹² QUiPP app is a free UK Conformity Assessed-marked, mobile phone clinical decision support tool (quipp.org), which provides clinicians with a rapid individualised risk score for spontaneous preterm birth, based on the woman’s risk factors, the cervical length measurement and the level of quantitative fetal fibronectin (qfFN) in the cervicovaginal fluid.13,17 qfFN testing consists of a vaginal swab taken during a speculum examination. This step, however, causes significant delays in the review pathway due to the tacit rule that only doctors perform speculum examinations in women less than 37 weeks’ gestation. Midwifery scope of practice does not make this distinction,¹⁸ but this appears to be at odds with common practice.

This practice change was identified by our triage lead midwife at a QUIPP app toolkit stakeholder event as an opportunity to tackle the long waiting time women often experience in local maternity triage until they are reviewed by one of the on-call doctors.

Methods

Context and intervention

The project was conducted within the maternity triage at Queen Elizabeth Hospital, Lewisham and Greenwich NHS Foundation Trust, a public District General Hospital (DGH) in Southeast London, UK, with 3980 deliveries in 2021 >28 weeks’ gestation.¹⁹ The maternity triage is midwifery-led, and all women presenting there are first assessed by the triage midwife, who escalates the case to the obstetric team, if appropriate, as per internal guidelines. In the preintervention period, the examination of women presenting with TPTL and/or preterm premature rupture of the membranes (PPROM) was carried out exclusively by a doctor (GP or O&G trainee). Examination included a sterile speculum examination to assess the length and position of the cervix, the dilatation of the cervical os and the presence of any vaginal fluid (blood, liquor). During the speculum, two point-of-care tests are taken to aid risk assessment of imminent preterm birth and PPROM.

At the time of the QIP, the point-of-care test used to assess the risk of imminent preterm birth was based on the level of fetal fibronectin measured in the cervicovaginal fluid. The level, along with the women’s obstetric history and details about the current pregnancy, was input in the QUiPP app to calculate the individualised risk for a woman of going into preterm labour to help decision-making (admission vs discharge). At the time of publication, availability of the qfFN test had become limited in the UK due to commercial issues,²⁰ but alternative preterm prediction tests are available (eg, insulin-like growth factor binding protein-1), which are taken using the same speculum method and in the same clinical setting.²¹

The intervention—initiation of midwifery-led preterm birth risk assessment—commenced on 1 April 2021. Women were included if they had signs of preterm labour (abdominal pain, contractions or vaginal discharge suspicious of PPROM) between 22⁺⁰ and 36⁺⁰ weeks’ gestation. Women with vaginal bleeding, in active labour (cervical os >3 cm dilated) or those without record of which healthcare professional examined them, were excluded. In the postintervention period (from April 2021 onwards), women deemed at low risk of preterm labour by the midwifery speculum assessment (os closed, PPROM excluded and low chance of going into preterm labour according to the QUiPP app) would still have a clinical review by a doctor to assess for other potential causes of abdominal/back pain prior to discharge.

Study of the intervention, measures and analysis

The project used the Plan-Do-Study-Act (PDSA) method to improve the management of women presenting with TPTL in maternity triage. While one PDSA cycle is reported here, further cycles are planned within the Trust’s audit processes.

Plan

The key improvement identified was to allow triage midwives to perform speculum examinations on preterm women, aiming to reduce the long waiting times for doctor reviews. This intervention was selected because triage midwives were already trained to perform speculum examinations and high vaginal swabs on pregnant women beyond 37 weeks’ gestation—procedures technically similar to those needed to assess the risk of PPROM and imminent preterm birth in preterm women. The project team consisted of the project lead, the maternity triage midwives and support workers implementing change and collecting data and academic clinicians performing data analysis. A standard operating procedure (SOP) was developed and approved to safely introduce triage midwives performing preterm speculum exams and point-of-care testing.

Do

Training materials were created, and the triage lead midwife delivered face-to-face training to all five core triage midwives. Training was completed by June 2021, with 80% of staff trained within the first month of implementation. The intervention was rolled out according to the SOP, ensuring that triage midwives could perform the speculum examinations and relevant tests safely.

Study

The intervention was monitored for safety and effectiveness. No adverse events, such as iatrogenic rupture of membranes, were reported during implementation. Due to a lack of capacity, no run charts were created to monitor the progress (see Discussion). However, data analysed 9 months after the intervention confirmed the project’s success, with waiting times reduced by half.

Act

Given the successful and safe implementation of the intervention with positive outcomes, no further PDSA cycles were deemed necessary. The change in the review pathway was fully incorporated into the Trust maternity triage internal guidelines.

Additionally, informal feedback from the newly trained staff indicated that they found it easier to visualise the cervix in patients under 37 weeks’ gestation compared with those beyond 37 weeks, when the cervix is ripening in preparation for childbirth.

Baseline measurement from January to March 2021 was the women’s waiting time when presenting with TPTL and/or PPROM. The primary outcomes were the interval between arrival and performance of qfFN and the interval between arrival and admission/discharge. Secondary outcomes were speculum findings, qfFN results, administration of ANCS, antibiotic use and the decision to admit or discharge. We prospectively recorded participant baseline characteristics: parity, gestation and single versus multiple pregnancy. The data were collected by the clinical team using the digital maternity notes (iCare) and the maternity triage admissions record.

Project team members not involved in direct clinical care performed the unpaired t-test to assess whether the differences in the mean waiting time between both groups were significant and the χ² test to assess the impact of the type of examiner on the decision to admit/discharge. Data were analysed using Stata V.15 software (Stata Corp, College Station, Texas).

Results

88 eligible women were identified. 44 cases had their initial assessment by the triage midwife, while 44 cases (control group) had their initial assessment by a doctor. The baseline characteristics were similar in both groups (table 1).

Table 1. Baseline characteristics of women in preterm labour assessed in triage by midwives versus doctors.

	Controls (doctor performing qfFN* test)	Cases (triage midwife performing qfFN* test)
Parity	n (%)	n (%)
0	19 (43)	17 (39)
>1	25 (57)	27 (61)
Gestation (weeks’ gestation)
Mean	30.3	28.8
<28	11 (25)	16 (36)
28–31+	17 (39)	18 (41)
32–35+0	16 (36)	10 (23)
Multiple pregnancy	1 (2)	2 (5)

^*qfFN = quantitative fetal fibronectin

qfFN, quantitative fetal fibronectin.

The mean waiting time between arrival and performance of qfFN in the intervention group was 67 min (SD=42.7), compared with 127 min (SD=61.2) in the controls (p<0.001) (table 2). However, there was no significant difference in the waiting time between arrival and discharge/admission, with a mean waiting time of 157 min (SD=84.7) for the midwife group (intervention) versus 181 min (SD=66.6) for the doctor group (control) (p=0.14) (table 2). Secondary outcomes (speculum findings, qfFN results and ANCS/antibiotics given on this presentation) were broadly similar. The proportion of women being admitted or discharged was not significantly affected by who was performing the speculum examination, X2 (1, n=88) = 0.39, p=0.053 (table 2).

Table 2. outcomes of women in preterm labour assessed in triage by midwives vs doctors.

Primary outcome	Controls (doctor performing qfFN* test)	Cases (triage midwife performing qfFN* test)	P value
Interval between arrival and qfFN test (min)	127	67	< 0.001
Interval between arrival and admission/discharge decision (min)	181	157	0.14
Secondary outcome	Doctor	Midwife
Speculum findings	n (%)	n (%)	n/a
Os closed	42 (95)	42 (95)	n/a
Multiple os	0 (0)	1 (2)	n/a
1–2 cm open	0 (0)	1 (2)	n/a
Os not visualised	2 (5)	0 (0)	n/a
qfFN	n (%)	n (%)
<10	31 (70)	40 (90)	n/a
10–49	10 (23)	2 (5)	n/a
50–199	1 (2)	2 (5)	n/a
>200	2 (5)	0 (0)	n/a
Invalid	0(0)	0(0)	n/a
Steroids given on this presentation	1 (2)	1 (2)	n/a
Steroids given on previous presentation	1 (2)	1 (2)	n/a
Antibiotics given	0 (0)	2 (5)	n/a
Decision to admit	7 (16)	5 (11)	0.53
Decision to discharge	37 (84)	39 (89)	0.53

^*qfFN = quantitative fetal fibronectin

qfFN, quantitative fetal fibronectin.

Discussion

Summary

Our project demonstrates that women presenting with symptoms of TPTL are, on average, reviewed twice as quickly by a triage midwife compared with a doctor. As a result, the vast majority of women in our cohort received reassurance sooner, which is likely to have had a positive impact on their anxiety levels. However, earlier performance of the speculum examination did not improve women’s overall waiting time in triage. Women still had to wait for a doctor review before being discharged or admitted (even after exclusion of TPTL/PPROM by the triage midwife), due to guideline requirements for all women with abdominal pain needing medical review. This policy is to ensure non-pregnancy-related causes of pain are considered, which midwives are experienced at managing independently at term. We are planning further stakeholder discussion regarding whether it is appropriate to amend this guidance, which could further reduce women’s waiting times. Nevertheless, an initial review by the triage midwife led to optimisation of the triage space by allowing women deemed low risk to return to the waiting area while awaiting review by a doctor.

Interpretation

The strength of our QIP lies in its pragmatism and task-shifting component, relevant to all preterm predictive tests needing a speculum examination. The intervention has remained sustainable, as the change in review pathway has been incorporated in the maternity triage internal guidelines, even as the predictive tests themselves have changed.²⁰ The following points are likely to have contributed to the QIP success:²²

• Positive leadership and commitment from key workers:²³ the project lead, who was herself the triage lead midwife, played a key role in the success of the project. She did not only suggest the intervention but also ensured its application and maintenance on the front line by incorporating it during her shifts in maternity triage as well as by leading teaching sessions for the triage staff.

• Positive value-perception by staff of the change:²⁴ triage midwives were empowered to finally tackle a long-standing issue that had been causing both patients’ dissatisfaction and suboptimal use of the triage capacity.

• Consistency among staff participating in the intervention:²⁵ although some of our units, such as the labour ward and postnatal ward, are regularly unable to facilitate sufficient cover due to chronic understaffing among the midwifery team, the fact that only a handful of midwives are trained to cover maternity triage allowed consistency within the team involved in the change of practice and, subsequently, contributed to its sustainability on the front line.

The task-shifting component of our QIP becomes particularly evident when looking at the quality of the speculum examination and qfFN performed in the midwife group: the cervical os was visualised in each case, and no invalid qfFN test has been reported. This might indicate that midwives can transfer their skills of performing speculum examination in term women onto preterm women. While both groups (doctors and midwives) learn how to perform speculum examination during their training, GP trainees—who perform the majority of the reviews in maternity triage—undergo only a 4-month rotation in O&G and have significantly less exposure to speculum examination than an experienced band 6 triage midwife, who, following their qualification, has worked independently for at least 1 year in maternity services before starting their activity in maternity triage. However, despite any mention by the Nursing and Midwifery Council¹⁸ that a speculum examination performed by a midwife on a woman before 37 weeks’ gestation would ‘fall out’ of the normal range of midwifery care, only a few units permit midwives to perform a speculum examination on a preterm woman. This mindset might be a remnant from the time when the population of childbearing women was not as high-risk, and therefore midwives were less involved in medicalised pregnancies.²⁶ In modern maternity care, it is time to reconsider this arbitrary rule and, instead, aim to provide women with an examination by the most experienced front-line clinician available, that is, the triage midwife.

Among the women who were identified as high risk and were admitted (n=12), only one delivered within 7 days following admission (at 34+).¹⁷ The low event rate of preterm birth in our cohort is representative of the rate in the general population (8% in the UK),²⁷ supporting the existing evidence that imminent preterm birth following TPTL is a rare event and stresses the need to reassure the majority of women who are low risk. In our entire cohort, only two women received ANCS—due to PPROM and high qfFN level, respectively—and neither of them delivered within 7 days. This suggests that our QIP was mainly beneficial to the low-risk population by improving the women’s experience and highlights an overall safe management of our high-risk population even before the QIP took place.

Limitations

One limitation of this project is a time bias, leading to potential unforeseen changes in management from other factors. However, the relatively short period of time during which the project was conducted (12 months) and the use of internal guidelines should have minimised this effect. In addition, the fact that the project was conducted after the COVID-19 pandemic ensured a standardised environment and working conditions. There was also a selection bias, as not all cases in the postintervention period were first examined by the triage midwife. Indeed, some of them might have been examined by a doctor because the midwife did not have the capacity to perform the speculum examination due to the acuity in triage, or conversely, were only performed by triage midwives with more confidence to take on new tasks. This reflects the limitation due to human factors of implementing a new protocol (staff training, culture change, building up confidence in undertaking the examination). However, even if the women who were reviewed quicker underwent a selection bias, the enhanced collaboration of midwives in reviewing preterm women has significantly improved the quality of their care.

Even if women were reviewed on average twice as quickly by the triage midwife compared with a doctor, they still had to wait another 80 min until a final review by a doctor took place before they could be discharged home. Allowing triage midwives to assess whether symptoms in women under 37 weeks’ gestation are pregnancy-related—as they already do for those over 37 weeks—or require a doctor review for concerning signs (eg, severe pain, diarrhoea, vomiting and fever) would significantly reduce waiting times and enable the on-call team to focus on higher-risk cases. A selective rather than a routine doctor review would also offer continuity of care by the midwife, which in turn is known to decrease the risk of preterm birth.²⁸ Since the performance of a speculum examination on preterm women by midwives is currently confined to maternity triage, another way to improve the impact of this transfer of task would be to extend it to the Day Assessment Unit where women sometimes require a speculum examination to rule out PPROM.²⁵

Although the audit was registered by the Trust audit committee in July 2021—allowing official data collection to begin only 3 months after the project had started—the impact of the intervention was monitored by the project lead via the maternity triage record system, where the times of presentation, review and discharge of all women presenting to maternity triage are recorded. Due to a lack of capacity, this did not lead to the creation of an official run chart to monitor the progress of the project. This reflects the reality of conducting a QIP in the immediate post-COVID-19 period in a chronically understaffed DGH, where, at that time, the maternity department had not yet appointed a dedicated midwife for QIPs, who could have performed this task. Nonetheless, the observations made by the project lead at the initiation of the project pointed to a positive impact of the intervention on waiting times, which was confirmed statistically after data collection was completed.

Following the successful implementation of the project in our unit in 2021, the next step was to apply the change in the maternity triage of our sister hospital, University Hospital Lewisham. However, despite the apparent generalisability of our project—conducted in a DGH, making it more representative of the general population with no additional cost for staff or consumables—the fact that it has not yet been implemented within the same Trust highlights the challenge of cultural changes.

Conclusion

In summary, the findings of this study showed that task shifting can have a beneficial impact on the care of women presenting with TPTL, regardless of the type of bedside tests used to assess the risk for imminent preterm labour.²¹ Despite this significant improvement in women’s care, the overall waiting time in triage was similar in both groups, as women in our unit currently need to be seen by a doctor before being discharged home. Rethinking this pathway and allowing the triage midwife to discharge women home who were deemed low-risk for imminent preterm birth would maximise the potential of this QIP. Although a protocol amendment will need to be undertaken prior to implementing this project in other centres for validations, our project shows that task shifting can be pivotal in improving both women’s care and maternity triage services.

Data availability statement

Data are available upon reasonable request.