Why was the service needed?

General anaesthesia (GA) for first-trimester surgical abortions is not routinely recommended due to its safety risks and financial costs.1 2 Low-dose intravenous (IV) midazolam and fentanyl to achieve a state of moderate (conscious) sedation (CS) is listed in (inter)national guidance1–4 as an alternative to GA or local cervical anaesthesia (LA) alone. During CS, depression of the central nervous system enables treatment while maintenance of verbal contact protects the airway. Oral analgesia (eg, ibuprofen) and a paracervical block are typically used alongside the IV medications. Compared with LA alone, CS is associated with greater satisfaction and lower pain scores with first-trimester surgical abortion.4 British Pregnancy Advisory Service (BPAS) provides approximately 90 000 abortions annually; 98% under National Health Service (NHS) contract. The cost of CS at BPAS was estimated to be 40% lower than GA because it does not require an anaesthetist, operating department practitioner or operating theatre with positive pressure ventilation, and requires only one stage of recovery. The ability to provide CS from a treatment room also meant more clinics could provide CS than GA, lowering financial and travel burdens for clients. We sought to implement CS to improve accessibility, safety, and cost-effectiveness of surgical abortion.

How was the service established?

Staffing, procedural steps, eligibility and competency-based training were informed by the Academy of Medical Royal Colleges (AoMRC) safe sedation standards5 with the support of an external expert (DC) and BPAS’ Clinical Advisory Group which included an experienced anaesthetist and a surgeon skilled in CS for oocyte retrieval. Most BPAS services are delivered from free-standing clinics, thus eligibility is limited to healthy patients or those with mild systemic disease (American Society of Anesthesiologists (ASA) I or II)6 without concerns about airway management during sedation (eg, through airway assessment including Mallampati score). An upper body mass index of 35 kg/m² is applied routinely but procedures are permitted up to 40 kg/m² on a case-by-case basis. A three-person model is used: a surgeon, assisted by a healthcare assistant (HCA), administers the drugs and performs the procedure while a nurse/midwife monitors the client’s observations and verbal responsiveness. Clinicians hold Immediate Life Support (ILS) certification so that the airway can be managed if needed and the HCA has basic life support training. Maximum drug dosages (fentanyl 100 μg, midazolam 2 mg) were determined from published studies and abortion care guidance.4 Pre-procedural fasting is not required based on evidence of no anaesthesia-related complications in a case series of over 47 000 surgical abortions performed under CS.7 One stage of recovery is required, staffed by an ILS trained nurse/midwife and an HCA.

In January 2015, with agreement from BPAS’ Clinical Governance Committee, one pilot site that only offered LA was identified for service implementation up to 12+0 weeks’ gestation. A second site, that offered GA or LA, was added mid-year and the upper gestation limit was raised to 13+6 weeks. No sedation-related complications were reported in 490 procedures carried out that year. From September–December 2015, an anonymous self-completed feedback survey was distributed to all 261 clients undergoing CS of which 191 (73%) responded (table 1).

Following evaluation, guidance was revised to improve client selection for CS and increase the upper midazolam dose to 4 mg to achieve better sedation, but the overall reassuring findings supported expansion of CS organisation-wide. Viewing GA as the gold standard, staff were initially concerned that the change would negatively impact client experience. Furthermore, the smaller team size worried some who previously worked in anaesthetist-led, larger teams. ‘Road shows’ involving medical (PAL), nursing, financial and operation directors were undertaken. During these, CS was discussed openly with internal and external data presented to demonstrate the safety of CS and benefits including, but not limited to, economic ones. Videos of clients undergoing CS were presented as well as the planned package to support implementation.

A four-stage training programme tailored to roles in line with the AoMRC standards was delivered. Self-guided reading formed the backdrop of a full-day theory session. Following this, supervised clinical experience required assessment in treatment and recovery areas until competent. After 20 cases, a competency consolidation took place with a trainer. Upskilling in cannulation and resuscitation was provided. Simulation exercises in over-sedation were disseminated to support skills maintenance, an audit tool to ensure policy adherence implemented, and new clinical incident codes per AoMRC standards added (use of flumazenil and/or naloxone; sustained drop in oxygen saturation <90%; unplanned instrumentation of the airway; intervention of an anaesthetist/airway expert; unplanned admission to hospital following sedation).

Some units required additional equipment such as pulse oximeters, medication trollies, controlled drug destruction (Doop) jars, and a controlled drugs register. Some clinic layouts needed modification to ensure the recovery room was near the treatment room, permitting rapid access by the surgeon if needed. Procedures were estimated to take 20 min therefore recovery rooms needed space for at least three recliners, allowing unhurried recovery times without slowing or halting lists. Centres struggling with this efficiency 'buddied up' with other centres to share good practice.

The plan for implementation of CS was communicated to commissioners who largely welcomed the practice, appreciating the greater number of local services.

What have been the outcomes so far?

Uptake of CS increased rapidly as staff were trained and clinics equipped. Positive staff feedback was received commending the patient-centredness of CS and their own appreciation for greater interaction with clients during treatment. Service outcomes from 2015 to 2019 are summarised in table 2, confirming the safety and efficacy of the model.

From 2016 to 2019, we received 13 076 feedback surveys from clients undergoing vacuum aspiration abortion under CS. Satisfaction remained stable at 9.5/10. Comparative scores were 9.6 for LA (n=11 409) and 9.4 for GA (n=19 068).

What lessons have been learned?

We have found CS to be a suitable, safe and highly satisfactory means of pain and anxiety management during vacuum aspiration for most women. Furthermore, staff value the greater engagement with clients that CS allows.

Success in implementation was aided by early staff engagement and open and honest dialogue, supported by evidence and the chance to see procedures performed at BPAS on video.

BPAS’ audit scheme has helped further refine the service, for example, by highlighting the need for descriptors of CS for non-clinical staff advising clients on treatment options and further staff education on setting expectations around the potential for some discomfort despite CS and CS failure.

The future – where do we go from here?

Moderate (conscious) sedation will continue to be used at BPAS in the setting of first-trimester surgical abortion and other indications for vacuum aspiration. We also use it selectively for osmotic cervical dilator insertion or feticide before second-trimester surgical abortion when patients are unable to tolerate these procedures under LA. It was a particularly valuable during the COVID-19 pandemic with anaesthetists redeployed, shortages of propofol, and the closure of theatres in hospitals from which BPAS provides some services. Against this background and to protect access to services, BPAS successfully accelerated an expansion of the use of CS in dilatation and evacuation abortion up to 17 weeks+6 days.8 In late 2021, BPAS will launch its fertility service and will employ CS for egg collections.9 BPAS is interested in exploring the potential for CS in other painful or anxiety-producing interventions that are often provided without anaesthetic (eg, cervical smears) or under LA (eg, intrauterine device insertion), in addition to nurse-delivered sedation.10