Making it easier to access services

Good evidence from the qualitative review showed that the process to obtain an abortion is complicated and is not transparent, that there is a lack of information available about how to access services and that more integrated, streamlined services and centralised referral would improve access to abortion services. However, there was no evidence available for the strategies to improve navigating the healthcare system that were included in the quantitative review protocol (namely centralised booking systems/single point of contact, public and/or professional awareness campaigns and school-based/youth group education programmes). The committee agreed that providers and commissioners of abortion services need to ensure that women are aware of how to access services and that information about services is widely available. Whilst there was not any evidence to specify how and where information should be available, the committee agreed that providing GPs, sexual health and contraception services and schools with written information about local abortion services, and ensuring such information is available through trusted websites, may be beneficial. The committee agreed that women’s’ choice should be prioritised, but that it is not feasible for surgical abortions to be available from all services or in all locations and that some services may lack the expertise or resources to perform abortions after a specific gestational age. Therefore, they recommended that timely onward referral should be made if services cannot offer an abortion after a specific gestational age or by the woman’s preferred method.

There was evidence that more women who self-referred had their abortion within 7 days of referral compared with women who were referred by a healthcare professional; however, this was based on very low quality evidence and there was no difference in the proportion of women who had their abortion within 14 days of referral. High quality evidence showed that decreasing waiting times was an important avenue for improving care and there are delays in getting GP appointments; further, there was high quality evidence that negative physician attitudes and conflicts with personal beliefs can impact provision of services and obtaining referrals for abortion. The committee also noted that women having an abortion tend to be young, healthy women who may not have an established relationship with a regular GP. Therefore, the committee agreed that direct access was an important mechanism for overcoming these barriers. No evidence was identified that compared different methods of facilitating direct access so it was not possible to recommend a specific approach. However, the committee noted that several systems were effective in practice including dedicated booking systems with extended opening hours or call back services, drop-in open access services and online booking, and there was high quality evidence that centralised referral would improve access.

There was good evidence that physicians’ personal beliefs can create a barrier to delivering abortion services and providing referrals. Therefore, the committee agreed that healthcare professionals should ensure that their views do not create a barrier or delays to providing or arranging an abortion.

A number of themes highlighted the difficulty of accessing abortion services for women living in remote areas. The committee were aware that the NHS Healthcare Travel Costs Scheme reimburses travel costs for people receiving benefits or qualifying for the NHS Low Income Scheme. However, costs for women who need to travel to reach a provider that is not locally commissioned in order to receive the necessary treatment are likely to be much greater than normal costs associated with accessing healthcare. Therefore, the committee agreed that help with funding for travel and accommodation would be beneficial for these women even if they didn’t have low income. The committee noted that the scheme only reimburses costs after the appointment. Women having an abortion often have to travel at very short notice, compared with women having treatment for other conditions that may have several weeks’ notice before an appointment, and may have difficulty arranging funds before the appointment. Therefore, the committee agreed that upfront funding of women’s travel and accommodation costs could improve access for women with low-income and women travelling to a provider that is not locally commissioned.

Waiting times

There was good evidence that there were insufficient resources and appointment times available for abortion services. The committee agreed that increased staffing and hours would improve access to abortion services, but that increased appointment times may not always be feasible. Therefore, the committee agreed that providers should ensure they have the capacity to deliver services with minimal delays, which may or may not involve increased appointment times. There was good evidence that there are long waiting times and delays when accessing abortion services and that decreasing waiting times is an important avenue for improving care. Further, the committee agreed that the earlier women are referred for abortion services, the more choice there will be regarding the type of procedure they have and a greater proportion of women will be able to have a medical abortion with expulsion at home, which would reduce associated costs and resource use. Therefore, the committee agreed that waiting times must be kept to a minimum. The committee were also aware of evidence that mortality from abortion, whilst remaining very low in absolute terms, increases for every additional week of gestation (Bartlett 2004) and therefore recommended that providers should ensure there is minimal delays throughout the abortion pathway and ideally undertake initial assessment within 1 week of requesting an abortion, and treatment within 1 week of assessment. The committee acknowledged that assessment and treatment may need to be expedited after 22⁺⁰ weeks’ gestation to allow treatment within the legal restrictions for abortion (up to and including 23⁺⁶ weeks’ gestation unless there is a fetal anomaly or risk to the life of the women). The committee noted that these recommendations would be broadly consistent with current RCOG guidance which recommends that women are assessed within 5 days of referral and have the abortion procedure within 5 days of making a decision to proceed (Royal College of Obstetricians and Gynaecologists 2011). The committee agreed that it was more appropriate to specify the time between assessment and abortion, compared with decision to proceed and abortion, as evidence shows the majority of women are certain of their decision to proceed at the time of the assessment (Cameron 2013). However, the committee acknowledged that it is not practical, or economically viable, for abortion services to be available from all providers every day, and that some services, particularly in rural areas, may only offer abortions on certain days of the week. Therefore, sometimes travel may be required to receive an abortion within the recommended timeframes. In these circumstances, the committee agreed that women need to be provided with sufficient information about the risks associated with delaying an abortion, and how this may affect the options available to them, to enable them to make an informed decision between travelling to access services within the recommended timeframe and having a longer wait to receive them locally. The committee also agreed that some women might want additional time to consider their decision after the assessment and that they should be given the time to do this but also be informed about the risks as stated above.

There was good evidence that difficulty organising time off work and childcare can cause delays accessing services. The committee did not think is was feasible to make recommendations about childcare as this may be required overnight, depending on gestational age and distance travelled, and it was not in the scope of this guideline to make recommendations about time off work. The committee agreed that the provision of same-day services where possible may be more convenient and minimise delays. However, it was not possible to recommend same-day services as we did not review evidence comparing same-day and multiple-day services.

There was some evidence that legislation and local policies, such as the use of ethics committees and state imposed waiting periods can cause delays in accessing abortion services. Whilst some of these restrictions may not be applicable to UK clinical practice, the committee agreed that women who are certain of their decision should be able to access services immediately without the need for compulsory counselling or enforced delays. The committee also agreed that women should be provided with or referred for support making a decision if they request this, feel free to change their mind and should be given information as to who to contact if they wanted further discussion or to cancel an appointment or procedure.

Location of services

There was evidence that there was no difference in the rate of women who were very satisfied and would recommend the service to a friend between abortion services delivered by telemedicine compared with face-to-face services. It was unclear whether or not there were clinically important differences in rates of women who were somewhat satisfied or dissatisfied between those services. There was either no difference in patient satisfaction with community and hospital services, or women preferred community services to hospital services. However, the evidence was very low quality and the committee agreed that women can have low expectations for abortion services so this outcome may not be very sensitive to differences between services if satisfaction is rated highly in both instances, compared to expectations. There was good evidence that making travel arrangements can cause delays to accessing abortion services and that community prescribing for medical abortion and telemedicine either has, or would, improve access to abortion services, increase flexibility and facilitate a more woman-centred approach to care. There was also evidence that the greater number of appointments needed for a medical abortion compared with a surgical abortion is a barrier to women choosing a medical abortion. When the quantitative and qualitative evidence were considered together, the committee agreed that community services and the use of remote assessments via telephone or videoconference may improve access to abortion services, but these methods may not be suitable for all women and more traditional hospital-based and face-to-face services should also be available. The committee agreed that appropriate methods of remote assessment may also include online services and may expand with future advances in technology. The committee acknowledged that current regulations in England would prevent some aspects of the care pathway being delivered by telemedicine but the circumstances in which services could be delivered remotely may expand following the legalisation of home use of misoprostol up to and including 9⁺⁶ weeks’ gestation in England at the end of 2018. These recommendations have the potential to reduce inequalities associated with certain groups who find it particularly difficult to travel to abortion services. For example, there was moderate quality evidence that the threat of violence, controlling circumstances and cultural backgrounds that accepts honour-based violence can be barriers to accessing abortion services as women may have difficulty leaving the house or be worried about the consequences if people knew they were having an abortion. There was also very low quality evidence that teenagers are more affected by logistical barriers than other women.

There was good evidence that arranging for someone to drive women home after an abortion can cause delays and necessitate unwanted disclosure. There was also evidence that lack of social support is a barrier to accessing abortion and can worsen other barriers. It would not be feasible to recommend that social support and transport are available for women having an abortion and needing someone to drive them home may sometimes be unavoidable, such as following surgical abortion with sedation. However, the committee agreed that these factors may improve as a result of the above recommendations as travel distances may be reduced.

Workforce and training

There was good evidence that women preferred nurse-midwife-led services over physician-led services and that there was no difference between these services in a number of patient satisfaction domains. There was also very low quality evidence that there was a shorter time between referral and assessment in nurse-led services compared with physician-led services and a greater proportion of women were seen within 5 days of referral, which is supported by moderate quality evidence that expanding the role of nurses in medical abortion would improve access. The extent of nurse involvement varied across included studies and regulations in England would restrict nurses and midwives from delivering some aspects of the care pathway; however, the committee agreed that, with appropriate supervision and restrictions as required by law, expanding the role of nurses and midwives in abortion services would help improve access to services.

There was good evidence that GPs were confused or unclear regarding details of services such as routes for referral and gestational limits, that further education was needed for the public and healthcare providers, and that the role of nursing staff in medical abortion would improve access but is hindered by shortfalls in training. Therefore, the committee recommended that healthcare professionals who may care for women requesting an abortion, such as nurses, midwives and GPs, should be able to gain experience in abortion services during training. The committee agreed that increasing exposure to abortion may be an important avenue for reducing stigma.

There was also evidence that a lack of knowledge and skills among healthcare providers was a barrier to performing certain abortion procedures. The core curriculum for obstetrics and gynaecology in the UK dictates that people have practical experience of medical and surgical abortion (Royal College of Obstetricians and Gynaecologists 2013, updated 2016); however, only 30% of abortion services in the UK are provided by the NHS (Department of Health 2018) and the evidence showed that NHS hospital-based providers are losing their clinical skills due to abortions occurring mainly in the independent sector. The committee noted that the amount of exposure to abortion services gained during training will depend in part on geographical location, which is supported by evidence that rural physicians lack professional support, continued professional education and that the lack of volume of abortions in the rural setting was a deterrent to the local training of abortion providers. The evidence showed that clinicians were more likely to provide, or intend to provide, abortion services when they had abortion training available to them during training and where curriculums were organised using an “opt-out” approach, such that trainees gained experience in abortion care unless they specifically opted-out for reasons of personal belief. Therefore, the committee agreed that all clinicians training in specialities where abortion training is part of the core curriculum should be enrolled in training, unless they opt out due to conscientious objection, and receive practical exposure to abortion services during training, either within the NHS or the independent sector. The committee agreed it was important for curricula to adopt this “opt-out” approach otherwise abortion care can be viewed as optional and unimportant, whilst for women it is a common procedure, with 1 in 5 pregnancies in England and Wales (excluding miscarriage) ending in abortion (Office for National Statistics, 2018). Further, the committee agreed that if abortion training is seen as optional, this may perpetuate the stigma surrounding abortion.

The committee agreed, based on their knowledge and experience, that, in order to ensure the future longevity of abortion services, abortion training needs to remain in the core curriculum for obstetricians and gynaecologists and sexual and reproductive health specialists; however, they could not make recommendations in this area as it is beyond the scope of this guideline.

Complex comorbidities

No evidence was identified for strategies that improved accessibility or sustainability of services for women with comorbid medical conditions. There was some evidence that accessing services is difficult for women with anxiety problems and the committee agreed that they would benefit from increased availability of telemedicine. However, the committee agreed that with many services being delivered in community settings and outside of the traditional hospital network, women with complex needs faced difficulties in accessing adequate care. The committee were particularly concerned that there may be delays in accessing treatment due to the need for referral and that services many not have adequate skills to safely deliver all treatment options. The committee were aware that, at the time of the development of this guideline, NHS England were in the process of developing the specification and commissioning framework for delivery of complex abortion care within each of the seven regions of England. Therefore, the committee did not define who should be considered as having complex needs or significant comorbidities or the requirements of the specialist service as these factors will be defined in the service specification. The committee agreed the importance of specialist centres for women with complex needs or significant comorbidities as safety has to be the priority and if such services are unavailable, then women with complex needs may be forced to continue with the pregnancy. There also needs to be robust pathways for referral that minimise delays, when care is not available locally. However, the committee agreed that safe care for this population may require some delays and travel.

Avoiding stigma

There was good evidence that there is a perceived stigma associated with abortion and some evidence that prior negative experiences with staff in abortion services may put some women off having another abortion and cause delays in presentation. Therefore, the committee agreed that healthcare professionals should be aware of the impact of their communication on women seeking abortion services.

Evidence showed that provider workshops were ineffective at reducing stigma; however, the evidence was low quality and from a study validating the Abortion Provider Stigma Survey, which has since been refined. Therefore, the committee agreed that there was insufficient evidence to recommend a specific approach to reduce stigma associated with abortion. The committee noted that awareness campaigns have been successful at targeting driving while intoxicated and stigma associated with HIV and mental health (Evans-Lacko 2014, Stang 2013, Yadav 2015); however, there was no evidence available for the effectiveness of such campaigns in relation to abortion.

Good evidence showed that women have concerns about the privacy and confidentiality of abortion services, reactions and judgements from others, and the need to disclose their abortion to unwanted people in order to overcome logistical barriers. Therefore, the committee agreed that abortion services should be sensitive to these concerns and that information should only be disclosed if there is a compelling need and is in the woman’s interest. Evidence for methods of maintaining confidentiality were not reviewed and specific methods could not be recommended. Recommending that direct access to abortion services is available may improve privacy by minimising the number of people involved in the referral process.

The committee were aware that women having an abortion being near other women continuing with pregnancies has been raised as a patient experience issue and can be distressing. The committee discussed that one method of addressing this concern could be to have wards or clinics for women having an abortion that are separate from other maternity wards. However, the committee agreed this may not be possible in rural areas where there would likely be insufficient resources to have separate clinics. The committee also agreed that at later gestations maternity wards are likely to be the safest place for women having an abortion. There were also concerns that separating women might actually perpetuate the stigma around abortion and may risk inadvertently identifying women as having an abortion if they are accessing an area that only provides this service. In the absence of evidence to recommend a specific approach to address this issue, the committee did not recommend separating these groups.

Future research

As there was sufficient evidence to inform the recommendations, the committee decided to prioritise other areas addressed by the guideline for future research and therefore made no research recommendations regarding strategies that improve the factors that help or hinder the accessibility and sustainability of a safe abortion service.

Making it easier to access services

Whilst the committee recommended that women should be able to directly refer for abortions they were unable to recommend a specific method given a paucity of effectiveness and cost effectiveness evidence. The committee agreed that enabling direct access may require changes to commissioning, as some commissioners currently require a referral for an abortion, and services, if they do not currently have a system for receiving direct referrals. Dedicated booking systems with extended opening hours, call back services, drop-in open access services and online booking will all incur costs to set-up where they are not already available. There was high quality evidence from the qualitative review that centralised booking would increase access (see the benefits and harms section on making it easier to access services) and from the results of the bespoke economic model on waiting times substantial cost savings could be achieved through women presenting earlier for abortion.

The NHS Healthcare Travel Costs Scheme already reimburses travel costs for women receiving benefits or qualifying for the NHS Low Income Scheme. Therefore, providing upfront funding for this population will not result in an absolute increase in costs, rather there will be a difference in timing of when funding is provided. The committee noted that some local commissioners and providers have informal processes in place where they will arrange and fund travel for women before the abortion, but this is not a common occurrence. Therefore, providers will need to introduce processes that allow them to provide upfront funding, which may have associated costs and resource use; however, these costs are likely to be one-off and may be offset from women having earlier abortions. There will be an increased cost associated with providing funding for women who do not have low income but are travelling to a provider that is not locally commissioned, as these women are not covered by the national policy. However, where locally commissioned services are not available, abortions could be delayed due to women trying to arrange travel which would lead to increased costs as the abortion would take place at a later gestational age, or could even result in the woman being unable to have the abortion if she is unable to travel.

Waiting times

Bespoke economic modelling was undertaken around waiting times to look at the potential cost savings from earlier access and reduced waiting times. The model assumed that with shorter waiting times women would access an abortion at an earlier gestational week and that cost savings would be realised through the lower NHS tariff, lower rates of complications for earlier gestational weeks and an increase in the number of medical abortions. All recommendations in this area are likely to increase access and either reduce waiting times or the time until initial presentation. The model made no distinction between reductions in time to procedure for either.

All recommendations in this topic are aimed at decreasing waiting times or the time until accessing abortion procedures. The economic model did not attempt to consider any particular intervention and its associated costs as there were no evidence that could quantify reductions from suggested interventions. These costs savings were interpreted by the committee in the context that there would be likely costs to enacting interventions to achieve these outcomes, such as reconfiguring services so that they are available on a greater number of days a week and have processes in place for self-referral. The economic model showed that even with very small decreases in waiting times that significant cost saving could be made. A reduction in 1 day for all abortions could save £9 per procedure or £1.6 million across all abortions per year. A decrease of a week would save £61 per procedure or £11.5 million across all abortions. The majority of these costs savings (>80%) came from women switching from surgical to medical abortion with nearly all the rest from women dropping between the 14 week and 20 week cuts in the NHS Reference Costs. A reduction in adverse events had minimal impact on costs as a result of these being rare events at all gestational weeks.

The committee pointed out that the model may underestimate the true cost savings from reducing waiting times as the model used a tariff based on inpatient procedures. It is very likely that if women are able to access abortions earlier either through shorter waiting times they may be able to have an abortion on an outpatient basis saving hundreds of pounds on those procedures. The cut point for 14 weeks and 20 weeks are also very insensitive and would only pick up cost savings when women passed those particular weeks. In reality even small differences in gestational weeks lead to less medically and resource intensive procedures.

It was the committees belief that even small reductions in waiting times would lend support to even relatively expensive interventions being cost saving. Despite being unable to look at it, explicitly reducing waiting times would almost certainly also lead to improvements in quality of life through a greater choice of procedures, a less intensive procedure and a reduction in time a woman had to continue with a pregnancy. The committee thought it appropriate to set an ideal maximum waiting time to encourage providers to keep times as short as possible. Whilst some areas are already meeting these targets others are quite a significant way from doing so and may incur significant costs in providing systems and interventions to achieve them. The committee strongly asserted that any effort to do so would be an efficient use of NHS resources and also potentially cost saving.

Location of services

There is unlikely to be any increase resource use around the recommendation for telemedicine as telephone and/or videoconferencing facilities will already be available in the vast majority of settings. Where these are not available they can be set up relatively inexpensively through for example ‘off the shelf’ videoconferencing software. There is potential with increased telemedicine to increase the number of women seen in a set amount of time and also potentially through a reduction in the number of missed appointments. It is expected that any upfront cost will be regained quickly.

A greater use of community services through telemedicine and as a result of the announcement to legalise home use of misoprostol is also likely to be cost saving through a reduction in the number of appointments, reduction in travel costs for those in remote areas and through reduction in barriers to presenting earlier for abortion allowing for less costly procedures.

Workforce and training

Increasing the responsibility and roles of nurses in the provision of abortions, within the constraints of the law (see the benefits and harms sections on workforce and training), will decrease the costs through less time spent by doctors (with a higher wage rate) providing these services. Increased access will also lead to women presenting earlier for less costly abortions. The evidence did identify a need for training in this area for nurses, but any upfront costs associated with providing training will likely be reimbursed though improved access.

It is expected that the recommendations around training needs, if included as part of the core training curriculums will be cost neutral.

Complex comorbidities

The development of specialist commissioning services is currently in progress and has a tariff structure in place. Therefore, there should not be a cost impact of recommending specialist centres are available.

Recommendations to minimise delay may produce some cost-savings due to less complicated procedures being required. However, medical abortions are often more costly than surgical abortions in this population, as they require admission for a number of days. Therefore, cost savings associated with reduced delays will not be as large as for women without comorbid medical conditions, where greatest savings come from more women having a medical rather than surgical abortion if they present to services earlier.

There is likely to be some savings associated with avoiding repeated assessments or investigations.

Avoiding stigma

The way healthcare professionals communicate with women having an abortion, and issues of confidentiality and sensitivity, should already be embedded in their training and, therefore, these recommendations should not be associated with any increased resource impact.

Evidence was identified that prior negative experience around abortion discouraged some women from seeking an abortion and resulted in later presentation. Whilst not explicitly exploring stigma, the economic model highlighted both the increased cost and increase in adverse events from women receiving an abortion at a later gestational week and even reductions of a few days could lead to potentially large cost savings.

There was a paucity of evidence for specific methods of reducing stigma associated with abortion. Therefore, no specific methods of reducing stigma or raising awareness were recommended. However, training programmes or information campaigns aimed at both people involved in providing abortion services and the wider public could have a significant resource impact depending on their medium.

Other considerations

There was some evidence that a lack of financial input into abortion services is impacting the care that can be provided. Whilst the committee could not make recommendations specifying the funding required for abortion services, they agreed that it is the responsibility of commissioners to ensure services are available to deliver the recommendations made in this guideline. There was also evidence that there is insufficient funding for abortion services for women ineligible for free NHS services; however, the committee could not make recommendations in this area as it is outside the scope of the guideline.

As the focus of these review question were accessibility and sustainability, evidence on the safety and efficacy of abortion services was not included in this review. However, the committee were aware of evidence from a Cochrane review (Barnard 2015) that showed no difference in the efficacy of medical abortion, or the complication rate following surgical abortion, conducted by mid-level providers (nurses, midwives and physician assistants) compared with physicians. There was some evidence from observational studies that the risk of failure or incomplete surgical abortion was higher for mid-level providers compared with physicians, but this effect was not observed in the RCT evidence. No complications following medical abortion were reported in either arm of any of the included studies. Therefore, the committee agreed that mid-level providers (nurses, midwives and physician assistants) can deliver abortion services safely and effectively.

The committee agreed that the recommendations made (particularly those related to location of services, making it easier to access services and comorbid medical conditions) have the potential to reduce current inequalities in accessing abortion services for the following groups by improving referral pathways, minimising travel and decreasing the number of appointments that women need to attend in person: women living in remote areas, women with low income, women with comorbid physical and/or mental health problems, vulnerable women, and girls and younger women.

Number of abortions of pregnancy

The total annual number of abortions of pregnancy were taken from The Abortion Statistics for England and Wales: 2017 (Department of Health 2018). In 2017, 189,859 abortions of pregnancy were recorded in England and Wales; 189,614 of these were performed before 24 weeks’ gestation and were eligible for inclusion in the economic model. Table 20 presents the number of abortions performed in England and Wales in 2017, by gestational age. Over half of abortions of pregnancy were performed before the end of the 7^th week of pregnancy and over 90% were performed before the end of the 12^th week. The total number of abortions of pregnancy has remained between 190,000 and 200,000 per year since 2009 and there is no anticipation that these numbers would change significantly in future years.

Table 20. Number of abortions performed in 2017 by gestational age

Method of abortion

Figure 6 shows the percentage of abortions of pregnancy by gestational week and method of procedure. As all abortions were carried out either surgically or medically the data lines are inversions of each other. Medical abortion is the most popular method up to and including 10⁺⁰ weeks’ gestation with over 90% of abortions before 7 weeks’ gestation being medical abortions. After 10⁺⁰ weeks’ gestation, surgical abortions are more common than medical and the percentage of surgical abortions continues to rise up to a gestational age of 19⁺⁰ weeks. After 19⁺⁰ weeks, surgery as a percentage of all abortions decreases but remains the predominant method. There is a 20 percentage point increase in medical abortions between 18⁺⁰ and 20⁺⁰ weeks’ gestation. This is most likely as a result of abortion on grounds other than ground C (which make up a larger proportion of abortions in later gestational weeks) which are often performed medically to allow for examination of the fetus.

Figure 6. Percentage of abortions of pregnancy by method and gestational week

Given that the proportion of medical abortions increases as gestational age is reduced, under assumption 2 of the model, any decrease in the time between initial presentation and procedure will lead to an increase in the overall proportion of medical abortions.

Adverse events

The number of adverse events (Table 21) were taken from The Abortion Statistics for England and Wales: 2017. The complications were reported in aggregated form and primarily consisted of haemorrhage, uterine perforation and sepsis, reported up to the time of discharge from the abortion provider. From the data, only a small proportion of abortions resulted in adverse events with a combined percentage of 0.12% and 0.18% for medical and surgical abortions, respectively. The percentage of adverse events increases as gestational age increases; therefore, decreasing the time between initial presentation and procedure will result in a reduction in the number of adverse events.

Table 21. Percentage of adverse events by method of abortion and gestational age

It is likely that the true number of adverse events will be underestimated in these figures given the narrow definition and tight timeframe for reporting. Carlsson 2018 considered complication rates at one Swedish hospital between 2008 and 2015 and estimated that complications occurred in 6.7% of all abortions. However, this study included incomplete abortion as a complication and recorded any visits up to 2 months after discharge. Other Scandinavian studies (Larsson 1992, Charonis 2006) with similar definitions of complications have reported complication rates of between 2.8% and 4.9%. A UK study of 28,901 women undergoing a medical abortion at a British Pregnancy Advisory Service (BPAS) clinic between May 2015 and April 2016 reported an adverse event, excluding incomplete abortions, in 0.2% of abortions occurring up to and including 9⁺⁰ weeks’ gestation, which is higher than the number of events estimated above (Lohr 2018). However, this definition of adverse events (hospital admission, haemorrhage and intravenous antibiotic administration) was very narrow. When need for surgical intervention for an incomplete abortion is considered, this adverse event rate rises to figures similar to the Scandinavian studies.

As the number of complications reported in The Abortion Statistics for England and Wales: 2017 are potential underestimates, and do not include costs associated with surgical intervention, the proportion of any cost savings associated with adverse events was explored.

Mortality is an extremely rare adverse event with abortions and was not considered by the economic model. No deaths were reported in 2017 and only 2 deaths occurred in the previous 5 years, during which there were over 1 million abortions (Department of Health 2018).

Costs and resource use

Quality of life

The economic model did not attempt to make any quantitative estimations around changes in quality of life. Regardless, no evidence was identified which compared quality of life between abortions at different gestational ages. However, the committee agreed that any reduction in time between initial presentation and procedure would improve quality of life, as long as this timing was the preference of the woman and care was taken that they were not unduly rushed. This is because women will have to carry an unwanted pregnancy for a reduced period of time, potentially have a greater choice of type of procedure and also receive a less intensive procedure with a lower probability of adverse events. Therefore, the committee agreed that any intervention to reduce the time between initial presentation and procedure would result in an overall increase in quality of life.

Time horizon

The model only estimates a reduction in cost savings for the latest year, for which data was available. However, it is likely, especially in the case of structural changes to services or where there is a large initial investment in services, that any cost savings achieved would go beyond this time horizon potentially perpetually. The time horizon from this economic model may therefore not capture all benefits from some potential interventions.

Discounting

The economic model only had a time horizon of 1 year and therefore clinical outcomes and costs were not discounted at NICE’s preferred 3.5% per annum. Potential cost savings beyond the time horizon of the model would need to take account of discounting.

Combined assumptions of the model

The alternative assumptions discussed above lead to 6 combined assumptions as listed below:

• Assumption 1a: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) differs between the arms based on the proportions reported in the England and Wales Abortion Statistics 2017. Differences in adverse events are not included.
• Assumption 2a: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) does not change from the base-case. Differences in adverse events are not included.
• Assumption 1b: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) differs between the arms based on the proportions reported in the England and Wales Abortion Statistics 2017. Differences in adverse events are included at the lower estimate of costs.
• Assumption 2b: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) does not change from the base-case. The total proportion of each abortion method will not change. Differences in adverse events are included at the lower estimate of costs.
• Assumption 1c: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) differs between the arms based on the proportions reported in the England and Wales Abortion Statistics 2017. Differences in adverse events are included at the higher estimate of costs.
• Assumption 2c: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) does not change from the base-case. The total proportion of each abortion method will not change. Differences in adverse events are included at the higher estimate of costs.

A further sensitivity analysis was performed that assumed all medical abortions performed before 9⁺⁰ weeks’ gestational age would incur the NHS reference cost for ‘Medical Abortion, less than 14 weeks’ gestation in an outpatient setting’ as discussed above. This sensitivity analysis was performed only for assumption 1a and assumption 2a:

• Sensitivity analysis assumption 1a: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) differs between the arms based on the proportions reported in the England and Wales Abortion Statistics 2017. All medical abortions performed before 9 weeks’ gestational age will be on an outpatient basis. Differences in adverse events are not included.
• Sensitivity analysis assumption 2a: The gestational age at the time of abortion reduces by the assumed number of days in the comparator arm compared to the base-case. The method of abortion (surgical or medical) does not change from the base-case. All medical abortions performed before 9 weeks’ gestational age will be on an outpatient basis. The total proportion of each abortion method will not change. Differences in adverse events not included.

Sensitivity analysis for outpatient assumption

Under the assumption where all abortions before 9⁺⁰ weeks’ gestational age are performed on an outpatient basis, total costs decrease for all assumptions by over £40 million (Table 28) and overall cost savings from reducing time between initial presentation and procedure are almost double that of the base case estimates. Over 1% of abortions of pregnancy can change to an outpatient procedure with only a 1 day reduction. This increase to just under 1 in 4 abortions of pregnancy when a 21 day reduction is assumed. (Table 29)

Table 27. Total cost saving and cost savings per procedure

Table 28. Total costs savings under outpatient assumption

Table 29. Increase in medical abortions of pregnancy as a result of a reduction in days