As sociological researchers of abortion experience, we welcome the invitation to reflect on the social consequences of legal regulation of abortion care with Prof Ruth Fletcher and others engaged in the legal witnessing project. The quantitative and qualitative data we have gathered show how important it is to question the legal assumptions underpinning policy positions on implementation of abortion provision, and to look at questions of abortion access across disciplinary boundaries. Thinking about particular policy problems, such as when follow-up abortion care is deemed unlawful under s.12 of the 2018 Act in cases where a certified early abortion with pills has failed and the pregnancy continues post 12 weeks concretely, and in light of their lived consequences, is vital. For our contribution to the legal witnessing blog series, we draw from empirical research we have been involved in since the 2018 Act was passed. This research collected both qualitative data from women and pregnant people who have had or sought abortion care under the Act, as well as quantitative and qualitative data with GPs providing abortion care under the Act (Conlon and Antosik-Parsons, 2022; Conlon, Antosik-Parsons and Butler, 2025; Campbell, 2025; Bratt, Young et al, 2026; Bratt, 2026).
Data Collection on Abortion Service in Ireland
The World Health Organisation (WHO) 2022 Abortion Care Guidelines state that “effective monitoring and evaluation (M&E) are essential for measuring abortion quality and trends, as a basis for policy dialogue and evidence-based decision-making to further improve service delivery and quality” (p18). For the first six years of the service, the Health Service Executive (HSE) had no national data-base on abortion care provision. The only data reported for abortion care was annual notifications to the Minister specified under the 2018 Act confined to reporting numbers of abortions under each section of the Act and the month and county of residence of people accessing abortion care not broken down by section. This means we have no data infrastructure for those years for effective monitoring and evaluation of abortion care at this critical juncture of implementation and design of a new service and model of care.
It is welcome that the Health Service Executive (2026) Termination of pregnancy services in Ireland report 2019-2025 report states that
“following the O’Shea Review and the Regan Review, NWIHP, in partnership with the National Perinatal Epidemiology Centre (NPEC), developed a national electronic data collection system for termination of pregnancy services. The system is initially focused on hospital-based termination of pregnancy services and is designed to support clinical governance, service planning, and continuous quality improvement” (pp 23-24).
Critically here this data collection system is confined to secondary care services which an informed estimate suggests represents about 2% of abortions annually. The more innovative abortion provision in primary care providing c.98% of abortions annually is not included in this national electronic data collection system of pregnancy services. The report should specify that what is referred to is a ‘national electronic data collection system for termination of pregnancy services in secondary care settings’
The report states that a programme of work is underway to ensure that all nineteen maternity units are represented within the national dataset. Work is scheduled to commence in 2026 to explore the development of a similar national data collection framework for termination of pregnancy services delivered in primary care settings, with the aim of establishing a comprehensive, whole-system view of termination of pregnancy service provision. Thus a data infrastructure for abortion care provided in primary settings, estimated at 98% of notified abortions under s.12, is only being explored seven years on. For the purposes of this project, service data is not being collected that allows accurate and comprehensive knowledge of instances where a certified early abortion with pills has failed, the pregnancy continues post 12 weeks, and there is a need for follow-up abortion care.
So what do we know of failed medication abortion presentation at or after 12 weeks gestation? We know that for each year the numbers of notified abortions has increased year on year with e.g. 6,666 in 2019; 8,156 in 2022 and 10,852 in 2024. And that most abortions provided are under s.12 of the 2018 Act,, representing 98-99% of abortions in Ireland each year (HSE, 2026). Preliminary data from the HSE’s pilot ‘national dataset’ reports on abortion care in hospitals under s.12 of the Act only from only a subset of 14 of the 19 providing hospitals for 8 months from May to December 2024. This demonstrates how incomplete this ‘national electronic data collection system for termination of pregnancy services’ is. The report of these 14 hospitals over 8 months in 2024 represents only 305 cases (2.8%) of the 10,711 s.12 abortions notified that year.
Activity Data for Early Medical Abortion Care in General PracTice (ADAPT)
The absence of health service data on abortion provision in primary care has concerned GPs regarding assuring quality, efficacy and access to abortion care for some time (Horgan et al, 2021) and led to a collaboration between one of the authors, Conlon, and START General Practitioners (GPs) to address this data gap. We secured funding from the Irish College of General Practitioners (ICGP) and developed the Activity Data for Early Medical Abortion Care in General PracTice (ADAPT) project to collect activity and clinical data on EMA provision from providing GPs who volunteer to participate. Ethical approval and oversight of data management is provided by Trinity College Dublin. By the end of its first year July 2025-26, 124 (25.6%) of providing GPs consented to return data systematically to the project and 1160 returns were received to the study, an estimated 10% of s.12 notifications for the year.
For the purposes of this contribution, we are able to refer to this data to show that mean gestational age at presentation for abortion care for 1160 cases presenting for abortion in General Practice was at 42 days or 6 weeks, well within the 12-week limit. 1% presented after gestational age that would allow for care in primary setting, 9 weeks + 6 days. In this dataset 95.7% of EMA’s provided were complete with the first round of Early Medication Abortion. This means that the pregnancy had ended and all of the pregnancy tissue was expelled within the medical abortion care process.
Of the remaining, 1.7% of medication abortions administered by GPs failed. Failed termination may not be identified until administration of the low sensitivity pregnancy test provided to women to self-administer fourteen days after taking the abortion medication. This carries a distinct risk that the woman will continue to be pregnant after 12+0 cut-off point.
Managing Failed Termination under s.12
Management of failed termination of pregnancy in Primary Care is governed by National Clinical Practice Guidelines on the Investigation and Management of Complications of Early Termination of Pregnancy (Boyd et al, 2022). The Guidelines state that
“Should a primary care GP provider suspect ongoing pregnancy post early TOP, referral to secondary care TOP services to confirm/out rule ongoing pregnancy should be arranged in a timely fashion. Provided the legal gestational limit of 12 weeks’ (84 days) for termination has not been exceeded, options for medical or surgical repeat termination should be discussed with the woman and the decision of which option to proceed with should be at their discretion and dependant on local resources. If the woman chooses to repeat medical treatment, a complete course of treatment should be given; oral mifepristone followed >24 hours but >48 hours later by buccal misoprostol. Consideration should be given to enhanced follow up to ensure the repeat termination procedure is completed. Retention in secondary care service to confirm passage of pregnancy tissue or scheduled ultrasound follow up in secondary care service, for those suitable for community repeat medical treatment, may be considered” (p.21)
The Guidelines further state that
“For women undergoing STOP, 12 weeks is the last date STOP may be performed. For women undergoing medical termination, 12 weeks is the last day that mifepristone may be administered. The Chief Medical Officer of the Department of Health has acknowledged that in some circumstances it may be necessary to extend beyond 12 weeks in order to complete the process of termination. “ (p.17)
This first phase of the legal witnessing project is concerned with denial of follow-up care on the grounds that it is illegal after 12+0 days. It is important to note that the guidelines specify that it is only mifepristone that cannot be administered after 12 weeks. No constraint applies to continued administration of misoprostol in cases of incomplete abortion. Incomplete abortion refers to where the pregnancy has been terminated following administration of mifepristone but all tissue has not expelled, which is the effect of misoprostol. Women presenting for follow-up abortion care can either present with a positive pregnancy test and symptoms or just symptoms e.g. bleeding. They will be monitored for hCG levels, the hormone indicating pregnancy, to establish whether or not the pregnancy has ended. If hcG levels are coming down, this indicates the pregnancy has ended. If hCG levels remain elevated, this indicates the pregnancy is ongoing and mifeprostol needs to be administered again to terminate the pregnancy and misoprostol to eliminate the related pregnancy tissue. If the pregnancy has ended but the woman is still symptomatic of bleeding or a scan shows retained products, then only misoprostol needs to continue to be administered or a surgical procedure performed to expel that tissue. But this is not a repeat termination of pregnancy. It is equivalent to management of miscarriage.
We see in the HSE report on 305 cases of s.12 abortion in secondary care in 2024, in only 15% cases was one dose of misoprostol sufficient to expel all of the pregnancy tissue while 33% required two doses, 23% required 3, 11% required 4, 15% required 5 and 3% required 6-9 doses of misoprostol to complete the medical abortion. This form of continued care under is not prohibited by the guidelines under 12 weeks for s.12 of the Act.
In the HSE (2026) dataset, 220 of the 305 cases reported related to medical abortion. Just under 2% required a repeat dose of mifepristone (p.28). This corresponds with the 1.7% of cases of failed abortion in the ADAPT dataset of 1160 cases from GPs. As outlined above, it is only in the context of a failed termination that a repeat of the full regimen of mifepristone and misoprostol or access to surgical termination instead is required. If 2% of those seeking medical abortion under s.12 each year have a failed abortion, this means about 180 people will need to have the full medical termination of pregnancy repeated or a surgical termination instead administered. Evidence shows a high risk of failure of medical abortion as gestation proceeds 1.8% at 57-63 days and 2.9% at 64-70 days compared to less than 1% under 26 days. In other words, those accessing medical abortion under s.12 between ten and twelve weeks gestation have greatest risk of failed abortion whilst moving closer to the gestational limit to access repeat medical or surgical abortion, making this project all the more important.
The requirement for repeat administration of mifepristone again (MTOP) or a surgical termination (STOP) following a first administration of the full medical abortion regimen is where the guidelines indicate that the 12 weeks + 0 rigid-time cut-off point applies. Two issues arise here, first the gestational time-limit does and should not apply in the case of administration of further doses of misoprostol to expel the tissue as this is for the purposes of completing expulsion of tissue from a pregnancy whose termination has been given effect by the mifepristone. This is reflected later on p.25 of the guidelines (Boyd, 2022) which state “For women who opt for medical management of incomplete termination less than 14 weeks gestation, administration of misoprostol should be 400 micrograms sublingually, buccally or vaginally or 600 micrograms orally. In cases of a missed abortion (retained non-viable fetus), mifepristone 200 mg orally should be administered 24-48 hours before misoprostol.” (p.25)
The second issue is the discretionary facility by the Chief Medical Officer of the Department of Health referred to in the Boyd guidelines that ‘in some circumstances it may be necessary to extend beyond 12 weeks in order to complete the process of termination’ (p.17). This clearly refers to administration of mifepristone to terminate the pregnancy beyond 12 weeks. If there is capacity for such discretion in some circumstances, suggesting the law can tolerate extension of administration of medical or surgical abortion ‘beyond 12 weeks in order to complete the process of termination’, can this not be a clear, established position? (see also Fletcher, 2025 137-139).
Abortion Care Experiences following 2018 Act Providing for Abortion Care (UnPAC)
Between 2019 and 2022, Conlon led the Unplanned Pregnancy and Abortion Care (UnPAC) study funded by the Health Service Executive and designated the user data strand for the O’Shea review which involved collecting qualitative interviews with women navigating abortion care under the 2018 Act, 48 of whom sought care under s.12. Within the dataset, women relayed stresses of failed abortion coming near the 12-week cut-off point detailed below, using pseudonyms.
Jade returned to her GP after an unsuccessful EMA close to the 9 week, 6 days, the point at which the model of care advises is the end point for care in primary setting and referral to hospital-based care is required. Jade described having to be very proactive to get the referring documentation from her GP to the hospital by hand using public bus transport on a Friday evening after 4pm in order that she would be seen at the earliest available time there so she could access repeat EMA within the 12-week limit if needed.
At about six weeks gestation, Grace had disclosed to her family GP that she was pregnant following a sexual assault. The GP advised her to attend pregnancy counselling and had her return for a further two face-to-face appointments costing €180 in total over a 3-week timeframe before telling her he was not a provider and reluctantly giving her a My Options leaflet. This is despite the abortion service being fully funded and free at the point of care for service users. Within the study we identified this GP as an obstructing GP who had not adhered to Irish Medical Council Ethical Guidelines or s.22(3) of the 2018 Act requiring a GP unable to provide care requested to refer a patient to a doctor who can do so. Jade explained
“At the time I was so confused and so feeling like vulnerable that he sent me away again [on my second consultation requesting abortion care] … he said, “try to, like, think it over and come back to me” and I was like “I’ve thought it over,” you know. So, I came back again, the third time. This was another €60 like out of my, you know. … he gave me like [the My Options leaflet], it was only the third time, and he was very hesitant to do it as well, I really think he was.”
For a small number of women interviewed for the UnPAC study, post-EMA complications eventually culminated in a hospital referral at close to the 12-week cut-off. Naomi recounted the strain of this time-limit
“And I [returned to the hospital for repeated hCG blood tests] for a couple of nights, Jesus I don’t know how many, I must have gone down about four times, three or four times. And then [the doctor] was like ‘no you’re going to have to, if [hCG level] doesn’t go down this time, you’re going to have to go in [to hospital admission for repeat EMA].”
Victoria described how protracted undergoing repeat medication abortion was, inducing pressure and stress as the limit for timing-out approached.
“So, the [hospital] admitted me in on a Saturday morning early and then we did the first medication again. And then repeated dosages of the second medication to induce the cramping and get the material to expel. And in that context, they gave me five doses of the second medication”.
After taking these multiple doses of the medication, Victoria described feeling like her body had aborted the pregnancy but that visual inspection had not confirmed it, making for a stressful episode for her.
“It really felt like that at least for me individually, like it took a really long time for the cramping to come on, it wasn’t as intense as you were warned it would be by the nurses and the doctors… They give you like a little cardboard tray thing to catch the material and then some nurse will inspect it to ensure that the material is there. So, the nurses would look at it and be like “no that’s not it, it’s just blood clots, just blood clots”. … after the fourth dose of the second medication, like they were still telling me that I hadn’t expelled the material. So, I was getting like really super stressed. And by that point my body had stopped cramping. Like I had this feeling of like physical, like my body was telling me you’re done, you know. … Whereas the nurses and doctors were saying no that “you’re not done, that’s not it. Here we’ll give you, this is the last dose of this medication, and you know if nothing happens by tomorrow then we start again.”
In Victoria’s case, a scan the next day confirmed that the tissue had been passed. The protracted account of repeat medication abortion within the hospital recounted by Victoria demonstrates the limited availability, use or offer of surgical abortion to women attending secondary care hospitals for repeat abortion. This is despite the risk of ‘timing-out’ this raises along with considerable adverse personal impact.
Jade attended the hospital after a failed EMA was confirmed by a scan at the hospital. By now she was concerned about being close to the twelve-week gestational limit. Following consultation with the doctor it was decided she would have a repeat EMA under supervision of the hospital. Mifepristone was administered and she returned two days later for misoprostol, assuming that it would just be a day case. However, it was discovered that the doctor administering the first medication had not signed off her certification as required and she could not proceed with the second medication until this was resolved. Eventually it was resolved but Jade ultimately had a surgical procedure after the additional round of medication abortion failed to fully expel the remaining tissue.
Navigating EMA Time-limits in Circumstances of Deprivation
Finally, a current study by two of us (Bratt and Conlon) developed out of the ADAPT project looks at challenges providing EMA under s.12 of the 2018 in an Inclusion Health or socially deprived context. The EMA pathway and the various ways time features therein present particular and specific challenges for people in circumstances of vulnerability, marginalisation and deprivation. Bratt’s research from the ADAPT study engages Lipsky’s (1980) concept of street-level bureaucracy to portray how committed abortion providing Inclusion Health and ‘Deepend’ GPs do not merely implement s.12 of the 2018 Act, but through the accumulation of daily discretionary decisions, effectively make it work and bend it to align with the circumstances of people living in challenging situations while seeking abortion. Bratt’s data shows the structural tension between providing EMA within the confines of lived circumstances that do not align with the ‘prototype lifestyle/circumstances’ anticipated by the model of care. This strand of the ADAPT study shows service challenges being resolved, case by case, not through policy or institutional flexibility or accommodations but through committed abortion providing GP’s own discretionary labour carving out individualised, attentive care.
In the Irish EMA primary care service, providing GPs have made a deliberate choice to provide abortion care and their commitment to the service is strong. Notwithstanding, the data shows that strong identification with the policy’s purpose does not, on its own, supply the institutional mechanisms, a formal exception process including in relation to time or a recognised discretionary care pathway, that would let a GP extend care to a patient outside normal boundaries without personally bearing the entire administrative and ethical risk of doing so.
Material hardship, unstable housing including living in a tent, language barriers, coercion, addiction, irregular immigration status, and stigma all featured in the dataset and had compounding effects, often narrowing the margin of time, privacy, or safety available to the patient and the providing GP navigating a pathway designed around a uniform, unhurried timeline. GPs providing EMA in Inclusion health or deprivation contexts demonstrate initiative and troubleshooting in relation to s.12 of the 2018 Act quietly and routinely register, embedding it into everyday administrative practice rather than reserving it for exceptional cases which characterises ‘mainstream’ EMA provision. “Being in addiction doesn’t align very well with medical abortion, because medical abortion is not an event; it’s a process over time, requiring multiple visits and requiring scheduling and being punctual and turning up… when they’re in their addiction, turning up for appointments is actually really difficult.”
GPs absorbing and trouble-shooting the EMA care pathway is due to systemic, structural facets of the service: absence of any obligation on secondary care to respond flexibly by e.g. performing surgical abortions on days other than the hospitals routine ‘set days’ thus risking someone ‘timing out’; failure of the pathway to formally authorise workarounds that adhere to clinical good practice GPs innovate but are not anticipated by the service design or models of care.
In these circumstances GPs commitment is required to substitute for flexible, responsive infrastructure and pathways not currently provided. Exceptions to normal practice boundaries are being made particularly by GPs in primary care, but at the cost of the individual GP/doctor bearing responsibility for a decision the pathway does not formally authorise. Challenges to meeting the needs of people seeking abortion care in Inclusion health or Deepend settings are not a series of unrelated implementation problems but rather of three interlocking manifestations of a single structural pattern. EMA’s clinical pathways core assumptions about time, privacy, language, and safety are calibrated to a materially secure patient, producing the mismatch Hart’s (1971) inverse care law predicts between where need is greatest and where a pathway functions smoothly.
Read together through a reproductive justice lens (Ross and Solinger, 2017), findings from Bratt’s strand of ADAPT suggests that formal legal entitlement under the 2018 Act has not, for a meaningful subset of patients, translated into the practical capacity to exercise that entitlement with the same safety, dignity, and autonomy available to a more securely resourced patients. Moreover, the gap between entitlement and practice is currently being closed by individual GP discretion and goodwill rather than by design. This is consistent with what Ireland’s own Irish Human Rights and Equality Commission told CEDAW in 2025: that policy commitments on abortion access exist on paper without being consistently implemented (OHCHR, 2025). The present study adds to that account a granular, practitioner-level explanation of how that gap is actually being managed day to day, at considerable cost and burden to the GPs managing it.
Eight years after legalisation, the evidence, from this study and from the literature it engages with, points toward the same conclusion, that the legal right to early medical abortion in Ireland is well established. However, GPs in Inclusion Health or contexts of deprivation perceived access to abortion service was persistently undermined for their patients by economic hardship, homelessness, language barriers, addiction and chaotic circumstances, the constraints of Direct Provision, and the mandatory three-day wait and visit structure, compounded further by stigma. Participants broadly perceived the policy and service framework as not designed with these barriers in mind, being inflexible to their patients’ social contexts, and falling short of the integrated, needs-responsive model that equitable provision would require.
Conclusion
Implementation of s.12 of the 2018 Act allowing for abortion up to 12 weeks gestation has operated to date to a rigid, least-time model of care specifying that 12 weeks refers to 12+0 days as an absolute cut-off point for access to s.12 abortion. There is clearly scope for 12 weeks to be interpreted as referring to 12 weeks + 6 days. How this rigid, least-time application model is applied in the context of incomplete and failed abortion following administration of the full medical termination medication regimen is a critical concern. Given the extensive reliance on medication abortion including at later gestational stages when, albeit low, failure rates are increased, it is inevitable that incomplete and failed abortion will feature in the service and encounter this rigid time limit. Reference to Chief Medical Officer discretion in Boyd’s 2022 Guidelines suggests capacity for less rigidity but absence of clarity and certainty does not serve the those seeking or providing abortion care well.
Evidence from UnPAC and ADAPT, the largest studies to date of abortion use and provision in Ireland respectively, shows that women are presenting in good time for EMA care, at a median of 44 days, well within the time limit for primary care providing medication abortion. Notwithstanding, later presentation closer to the time limit of 12 weeks (84 days) since a woman’s last menstrual period is inevitable for a host of reasons. Some are general and everyday reasons – uncertainty of menstruation dates, ambivalence in decision-making, or obstacles to finding an abortion provider. Others confirm global research findings that people in vulnerable and marginalised circumstances, including those served by Inclusion Health GPs and GPs working in areas of deprivation, are more likely to present later. For this group, navigating the primary care EMA pathway, identified in UnPAC as placing a heavy burden on service users, is more challenging. The risk of timing out when seeking abortion care after a failed first EMA is exacerbated by the rigid, ‘least-time’ approach adopted to implementing s.12 of the 2018 Act.
Testimonies of challenges including delays handing over from primary to secondary care; obstructing GPs/midwives/hospital doctors; protracted repeat rounds of misoprostol to complete expulsion of all pregnancy tissue due to limited availability, use or offer of surgical abortion, all increase the risk of ‘timing-out’ under the rigid, ‘least-time’ approach to operating s.12 of the 2018 Act. Considerable adverse personal impact of both having and providing abortion care is the key effect of this rigid, ‘least-time’ model. The legal right to early medical abortion in Ireland is well established, but whether it is equally exercisable in practice continues to depend on the deprivation of the area a patient happens to live in, and on the willingness of an individual GP to personally close whatever gap that deprivation leaves. And one key more careful consideration is how time does, but also could, operate under the law in action. The argument above shows that s.12 can tolerate continued administration of abortion medication, including mifepristone, and administration of surgical termination when continued pregnancy is detected after a woman has been administered medication abortion before the twelve weeks cut-off point set down by the Chief Medical Officer and referred to in all Guidelines. If the CMO has stated the Act can tolerate repeat medication or surgical termination ‘in some circumstances’ after 12 weeks gestation, surely a coherent, fair and equitable legal model entails that it can tolerate it in all?
Dr Catherine Conlon, Katelin Bratt and Danielle Young School of Social Work & Social Policy, Trinity College Dublin
Bibliography
Bratt K, Young D, Favier M, et al Early medical abortion in Irish general practice: the Activity Data on Abortion in General Practice (ADAPT) data collection project and telemedicine as routine care. BMJ Sexual & Reproductive Health Published Online First: 28 July 2026. doi: 10.1136/bmjsrh-2026-203296
Boyd S, Feeney S, Harte K, Hayes S, McCarthy C, Hayes-Ryan D. National Clinical Practice Guideline: Investigation and Management of Complications of Early Termination of Pregnancy. National Women and Infants Health Programme and The Institute of Obstetricians and Gynaecologists. December 2022
Conlon, C., Antosik-Parsons, K., & Butler, É. (2025). Experiences of the Irish model of community medical abortion: adherence to self-managed, people-centred abortion care. Irish Political Studies, 40(1), 6–28. https://doi.org/10.1080/07907184.2024.2369335
Conlon C, Antosik-Parsons, et al. (2022) Unplanned Pregnancy and Abortion Care (UnPAC) Study. Available at: http://www.tara.tcd.ie/handle/2262/101813.
Fletcher, R. (2025) Witnessing legal sources of time for better abortion care Irish Journal of Sociology 33(1-2) 129-148 https://doi.org/10.1177/07916035251342137
Hart, J. T. (1971). The inverse care law. The Lancet, 297(7696), 405-412.
Horgan P, Thompson M, Harte K, Gee R. (2021) Termination of pregnancy services in Irish general practice from January 2019 to June 2019. Contraception. 2021 Nov;104(5):502-505. doi: 10.1016/j.contraception.2021.05.021.
IFPA (2025) Annual Report 2024. IFPA: Dublin.
World Health Organization. (2022). Abortion care guideline. Geneva: World Health Organization
Discover more from Doing Feminist Legal Work
Subscribe to get the latest posts sent to your email.