By Angela Dawson
Early medical abortion is safe, effective and increasingly central to abortion care. As services expand through primary care, telehealth and self-managed follow-up models, clinicians need practical ways to identify the small proportion of patients who may be at higher risk of complications and who may benefit from additional counselling, analgesia planning or follow-up.
Our recently published scoping review mapped the evidence on predictors and prediction models for complications following early medical abortion, defined as medical abortion at up to 12 weeks’ gestation using mifepristone followed by misoprostol. We found that, despite a large literature on the safety and effectiveness of early medical abortion, evidence on who is more likely to experience complications remains limited and uneven.
Thirteen studies met the inclusion criteria. They addressed outcomes including pain, retained products of conception or incomplete abortion, continuing pregnancy, heavy bleeding requiring transfusion, infection and need for surgical intervention. Across these outcomes, one factor stood out most consistently: gestational age. Higher gestational age was repeatedly associated with increased likelihood of incomplete abortion, continuing pregnancy, heavy bleeding and more severe pain.
Other predictors were more outcome specific. Pain was associated with factors such as pre-procedure anxiety, a history of dysmenorrhoea, nulliparity, prior caesarean delivery and higher gestational age. Retained products of conception and incomplete abortion were associated in some studies with older age, parity, previous spontaneous abortion, previous termination of pregnancy and prior caesarean delivery. Continuing pregnancy was linked with gestation of eight weeks or more and previous abortion in some studies.
However, the review also highlights a major gap. Very few studies developed or validated clinical prediction models. Two studies included models to predict surgical intervention following early medical abortion, but their performance and applicability were limited. One model showed moderate discrimination, while another relied on ultrasound-derived gestational sac measurements, which may not be feasible or necessary in many early medical abortion pathways.
This matters because models of care are changing. Routine in-person follow-up is no longer required for all patients, and many services safely use telephone follow-up, symptom checklists and low-sensitivity pregnancy tests. In this context, better evidence about predictors of complications could help clinicians move beyond a one-size-fits-all approach. Risk assessment could support clearer counselling, more tailored pain management, targeted follow-up and earlier escalation when needed.
At the same time, prediction must be used carefully. Early medical abortion remains highly effective and complications are uncommon. Risk prediction should not create unnecessary barriers to care, over-medicalise a safe procedure or increase reliance on ultrasound where it is not clinically indicated. Instead, its value lies in supporting person-centred care: helping patients know what to expect, when to seek help and how services can respond promptly to emerging problems.
The next step is stronger evidence. Future studies need larger prospective datasets, standardised definitions of complications, clear reporting of mifepristone–misoprostol regimens, and external validation of prediction tools across diverse service settings. This is particularly important as abortion care continues to shift towards community, primary care and telehealth models.
For clinicians and services, the message is practical: gestational age is the most consistently reported predictor, pain deserves proactive assessment and management, and current prediction tools are not yet robust enough for routine implementation. For researchers, the priority is clear: develop and validate simple, clinically useful models that improve safety and support access, without adding unnecessary steps to care.
Angela Dawson is a Professor of Public Health in the Faculty of Health at the University of Technology Sydney and an adjunct Professor at the School of Public and Populational Health at the University of New South Wales. She leads the reproductive theme in the Women and Children’s Health Collaborative at INSIGHT, the UTS Research Institute that seeks multidisciplinary solutions to complex issues. Her work spans applied research health services, health promotion, and prevention research for under-served populations including Indigenous Australians, refugees and migrants, people experiencing mental and substance use disorders and people have been incarcerated. She has undertaken research into the delivery of reproductive health services in humanitarian emergencies, the management and referral of women who have experienced domestic violence and FGM/C and access to abortion and emergency contraceptive pills in Australia and internationally.