Telemedical Abortion - Effective, safe and acceptable?
- Healthcare for Both Lives
- Jun 12
- 6 min read
Effective, safe, and acceptable. These are the three words with which Aiken et al conclude their assessment of telemedical abortion - also known as ‘Pills by Post’ - published in 2021 in BJOG. In an accompanying editorial, Prof Lesley Regan, a past president of the RCOG (2016-19) expressed the same opinion, using the same three words – safe, effective and acceptable.
The same study was quoted – five years later - by Baroness Blackstone in the House of Lords, as part of the recent debate over the decriminalization of abortion up to birth.
Given how frequently this study is cited in defence of telemedical abortion policy, its conclusions demand careful scrutiny. The question must be asked: does this three-fold conclusion – safe, effective and acceptable – have a strong evidentiary basis? And do current reporting systems even have the capacity to consistently and representatively collect such evidence?
An examination of the current systems for identifying and recording abortion complications, for collecting follow up data, and for linking records between different NHS services, suggests that the answer may well be ‘no’.
The UK does not collect linked follow-up data after abortion
One challenge in collecting robust evidence is the difficulty of following up women who have had induced abortions. Sometimes abortions are concealed, and this means subsequent complications and side effects may be either miscoded or misclassified. In some countries, ways have been found to improve the quality and quantity of follow up data. Denmark and Finland, for example, have established registries which enable anonymised abortion data to be linked to a woman’s medical record.
Collecting better abortion data in this way allows for more accurate understanding of the associated risks to women. A publication from the Finnish dataset, following 42,619 women (2000-2006) after having an abortion before 9-weeks’ gestation, showed a nearly 8-fold increase in risk of haemorrhage for medical abortion compared to surgical abortion. 5.9% of women undergoing medical abortion required surgical (re)evacuation. Women with adverse events were 20% of the medical abortion group compared to 5.6% of women undergoing a surgical abortion.
A Danish study reported similar findings. Of 86,437 early medical abortions over a 10-year period from 2005, 5320 (6.2%) underwent a surgical intervention within 8 weeks. The proportion of surgical interventions increased from 3.5% in the 5th to 6th gestational week to 10.3% in week 9.
A comparable Swedish study looked at all abortions (both medical and surgical) from one hospital from 2008 to 2015, reporting overall complications (bleeding, infection or incomplete abortion). For medical abortion, one cohort took both sets of pills in the hospital, and another took the Mifepristone in the hospital and the Misoprostol at home. For women whose gestational age at the time of the abortion was <7 weeks, the complication rate was 2.4% in the hospital cohort, compared and 7.3% at home.
As can be seen, collecting high quality follow-up data allows for much greater insight into health outcomes following abortion. This quality of data is lacking in the UK, and so we are left reliant on estimates. In 2021, Kevin Duffy used freedom of information requests to try and capture lost follow-up data from NHS trusts. His analysis estimated that (conservatively) 5.9% of women having an induced medical abortion are subsequently treated at an NHS hospital for complications arising from an incomplete abortion with retained products of conception.
The focus of the Aiken study – “safe, effective and acceptable” – was on outcomes for women having face to face abortion appointments compared with those for women accessing abortion advice by telemedicine, receiving abortion pills either in the post or from a collection point. As the authors acknowledge, ‘The main limitation of this study is that we were unable actively to follow up patients after their abortion. There is a potential gap in the consistency of reporting incidents, due to some complications not meeting the threshold of serious incidents, multiple routes of entry into the NHS and informal communication between the NHS and abortion providers.’
Their definition of ‘serious incidents’ was ‘haemorrhage requiring transfusion, significant infection requiring hospital admission, major surgery and death.’
If the Nordic studies cited above are assumed to be reliable given their high-quality data collection, we may conclude that the Aiken study indeed significantly underreports abortion complications in both the telemedicine group and the face-to-face group. The study did not and could not measure any true incidence of short-term post abortion complications. Consequently, its much-cited conclusions – that telemedicine is ‘effective, safe and acceptable’ - risk being unjustified.
The HSA4 (Abortion Notification) Form does not allow for the collection of linked follow-up data
In England and Wales, practitioners carrying out an abortion must complete an abortion notification form (HSA4) and submit it to the Chief Medical Officer within 14 days of the procedure. It is not required to include the NHS number of the woman on this form. As can be seen, the timeframe for submitting the HSA4 necessarily excludes complications that lead to surgical interventions more than 14 days after the abortion.
Further, there is no UK equivalent to the Finnish or Danish abortion registries, so no long-term record linkage research can be carried out. This is a missed opportunity, as the UK would potentially have one of the world’s largest abortion datasets if NHS numbers were used in the same way as their equivalents in Finland and Denmark.
The NHS number is the primary way the NHS collects data for use in examining patient outcomes. In the programme implementation guidance of 2010, supporting NHS number information standards 2008, it is stated that ‘all NHS procedures should be followed to scrutinise outcomes (including procedures outsourced to private clinics) on behalf of the NHS.’ The presence of an NHS Number on all patient records, both paper and electronic, is now mandatory. In 2008, the National Patient Safety Agency (NPSA) published a Safer Practice Notice (SPN) for the NHS Number which mandated the use of the NHS Number as the national patient identifier or alongside any local identifier used. It must be used alongside other demographic information as part of safe practice to link together the correct records to a particular patient. Given these commitments, the absence of the NHS number as a mandatory inclusion on the HSA4 form seems indefensible.
However, despite this weakness in the HSA4 form as a data collection tool, it has recently been weakened even further. In April 2023, the form was amended to reflect the change in legislation permitting both pills to be taken at home. In these circumstances, the name and address of the place of termination is no longer required. Instead, there is an additional mandatory question asking whether any part of the consultation or treatment was provided face to face (in person) by a registered medical practitioner, nurse or midwife. So in addition to not allowing for record linking of the woman receiving the abortion, the HSA4 form now also does not record the place of the termination either.
The current situation is unsatisfactory for the following reasons:
1 The HSA4 form is not compliant with NHS number policy of 2008[1] and compliance is not being enforced.
2 The HSA4 form is not compliant with Caldicott principles – using a woman’s name instead of an NHS number risks breaching confidentiality.
3 The HSA4 form is not complying with Caldicott principle number 7, which states: ‘The duty to share information can be as important as the duty to protect patient confidentiality.’ The form undermines accurate audit and research purposes.
4 The Data Protection Act 2018 does allow for the processing of data for medical research, and this opportunity is currently being lost. The UK is lagging behind other countries where linked data is used by epidemiological researchers concerned about reproductive and other long-term risks to women. As shown above, large record linkage studies have been performed in Finland and Denmark which established their computerised abortion registries in 1983 and 1973 respectively.
5 No explanation has yet been offered as to why audit and research into the costs and consequences of abortion should not be facilitated by proper use of the NHS number on the abortion notification forms to create an abortion registry.
6 All epidemiological research is confidential, and abortion should be treated in the same way.
Making inclusion of the NHS number on the HSA4 form mandatory would enable data to be linked, and evidence of post abortion effects for female and neonatal health to be established. This data has been missing in England and Wales for decades, hindering research into both short and long-term outcomes, and limiting meaningful audit of abortion practice. If abortion policy and clinical guidance are to be based on robust evidence rather than incomplete reporting, proper use of NHS numbers to facilitate confidential record linkage is essential. Without that high-quality reporting, the claim that abortion by telemedicine is “effective, safe and acceptable” will remain no more than a claim.
Aiken A, Lohr PA, Lord J, Ghosh N, Starling J. Effectiveness, safety and acceptability of no-test medical abortion (termination of pregnancy) provided via telemedicine: a national cohort study. Bjog 2021;128(9):1464–1474.
Reardon DC. The abortion and mental health controversy: A comprehensive literature review of common ground agreements, disagreements, actionable recommendations, and research opportunities. SAGE Open Med 2018;6:2050312118807624.
Niinimaki M, Pouta A, Bloigu A, Gissler M, Hemminki E, Suhonen S, et al. Frequency and risk factors for repeat abortions after surgical compared with medical termination of pregnancy. Obstet Gynecol 2009;113(4):845–852.
Meaidi A, Friedrich S, Gerds TA, Lidegaard O. Risk factors for surgical intervention of early medical abortion. Am J Obstet Gynecol 2019;220(5):478 e1–478 e15.
Carlsson I, Breding K, Larsson PG. Complications related to induced abortion: a combined retrospective and longitudinal follow-up study. BMC Womens Health 2018;18(1):158.
Duffy K. FOI Investigation into Medical Abortion Treatment Failure (2021). In: FOI Investigation into Medical Abortion Treatment Failure (2021). 2021.
https://standards.nhs.uk/published-standards/nhs-number
Comments