Telemedicine early medical abortion at home: evaluation

Methods

The evaluation comprised four work packages (WP), using multiple methods (quantitative and qualitative) to enable systematic triangulation of data sources which would address the evaluation objectives and questions. The methods used in the four work packages are detailed below.

Work Package 1: Effectiveness and safety

This WP was designed as a comparative retrospective cohort study to compare rates of success, adverse outcomes, and serious complications with EMAH in the six months pre- and 12 months post- the introduction of the COVID-19 approval to EMAH delivery. This WP was designed to provide data to address research questions 1, 2 and 4.

The evaluation team contacted the clinical leads of all NHS abortion services in Scotland and asked them to report on: outcomes of abortion, as per the existing definitions set out in the Medical Abortion Reporting of Efficacy (MARE) guidelines^[16]; any severe complications; and other agreed adverse outcomes of interest. Data were requested to cover two time periods, before and after the introduction of the changes to EMAH. The outcomes and complications that they were asked to report on were agreed at the outset of the project in conjunction with the clinical leads of the abortion services. The time periods chosen were the six months prior to the COVID-19 approval (October 2019-March 2020) and 12 months following this (April 2020-March 2021).

The abortion outcomes, severe complications and other agreed adverse outcomes were:

• Outcomes of abortion:
  • complete abortion (without the need for surgical evacuation)
  • ongoing pregnancy after treatment
  • incomplete abortion (requiring surgical evacuation)
  • medical management of retained products of conception (treated with additional mifepristone and/or misoprostol)
• Serious complications:
  • haemorrhage (1 litre or more, or requiring transfusion)
  • severe infection (requiring intravenous antibiotics)
• Other adverse outcomes of specific interest in relation to the changes made with respect to pre-COVID 19 practice:
  • diagnosis of ectopic pregnancy after treatment commenced
  • expulsion of a pregnancy at an advanced gestation (over 12 weeks)
• Other outcomes of interest:
  • hospital admission for pain management
  • unscheduled hospital attendance within 24 hours of misoprostol
  • reattendance to the abortion service within six weeks for a concern related to abortion

Clinical leads were also asked to provide clinical details (pseudonymised) of all these cases, sufficient for the purposes of verifiying that the outcome/complication met the agreed criteria. In addition, as a further quality check to ensure that data on safety were as accurate as possible and to minimise underreporting of adverse outcomes or complications, the evaluation team undertook a validation exercise. For this, Public Health Scotland (PHS) provided data on the numbers of cases of re-admission to hospital within one month of EMAH due to specific adverse outcomes and complications, together with some pseudonymised details. These data were then provided to the abortion lead in each Health Board so that they could cross-check that the data they had submitted were as accurate as possible.

Data was provided from eight of the eleven mainland NHS Health Board areas (Ayrshire and Arran, Borders, Dumfries and Galloway, Fife, Grampian, Greater Glasgow and Clyde, Lothian and Tayside). The participating Health Boards included those providing the greatest number of abortion procedures, and those performing the highest proportion of EMAH procedures. A small number of women underwent EMAH who resided in the three Island Board areas, and these cases were included in the data from the mainland support board (Grampian for Orkney and Shetland, and Greater Glasgow and Clyde for Western Isles).

The remaining Health Boards were either not able to submit data (one board) or else submitted incomplete data without sufficient case details to permit verification of the outcome or complication (two boards). This was mainly due to staffing constraints and time pressures.

Data on the total numbers of EMAHs within each time period in each Health Board were provided by PHS. These data are publicly available and routinely collected by PHS, based on the monitoring data provided to the Chief Medical Officer (1).

The sample size for WP1 was based upon annual figures for abortions in Scotland, and the proportion of these that involved EMAH. A sample from the six-months prior to March 2020, and a 12-month period after the changes were introduced, was expected to give a reasonable sample size to compare rates of effectiveness (complete abortion rate) of EMAH over the two time periods. However, given that adverse outcomes of abortion are uncommon, and serious complications of medical abortion are known to be rare (for example, the rate of haemorrhage is 1 in every 1000 patients), it was accepted that a much larger sample size (tens of thousands) in each group that would take many years of future data collection would be required to compare the rates of rare complications with statistical significance.

Given the small numbers of complications and adverse events, outcomes of EMAH, serious complications and adverse outcomes were reported as percentages. For comparisons of complete abortion rates in each time period, a Chi squared test of statistical significance was used (Graph pad, USA).

An application to conduct WP1 was submitted to Public Benefit and Privacy Panel for Health and Social Care (HSC-PBPP) in May 2022 and approval was given in August 2022.

Work Package 2: Mapping clinical abortion care pathways

WP2 was developed to provide context for understanding the models of abortion care in place across Scotland, and provided information relevant to addressing all of the research questions, with a particular focus on helping to respond to research question 4. We aimed to map abortion care pathways for all NHS Health Boards with a particular focus on EMAH. A list of key points in the abortion care pathway was developed and refined in consultation with the evaluation team. These included: referral pathways; key elements of service delivery; treatment options offered; and safeguarding processes. They also included wider elements of sexual and reproductive health provision, such as contraception provision and testing for sexually transmitted infections.

A questionnaire gathering these items (see Appendix 1) was sent to lead clinicians in abortion care at each Health Board and an initial map was developed based upon these responses. The initial maps were sent to the Scottish Government policy and analytical teams for feedback. Following this feedback, further items of interest were added and the lead clinicians were contacted twice more to check the maps for accuracy and to add requested additional information. The mapping activity started on 23^rd January 2022. The final versions of pathways were confirmed on 21^st October 2022. The maps produced therefore reflect pathways in place at this time and not necessarily those that were in place during the timeframes covered by the data collected for WP1.

The service maps were used to contextualise the data collected in other work packages within this report. From the maps, we also generated a list of recommended processes and procedures for services around EMAH.

Work Package 3: Patient experiences of abortion care

This work package was undertaken to provide data to address research question 3 and 4 as well as aspects of the other research questions. An online quantitative survey (anonymous) was conducted of those who had undergone EMAH in any Scottish Health Board in the preceding three weeks. The online survey tool used conformed to NHS security standards.

Patients received information about the survey from clinic staff in their pre-abortion consultation and were given a leaflet in their home medication pack. The leaflet included a URL link and a QR code to the online survey. Those who wished to participate were advised to use the link (or QR code) provided to complete the survey two to three weeks after treatment. They could opt out of the survey at any time and also choose not to complete questions if they wished. Patients were also given the option to respond to the same survey conducted via telephone with a researcher (clinical research nurse/midwife) at the same time point, although no patients took up this option.

The survey was designed to determine views on patients’ experience and satisfaction with care overall, and specific aspects of care, including:

• the type of pre-abortion consultation received
• the privacy of the consultation
• the information provided on EMAH
• contraceptive counselling and provision of contraception
• the offer of testing for sexually transmitted infections
• questioning on gender-based violence
• views on having/not having an ultrasound scan
• access to abortion medications
• views on continuation of elements of EMAH service delivery

The survey also requested some demographic characteristics of respondents (age range, ethnicity, Health Board area). The questions were developed in close collaboration with staff in abortion services across Scotland (Scottish Abortion Care Providers network).

The final survey comprised 34 questions. Four initial questions ensured the individual was giving informed consent and was eligible to participate. The subsequent 30 questions on their experience and demographics required tick box responses to a range of options (see Appendix 2). Respondents also had the opportunity to provide free text comments at the end of the survey.

Descriptive analysis of each question was generated by the online survey tool, including numbers and percentages of total responses to each question. Free text comments were categorised by the research team as positive or negative and grouped into themes.

The NHS Lothian Information Technology Governance team approved the data protection impact assessment that detailed the plan for data capture, handling, and storage of the survey data. NHS Lothian Caldicott gave approval for the survey in June 2022. Approval at each NHS Health Board was also given. The need for local Quality Improvement Team approval at each board resulted in different timing of starting the survey, as well as significant delays in starting at all boards (see Appendix 2). The duration of the survey ranged between 1 to 4 months at each board. The survey commenced on 4^th July 2022 with others Health Boards starting later and all surveys closed on 30^th November 2022.

Work Package 4: Staff perspectives on provision of abortion care

This work package utilised findings from an ongoing project conducted by one of the co-investigators (NB) called The ENHANCES Study (undErstaNding provider perspectives on High-quality AbortioN CarE in Scotland. This was undertaken as part of a THIS (The Healthcare Improvement Studies) Institute-funded fellowship that examines NHS staff (abortion care providers) perspectives on provision of abortion care, and the integration of patient experience in quality improvement. As such, WP4 was conducted seperately, although in parallel to, the main EMAH evaluation study.

One of the key aims of this work package was to explore abortion care providers’ perspectives on what constitutes safe, high-quality abortion care in Scotland, and gain insight into approaches used to inform service improvements in abortion care. The study also explored barriers and facilitators to service provision and implementation (including telemedicine and home use of both abortion drugs, mifepristone and misoprostol) and staff perspectives on the integration of patient experience (including participatory approaches) into service improvement activities. The ENHANCES study findings provide insights on staff perspectives on key areas related to research question 4, but also contributed to the other three research questions.

Between February and October 2022, staff working in abortion services (nurses, midwives, and doctors) in Scottish NHS territorial Health Board areas were invited to participate in a qualitative study. The primary inclusion criterion was being involved in the provision and delivery of abortion care (including direct provision, leadership and management) in Scotland.

Staff were recruited from nine Scottish NHS territorial Health Board areas (NHS Lothian, NHS Greater Glasgow & Clyde, NHS Grampian, NHS Western Isles, NHS Forth Valley, NHS Shetland, NHS Ayrshire & Arran, NHS Tayside, NHS Fife). The sample comprised 27 individuals with a variety of experience and skill levels involved in the provision of abortion care: consultants/abortion service leads (n=13); nurses/midwives (n=11); specialist trainees in sexual and reproductive health or obstetrics and gynaecology (n=3).

The study employed a purposive, network-based and snowball sampling approach to provide representation across occupational groups and level of experience in abortion care provision. There were no predetermined participant numbers within each category of the sampling frame.

Information about the study was disseminated using: (1) email advertisements shared via the Scottish Abortion Care Providers network (with permission from co-chairs of the network); (2) through professional networks; and (3) snowball sampling (colleagues sharing information about the study). Although the study was with clinical staff, recruitment did not take place directly through NHS boards. Staff who chose to participate did so in their own time. Staff were provided with information on the study and invited to contact the researcher if they were willing to participate in an interview. Consent for study participation was obtained prior to interview. Participants were not paid for their participation.

One of the investigators (NB) conducted the interviews using a topic guide which covered:

• staff experiences of the introduction of telemedicine for early medical abortion during COVID-19
• practical dimensions of abortion service provision
• perspectives on what constitutes high-quality abortion care (including patient experience)
• barriers and enablers to high-quality abortion care
• views on the future of early medical abortion, including telemedicine care

Interviews were conducted online (n=23) or by telephone (n=4), were digitally recorded, and lasted between 40 and 80 minutes. Interviews were transcribed verbatim by a University of Edinburgh-approved General Data Protection Regulation compliant external transcription service.

Interview data were analysed thematically by NB. NVivo Qualitative Data Analysis Software (Version 12, 2018, QSR International Pty Ltd) was used for data coding and management. An initial coding framework, which captured both the original research questions and emergent issues, was developed and applied to the interview data. The coding scheme was revised and refined (iterative cycle) as analysis progressed. Recurrent themes were identified through a process of cross-comparison. NB met regularly with members of the wider research team to discuss analytic interpretations; the composition of the team allowed discussion of the data from different disciplinary perspectives. At the time of writing, further in-depth analysis is ongoing.

The study was reviewed, and received favourable ethical opinion, by the Edinburgh Medical School Research Ethics Committee (EMREC) at the University of Edinburgh (Application 21-EMREC-052; 25 January 2022).