[341 words]
Methodology :
Clinicians in two tertiary level university teaching hospitals (National Maternity Hospital (NMH), Dublin and Cork University Maternity Hospital (CUMH)) and one secondary level unit (Wexford General Hospital (WGH)) were invited to participate. NMH and CUMH each have more than 9,000 births per year and are teaching hospitals at both undergraduate and postgraduate level. WGH is a peripheral centre of excellence with more than 1,500 births per annum and provides postgraduate teaching in Obstetrics and Gynaecology.
The current training scheme in Ireland comprises two parts: basic specialist training (BST) and the higher specialist training (HST). After completion of one year of internship trainees enrol into BST which comprises two years working as a senior house officer and one year as a junior registrar. While some trainees would start performing OVD under supervision during BST1 or BST2, the majority would only start performing OVD on a regular basis during BST3. Following completion of BST and successful pass in the Royal College of Physicians membership examinations, trainees then interview for HST, a five-year structured training program. On successful completion of this, trainees are awarded a certificate of completion of specialist training (CCST).
Training in OVD comprises of two parts: firstly a theoretical introduction (as part of BST years 1 and 2 and again prior to the start of the third year) and secondly an on the job apprenticeship under close supervision. For the final year of BST trainees are supervised on the labour ward by a senior obstetrician (year 4/5 HST or consultant) for every OVD they perform. A logbook of every OVD is required and reviewed every year at the end of training year assessment.
For this study, clinicians at all levels of obstetrics training from BST3 level to final year specialist trainees and consultant obstetricians were invited to participate in the study. Following consent, participants were interviewed using open ended non-directive questions (Table I). As interviews progressed and new themes developed this guide was modified to allow further exploration. The interviews were recorded as audio (Dictaphone) and later transcribed (Express Scribe Transcription). Audio files were securely stored after transcription to allow review, and will be destroyed in the usual manner. Thematic analysis was performed until saturation.The qualitative research method was Thematic Analysis19. This was picked as it is a “flexible and useful research tool ” which can “potentially provide a rich and detailed, yet complex, account of data19
The process was as follows: As the interviews progressed and there was increasing saturation of data, the interviewers took care to carefully compare findings by checking with new participants and triangulation of data. Audiotapes were transcribed verbatim by the researchers and the transcripts were entered into NVivo and coded. Transcripts were read and reread as codes were assigned by two coders (ZA, LC). The data was first divided into the sections outlining the interview guide. This helped with the initial assessment of the data using a deductive approach, as the pre-existing structure of the interview guide provided base concepts, ideas and topics that acted as scaffolding for analysis. Data was then coded following the individual guiding question of the interview guide, narrowing, and defining the codes that presented in the data initially. At the end of stage two, codes were identified. Theme development occurred during stage three: codes were examined inductively, looking at what the participants were saying and whether some topics were continuous throughout the interview. The codes were then observed and grouped several times while the themes were identified by the researchers. Themes were developed from data that were prevalent across interview, important and substantial. Saturation of themes was defined when refinements did not add any new substantial themes.
Trustworthiness was enhanced by two methods: method checking and triangulation20,21. A variation of member checking was used, where instead of reviewing themes with the original participants, themes were instead reviewed with a different participant to see if they saw them as authentic. Triangulation was via researchers where two researchers initially analysed data (ZA, LC) and two reviewed identified themes (FMcA, MH).
The research team had previously reflected on possible personal sources of bias: MH and FMcA had participated in training in OVD and worked in clinical academia; ZA is a trainee in Obstetrics and Gynaecology; these were identified as a possible bias in the research and as such both identified this bias and worked with the other member (LC) who would not have this bias. This study is reported following the COREQ standards22.
The participants were also asked to provide basic demographics including years of clinical work (both in the speciality and in performing OVD), numbers of OVD, and preferred instrument. Ethics Committee permission was approved by the Ethics Committee of the National Maternity Hospital. There were no Patients involved in this study. A core outcome set (COS) was not used in the development of the trial; the CROWN database was checked and no relevant COS was found to be published or in development.