We used a qualitative approach as the evolving and flexible nature of a qualitative approach helps explore the inner experiences of the participants and how they form and transform meanings with regard to the embryos.

Study site: Our study was conducted in Dakshina Kannada (South Canara), a southern coastal district of the Indian state of Karnataka. The total population of the region is 20,89,649, with 88.57 per cent literacy as per the 2011 census (Dakshina Kannada District Administration 2024).

Inclusion criterion: Residents couples of Dakshina Kannada (DK) who underwent IVF procedures at any point in their life. Factors like the location of the IVF centre, the outcome of the IVF procedure, the involvement of a donor gamete, and the age of the participants, amongst others, did not affect the participant’s inclusion.

Exclusion criterion: Couples still undergoing IVF treatment, i.e. those who are currently in some stage of an IVF cycle. This is because people on the IVF journey are likely too preoccupied with the outcome. Also, this population subgroup would be harder to recruit unless one collaborates with IVF clinics.

Recruitment: The process of recruitment consisted of three stages—identifying the potential participants, approaching them, and obtaining consent.

Identifying the potential participants: Convenient sampling and snowballing were the methods used to identify and recruit potential participants. We asked health professionals, IVF providers, laypersons like auto-rickshaw drivers, students, gym instructors, and shopkeepers whether they knew someone in a personal capacity who meets the inclusion criterion of our study. If they did, then flyers/participant information sheets (in English and/or local language) were handed to them to forward to the potential participant.

Approaching the participants: Only if and when the potential participants were willing to be contacted by our research team did the Principal Investigator (PI) contact the participant. Our study design allowed the participants to choose the date, time, and mode (telephone, video call, or in-person) of the interview.

Obtaining consent: Written informed consent was obtained from all participants. For participants who did not understand English, we developed the informed consent document in the local language with the help of a professional translator. Consent for audio recording was obtained separately.

Study tool: We conducted interviews with the help of a set of guiding questions and a demographic sheet. We developed open-ended questions on certain domains of inquiry using the existing literature on this subject—the informed consent process, views on embryos, preferences regarding embryos, concerns during the process of IVF, decision-making process, knowledge and views on research using embryos, amongst others. More domains emerged as interviews were recorded- seeing the embryos, prior research on IVF, embryo adoption, relationship with the IVF provider, and eugenics, amongst others. To minimize response bias, the questions designed were checked for inherent bias by reviewing them several times. We ensured that the question did not lead the respondent to a definite response.

The demographic sheet captured relevant information like- the number of IVF cycles, religious adherence, education, employment status, and use of donor gametes. It was indicated clearly on the sheet, as well as informed orally, that answering questions is not mandatory.

Interview: The language of the interview was either English or Kannada, depending upon the participant’s preference. In the case of Kannada, a bilingual interpreter would accompany the PI during the interview with the participant’s permission. One interview was obtained per couple, and they were given full authority to decide whether they wanted to be interviewed together, separately, or only one person wanted to be interviewed. Post-interview, the participants were requested to suggest other participants for the study (snowballing). Participants had the full right to discontinue and/or withdraw their participation from the study at any point.

All the interviews were transcribed in English from the audio recordings. For interviews in the local language, the interpreter would translate the participant’s words into English during the conduct of the interview. All interview transcripts were anonymized and only the research team had access to identifiable information.

Analysis: We used the grounded theory (GT) approach for the analysis of data. GT allows us to look at the data from various angles and to come up with comprehensive explanations. It is useful to explore beliefs, meaning, rationale, logic, emotion, response to the problem, and action interactions among others. Besides, these steps of GT have been proven to be culturally sensitive and the theory can always be revised and updated as more information comes in (Corbin and Strauss 2014).

The following steps were involved in the analysis-

(i)

Open coding: The interview transcript was read line by line. All transcripts were reviewed, and the data was broken into smaller parts to examine, compare for similarities and differences, and conceptualize open codes.

Open codes are the words used by the PI to denote the interpreted meaning of the data (Corbin and Strauss 2014). If a statement could mean more than one thing, it was coded into more than once generating more than one open code for a similar statement. Memos were also written in this process. Memos are the questions, thoughts, and inputs which come into the researcher’s mind while doing the process of coding. Memos subsequently help to relate the codes with each other.

The free software WEFT QDA was used to perform open coding. We made a sincere attempt was made to cover every bit of the data while coding.

The process of open coding in WEFT produced a total of 319 open codes across seven interviews.

This is how a typical code at the stage of open coding looked like in WEFT- Fig. 1.

Fig. 1

A typical open code in WEFT

(ii)

Axial coding: The open codes were related to each other to form explanations via categories (axial codes). To make a large number of open codes easy to handle these codes were initially sorted into nine broad categories—(1) views of embryos, (2) reasons/motivations for IVF/IVF centre, (3) factors motivating/affecting decisions, (4) concerns/fears/inhibitions, (5) constraints/restrictions/mandatory procedure, (6) unaware/lack of awareness, (7) openness and satisfaction, (8) ambition/desire and (9) miscellaneous. These categories served as a precursor to axial codes.

Based on the categories, the open codes were further grouped using coloured bookmarks and chart papers to create axial codes.

The purpose of using coloured bookmarks and chart paper was to have a visual frame that would help in arriving at the axial codes. Once the axial codes were created all the open codes were rechecked to ensure nothing was ignored. At this point, the open codes which did not fit anywhere were classified as “Miscellaneous.”

The axial codes were further refined by merging similar open codes.

Finally, using Microsoft PowerPoint, a refined axial code with all its associated open codes and its strength was produced.

(iii)

Selective coding: The “core axial code” was selected and related to all other codes producing a tentative GT (Corbin and Strauss 2014) (Noble and Mitchell 2016).