A community-based study on induced abortions: some unanswered questions

In their article on community-based research on induced abortion, Ganatra and Hirve (1) provide a candid documentation of the many ethical dilemmas confronting those engaged in research on sensitive issues related to reproductive and sexual health and rights. The effort made by researchers to ensure adherence to ethical norms is indeed commendable and exemplary. There are, however, a number of questions that remained unanswered in the article. We raise these below, mainly with a view to understand better the reasons why certain procedures and processes were adopted, so that these may be emulated by future research studies on abortion.

1. What were the research questions? Paragraph 2 of the article (Our own ….. socio-cultural context) suggests that the research questions were as follows: 2. Where and how do rural women access abortion services?

What are the factors influencing women’s access to abortion services? How unequally is access to services distributed across different population groups (by marital status, parity, age, class, rural/urban residence, nature of health facility accessed, etc.)?

How are women’s choices of where to access abortion services modified by their socio-cultural context?

If these were indeed the research questions, then the need to use a case-identification method is not clear. For example, answers to questions 1 and 3 could well be gathered through focus group discussions and in-depth interviews with key informants. Gathering case studies of detailed reproductive histories of women representing different population groups would provide insights into a range of factors that influence women’s access to abortion services. Information on sensitive issues like places from which single women access induced abortion services may be had from key informants such as traditional midwives and medical practitioners and paramedical workers servicing the area. Reproductive health surveys of women of reproductive ages could also have achieved this end (see point 4 ).

In all these cases, there would be no need to identify specific persons and hence, elaborate procedures to mask users of abortion services could have been avoided entirely. And we would still be in a position to answer many of the questions we started with.

Since the research included a three-month period to gauge the sensitivity to stigmatisation and did identify different levels of stigmatisation corresponding to different sub-groups of abortion-seekers, one wonders if it may not have been possible to document case-studies from each of these. This would have helped understand issues related to access to safe abortion services for these different groups.

2. Cases identified through clinic-based providers were enrolled prospectively. The power equations between a medical professional and rural women being heavily tilted in favour of the former, is it not likely that some women gave consent believing this is what was expected of them, and owing to reluctance to offend the service provider?
3. The role of community-based ‘informants’ is not clear. Did they provide information on ‘cases’? Would this not represent breach of confidentiality? Why was it so important to identify women who had recently undergone abortions through third-party informants? Since the researchers have carefully considered this issue, it is possible that their decision was based on the assessment that the ‘benefits’ of this potential breach of confidentiality outweighed the ‘costs’. However, why they reached this conclusion is not very clear from the article.
4. The researchers note that ‘for most married women who had undergone an abortion with the knowledge and support of her husband, the issue was neither overly sensitive nor stigmatising’ ( para 5). This is no doubt an accurate observation. So the question is, why not carry out a general reproductive health survey covering all women of reproductive ages (designed to capture an adequate sample of induced abortions, based on an estimated induced abortion rate) in which women report their induced abortions and answer open-ended questions pertaining to where these were accessed, how they made their decisions and so on? In fact, this is what appears to have been done even with the knowledge that a woman had undergone induced abortion (para 1-2, p.8).
5. Data were collected prospectively and retrospectively; in the clinic setting and in the community setting, from women who could not choose to deny their induced abortion (because this information was obtained from the person who provided them these services), and those who could choose to do so. These two sets of data may not be comparable, and one wonders how this issue was reconciled for data analysis.
6. Discontinuing the interview if a woman reported her induced abortion as spontaneous may be sensitive, but is it ethical? Is that not a reason to continue the interview to find out if there was any related morbidity, its nature and severity and the source of health care sought by the woman? These questions would not only help assess whether the abortion was ‘probably’ or ‘possibly’ induced (2), but are important in themselves to assess women’s access to relevant reproductive health services, and to provide or refer them in case of lack of access. This is particularly important if one considers the possibility that women who have undergone unsafe induced abortions may be at high risk of related complications.
7. In observing that some women who had ‘in fact’ had an induced abortion chose to report it as ‘spontaneous’, the authors appear to have assumed that it is their informants who reported the ‘truth’, and that the women did not. The basis of this assumption is not clear. Could it be that the informants want to be helpful and eager to be of service, and include among cases those who they only suspect to have had an induced abortion?
8. The article states that when the woman was at risk of being stigmatised, she was not interviewed at all, unless it was possible to do so in a clinic setting. In our experience in a rural poor setting in Tamil Nadu, this is the group least likely to access safe abortion services in a clinic setting, and most at risk of an ‘unsafe’ abortion. Was the researchers’ experience in Maharashtra different, and were they able to interview this group of women in a clinic setting? If yes, could they speculate on why their setting is different?
9. The use of ‘dummy’ interviews seems to be an unsatisfactory strategy to protect the ‘case’ for several reasons. One, having a group of interviewers conducting several simultaneous interviews could have been intimidating for the household concerned. Did all these members give informed consent to be interviewed? If they did, then using them as dummies amounts to misleading respondents who so willingly give their time to answer questions, and appears to be a breach of trust. Two, it must be possible to find other ways of ensuring privacy even in a rural community, such as finding a room (the balwadi on a non-working day or in the evening, the panchayat office, etc.) in which respondents are interviewed in private, one by one. We have done this as part of several research studies in villages in Tamil Nadu, and wonder what made this impossible in the researchers’ setting. Three, were the considerably higher costs that this strategy is likely to have entailed justified? And finally, if the issue was not very sensitive (because it is not sensitive for most married women and other women were in any case interviewed only in a clinic setting), why was it even necessary to have dummy clients?
10. There are many instances in the article where the authors single out qualitative methods as intrusive, potentially threatening, leading to ‘coercive participation’ and posing ‘complex ethical dilemmas’, implying that quantitative methods can be absolved of these traits. (Paragraphs 1 and 4 on page 7 and the concluding paragraph on page 8). Or have we misunderstood their stance?
11. What is the researchers’ responsibility towards those in need of services for abortion-related morbidity? Referring them to a medical facility may not be meaningful, because women may not be able to access these for the same reasons that prevented their using these facilities for a safe abortion in the first place.
12. It would be of value for researchers like ourselves to learn about the ways in which findings from this study have been (or will be) utilised to enable equitable access to abortion services for future abortion seekers.
13. We find the authors’ concluding remarks rather disconcerting. Why strive for an accurate estimate of rates of utilisation or of morbidity at the cost of participants’ dignity and autonomy? Is not this concept rooted in benign paternalism that assumes ‘we know what is good for you (even if you don’t).’?

To conclude, we agree with the authors that we researchers have a responsibility to be up front about the ethical dilemmas we confront. The questions above have been raised with a view to continue the debate in the same spirit.