Case 7: Confidentiality in a Study of Abortion at the Community Level*

Background

The Medical Termination of Pregnancy (MTP) Act "made abortion in India legal on a wide range of social and medical grounds yet the availability, geographic accessibility, and functioning of approved MTP centers remains inadequate and it is estimated that abortions outside of the legally recognized sector are two to five time's more frequent than legal ones. While the demographic profile of the abortion seeker, the dangers associated with illegal abortions and the links between abortion and fertility control have been well documented, much of the information comes from urban hospital/clinic based studies, many of which are based on review of hospital records rather than actual interviews with abortions seekers. Because of their self-selective nature, even well designed hospital studies cannot adequately address issues of provider choice. Community-based studies are few and mostly in the nature of knowledge, attitude, and practice (KAP) surveys. Problems of deliberate underreporting or the inadvertent classification of induced abortions as spontaneous events make case findings, especially in survey settings, a challenge."

To determine the level of complications post-abortion, a local non-government research group in Maharashtra State of India, designs a study which uses a combination of qualitative and quantitative approaches to data-gathering to examine provider choice, the abortion experience, and consequences from a women's perspective. The study will cover 140 villages (population 320,000) in an area where government abortion services are provided through the district hospital, some rural hospitals, and a few primary health care centers. In addition to these facilities in the public sector, a number of small private hospitals and some NGOs offer varying levels of abortion services.

Case Finding Methods

During an initial three-month preparatory phase, discussions were held with key informants in the community and abortion service providers. These interviews helped to delineate the range of providers, determine case finding methods, and define areas of concern to the community and providers. The study team proposes to set up an information system which will use two main sources to determine who has had an abortion in the study area: (1) community-based health workers and women's groups; and (2) health workers [in both the formal and informal sectors] who are providing abortion services.

Cases will be enrolled prospectively over an 18-month study period. All health providers who perform abortions and who agree to be part of the case finding process will be given an instruction sheet explaining the purpose and methods of the project. They will discuss the project with abortion seekers and obtain their consent to be interviewed at home by the research team. Verbal consent will be obtained as written consent is thought to be threatening in the culture of rural Maharashtra. Health providers will give the researchers the names of those who agree to participate but will not provide a list of those who refuse to be interviewed. Information sources within the community will be asked to serve as intermediaries to schedule an interview between the researchers and the consenting abortion seeker not identified by the facility-based system. No incentives will be offered for case identification. Interviewers will go to the villages of those who have had an abortion no more than three months after the procedure.

As some of the women may have changed their mind about being interviewed after their original consent, a second layer of protection from unwanted attention is afforded the abortion cases by adopting the following measures: (1) during the interview attention will be focused on gynecological problems, past pregnancy outcomes, and health complications rather than the abortion episode per se; (2) interviews with abortion cases will be conducted only when there is a cluster within a community and dummy interviews will be conducted in the same village using the same questionnaire with women not known to have had an abortion; and (3) artificial privacy will be created during the interview by using a "team" of interviewers where one person conducts the actual interview while the other members of the team engage family members in dummy interviews. Women will be free to discontinue the interview at any time without prejudice. The study is submitted to an India-based foundation, which agrees to fund it.

Deliberations of the Ethical Review Committee

The study is presented to the ethical review committee of the international funder following its approval by the state ethical review board. All but one member of the ethical review committee, an anthropologist who has worked extensively in rural India, approve of the study. She is concerned that there is some chance that the confidentiality of the abortion seeker may be compromised and wants assurances that the records that identify the research subject will be kept confidential. The study group responds that all records will be kept under lock and key in the main offices of the NGO, which is nowhere near the study site.

Questions:

Please comment/elaborate on the following questions:

• Does the process of identifying individuals who have had an abortion insure confidentiality?
• Should oral consent substitute for written consent in a population that has high rates of illiteracy and is concerned with putting any signature/thumb print on a written document that they may not understand?
• Does the interview process insure confidentiality?
• Comment on the procedure that was used to protect the women who had consented to be interviewed at the time they underwent an abortion but who may have subsequently changed their minds (dummy interviews in the community, clustering of interviews, and dummy interviews with other family members during the interview with the woman).
• What other approaches might be suggested to protect the interviewee from the possibility of unwanted attention?