Ethical and definitional considerations in research on child sexual violence in India

This study draws from a broader systematic review which aimed to examine the prevalence, determinants and consequences of CSV in India (under review). The review explored articles published in peer-reviewed journals, indexed in PubMed, POPLINE, and PsycINFO, that described CSV experience or perpetration in India. To be included in the review, the articles had to meet the following criteria: i) be written in English, ii) be published between January 1, 2006 and January 1, 2016, iii) involve human subjects, and iv) collect original data on experience, perpetration or response to CSV in India. Searches were conducted using 55 search terms, each paired with the term, “India” (see Table 1).

The initial search yielded 4186 articles. After removal of duplicates, 3725 were left and screened for potential relevance by title and abstract review. Through title and abstract screening, 2760 studies were excluded based on irrelevancy, leaving 965 for full-text screening. Of the 965 articles, 762 articles were excluded because they did not meet eligibility criteria, 131 were excluded as insufficient information was provided to assess eligibility (with two failed attempts to contact authors for additional information), and 21 were excluded because the full text was not found through our institutional libraries nor through repeated attempts to contact authors. The final list of included articles consists of 51 studies. Figure 1 depicts the PRISMA Flow Diagram [29].

To inform the development of the ethics checklist a web-based search was conducted for guidelines, recommendations, and checklists from public health authorities guiding the design, conduct, and/or reporting of research on child abuse, gender-based violence, and sexual violence. Additionally, we sought input from India CSV experts for guidelines of relevance. Both methods led to author consensus that the following six documents were of greatest relevance for developing an ethics framework to evaluate CSV research in India: the UNICEF review “Ethical Principles, dilemmas, and risks in collecting data on violence against children,” the WHO “Putting Women First: Ethical and Safety Recommendations for Research on Domestic Violence Against Women,” the Sexual Violence Research Initiative “Ethical and Safety Recommendations for Research on the Perpetration of Sexual Violence,” the UNICEF “Guidelines on the Protection of Child Victims of Trafficking”, the Indian Council of Medical Research (ICMR) “National Ethical Guidelines for Biomedical Research Involving Children,” and the WHO clinical guidelines “Responding to Children and Adolescents Who Have Been Sexually Abused.” [4, 7‐11]. To develop the Ethics Checklist, key themes from each of the four documents were first extracted and commonalties between documents noted. The key themes were then discussed among the entire research team for inclusion based on their relevance to CSV research until consensus for inclusion was reached. The Ethics Checklist is presented in Table 2. The checklist takes into consideration the following key ethical principles: i) informed consent, ii) confidentiality, iii) selection, training, and protection of study team members, iv) validity of CSV measurement methods, v) measures to minimize participant harm and vi) participant compensation. Each of the 51 articles were reviewed by a single author (RD) and information regarding these key ethical principles was extracted. When questions arose or clarifications were needed, RD consulted a second author (AK) to reach consensus. If the information was not reported in a manuscript, we stated so.

Informed consent is the process by which researchers convey to potential participants and/or their guardians the purpose and activities of a study, its associated risks and benefits, and right to withdrawal. Further, it provides participants an opportunity to ask questions before deciding whether to participate. Domestic and international guidelines recommend obtaining consent from both the parent or guardian and assent from the children in accord with their developmental level and decision-making capacity, in research involving minors [2, 4, 9‐11]. In cases where parental consent cannot be obtained due to children being from shelter or runaway homes, street children, or in circumstances when caregivers are themselves perpetrators, guidelines recommend consent be obtained from the local guardians (i.e. non-governmental organization (NGO) staff, police officials). The ICMR guidelines discuss that for research related to child abuse, parental (or legal guardian) consent can be waived but that ethics committee should suggest an alternate mechanism to safeguard the child [10].

In our review, equal frequencies of quantitative (65% or 26/40) and qualitative studies (64% or 7/11) reported obtaining informed consent. Out of these studies, 42% (13/31) studies specified whether the consent was obtained verbally or in writing. As recommendations for obtaining informed consent differ by age, we categorized and then analyzed the studies in our review according to the age ranges of the populations they surveyed: i) participants ≤18 years; ii) participants ≥18 years and iii) participants spanning the age range of 8–49 years.

Among the 25 studies that included participants of age ≤ 18 years, 24% (6/25) obtained consent from both the study participant and parent or local guardian, and 44% (11/25) obtained consent from the study participants only. The latter studies were largely conducted in schools and observational/shelter homes, where prior permission to conduct the study was obtained from the respective institutional authorities. Only one study (of street children) reported taking informed consent solely from a guardian (i.e. police personnel and NGO staff members) [12]. In 20% (5/25) studies, whether informed consent was obtained was not reported, and in 8% (2/25) informed consent was deemed not necessary as the research was limited to data extraction from medical records [13, 14].

Among the 17 studies that included participants in the age ≥ 18 years, 71% (12/17) reported consent was obtained from the study participants, while 29% (5/17) did not report information about consent provision. Of note, all of the studies that did not report obtaining consent were cross-sectional and obtained CSV data from FSWs.

Lastly, seven studies included participants spanning the age range of 8–49 years. Out of the seven, three studies did not report information on obtaining consent, one study reported obtaining consent from both the parent and the participant [15], and one reported obtaining consent from the participant alone [16]. Of note, none of the 25 studies surveying CSV in participants age ≤ 18 years, nor the seven studies spanning age 8–49 years, reported discussing with the participant and/or guardian the limits of confidentiality given the mandatory reporting laws for child abuse in India.

Another key principle is maintaining privacy and confidentiality of data to protect participants from potential stigma and reprisal from the perpetrator. Sixty-three percent (63% or 32/51) assured participants about maintaining confidentiality of information. Out of the 32 studies, 24 (75%) reported measures that were used to enhance data confidentiality. Twenty-nine percent (29% or 7/24) reported use of secure environments in addition to data confidentiality. Forty-six percent (46% or 11/24) reported changing the names of the participants or not collecting identifying information in the questionnaires and 3/24 studies reported using a secure environment like conducting interviews in private settings or a comfortable safe place where confidentiality could be maintained. Two (2/24) reported de-identifying study documents (i.e. collecting and editing data the same day or destroying audio files). Nineteen (19/51) studies did not report information on confidentiality. Out of these, 5/19 studies were based on secondary data analysis and 14/19 did not report any information. None of the studies reported using computer-based technologies to collect de-identified data to circumvent mandatory reporting laws. Also, none reported whether or not they contacted legal or law enforcement authorities if ongoing CSV was noted.

Key to the conduct of CSV research is the assurance of trained (i.e. in research ethics, nonjudgmental and respectful communication, facilitating referral to support services, minimizing re-traumatization)and qualified research team members who are involved in the collection of data [4, 9‐11]. Additionally, protections should be in place for the research staff to assess for emotional trauma and physical harm they may incur as a result of collecting the data, with referral to counseling or other support resources as necessary.

About 69% (35/51) of studies in our search reported that study team members were trained in the collection of CSV data, of which 74% (26/35) discussed specifics of the training received. Only 11% (4/35) of the studies reported the study team received formal training in ethics, with all four providing an online ethics training module [17‐19]. Twenty-three (23/51) reported the qualifications of the research staff, with many using junior public health nurses, clinical psychologist or psychiatric social workers, medical officers, and peer researchers to collect data. No studies discussed safeguards they may have used to assess and address the impact of the research on the emotional and/or physical wellbeing on the study team.

Existing literature stresses the ethical obligation by investigators to ensure the validity of the tools and procedures they use to measure violence. This includes surveying a range of CSV behaviors, classified by WHO as “non-contact sexual abuse,” “contact sexual abuse,” and “forced sexual intercourse,” [20] using validated tools, and avoidance of loaded terms like ‘rape’, ‘abuse’ and ‘violence’.

Twenty-two percent (22% or 11/51) studies reported surveying all three WHO classifications in their assessment of CSV. An equal number (18% or 9/51) limited their surveillance to two categories of CSV. Almost half (43% or 22/51) either limited their surveillance to only one form of abuse or used a loaded term like “sexual abuse” to assess CSV. The remaining studies (18% or 9/51), a mix of qualitative and quantitative manuscripts, did not specify how they assessed CSV.

Among the 40 studies using quantitative methods, only 16 studies (40%) reported use of standardized tools. The tools used included the ISPCAN Child Abuse Screening Tool (ICSAT)–C and ICSAT- R, the Childhood Trauma Questionnaire, adapted versions of the questionnaire used in the MoWCD study, Finkelhor’s Sexual Abuse Scale, and the Sexual Abuse Screening Tool. Out of the 11 studies that used qualitative methods, the most common tools used to explore CSV experiences were focus-group discussion and in-depth interviews.

Guidelines on related topics discuss obligation by the researcher to minimize possible distress or harm caused to the study participant that may result from his/her participation in the study. This includes assessment of risk and provision or referral to counseling and medical services as necessary.

Only 25% (13/51) of studies discussed such methods. Among them, 54% (7/13) reported providing counselling services to participants as necessary, 31% (4/13) reported referring participants to psychiatric or general services if they appeared in distress [21‐23], one reported specific efforts to not re-traumatize participants [24], and one qualitative study requested participants to participate in a one-on-one in-depth interview of convenience to them if they were apprehensive in participating in the focus group discussion [25].

Guidelines warn against incentives or inducements to participate in research, particularly for youth [10]. Among the 51 studies included in the review, half (49% or 25/51) included participants of the age ≤ 18 years. Of these, more than half (19/25) did not report any information regarding compensation or incentives given to participants, two reported giving refreshments as incentive to participate [26, 27], one reported giving monetary compensation (Rupees 100 or USD $2) to participants (street children) for the loss of working hours [12], and 2/25 did not give compensation as the study was limited to secondary data analysis. Among the 51 studies, 17 studies included participants of age ≥ 18 years. About half (8/17) of them did not report any information regarding compensation or incentives. Among the remaining (9 studies) the vast majority (7/9) provided participants with monetary compensation (USD $2–$4), one provided reimbursement for transportation [23], and one reported giving ‘health kits’ to participants (FSW’s) that included soap, hand towels, sanitary napkins, toothpaste, toothbrush and a comb [28]. About 14% (7/51) included populations spanning the age group between 8 and 49 years, none of which reported information on incentives given to the participants.