Reproductive coercion (RC) is a prevalent form of gender-based violence that interferes with female reproductive agency and is highly associated with intimate partner violence (IPV). Women’s self-help groups are a promising avenue to increase support and coping for violence, but little research exists to describe how survivors participate in women’s self-help groups nor how they may discuss RC within these groups. This study aims to quantitatively assess the potential of women’s-only self-help groups as a platform for reaching women in need of support for RC and IPV among a sample of reproductive-aged women seeking contraceptive services from private, community-based clinics in Nairobi, Kenya (n=659). To do this, we (1) described women’s use of self-help groups and (2) examined associations of women’s experience of RC and IPV with self-help group participation and discussion of RC within self-help groups. We also assessed differences in discussion of RC by group type (loans and savings clubs, charitable/social work, religious group, sports/hobbies, others). We found that more than half of women participated in self-help groups, with most of these women participating in economic self-help groups (63%), and that self-help group participation was not associated with prior experience of RC or IPV. Women who had ever experienced RC were more likely to report discussing RC within women’s self-help groups (AOR 1.82 95% CI 1.06-3.14), and RC discussion was less commonly reported in economic groups compared to charitable/social work groups.  Findings highlight that participation in and discussion of RC within women’s-only self-help groups is common and acceptable to survivors, and supports the need to integrate RC and IPV programming, particularly within the common, sustainable, and scalable structure of economic women’s self-help groups.

Data Source

This study utilizes cross-sectional data from women seeking family planning services from six private clinics in Nairobi, Kenya between July 2018 and December 2018, collected as part of the baseline questionnaire of the ARCHES (Addressing Reproductive Coercion in Health Settings) Kenya adaptation and evaluation study. ARCHES is a brief intervention delivered within routine family planning services to increase women's ability to cope with experiences of IPV and RC and improve family planning use. Details on the study protocol and design have been published elsewhere (Uysal, Carter, et al. 2020).

Briefly, the baseline survey included women aged 15 to 49 years attending study clinics who reported interest in receiving family planning services at their appointment that day and not being currently pregnant or sterilized, were able to communicate in English or Kiswahili, were planning to remain in the area for a period of six months (duration of the study) and had a phone that would allow them to be re-contacted for follow-up surveys. Women completed a baseline survey immediately prior to their appointment and a short exit interview immediately following their appointment before leaving the clinic; any women who reported not speaking with their provider about family planning during their appointment were excluded from the study. We included all women who provided baseline survey responses and reported any family planning discussion at their appointment at the exit in the present analysis.

Measures

We examined two aspects of self-help group involvement as our outcomes of interest, self-help group participation and discussion of RC topics within self-help groups. Self-help group participation was asked directly: “Do you attend any group meetings or activities that are for women only?” Among those who responded ‘yes’, they were asked how many times they attended this group in the prior three months. Women who indicated participation in a group were also asked the type of group they attend most frequently, with answer options: religious group, charitable/social work, sports/hobbies, loans/savings clubs, or other. Discussion of RC topics within self-help groups was assessed among women who responded that “yes” they attended group meetings or activities for women only. To assess discussion on RC topics participants were asked whether they discussed the following topics in any of the group meetings they attended within the past three months: 1) Male partners or in-laws that pressure or force women to stop using family planning (Family planning related RC), 2) Male partners or in-laws that pressure or force women to become pregnant (Pregnancy-related RC), 3) How to use family planning without a partner or other family member knowing (Covert family planning use). Responses were binary (yes/no). Women who indicated ‘don’t know’ were coded as ‘no’ for analyses (n selecting ‘don’t know’ <1%). We examined these items separately and as a combined ‘participated in any of the three discussion topics’ (yes/no) variable.

We examined participation in self-help groups and discussion of RC across two independent variables, ever and past-year experience of reproductive coercion (RC) victimization and intimate partner violence (IPV) victimization. RC was evaluated based on 11 items capturing contraceptive sabotage and pregnancy coercion (Uysal, Carter, et al. 2020). RC was indicated if the respondent indicated ‘yes’ to any item. Women who responded yes to any RC item were also asked when the last time a partner did any of these things to them was; we used this to determine past year experience. IPV was assessed as experience of physical or sexual violence, including whether a male partner had ever: “pushed you, shaken you, or thrown something at you,” “slapped you,” “twisted your arm or pulled your hair,” “hit you with his fist or with something that could hurt you,” “kicked you, dragged you, or beat you up,” “choked you or tried to burn you,” or “forced you to have sex or do something sexual with you when you didn’t want to.” IPV was indicated if the respondent indicated ‘yes’ to any item. As with RC, women who indicated every experience of any form of IPV were asked when the most recent experience was, from which past year experience was determined. Women missing information on RC or IPV items were excluded from the respective analyses.

Demographic characteristics included women’s age (years), education (primary or less, secondary, higher than secondary), literacy (yes/no), marital status (married or cohabitating with a partner, not married), parity (0, 1, 2+), past year employment (yes/no), and past 30 days food insecurity (yes/no).

Statistical Analyses

We first present descriptive statistics regarding demographics and outcomes of women’s self-help group participation ever, and within the past three months. Chi-square tests were used to evaluate differences in participation for categorical and binary measures, and t-tests (assuming unequal variances) were used for continuous measures. We then present descriptive statistics regarding the frequency of participation, group type, and discussion within groups. Next, we present unadjusted bivariate comparisons (using chi-squared tests) and unadjusted and adjusted mixed-effects logistic regression models to examine the relationship between women’s self-help group participation ever and outcomes (RC ever and past year, IPV ever and past year); these analyses utilize the full sample. Next, we repeat these unadjusted bivariate comparisons and unadjusted and adjusted mixed-effects logistic regression models with any discussion of RC in the prior three months as the predictor of interest; these analyses were limited to women who reported participating in self-help groups within the past three months. These analyses were conducted for discussion of any RC topic, as well as for each of the three specific topics. Adjusted models accounted for clinic-level similarities via clinic random intercept specifications, and controlled for age, education, literacy, marital status, parity, employment, and food security via fixed effects.  Finally, we assessed discussions of RC (any and specific topic) by women’s self-help group type, using bivariate chi-squared test comparisons; due to small numbers of sports/hobbies group participants (n=3) and ‘other’ type group participants (n=7), only religious groups, charitable/social work, and loans/savings clubs were included in these comparisons. All analyses were completed on non-missing data (complete cases); no imputation was performed.

Significance was set at p<0.05 for all comparisons and marginal significance at p<0.1; odds ratios (ORs) and adjusted odds ratios (AORs), along with 95 percent confidence intervals (CIs) are reported for regression results. All analyses were conducted using STATA 15.1.