A Risk & Safety Protocol to Mitigate Intimate Partner Violence Among Young South African Women Who Test HIV+ during Mobile HIV Counseling and Testing

Background:

South Africa has the worst HIV epidemic¹ and violence against women² with young women bearing the greatest burden of care: they are disparately affected by the HIV-IPV link³, less likely to engage in care⁴, and consequently suffer impaired health outcomes⁵. In response to the dearth of rigorously tested practices addressing the HIV-IPV intersection^6,3, the purpose of this study was to develop an HIV-IPV risk and safety protocol and determine is safety, effectiveness, feasibility and acceptability.

Methods:

This study was a Phase II randomized clinical trial over a risk and safety protocol with women experiencing IPV and recently testing HIV+ during mobile testing in South Africa. Participants were randomly assigned to one of three groups: the Standard of Care (SOC) or an Experimental group (Exp.) with two dosage levels. Those in the experimental group received a psycho-educational intervention and safety plan with the second dosage level receiving an additional safety plan.

Intention-to-treat analysis was conducted; independent sample t tests and Pearson’s Chi square tests were used to test for group differences; dimensionality and reliability analyses were done on the assessment instruments; and data were collected from June-December, 2015 in Gauteng province.

Results:

446 people were recruited, 255 met eligibility criteria, and 97.65% were retained by posttest. Research results focused on youth (< 34 years old; M= 26; n=142) as a sub-sample.

The Danger Indictor was developed to assess for IPV, violence vulnerability, and HIV-IPV risk.  A sufficiently unidimensional factor was extracted in an EFA and merits an overall summative danger indicator score with high reliability (Cronbach alpha= .882). An HIV-IPV risk instrument was also constructed with good reliability (Cronbach alpha=  .824). At a 30 day follow-up, there was no significant difference in participant safety scores (t(1,140)= -.710; p= .677) suggesting the experimental component did not compromise participants’ safety. Significantly more individuals in the SOC experienced violence upon partner notification of positive serostatus (X²₍₁₎ = 7.59, p=.016). Significantly more individuals in the Exp. group had used consultation strategies as part of a safety plan (X²₍₁₎ = 76.08, p=.000). About 96% reported it was helpful to be asked about their relationship difficulties, with 93.66% reported no regrets associated with participation. Although a slightly higher percentage of individuals younger than 34 years in the Exp. group  linked to care, compared to those in SOC, the difference was non-significant. Conversely, those aged 33-43 years had significantly more individuals in the Exp. group link to care.

Implication:

Study results show an overwhelmingly well-received protocol that was effective in reducing assault upon partner notification of serostatus, increasing the use of safety planning strategies and promoting safe linkage to care. These results will be used to inform a Phase III clinical trial on HIV-IPV with a focus on linkage to and maintenance in care. This study also contributes to the validation of two new psychometric instruments targeting the HIV-IPV intersection. Consistent with the literature, individuals aged ≤23 years had the lowest linkage to care rate, and future research should focus on improving outcomes among this population.