Global Health Action (Dec 2023)

Exploratory validation of relationship functioning and non-pregnant partner behavior scales in pregnant people living with HIV in Mpumalanga Province, South Africa

  • Daniel E Sack,
  • Tshegofatso M Seabi,
  • Michael B Frisby,
  • Matthew A Diemer,
  • Godfrey H Ndlovu,
  • Ryan G Wagner,
  • Carolyn M Audet

DOI
https://doi.org/10.1080/16549716.2023.2210882
Journal volume & issue
Vol. 16, no. 1

Abstract

Read online

Partner engagement in antenatal care can improve care for pregnant people living with HIV. However, concerns about engaging unsupportive non-pregnant partners warrant further study to avoid engaging partners who pressure their pregnant partner to refuse testing or treatment and/or perpetuate intimate partner violence. We adapted established relationship functioning and partner behaviour questionnaires among pregnant people living with HIV initiating antenatal care in rural South Africa. We identified 13 previously validated psychometric scales with 255 items that assess relationship functioning and partner behaviour, but, to our knowledge, had not been used in Southern Africa. After item translation and cognitive interviewing with 30 pregnant people, we recruited an additional 208 pregnant people living with HIV receiving antenatal care. We conducted an exploratory factor analysis with maximum-likelihood extraction and oblique promax rotation with the 58 items and 10 scales that remained after translation and cognitive interviewing. We used parallel analysis, scree plots, and the Kaiser criterion to guide factor retention and assessed internal factor consistency via Cronbach’s alpha. Of the 208 participants recruited, 197 (95%) answered each question and were included in the analysis. Exploratory factor analysis revealed 7 factors that assessed partner social support, sexual relationship power, emotional intimacy, threatened or enacted violence, sexual intimacy, violence in relationships, and partner engagement in pregnancy care via 37 items. Factor absolute Spearman correlations ranged from 0.012 to 0.518 and Cronbach’s alpha ranged from 0.84 to 0.92. This preliminary analysis will guide further scale development. Future developments will also include relevant clinical outcomes to assess the predictive validity of the resulting measures. These steps will further refine these questions into a succinct screening tool to assess relationship functioning and partner behaviour. This screening tool may eventually guide the selection of partner-based interventions during pregnancy to improve outcomes for pregnant people and their partners.

Keywords