Translation and evaluation of the Perinatal Grief Intensity Scale in Kenya and Uganda

Stillbirth or neonatal death are distressing life experiences for parents. In many sub-Saharan African countries, bereavement care and support are inadequate. Context-appropriate interventions are needed to improve perinatal bereavement care. To assess whether these interventions work, reliable assessment tools are required. This study aimed to translate and evaluate the Perinatal Grief Intensity Scale [PGIS] for use among bereaved women in Kenya and Uganda.

A mixed-methods study, involving cross cultural translation, adaptation and psychometric evaluation in Kenya and Uganda. The PGIS was translated into Kiswahili and Luganda with input from community engagement groups and stakeholders. Pilot-testing was conducted in 2 stages; 'Think aloud' interviews (n = 16) and test-retest (n = 20). The finalised tools were tested with 160 women who had experienced stillbirth or neonatal death, 20 in the test-retest phase and 140 in wider testing. Analyses included test-retest reliability, internal consistency and exploratory associations with participant characteristics using single variable linear regression.

Agreed translations of the PGIS in Kiswahili and Luganda were produced. In pilot testing, test-retest consistency was high with an intraclass correlation coefficient of 0.840 (95% CI, 0.590-0.937). Wider testing demonstrated acceptable internal consistency, Cronbach's alpha of the total PGIS score was 0.737, confirming overall reliability, with similar results in the pilot testing and wider testing subgroups. Single variable linear regression showed that PGIS scores were higher for women in Uganda (coefficient 0.123, 95% Confidence Interval (CI) 0.042 to 0.205, p value = 0.003) compared to Kenya, and higher for caesarean births (coefficient 0.113, 95% CI 0.014 to 0.211, p value = 0.025) compared to vaginal births.

Preliminary evaluation of the Kiswahili and Luganda versions of the PGIS demonstrated good reliability for use in Kenya and Uganda. The tools have potential for future use in research to assess the impact of bereavement care interventions and in clinical practice to identify women who would benefit from increased bereavement support. Further assessments, including factor analyses and predictive performance are required to establish validity in this context.