The impact of Cognitive Processing Therapy on stigma among survivors of sexual violence in eastern Democratic Republic of Congo: results from a cluster randomized controlled trial

Murray, S. M.; Augustinavicius, J.; Kaysen, D.; Rao, D.; Murray, L. K.; Wachter, K.; Annan, J.; Falb, K.; Bolton, P.; Bass, J. K.

doi:10.1186/s13031-018-0142-4

Conflict and Health

Table 3 Moderation of the effect of CPT on the mental health of women in Democratic Republic of Congo by stigma reported at treatment initiation, April 2011–February 2012 (n = 405)

From: The impact of Cognitive Processing Therapy on stigma among survivors of sexual violence in eastern Democratic Republic of Congo: results from a cluster randomized controlled trial

Outcome	Felt stigma^e b (SE) ^a	Felt stigma p-value	Enacted stigma^e (SE) ^a,b,c	Enacted stigma p-value
Depression and Anxiety ^d
Post intervention	−0.07 (0.17)	0.71	0.04 (0.02)	0.08
6 mo. post-intervention	0.01 (0.18)	0.94	0.03 (0.03)	0.36
Posttraumatic stress ^d
Post intervention	−0.18 (0.16)	0.27	0.02 (0.03)	0.56
6 mo. post-intervention	0.01 (0.20)	0.95	−0.004 (0.03)	0.88
Functioning^e
Post intervention	−0.17 (0.21)	0.42	−0.02 (0.03)	0.55
6 mo. post-intervention	0.18 (0.25)	0.48	−0.01 (0.04)	0.73

Note: CPT Cognitive Processing Therapy, SE standard error. ^ab is the beta-coefficient for the three-way interaction between treatment group (CPT or IS), assessment time point and moderating variable (felt-stigma or enacted stigma scale score) and represents the difference associated with a one unit increase in stigma score of the change in outcome over the time period between women in the CPT and IS arms. Estimated using longitudinal mixed-effect linear regression; random effects included participant, CPT group, and village. Model covariates were baseline age, marital status (currently married yes or no), currently pregnant, language, having lived in the current village for at least 10 years or less, total number of people living in the household, number of children for which the participant is responsible, and number of types of traumas experienced and witnessed. ^bSample size for the analysis of enacted stigma as a moderator is reduced to n = 383 due to some women missing more than 50% of the scale as one item was only relevant to married women (“rejected by husband”) and another to those with children (“forced to live away from your children”). ^cFor estimation of beta coefficients for the outcome of posttraumatic stress, 10 imputations were used instead of 11 due to a convergence failure in one imputed data set. ^dThe HTQ item “feeling detached of withdrawn from others” and the HSCL item “feelings of worthlessness, no value” were excluded from the average posttraumatic stress and combined depression and anxiety average scores respectively due to the inclusion of these items on the felt stigma scale. ^eLocally developed scale, based primarily on qualitative research with women in DRC

Back to article page

ISSN: 1752-1505

Contact us

Submission enquiries: journalsubmissions@springernature.com