Antepartum and labour-related single predictors of non-participation, dropout and lost to follow up in a randomised controlled trial comparing internet-based cognitive–behaviour therapy with treatment as usual for women with negative birth experiences and/or post-traumatic stress following childbirth

Abstract

Objectives Internet-based interventions are often hampered by high dropout rates. The number of individuals who decline to participate or dropout are reported, but reasons for dropout are not. Identification of barriers to participation and predictors of dropout may help improve the efficacy of internet-based clinical trials. The aim was to investigate a large number of possible predictors for non-participation and dropout in a randomised controlled trial for women with a negative birth experience and/or post-traumatic stress following childbirth.Setting A childbirth clinic at a university hospital in Sweden.Participants The sample included 1523 women who gave birth between September 2013 and February 2018. All women who rated an overall negative birth experience on a Likert scale, and/or had an immediate caesarean section (CS), and/or severe postpartum haemorrhage (≥ 2000 mL) were eligible.Methods Demographic, antepartum, and labour-related/postpartum predictors were investigated for non-participation (eligible but denied participation), pre-treatment dropout (prior to intervention start), treatment dropout, and loss to follow-up. Descriptive statistics and logistic regression were used in the data analysis.Results A majority (80.3 %) were non-participants. Non-participation was predicted by lower level of education, being foreign-born, no experience of counselling for fear of childbirth, multiparity, vaginal delivery (vs CS and vacuum-assisted delivery) and absence of: preeclampsia, anal sphincter injury and intrapartum fetal distress. Pretreatment dropout was predicted by the absence of severe haemorrhage. Treatment dropout was predicted by vaginal delivery (vs immediate CS), vertex presentation and good overall birth experience. Loss to follow-up was predicted by vaginal delivery (vs immediate CS or vacuum-assisted delivery) and absence of intrapartum fetal distress.Conclusions Mothers with no obstetric complications were more likely to not participate and dropout at different time points. Both demographic, antepartum and obstetrical variables are important to attend to while designing procedures to maximise participation in internet-delivered cognitive–behavioral therapy.Trial registration number ISRCTN39318241