Aims and hypotheses
The primary aim of this study is to evaluate the effectiveness of Video Interaction Guidance in parents of premature infants. We hypothesise that VIG enhances parental bonding after premature childbirth and prevents adverse parent-infant interaction.
The secondary aim is to further elucidate the bonding process between parents and their preterm infants. We hypothesise that the bonding process in parents with premature infants is delayed compared to parents with full term infants.
Research design
This is a multi-center randomised controlled trial (RCT) to examine the effectiveness of Video Interaction Guidance. The RCT will be conducted in the south of the Netherlands in 8 hospitals. In total 210 newborn infants with their parents will be included in this study. Healthy term infants (>37

weeks GA, n

=

70) and their parents and moderate term infants (32–37

weeks GA, n

=

70) and their parents will be recruited from maternity wards of 6 general hospitals. Extremely preterm infants or very low birth weight infants (<32

weeks GA, n

=

70) and their parents will be recruited by the NICU of 2 specialized hospitals The participant flow is displayed in Figure .
Recruitment
Parents will be recruited within 24 hours after delivery. They will be asked to participate in this study by a nurse/gynaecologist or pediatrician. The parents will be well informed through an information brochure and a letter about the aims, the implications of the study and the intervention in general (i.e. the amount of time and the home visits). When both parents agree to participate, they have to sign an informed consent form.
Randomisation
Once informed consent is obtained, the parents and newborn infants will be divided into 3 groups: a reference group, a control group or an intervention group. The reference group contains all healthy term infants (>37

weeks), whereas the moderate (32–37

weeks GA) and the extreme preterm infants (<32

weeks GA) will be randomly allocated to either the control or the intervention group. Randomisation will be arranged by using computer-generated random numbers. The allocation of the control or intervention group is performed by using sealed envelopes.
Trial procedures
All parents and their newborn infants will receive standardized hospital care after delivery (care as usual). Furthermore, depending on the group into which parents are divided, several interaction moments between the parents and their infant will be recorded on video during the first week after delivery. Parents from the reference group will be recorded on video once within the first 24 hours after delivery. Parents allocated to the control group, will be recorded on video twice within the first week, respectively at day 1 and 7. Parents allocated to the intervention group, will be recorded on video three times within the first week, respectively at day 1, 4 and 7. The parents from the intervention group will receive feedback on these recordings from the VIG- nurses, whereas parents from the reference and the control group will not receive feedback.
Measures
The data includes observations of parent-infant interactions, questionnaires, and semi-structured interviews. Data collection was started in September 2009 and will be completed in December 2011. The measurements from both parents will be conducted at day 1, 3 and 7 and month 1, 3 and 6. The time schedule of the measurements can be found in Table .
Primary outcomes
Parental bonding and parent-infant interactive behaviour are the primary outcomes of this study. To assess parental bonding the following tools will be used: the Pictorial Representation of Attachment Measure (PRAM) [
63], the My baby and I questionnaire [
64], the Postpartum Bonding Questionnaire (PBQ) [
65], the Yale Inventory of Parental Thoughts and Actions questionnaire (YIPTA) [
66,
67], the Mother And Baby Scale (MABS) [
68] and the Parental Bonding Instrument (PBI) [
69]. Interviews which are used are: the Clinical Interview for Parents of high-risk infants (CLIP) [
70] and the Working model of the Child Interview (WMCI) [
71]. Parent-infant interactive behaviour will be rated from videotapes which are recorded during interaction moments. To assess parent-infant interaction the Emotional Availability Scales (EAS) [
72] will be used. The EAS contains the following five subscales: sensitivity, structuring, intrusiveness, responsivity and involvement, measured on a 5 or 9 point Likert-scale.
Secondary outcomes
Parental secondary outcomes are
Stress, will be measured by the Traumatic Event Scale (TES) [
73], the Parental Stress Scale Neonatal Intensive Care Unit (PSS-NICU) [
74], the Perinatal PTSD Questionnaire (PPQ) [
75] and the Perceived Stress Scale (PSS) [
76].
Depression, will be measured by the Edinburg Postnatal Depression Scale (EPDS) [
77].
Anxiety: will be measured by the State-Trait Anxiety Inventory (STAI) [
78].
Anger: will be measured by the State-Trait Anger Expression Inventory (STAXI) [
79].
Satisfaction with hospital care and intervention: will be measured by the Nurse Parent Support Tool (NPST) [
80] and supplementary contentment questions about hospital care and if received: the intervention VIG.
Infant secondary outcomes are
Behaviour of the infant: will be measured by subscales of the Infant Behaviour Questionnaire Revised (IBQ-R) [
81] and the Ages and Stages Questionnaire: Social-Emotional (ASQ- SE) [
82].
Confounders
Parental confounders are
Various mood states: will be measured by the UWIST Mood Adjective Checklist (UMACL) [
83], the I Feel Pictures [
84] and supplementary questions about crying behaviour of the parents.
Coping: will be measured by the Ego Resilience scale (ER89) [
85], the Perceived Maternal Parenting Self-Efficacy questionnaire (PMP S-E) [
86], the Coping Inventory Stressful Situations questionnaire (CISS) [
87] the Parenting Sense of Competence Scale (PSOC) [
88] and the Soothing Methods Questionnaire [
89].
Personality: will be measured by the Quick Big Five (QBF) [
90].
Support: will be measured by the Family Assessment Measure III, subscale: Spousal Support Scale (FAM III) [
91] and the Relationship Questionnaire (RQ) [
92].
Background variables: additional (personal and medical) information.
Infant confounders are
Temperament: will be measured by the Infant Characteristics Questionnaire (ICQ) [
93].
Background variables: additional (medical) information
Long-term follow-up
In the Netherlands, with approximately 14.000 preterm births per year and an alarming number of young children victimized by maltreatment and neglect [
94] more insight into the process of parenting and potential risk or protective factors is badly needed [
2,
4,
5]. The purpose of the follow-up study is to increase understanding of the process of parenting a premature infant and the determinants of positive and negative parental and infant outcomes later in life. The primary aim of the follow-up study is to gain more insight into the process of parental bonding, attachment and parent-infant interactive behaviour after (premature) childbirth, in relation to the infant’s development at the age of two. Controlling for gestational age, it is expected that there will be a difference in social-emotional, behavioral and cognitive development between preterm infants and full term infants, in favor of the latter group. The secondary aim of the follow-up study is to examine the long term effects of a VIG, whereas it is expected that parent-infant dyads who received VIG have a better quality of attachment than dyads that did not receive the intervention. The final aim of the study is to assess the risk of abuse or neglect. Based upon literature the risk of infant maltreatment and neglect is expected to be higher in parents of preterm infants compared to parents of full term infants.
All families (N

=

210) participating in the first part of the study will be invited to take part in the follow-up study. Parents will receive a letter containing information about the follow-up study, which will be introduced as study of attachment and development of premature infants. A week after having received the letter, parents will be called and will be asked if they affirm to participate. Measurements will consist of questionnaires, interviews, computer tasks and video- recordings with both parents. Data collection will start in September 2011 till June 2013.
Sample size, power and statistical analysis
Sample size and power
A randomised controlled trial will be conducted to evaluate the effectiveness of VIG for parents with premature infants. In comparison with the control group (receiving CAU) it is expected that parents in the intervention group (receiving CAU and VIG) exhibit an enhanced quality in parent-infant interactions and they experience less bonding problems. Bonding problems will be measured with the Postpartum Bonding Questionnaire (PBQ) [
67] and the quality of the parent-infant interaction will be measured with the EAS [
72].
The sample size calculation is based on the identification of a minimal relevant clinical difference on the EAS subscales of 1.15-1.55 points. Considering a dropout rate of 20-30%, a standard deviation of 1.54 points, power of 80% and a significance level of 5%, an average sample size of 23 (17–29) in each group will be sufficient to detect a clinical relevant difference on the EAS subscales This number has been increased to 35 per group, to allow for the anticipated drop out during the first year.
Data analysis
Multilevel modelling will be performed [
95]; a flexible regression technique particularly suited for complex longitudinal datasets. To detect clinical significant differences and considering the maximum number of available pre-term infants, the inclusion of 70 infants per group is deemed sufficient. With approximately 100 (high-risk preterm) births a year in the separate specialised hospitals and approximately 150 low-risk premature and 2400 term births in the general hospitals per year, inclusion of 3 x 70 infants and their parents is feasible.
Ethical considerations
This study is funded by a Dutch foundation: “Achmea, Foundation Victim and Society”. Ethical and local feasibility approval was obtained for all eight participating hospitals. (Project number: NL24021.060.08, MEC approval: Catharina Hospital, Eindhoven, date: 21-10-2008). All parents will be provided with both verbal and written information about the study and written informed consent is obtained prior to enrolment in the trial.
A data supervising committee will monitor the progress of the study and the inclusion of the parents. No formal stopping rules are in place however funding will only be extended in case of sufficient participating families in the trial.