Advances in neonatal care

Improvement of prenatal and postnatal care over the past decades has led to increased survival of very preterm neonates born less than 32 weeks’ gestation.1 Among others, protective measures to promote health and subsequent neurodevelopment have been developed, including optimisation of nutritional support, better characterisation of neonatal stress and improved pain management. In the same perspective, developmental care was introduced in the neonatal intensive care units (NICU) since the 1990s with the intention of minimising the adverse consequences of prematurity on the developing brain. During a critical period of development, the preterm brain is highly vulnerable to injury, represented by cerebral haemorrhage and insult to the white matter. Additionally, beyond the injury, preterm neonates are prone to alteration of brain maturation with disruption of normal developmental trajectory of both grey and white matters.2–4

Impact of prematurity on parents’ well-being

Preterm birth and caring for a preterm infant may be distressing for parents, who often feel vulnerable and incompetent in the high-tech NICU environment.23–25 Parents may present with difficulties in understanding and capturing subtle cues from their infant.26 Parents show important signs of stress,27 and require more support during the first year after the preterm birth compared with parents of term infants.28 They may also experience mental health symptoms, including post-traumatic stress disorder (PTSD),29–35 anxiety and depression.25 36

Although the hospitalisation of a preterm neonate may affect both parents equally,36 most of the studies examining parental emotional distress so far focused on mothers’ experience and needs.24 37 38 After birth, the mother normally initiates specific behaviours towards her newborn, aimed at supporting the neonate who experiences high levels of stress during hospitalisation in the NICU,39 and at fostering the infant’s socioemotional development.40 41 However, mothers of preterm infants may present difficulties in developing these protective behaviours.42 Thus, parental stress may interfere with the infant’s socioemotional and cognitive development, and is associated with more difficulties in building positive parent–infant relationships due to disrupted interactions.40 43 44 However, a recent meta-analysis showed that mothers of preterm children were not less sensitive or responsive toward their children than mothers of full-term children.45

Perceived parental self-efficacy

Perceived parental self-efficacy is defined as ‘beliefs or judgements a parent holds of their capabilities to organise and execute a set of tasks related to parenting a child’.46 Self-efficacy includes two separate notions: first, the belief of efficacy in caring for one’s own child across several varied domains of functioning and successful incarnation of the parental role (general self-efficacy),47–49 and second, the perception of one’s own abilities to complete a specified task within a specific domain (specific self-efficacy).50 The present study will focus on the specific self-efficacy, which appears to drive actions and predicts parents’ behaviours.48 51

As demonstrated in previous studies, perceived parental self-efficacy appears to mediate the relationship between psychosocial risk factors and maternal competences.52 53 Thus, a perception of low self-efficacy is associated with parental depression,52 54–57 high levels of parenting stress,58 59 low family support,60 poor infant health59 61 and demanding infant temperament.62 63 In contrast, a perception of high self-efficacy is associated with sensitive and receptive parental behaviour, and is related to improved infant socioemotional development.56 64

Parents of preterm neonates face a complex challenge. While they might be responsive to their infant cues, preterm neonates might not be capable of engaging in sustained and responsive interaction, as they tend to be less attentive and reactive due to immaturity, and to show more negative behaviours and emotions, as well as less rewarding interactions than their term-born peers.56 65–67 In parallel, mothers of preterm neonates, who are at risk of experiencing depression, anxiety or post-traumatic disorder,25 35 may not be able to interact as adequately with their child, and could be less sensitive than mothers without mental health symptoms. Mothers of preterm infants may be at a higher risk of decreased maternal confidence,68 although the limited evidence available so far is mixed.57 69 The quality of care provided by parents is strongly influenced by the maternal perception of self-efficacy, and interventions promoting this may therefore help to increase parenting quality.49 70

To date, only few early interventions have focused on enhancing perceived parental self-efficacy. The interventions that are currently available in the early neonatal period mainly aim to decrease parental trauma and stress-related symptoms, and to improve parental responsiveness within the parent–`infant interaction.10 23 71–73 Thus, a recent meta-analysis identified only two interventions intended to increase perceived maternal self-efficacy.74–76 These two interventions concentrated on different techniques of parenting education, and one of the two demonstrated improved cognitive child development at 4 months of age.76 The present study focuses on the joint observation, which is an interdisciplinary intervention performed in the NICU soon after birth. The main aim of the study is to examine whether this early intervention increases perceived parental self-efficacy.

Joint observation

The joint observation (JOIN: Joint Observation In Neonatology77) was developed in line with the three objectives of the developmental care model. This early intervention programme in the NICU is carried out by an interdisciplinary partnership of professionals, thereafter called observers, including NICU nurses, paediatricians, clinical child psychologists or child psychiatrists. They all received a 20-hour training, delivered by the same experienced clinical child psychologist (AB) for consistency, and participate in regular supervision sessions during the study period.

The intervention combines elements issued from four distinct theories of neonatal and infant development. First, the evaluation of neonatal behaviour developed by Brazelton and Nugent78 underlines the importance of parents detecting the neonate’s competences and fragilities, and interpreting stress cues to adjust to the infant’s regulation needs. Second, the synactive model of Als79 proposes a programme of individualised care avoiding overstimulation in the NICU and supporting the neonate’s self-regulation and competences. Third, the sensorimotor approach elaborated by Bullinger80 consists mainly of the assessment of sensory dystimulations provided to the neonate, and the management of subsequent tonico-postural disturbances observed during the NICU stay with the long-term perspective of optimising the infant’s development. This approach builds a framework to adjust the care procedures to the neonate’s capacity to treat multisensory information, and to reach a sensoritonic balance that allows and supports interactive behaviours. Fourth, the interactive guidance is a model based on the observation and analysis of parent–infant interactions through the therapeutic use of video feedback.81–83 This approach aims to allow parents to become aware of their competences and resources, as well as the skills and needs of their infant. Video feedback has recently been studied as an intervention in the NICU.84 85 The authors postulated that the interactive guidance through video feedback reduces the negative impact of preterm birth on the parent–infant relationship, and the behavioural withdrawal of the parent. The results of this previous work have revealed increased parental sensitivity and positive effects on the developing relationship.84 A randomised clinical study implementing video feedback not only during the NICU stay but also during the first year of life specifically demonstrated positive effects on parents of preterm infants with a lowering of mothers’ post-traumatic stress symptoms and enhancement in maternal sensitivity and quality of mother–infant interactions.86

In order to address the main objective, the present intervention is focusing on following areas: (1) The neonate’s adaptive capacity and competences are highlighted, as well as interactive signals, in order to promote parents’ emotional involvement, awareness of the infant’s perspective, resources and needs. For instance, the neonate’s interactive initiatives, as well as the responses to parental touch and/or voice, such as eye opening or head turning, will be highlighted. (2) To reinforce parental responsiveness, the parent’s behaviours that are supporting the neonate’s signals are pointed out during the video extracts, highlighting the parental relational competences frequently unidentified by the parents themselves. For example, positive emotional interactions between the mother and her neonate will be emphasised, such as adapting the voice and facial expression in order to support the neonate’s alertness. (3) With the aim of developing individualised care, measures can be suggested acknowledging the specificities of each neonate (sensorial irritability, tonico-postural disturbances or withdrawal for instance), and adjusting the care to reinforce the neonate’s own capacity of autoregulation and to support sensoritonic balance development in a long-term perspective. For instance, the parent’s gestures of support according to the baby’s tonico-postural needs will be identified, such as supporting the baby’s neck or pelvis during the interaction or adjusting the rhythm to help the neonate developing autoregulation competencies.

The joint observation pragmatically consists of two phases. First, a video-recorded period of routine care for the preterm neonate (such as a nappy change) is carried out by both the parent (mostly the mother) and a NICU nurse for a duration of approximately 30 min.

There is no intervention by the observers during the videotaping. Second, before gathering with the mother, several short extracts of the period of care are carefully selected by the observers in order to reach the objectives of the intervention. During the discussion and for illustration purpose, the observers will play back 4–6 short extracts of 10–30 s each to the parents, showing short specific moments of interactive behaviours that usually escape awareness. This video feedback, which lasts about 60 min, is conducted in order to point out the quality of the relational and emotional parent–infant interactions, and highlights moments of attunement, adjustment, synchrony, reciprocity and mutuality.

Aims of the present study

The objectives of the present study are to measure the effects of the joint observation as an early intervention performed in the NICU on outcomes relative to parental perception and mental health, as well as to indices of the parent–infant relationship quality and of child development. The primary outcome measure will be the perceived maternal self-efficacy. Secondary objectives will be to measure the impact of this intervention on maternal mental health (including perceived stress, post-traumatic stress, anxiety, depression), on maternal perception of the parent–infant relationship, on maternal responsiveness and on the neurodevelopment of the infant at 6 months corrected age (CA). In addition, acceptability of the intervention and maternal satisfaction will be assessed.

Study design

We will conduct a monocentric randomised controlled trial testing an intervention compared with treatment-as-usual, in the level III NICU of a Swiss University Hospital.

Study population, recruitment, group allocation and blinding

All mothers of preterm neonates born between 28 and 32 6/7 weeks of gestational age (GA), admitted to the NICU and aged less than 8 weeks of life are eligible to participate. Exclusion criteria were set for ethical considerations and in order to avoid approaching mothers needing acute treatment, and include the following: maternal age <18 years; established intellectual disability or psychotic illness; insufficient French-speaking level to complete questionnaires due to impossibility to obtain valid translations to multiple languages for financial reasons; and cardiorespiratory instability of the preterm neonate (severe brady-apnoea syndrome, oxygen requirement >30%) to ensure survival during the study period. Regarding twins or triplets, only the first-born neonate or the one being more stable will be included in the study.

Recruitment will be performed by the study nurses, who approach the eligible mothers once their infants are stable enough, that is, after the critical period of the first week of life when cardiorespiratory stability is established, usually on non-invasive ventilation and with oxygen requirement <30%, which would also permit more active participation of the parents in the neonate’s care. The allocation ratio of randomisation is 1:1, using a computer-generated list of random blocks (https://www.sealedenvelope.com/simple-randomiser). The allocation sequence will be concealed from the principal investigator in sequentially numbered, opaque, sealed envelopes. Envelopes will be opened only after the enrolled participants gave signed consent and completed all baseline assessments. The principal investigator and the statistician will be blind to group allocation. All participant data will be coded to ensure confidentiality.

Control group

Participants in the control group will receive treatment-as-usual. They will be asked to complete questionnaires at the same time points as the participants in the intervention group: at recruitment, at 1 month after enrolment and at 6 months of their infant’s CA. At 6 months CA, a neurodevelopmental assessment of the infant and a 10 min filmed mother– infant interaction will take place.

Intervention group

Mothers assigned to the intervention group will be asked to complete self-report questionnaires at the three time points described above. The intervention in the form of the joint observation will be planned after enrolment depending on the infant’s clinical state and stability. The intervention is twofold: first, the observers are jointly observing a period of care administered to the neonate jointly by her mother and a NICU nurse. The care procedure is video-taped and an observation grid87 is completed by the observers. Second, the mother and the NICU nurse participate in a video feedback session with the two observers. The discussion is based on the principles of interactive guidance,84 as described above. At the end of the intervention, the mother will also be asked to complete a questionnaire regarding her satisfaction with the intervention.

At 6 months CA, a neurodevelopmental assessment of the infant and a 10 min filmed mother–infant interaction will be carried out.

Primary outcome

The primary outcome is the difference in perceived maternal self-efficacy between the control and intervention groups measured with the Perceived Maternal Parenting Self-Efficacy questionnaire 1 month after study enrolment.

Secondary outcomes

Data collection and visits

After enrolment, mothers will be asked to complete several questionnaires described below, and again 1 month after study enrolment and at 6 months CA. Infants will return at 6 months CA to the neonatal follow-up clinic for the neurodevelopmental assessment and the 10 min filmed mother–infant interaction. The study measures and timings are summarised in table 1 and figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow chart of the study. CA, corrected age.

Self-report questionnaires

Demographic and perinatal characteristics

Mothers will also report demographic information, including socioeconomic status, level of education105 and previous psychiatric disorder. Neonatal characteristics will be collected from the medical record on severity of morbidity (GA and weight at birth, Apgar score, complications—need for mechanical ventilation and respiratory morbidity, sepsis and cerebral lesions), as well as the Clinical Risk Index for Babies,106 which represents neonatal morbidity severity.

Assessment of maternal sensitivity and responsiveness

Maternal sensitivity will be assessed at 6 months CA by coding a session of free play between the mother and her infant with the Emotional Availability Scales.107 Six domains are evaluated, of which four relate to the mother’s behaviour (sensitivity, structuration, intrusion and hostility toward the infant), and two to the infant’s behaviour (reactivity to the mother and maternal involvement). Patterns of interaction and emotional availability can therefore be measured on separate scales.108

A second tool, the Care Index,109 will measure maternal sensitivity, by assessing the interactive behaviour within the mother–infant dyad according to seven scales (facial expressions, vocalisations, posture, expressed affection, turn-taking, control and activity). Three maternal (sensitive, controlling, passive) and four infant (cooperative, compulsive-compliant, demanding, passive) characteristic behaviours are coded on each scale.

Neurodevelopmental assessment of the infant

A standardised neurodevelopmental assessment will be conducted at 6 months CA by a developmental paediatrician, using the BSID-III,110 which entails three subscales (cognitive, language and motor) with a normative mean score of 100±15 SD.

Sample size calculation

Power calculation (G*Power)111 based on published means and SD50 55 related to perceived parental self-efficacy in a sample of preterm (M=58.51, SD=12.57) and of term-born (M=65.9, SD=8.2) neonates showed that 68 participants would need to be recruited (α=0.05, 1-β=0.80, unilateral hypothesis). This is based on the assumption that parental self-efficacy in mothers of preterm babies in the intervention group who benefited from the intervention would be comparable to that in mothers of term babies. Therefore, it is planned that 80 mothers will be enrolled to anticipate possible participant withdrawal.

Statistical analyses

For the primary outcome regarding the difference in the perceived maternal self-efficacy between the intervention and control groups, linear regression analysis will be employed, with maternal self-efficacy at 1 month as the dependent variable and group as the explanatory variable, with adjustment for baseline maternal self-efficacy. For secondary analyses, linear mixed model regressions will be conducted, with maternal self-efficacy at 1 and 6 months as dependent variables and group, time and the interaction group x time as independent variables, adjusted for baseline maternal self-efficacy. The sample principle will be applied to all other secondary outcomes. We will include potential confounding variables, if necessary. These include maternal age, sex of the children and socioeconomic status where applicable. For confirmatory analyses, a Bonferroni correction for multiple analyses will be applied. For initial exploratory analyses, no such correction will be used.112

Differences between groups will be adjusted for the respective baseline values in case they differ using potential confounding variables, if necessary. These include maternal age, sex of children and socioeconomic status where applicable. Variables will be transformed if residuals are not normally distributed.

Patients and public involvement

In the early 2000s, the joint observation was introduced in our NICU by two professionals as part of the care of the parent–infant dyads.77 Due to positive feedback from the parents, the intervention was more routinely performed and systematised to the point that a randomised controlled trial was needed to examine its validity.

Although patients and caregivers’ feedback was considered in designing and adapting the intervention, they were not directly involved in the design or recruitment of this study. However, results will be disseminated in written form to the participants and distributed to the public via social media and public events.

Adverse events

Expected and unexpected adverse events will be recorded during the study period. As the intervention does not involve medical or pharmaceutical treatment, the risk that an adverse event would occur is low. A child psychiatrist will be available for clinical assessment and follow-up if needed, particularly if significant psychological distress or psychiatric illness of the mother or her infant is detected during study participation.

Data management

All study data will be coded and entered by research staff (psychology assistant). The database will be regularly updated by the IT Service of the Lausanne University Hospital. Double data entry will be done for the primary outcomes. For the rest of the data, a random 5% will be double-checked. The principal investigator, the coinvestigators and the statistician will have access to the final trial data set. Individual participant data collected during the trial (after deidentification) on which publications from JOIN consortium are based will be available on reasonable request.

Ethics and dissemination

The local ethical committee approved the study protocol. Little to no risk is expected by participation of the mothers and their neonates in the trial. Signed informed consent will be obtained from all participating mothers. Participation in the study will not interfere with the typical care patients receive after childbirth and during NICU stay. Results from the study will be disseminated at national and international conferences, and in peer-reviewed journals. This randomised controlled trial is registered in clinicaltrials.gov (NCT02736136, Pre-results stage).

Significance and outlook

This study might result in an evidence-based early intervention aimed at reinforcing parental competences, in particular at increasing perceived parental self-efficacy. It would represent a brief, easily accessible and safe early intervention, which could be implemented in the routine care in the NICU, thus leading to significant changes in clinical practice. It will also help to characterise the relationships between perceived parental self-efficacy, maternal mental health, maternal perception of their relationship with their infant and their infant’s temperament and maternal sensitivity.

Due to its interdisciplinary nature, this research is of interest for clinicians, educators and researchers in the field of paediatrics and development, psychology, child psychiatry and public health.