Evidence reviews for involving and supporting parents and carers

Overview of methods

A decision-analytic model in the form of a simple Markov model was constructed to evaluate the relative cost-effectiveness of interventions with a focus on parent/carer involvement in the care of preterm babies requiring respiratory support. The interventions assessed were NIDCAP^® (in addition to standard care) compared with standard care only. The choice of interventions assessed in the economic analysis was determined by the availability of respective clinical data included in the guideline systematic literature review. The economic analysis considered effective interventions, as demonstrated by the systematic review of clinical evidence. The study population comprised of preterm babies requiring respiratory care (<27 weeks’ gestation). Clinical data were derived from 2 studies included in the guideline systematic review of clinical evidence and other published literature.

The measure of outcome in the economic analysis was the number of quality-adjusted life years (QALYs) gained. The perspective of the analysis was that of NHS and PSS. Resource use and cost data was based on the published literature and where necessary supplemented with the committee’s expert opinion. National UK unit costs were used. The cost year was 2017. Two methods were employed for the analysis of input parameter data and the presentation of the results. First, a deterministic analysis was undertaken, where data were analysed as point estimates and results were presented in the form of incremental cost-effectiveness ratios (ICERs) following the principles of incremental analysis. A probabilistic analysis was subsequently performed in which most of the model input parameters were assigned probability distributions. Subsequently, 10,000 iterations were performed, each drawing random values out of the distributions fitted onto the model input parameters. Mean costs and QALYs for each treatment option were calculated by averaging across the 10,000 iterations. This approach allowed more comprehensive consideration of the uncertainty characterising the input parameters and captured the non-linearity characterising the economic model structure. Results of the probabilistic analysis were also summarised in the form of cost-effectiveness acceptability curves, which express the probability of NIDCAP^® being cost-effective at various at various cost-effectiveness thresholds. Various deterministic sensitivity analyses were undertaken to test the robustness of the conclusions. Sub-group analysis was undertaken to explore the cost-effectiveness of NIDCAP^® in preterm babies 27-34 weeks’ gestation. Also, a secondary analysis was undertaken where the cost-effectiveness of NIDCAP^® was explored using a wider public sector perspective.

Findings of the base-case economic analysis

According to deterministic analysis, from an NHS and PSS perspective in preterm babies <27 weeks’ gestation NIDCAP^® (in addition to standard care) was a cost-effective option with a cost per QALY of £14,380 (versus standard care) that is below the lower threshold of £20,000 per QALY. According to the deterministic sensitivity analyses, the results were sensitive to the risk ratio of neurodevelopmental problems for NIDCAP^® with a potential for the ICER to increase above the threshold of £30,000 per QALY when using the upper confidence interval value for the risk ratio of neurodevelopmental problems (cognitive domain). The results were also sensitive to the utility value for moderate neurodevelopmental problems with a potential for the ICER of NIDCAP^®to be above £20,000 per QALY threshold. The conclusions wree robust to changes in other model inputs including cost inputs and baseline rates.

The conclusions of the probabilistic analysis were similar to those of deterministic analysis. At the lower threshold of £20,000 per QALY (NICE, 2008b) the probability of NIDCAP^® (in addition to standard care) being cost-effective was 0.673 and it increased to 0.843 at the threshold of £30,000 per QALY. NIDCAP^® (in addition to standard care) became dominant in preterm babies <27 weeks’ gestation from a wider public sector perspective.

The results of the sub-group analysis indicated that from an NHS and PSS perspective and also a wider public sector perspective NIDCAP^® (in addition to standard care) was unlikely to be cost-effective in preterm babies >27 weeks’ gestation.

A threshold analysis was undertaken which indicated that for NIDCAP^® to be cost-effective in preterm babies >27 weeks’ gestation at the threshold of £20,000 per QALY the public sector costs per child with neurodevelopmental problems would need to be substantially higher than expected and as a result, NIDCAP^® is unlikely to be cost-effective in this sub-group of babies even from a wider public sector perspective. Although, the cost-effectiveness of NIDCAP^® in preterm babies >27 weeks’ gestation may be improved when condsidering a longer lifetime horizon. However, clinical and cost data was insufficient to inform such analysis.

Strengths and limitations

This analysis attempted to estimate the cost-effectiveness of NIDCAP^® (in addition to standard care) in preterm babies requiring respiratory support with clinical data from the guideline systematic review. Clinical data on NIDCAP^® was limited and focused only on the neurodevelopmental mental delay. However, the effectiveness of NIDCAP^® in terms of reduction in neurodevelopmental mental delay was judged by the committee to be very important. Due to the lack of suitable data the NIDCAP^® intervention cost was based on the committee expert opinion. Also, there was a lack of cost data in children with neurodevelopmental problems.

Critical outcomes

Important outcomes

Critical outcomes

Important outcomes

Critical outcomes

Important outcomes

Critical outcomes

Important outcomes

The outcomes that matter most

The committee agreed that the aims of involving parents and carers in caring for preterm babies on respiratory support were to reduce the length of hospital stay and the incidence of BPD, and to improve neurodevelopmental outcomes, and the committee therefore prioritised these as critical outcomes. The committee agreed that neurodevelopmental outcomes were the most important of these because of the life-long impact on the affected baby and their parents or carers.

The committee were keen to see if there was evidence that parent and carer involvement reduced Mortality prior to discharge and rates of sepsis and so these were chosen as important outcomes. Infant growth (defined as changes in z scores for weight, height or head circumference) was prioritised as an important outcome as this would be a more immediate marker of the potential benefit of the involvement of parents and carers in a baby’s care. Finally, parental satisfaction was chosen as an important outcome to determine if involvement was felt to be of benefit to the parents and carers too.

Evidence from RCTs was available for four of the interventions stipulated in the protocol - kangaroo or skin to skin care, non-nutritive sucking, FIC and NIDCAP^®. Evidence was assessed using GRADE methodology.

For kangaroo care or skin to skin care no evidence was found for the critical outcomes BPD and neurodevelopmental outcome and for the important outcomes infant growth and parent/carer satisfaction.

For non-nutritive sucking no evidence was found for the critical outcomes BPD and neurodevelopmental outcome and for the important outcomes sepsis, Mortality prior to discharge, infant growth and parent/carer satisfaction.

For FIC no evidence was found for the critical outcome neurodevelopmental outcome and for the important outcomes sepsis, infant growth and parent/carer satisfaction.

For NIDCAP^® no evidence was found for the important outcomes infant growth and parent/carer satisfaction.

The quality of the evidence

The quality of evidence ranged from moderate to very low. It was most often downgraded because of the uncertainty around the risk estimate or because of risk of bias introduced by a high risk of contamination across treatment groups. This was due to the fact that blinding of parents and NICU staff to treatment allocation was not feasible for the NIDCAP^® intervention. There was heterogeneity within some meta-analyses. In most cases subgroup analysis according to gestational age accounted for this heterogeneity but otherwise a random effects model was used.

Meta-analysis and stratification of results by gestational age was performed for some NIDCAP^® outcomes although stratification as specified in the protocol was not possible from the available data.

Due to the lack of data for some interventions and outcomes the committee made a research recommendation, prioritising the need for more evidence on the impact of Family Integrated Care and parental involvement as part of NIDCAP^® on length of hospital stay and BPD.

Benefits and harms

The evidence that was included for kangaroo care and skin to skin care was limited to two small RCTs, creating uncertainty around the risk estimate for the three relevant outcomes presented. However, in combination with their own experience, the committee were aware of a large body of RCT evidence from developing world settings (mainly South America) that had established the benefits of this type of care, and of evidence from observational studies conducted in UK settings examining surrogate outcomes such as heart rate. The committee were also aware of other studies examining the mechanisms underlying the benefits of care, for example, the positive effect of kangaroo care in reducing cortisol levels and raising oxytocin levels, which aided breastfeeding and improved babies’ digestion. Therefore, although some of this evidence was for a different population, and not for preterm babies receiving respiratory support in a NHS setting, the committee considered that there would be very little risk of harm associated with this practice and it would be difficult to justify a “no treatment” comparison group in further research on kangaroo or skin to skin care in preterm babies receiving respiratory care in UK hospital settings. The committee identified that in their clinical experience the only risk from kangaroo or skin to skin care was the risk of accidently extubating a baby, but that this was extremely rare. The committee therefore made a recommendation to support parents and carers by advising them about the potential benefits of interacting with their baby using practices such as skin-to-skin or kangaroo care.

While the review did not identify any evidence on verbal interaction, from their clinical knowledge, the committee were aware of the benefits of early communication for the development of the preterm baby’s hearing.

Evidence from one small RCT demonstrated that there was a clinically significant reduction in length of hospital stay when non-nutritive sucking was offered at the onset of nasogastric tube feeding. The committee made a recommendation to explain the benefits of non-nutritive sucking to parents based on the evidence, the physiological rationale of feeding reinforcing the sucking reflex as these actions are simultaneous, and because they believed there would be no associated harm. However, a weak recommendation was made to consider non-nutritive sucking opportunities in between feeds if the baby showed an interest in sucking. The committee believed this might also improve feeding, but the evidence did not demonstrate a parallel reduction in length of hospital stay and the physiological rationale did not directly support a stronger recommendation.

Evidence from a large cluster randomised trial demonstrated no additional benefit with FIC compared to standard care for two critical outcomes and one important outcome. The committee discussed the limitations of the contributing study and the feasibility for and impact on parents and carers of a commitment to participate in the care of their baby for 6 hours in the neonatal unit on a daily basis. Although the committee agreed with the principles forming the basis of FIC, they chose not to make a clinical recommendation believing that these principles underlie many of the clinical recommendations made across the guideline.

NIDCAP^® is an intervention comprising a detailed neurobehavioural observation of the baby with recommendations then made for individualised care and interaction based on the baby’s cues of challenge or competence. This is delivered by a neonatal professional extensively trained over two years in neuro-behaviour. Whilst the evidence did not demonstrate clear benefit for most of the outcomes that were prioritised for review, there was no evidence of harm – and benefit was demonstrated for one of the componemnts of the key outcome of neurodevelopment at 18 months and subsequent follow-up, when used with infants under 27 weeks. The committee acknowledged the considerable expense of NIDCAP^® training but also noted that few NIDCAP^® professionals are required per unit and that being part of a NIDCAP^® network or having access to a NIDCAP^® professional to ensure the use of the NIDCAP^® approach would have beneficial effects.

Friends and family

• High quality evidence from 3 qualitative studies carried out among fathers and parents of preterm infants requiring respiratory support in the NICU found that practical support, including meal preparation, assistance with household tasks, and child care, from friends and family assisted the parents in involving themselves with their preterm infant in the NICU. Parents also found that family and friends who were familiar with the NICU and demonstrated empathy and understanding of the parents’ anxieties reduced the stress over the burden of educating and reassuring those in the social support network who were not familiar with the situation.

Counselling

• Moderate quality evidence from 2 qualitative studies carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that an interdisciplinary NICU team with professionals who are able to provide psychological and spiritual support was valuable and some fathers utilised online chat rooms with similar parents in order to guide their involvement in their child’s care.

Partners

• High quality evidence from 6 qualitative studies carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that being able to talk about the NICU experience with their partner and developing a routine around caregiving activities supported parents in coping with having their infant in the NICU.

Facilitating parents in participating in care

• High quality evidence from 5 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that staff acted as gatekeepers to their participation in their infant’s care. Participating in ward rounds, hearing information about their child, and caring behaviour facilitate and support parents in becoming involved with their infant’s care.

Facilitating the transition into the parenting role

• High quality evidence from 7 qualitative studies carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that parents felt more confident transitioning into the parenting role when staff provided encouragement and the parents felt they had the freedom to care for their child with the staff present to help if needed. Staff who provided informal and formal training on providing care and who acted as role models that the parents could observe were also welcome supports.

Communication to reduce stress

• High quality evidence from 8 qualitative studies carried out among parents and mothers of preterm infants requiring respiratory support in the NICU found that communication with staff was crucial for developing a trusting relationship with staff and minimising parental anxiety. Elements such as using transparent communication methods to provide personalised information, family meetings to facilitate shared decision making, and regular phone updates when the parents are not in the NICU, assisted the parents and mothers to reduce stress. Parents need to feel that their beliefs and concerns are respected and that the information they receive is shared at the appropriate time and is not too medical.

Interpersonal relationships

• High quality evidence from 7 qualitative studies carried out among parents and mothers of preterm infants requiring respiratory support in the NICU found that feeling a sense of rapport with staff gave the parents both self-confidence in their parenting role and that their infant was being cared for well in the NICU. Parents found it beneficial when staff facilitated friendships with other parents and NICU graduate parents, through activities such as coffee hours or scrapbooking sessions, as enjoyed interacting with people whose child was or had been receiving the same care.

Continuity of care

• High quality evidence from 6 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that having continuity in the staff caring for their infant facilitated a sense of trust and confidence in the care the nurses were providing. Parents felt that lack of consistency in care meant that staff did not always know the infant and would have different opinions on the type of care that was needed. Parents felt supported by having a contact or designated nurse or doctor.

Shared experiences

• High quality evidence from 3 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that having a parent-buddy who spoke the same language, was from the same ethno-cultural background, and had the same experience with an infant in the NICU enabled them to communicate their feelings and concerns and understand the preterm birth experience. Engaging with other NICU parents helped parents to cope because it provided them with information and perspective.

Observational learning

• Moderate quality evidence from 1 qualitative study carried out among fathers of preterm infants requiring respiratory support in the NICU found that being able to watch other parents in open-spaced NICUs as they cared for their own infants helped them to become more involved with their infant.

Need for privacy

• High quality evidence from 5 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that the lack of privacy, noise, and business in the NICU prevented parents from engaging in skin-to-skin care and feeling comfortable expressing emotions.

Friendly, homelike environment

• Moderate quality evidence from 2 qualitative studies carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that allowing 24 hour visiting access and an NICU environment with décor and furniture that resembled a home environment facilitated involvement in their infant’s care.

Feelings of security or insecurity

• High quality evidence from 4 qualitative studies carried out among parents, fathers, and mothers of preterm infants requiring respiratory support in the NICU found that in order to feel secure in the NICU environment they had to understand the different medical equipment and monitors. An open-room design made some mothers feel safer and more secure as they were in close proximity to medical staff.

Participating in care

• Moderate quality evidence from 3 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that the presence of respiratory equipment and lines in the NICU environment highlighted the severity of their infant’s health condition and limited their involvement in nurturing their infant. The cultural environment of the NICU, including policies, restricted visiting hours, and prevention from joining in ward rounds, hindered parents from being able to engage with their infant.

The outcomes that matter most

The committee agreed that the support valued by parents or carers of preterm babies receiving respiratory support had thematic outcomes relating to social and psychological elements, parent-to-parent relationships, staff interactions, hospital environment and employment. All of these thematic outcomes were considered useful once the evidence had been appraised. These thematic outcomes reflect what service users value as the evidence was identified from interviews with parents themselves.

The committee prioritised psychological support for parents or carers of preterm babies receiving respiratory support as being of primary importance due to the large effect that poor mental health can have on both the wellbeing of the baby and the rest of the family. The committee noted that it would have been beneficial to have had more specific evidence regarding the type of psychological support and counselling that parents valued.

The quality of the evidence

Evidence was available from 15 qualitative studies, with 2 focusing on the perspective of mothers, 2 focusing on the perspective of fathers, and 11 focusing on the perspectives of both mothers and fathers. No studies were identified that investigated the perspectives of other carers of babies receiving respiratory support. Evidence was found for all of the thematic categories identified in the protocol. The quality of the evidence in this review ranged from low to high, but the majority of the evidence was moderate to high, which meant that the committee could make strong recommendations.

The quality of evidence was most often downgraded because of methodological limitations affecting the risk of bias, inadequacy of the evidence and relevance of the findings.

Methodological limitations affecting the risk of bias were generally attributed to some studies not clearly reporting the sampling method or relationship between the researcher and participants.

The confidence of the adequacy of the evidence was downgraded in some instances as a result of data saturation not being reached. In these instances, the themes were under-developed and analysing further data would likely reveal new data and concepts.

The confidence in the relevance of the findings was downgraded due to indirectness in the study population with some babies not being preterm.

Benefits and harms

In considering the evidence presented, the committee acknowledged the principles set out in the NICE Quality Standard 4 (QS4) on Specialist Neonatal Care. Quality statement 5 of this document covers ‘Encouraging parental involvement in care’ and states ‘Parents of babies receiving specialist neonatal care are encouraged and supported to be involved in planning and providing care for their baby, and regular communication with clinical staff occurs throughout the care pathway.’

The committee noted that there was evidence that parents valued having friends and family who were informed of the realities of having a preterm baby and who provided practical support, such as meal preparation and caring for older children. There was also evidence that parents valued psychological support and counselling, and although there was no specific evidence about who should deliver this care the committee agreed that it should be a qualified professional.

There was evidence that parents wanted to be supported by staff in caring for their baby, and this again was in-line with quality statement 5. There was evidence that parents value participating in ward round discussions about their baby, help transitioning into a parenting role and being recognised as partners in their baby’s care. Parents also expressed the need for clear, consistent, timely communication, the development of good interpersonal relationships and continuity of care.

The evidence showed that parents valued having the opportunity to engage with graduate parents of preterm babies or to have parent-buddies that could help them cope and understand the experience of being a parent of a preterm baby.

Finally, there was evidence regarding the hospital environment, showing that parents valued having 24-hour access to the neonatal unit, a homely environment with comfortable furniture and that private areas to facilitate skin-to-skin care and difficult conversations were required.

Employment support (such as paternity leave) was valued by parents but recommendations were not made in this area as any recommendations would be beyond the remit of this guideline and would rely instead on the parental leave policies of parents’ employers.

The evidence identified potential benefits of implementing support valued by parents or carers of preterm babies receiving respiratory support, including improving the parent’s experience and family relationships and better breastfeeding rates. Although the purpose of the review was to identify support valued by parents, it was also noted that improved support to parents had a beneficial effect on staff too, with fewer staff absences. The committee noted that in some units the professionals providing support to parents also provided support to the staff.

The committee identified several potential harms associated with implementing these recommendations, including issues of confidentiality arising with parents participating in ward rounds (and who may therefore be present on the ward when other babies are being discussed), conflict between staff and parents who have been given more decision-making power and feelings of exclusion by parents who are not able to visit their baby. However, overall the committee did not think these harms were a major problem.

The committee agreed that the benefits of implementing the support valued by parents outweighed the harms. The committee noted that there are solutions to the potential harms. For example, some units already give headphones to parents to maintain confidentiality during ward rounds, and parents who are not able to visit their baby (for example mothers who are too ill to attend the neonatal unit) can still receive updates on their baby’s care through phone calls from the medical team or by receiving videos or photos of their baby from nurses, although the committee recognised that this was not as good as participating in care by being present with their baby. Enabling and supporting parents to participate in their baby’s care, and fostering a culture where parents are regarded as partners in their baby care, is key to reducing conflict and tension between parents and staff.

While there was evidence that parents and carers expressed the need for maintaining continuity amongst the health care professionals caring for their baby, the committee did not make a recommendation based on this evidence because they did not think that such a recommendation could be implemented given ongoing staff turnover.

Prenatal maternal and infant health

• Low quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that parents were given information, including morbidity and mortality for preterm infants born at different gestational ages. However, parents wanted more specific information on the treatments their infants would likely need after delivery.

Postnatal information

• Low quality evidence from 3 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that staff provided the most information at the beginning of the infant’s hospitalisation, but parents would have liked a delayed postnatal review of what happened prenatally and during the birth, as many mothers were still recovering from the birth when they received the majority of the information.

Understanding the infant’s medical condition

• High quality evidence from 3 qualitative studies carried out among fathers and parents of preterm infants requiring respiratory support in the NICU found that understanding their infant’s medical condition and care was crucial. However, parents found that staff did not always fully explain complex medical issues or would leave parents waiting for information about their infant’s illness, which caused them anxiety.

Receiving updates of the infant’s health status

• High quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that parents appreciated receiving clear information about their infant’s health status immediately after exam results or tests. Mothers did not like when they had to receive information from their husbands and would have preferred to receive updates from a physician.

Parenting activities

• High quality evidence from 6 qualitative studies carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that nurses were crucial in providing information in regards to caregiving practices, such as feeding and nappy changes. Informal and formal training provided by patient staff assisted parents in developing the confidence to participate in their child’s care.

Changes in care

• Moderate quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that parents insisted on receiving information in regards to changes in the infant’s medical treatment, such as changes in intubation, catheter, and location in the hospital. Parents preferred to receive this information from the neonatologist as opposed to the nurse.

Understanding behavioural cues

• Moderate quality evidence from 1 qualitative study carried out parents of preterm infants requiring respiratory support in the NICU found that mothers, more often than fathers, wanted explanations of the infant’s reactions and behaviours.

Breast feeding

• Low quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that mothers perceived information provided in breast-feeding programs as useful, as it helped them make decisions in regards to feeding their infant.

Skin to skin care

• Moderate quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that parents were reluctant and lacked confidence to engage in skin to skin care when nurses did not provide them with information on how to hold and transfer the infant without dislodging tubes and ventilator equipment.

Plans to have children in the future

• Low quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that one mother whose infant had died wanted more information on the cause of death and advice for pregnancies in the future. Mothers who knew someone who had an extremely premature infant who survived found that this information gave them hope for their child.

Decision making

• Low evidence from 2 qualitative studies carried out among fathers and parents of preterm infants requiring respiratory support in the NICU found that staff sharing information and providing opportunities to ask questions facilitated parents becoming involved in decision-making about the infant’s care. Adequate and clear information enabled parents to feel confident when physicians asked them to make a decision about their infant’s care.

Telephone

• Moderate quality evidence from 2 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that regular and ritualised phone calls were appreciated. Parents reported feeling reassured and linked to their child by receiving regular phone calls when they were at home and the infant was still in the neonatal unit. In contrast, receiving routine information at home through an unexpected phone call caused alarm, as it was assumed that an unplanned call was linked to bad news.

Medical team (member not specified)

• Moderate quality evidence from 2 qualitative studies study carried out among parents and fathers of preterm infants requiring respiratory support in the NICU found that information should be shared by staff members who are adequately trained to provide tailored medical information that is tailored to their emotional needs and technical knowledge and who provide parents with the opportunity to ask questions and recommend additional resources.

Nurses

• Low quality evidence from 2 qualitative studies carried out among parents and mothers of preterm infants requiring respiratory support in the NICU found that nurses assisted parents in understanding complex medical concepts and reduced feelings of anxiety. Due to nurses’ regular interactions with the infant, parents felt that primary nurses were most adept at providing day-to-day information and was the best source of information about changes in their baby’s medical condition.

Physicians or neonatologists

• Low quality evidence from 2 qualitative studies carried out among parents and mothers of preterm infants requiring respiratory support in the NICU found that the neonatologist was the preferred source of information for technical or complex information, even if parents required additional explanations from nurses afterwards. Physicians should provide as much information as is required to convey the complexities of the situation and allow the parents to ask as many questions as needed.

Timing and consistency

• High quality evidence from 4 qualitative studies carried out among parents of preterm infants requiring respiratory support in the NICU found that parents, especially mothers, struggled to absorb and understand information that was shared with them during prenatal consultations when they learned their infant would be premature. Many parents were overwhelmed by the amount of information they received during this emotional experience, which later prevented them from being able to recall information. Parents stated that their preferred time to receive information would be during clinical rounds as opposed to during the prenatal consultation, immediately after delivery or before discharge. Parents would be interested in receiving information at a time separate from rounds. Additionally, it is crucial for parents to receive honest information that is shared consistently by all the members of the care team to avoid having parents receive confusing and varying messages.

Other resources (including books, internet resources, friends and family)

• Moderate quality evidence from 1 qualitative study carried out among parents of preterm infants requiring respiratory support in the NICU found that the majority of parents received information from staff and the medical care team, although sources such as printed materials, friends and family or the internet were also consulted.

The outcomes that matter most

The committee agreed that the information valued by parents or carers of preterm babies receiving respiratory support had thematic outcomes relating to prenatal and postnatal information, caregiving information, information about the baby’s health status, information for the future and understanding the neonatal unit environment. In addition, the committee agreed that it was important to know the preferred format of this information. All of these thematic outcomes were considered useful once the evidence had been appraised. These thematic outcomes reflect what service users value, as the evidence was identified from interviews with parents themselves.

The committee prioritised the consistency, clarity and timely nature of information, as the evidence highlighted the importance of the adequate pacing of information, regardless of the type of information.

The committee noted that there was no evidence on formats of information using modern technology, such as apps, online resources or Facebook groups that many parents may utilise. However it was decided this was not a priority for a research recommendation.

The quality of the evidence

Evidence was available from 10 qualitative studies, with 2 focusing on the perspective of fathers, and 8 focusing on the perspectives of both mothers and fathers. No studies investigated the perspective of other caregivers responsible for babies requiring respiratory support. Evidence was not found for all of the thematic categories identified in the protocol, specifically formats including print, online resources or technology. The quality of the evidence in this review ranged from low to high, but the majority of the evidence was moderate to high, which meant that the committee could make strong recommendations.

The quality of evidence was most often downgraded because of methodological limitations affecting the risk of bias, inadequacy of the evidence, and relevance of the findings.

Methodological limitations affecting the risk of bias were generally attributed to some studies not clearly reporting the sampling method or relationship between the researcher and participants.

The confidence of the adequacy of the evidence was downgraded in some instances as a result of data saturation not being reached. In these instances, the themes were under-developed and analysing further data would likely reveal new data and concepts.

The confidence in the relevance of the findings was downgraded due to indirectness in the study population with some babies not being preterm.

Benefits and harms

The evidence showed that parents valued information on prenatal health issues (for both mother and baby) which included the likely morbidity and mortality at different gestational ages. However, parents had difficulty understanding some information about their baby’s prognosis and the birth of the baby if information was provided immediately after stressful events or if it was not given in a clear manner. The timing of information giving was therefore important, with some parents wanting it deferred in such circumstances.

Parents valued information about their baby’s medical condition and their medical care, and this should be provided in a timely fashion and delayed only where circumstances demanded it.

Parents valued information about how to care for their baby, as well as how to interpret their baby’s behavioural cues. Parents also valued information on breastfeeding and on providing skin-to-skin contact as part of the baby’s care, and information from the nurses on these areas greatly increased their confidence and willingness to be involved in their baby’s care. Parents valued receiving information that was tailored to their needs and that was delivered by the appropriate member of staff, with some parents valuing more technical information about changes in care coming from medical staff rather than nursing staff. Parents valued consistent information, and the committee agreed that it was important that whoever was delivering the information (nurse, doctor or other healthcare professional) should deliver it clearly and check for the parents’ understanding, and not rely on the information having to be re-explained by another member of the team later. Staff should however be aware that it may take time for parents to absorb information and they may require repeated encouragement before becoming confident in caring for their baby with reduced input from staff. The evidence also described how having an understanding of the medical equipment (for example, its purposes, and what alarms and buzzers meant) and being able to ask questions regarding their baby’s health and care enabled parents to become comfortable caring for their baby.

Parents valued information regarding the future (such as hereditary issues) and having sufficient information to be involved in decision-making.

The evidence addressed the importance parents placed on the format, timing and consistency of information provided, and the committee discussed the need for the medical and nursing teams to agree with parents the method of delivery and frequency of information. The evidence demonstrated that parents value information that is consistent between healthcare professionals so as to avoid confusion and mistrust.

Due to the complexity of medical information, the committee agreed that information shared with parents and carers should where appropriate be followed-up by high-quality written and online resources, and that parents are aware of key contacts on the neonatal unit.

The committee agreed that the potential benefits of the recommendations would include more accurate and consistent information, enabling parents and carers to feel more confident and improving relationships between staff and parents/carers.

The committee did not identify any harms related to these recommendations.

The committee discussed the value placed by parents on information for the future (such as hereditary issues) but did not make any recommendations as they felt this may require specialist information provision, would be on a case-by-case basis, and did not apply to the majority of babies requiring respiratory support.

Interventions assessed

The choice of treatments assessed in the economic analysis was determined by the availability of respective clinical data included in the guideline systematic literature review. The economic analysis considered effective treatments, as demonstrated by the systematic review of clinical evidence. The committee explained that interventions such as kangaroo care and skin to skin contact, and non-nutritive sucking have negligible intervention costs, and also clinical data was very limited for these interventions. According to the committee expert opinion, NIDCAP^® is the only intervention that is associated with high intervention costs and therefore should be pursued in the economic evaluation. NIDCAP^® model postulates that an understanding of the neurodevelopmental expectations of the preterm as expressed in the infant’s behavior will provide a reliable basis for the examination, and adaptation of traditionally delivered newborn intensive care, including a realignment of the parent and carer involvement (Als 2011). The model considered standard care treatment as a comparator.

Model structure

A simple Markov model was constructed using Microsoft Office Excel 2013. The structure of the model was determined by the availability of clinical data. According to the model structure, hypothetical cohorts of 100 babies born preterm (<27 weeks’ gestation) requiring respiratory care were initiated on either NIDCAP^®(in addition to standard care) or standard care only. Across the report NIDCAP^® will refer to NIDCAP^®in addition to standard care.

Babies initiated on NIDCAP^® were assumed to have continuous involvement from their NIDCAP^® professional for the duration of the initial hospital stay. The model included the following health states: ‘well’, ‘moderate neurodevelopmental problems’, and ‘severe neurodevelopmental problems’. The model included yearly cycles. At the end of each cycle a baby could remain in the ‘well’ state, move to ‘moderate neurodevelopmental problems’ state, or the ‘severe neurodevelopmental problems’ state. According to the committee expert opinion once a baby is in either the moderate or severe neurodevelopmental problems state they will remain in that health state for the duration of the model (that is, there are no transitions between moderate and severe states). In the model neurodevelopmental problems were defined as neurodevelopmental mental delay.

The half-cycle correction was applied in the Markov model to compensate for the fact that transitions between states, in reality, occur in the middle of each cycle on average.

Given the lack of long term clinical and cost data the time horizon of the analysis was 18 years. A schematic diagram of the model is presented in Figure 10.

Figure 10. Schematic diagram of the Markov model constructed for the assessment of the relative cost-effectiveness of NIDCAP® for preterm babies requiring respiratory support

Clinical input parameters and overview of methods employed for evidence synthesis

Clinical input parameters consisted of the risk ratio of developing moderate or severe neurodevelopmental problems with NIDCAP^®versus standard care. The guideline meta-analysis identified 2 RCTs assessing NIDCAP^® versus standard care that provided efficacy data (that is, moderate or severe neurodevelopmental problems). Both Peters 2009 (n =101) and Maguire 2009b (n=139) reported data at approximately 2 year follow-up. In both studies neurodevelopmental problems were assessed using Bayley Scales of Infant Development (BSID) II Mental Developmental Index (MDI).

Other clinical input parameters included the absolute risk of moderate or severe neurodevelopmental problems associated with standard care. The committee identified 1 UK-based prospective cohort study (Moore 2012) that provided the number of babies developing moderate or severe neurodevelopmental problems when using standard care treatment. In this study a community-based cohort of surviving babies (n=1,031) born in 2006 before 27 completed weeks of gestation was studied prospectively over a 3 year period. The study reported neurodevelopment disability on different domains including motor, hearing, vision, cognition, and communication. Neurodevelopmental outcomes were assessed in 576 preterm babies, with 501 of babies assessed using the BSID III, 39 using the Wechsler preschool and primary scales of intelligence, and 10 using only the cognitive scale of BSID III. The effectiveness review identified statistically significant effect of NIDCAP^®only on the MDI subdomain of the BSID II scale. The committee explained that BSID II MDI subdomain evaluates sensory-perception, knowledge, memory, problem solving, and early language. Thus, BSID II MDI measures a combination of early cognitive and language development. Based on the above the committee concluded that BSID II MDI equates most closely with the cognitive function subdomain reported in Moore 2012. For the purposes of modelling a 3-year cumulative probabilities reported in Moore 2012 were used to estimate annual probabilities of developing moderate or severe neurodevelopmental problems (on a cognitive function subdomain), assuming exponential function, which were subsequently attached to the standard care treatment.

Given the lack of longer term data the efficacy data was applied only over 2 years. The absolute risk of neurodevelopmental problems (mental delay) associated with NIDCAP^® was estimated by multiplying the respective relative risk by the baseline risks of moderate or severe neurodevelopmental problems (cognitive function subdomain) as calculated for standard care.

In the model the final membership in each health state in year 2 was carried over to the subsequent years for the duration of the model to estimate associated costs and outcomes.

The mortality was not considered in this analysis since, as indicated by the clinical review, there is no clinically significant difference in mortality with NIDCAP^® compared to standard care for preterm babies requiring respiratory care.

Utility data and estimation of QALYs

In order to express outcomes in the form of QALYs, the health states of the economic model needed to be linked to appropriate utility scores. Utility scores represent the health-related quality of life (HRQoL) associated with specific health states on a scale from 0 (death) to 1 (perfect health); they are estimated using preference-based measures that capture people’s preferences on the HRQoL experienced in the health states under consideration.

NICE recommends the EuroQol five dimensions questionnaire (EQ-5D) (Brooks, 1996) as the preferred measure of HRQoL in adults for use in cost-utility analysis. The standard version of the EQ-5D has not been designed for use in children. As a result an alternative standardised and validated preference-based measures of health-related quality of life that have been designed specifically for use in children can be considered (NICE, 2013).

Petrou (2013) estimated utility scores associated with neurodevelopment impairment using parents’ ratings of their children’s HRQoL around the child’s eleventh birthday on both the HUI2 and HUI3. The HUI is a family of preference-based multi-attribute utility measures (Torrance 1995). The HUI2 consists of 6 domains: sensation, mobility, emotion, cognition, self-care, and pain. A seventh domain of fertility can be added if relevant. The HUI3 health state classification has many similarities to the HUI2, but with the sensation domain expanded into 3 separate attributes of vision, hearing and speech, and additional response levels added to some domains.

Responses to HUI3 can be converted into utility scores using a published algorithm that was developed based on the principles of multi-attribute utility theory, following a valuation survey of members of the general population in Canada; respondents’ preferences were elicited using visual analogue scale and standard gamble (Feeny 2002).

In the analysis, HUI2 scores were used since unlike HUI3 it has an underpinning multi-attribute utility scoring algorithm that has been estimated on the basis of the preferences of members of the UK general population with respondents’ preferences elicited using standard gamble (McCabe 2005; Petrou & Kupek, 2009), which is a method recommended by NICE.

Cost data

Intervention cost for NIDCAP^® was calculated by combining resource use estimates with respective national unit costs. Intervention cost consisted of NIDCAP^® professionals’ time. The cost of a NIDCAP^® professional’s time was estimated by combining the mean total NIDCAP^® professional’s time per child treated, as advised by the committee expert opinion, with the national unit cost of a Band 7 hospital nurse (Curtis & Burns, 2017). According to the committee’s expert opinion, all babies receiving NIDCAP^® would have an initial observation that combined with the report preparation would take approximately 1 day. It was further explained that the NIDCAP^® professional would spend additional 3 hours per week for the duration of the initial hospital stay to assist with the implementation of the recommendations in the report which is prepared by the NIDCAP^® professional. This follow-up involvement supports the family and team with the integration of recommendations and adapting these to suit the baby’s changing developmental needs.

The duration of initial hospital stay was obtained from a recent study by Seaton (2018). In the study the authors predicted the length of stay in neonatal care for all admissions of singleton babies born at 24–31 weeks’ gestation from 2011 to 2014. Data were extracted from the National Neonatal Research Database in the UK. A total of 20,571 preterm babies were included. In the study the median length of stay was reported for each gestational age. Using the reported data a weighted average length of stay was calculated to estimate the duration of length of stay for the average preterm baby born 22-26 weeks’ gestation. The estimated duration of length of stay was used to approximate the duration of NIDCAP^® professional involvement beyond the initial observation.

The unit cost of a hospital nurse per hour of client contact was estimated based on the mean full time equivalent basic salary for Agenda for Change Band 7of the July 2016-June 2017 NHS Staff Earnings estimates, including salary, salary oncosts and overheads. The apportioned qualification costs per hour of contact were negligible and were not considered.

The intervention cost of standard care was zero given that it was administered in both arms.

The health and social care costs incurred by children with neurodevelopment problems were obtained from Petrou (2013). Like for health related quality of life, economic costs were extracted from detailed postal questionnaires completed by the main parent around the child’s 11th birthday asking about resource use over the previous year. The economic costs were estimated from an NHS and PSS perspective and included hospital inpatient care, hospital outpatient and day care, community health and social care, drugs and medications. The resource use estimates were combined with appropriate unit costs taken from national sources in order to estimate an overall annual health and social care cost incurred by children with moderate or severe neurodevelopment problems. Petrou (2013) also reported cost data for children in the control group (that is, school classmates who were born at full term and matched for age, sex and ethnic group). The above costs were used to estimate incremental NHS and PSS costs in children who are in the ‘moderate’ or ‘severe’ neurodevelopmental problems health state, respectively.

Petrou (2013) also estimated incremental public sector costs (inclusive of education costs), during the 11^th year of life for children but only for children with severe neurodevelopmental problems. The cost categories included in the public sector costs besides education costs were not reported. Given the lack of public sector costs in children with moderate neurodevelopmental problems, a ratio of incremental public sector costs to health and social care costs was estimated using cost data for children with severe neurodevelopmental problems. The resulting ratio was applied to health and social care costs for children with moderate neurodevelopmental problems to approximate costs from public sector costs in this population. The committee explained that the costs associated with neurodevelopmental problems are likely to be higher once the child starts school. As a result, in the secondary analysis, NHS and PSS costs were included up to the age of 5 years and wider public sector costs (inclusive of education costs) were applied at 5 years onwards for the duration of the model.

The analysis considered only costs associated with neurodevelopmental problems and did not include costs associated with children who are in the ‘well’ health state.

All costs were uplifted to 2016/17 prices using the hospital and community health services inflation index (Curtis & Burns, 2017).

Table 22 reports the mean (deterministic) values of all input parameters used in the economic model and provides information on the distributions assigned to specific parameters in probabilistic sensitivity analysis.

Table 22. Input parameters used in the economic model of NIDCAP® for preterm babies requiring respiratory support

Data analysis and presentation of the results

Two methods were employed to analyse the input parameter data and present the results of the economic analysis.

First, a deterministic analysis was undertaken, where data are analysed as point estimates; results are presented as mean total costs and QALYs associated with each treatment option are assessed. Relative cost-effectiveness between alternative treatments was estimated using incremental analysis: all options were ranked from most to least cost-effective. Options that were dominated by absolute dominance (that is, they were less effective and more costly than one or more other options) or by extended dominance (that is, they were less effective and more costly than a linear combination of two alternative options) were excluded from further analysis. Subsequently, incremental cost-effectiveness ratios (ICERs) were calculated for all pairs of consecutive options remaining in the analysis.

ICERs expressed the additional cost per additional unit of benefit associated with one treatment option relative to its comparator. Estimation of such a ratio allowed consideration of whether the additional benefits were worth the additional cost when choosing one treatment option over another.

The treatment option with the highest ICER below the cost-effectiveness threshold was deemed to be the most cost-effective option.

One-way sensitivity analyses explored impact of varying:

• the risk ratio estimate (using upper and lower CI);
• the baseline risk estimates (±20% around the base-case value);
• the utility values (±20% around the base-case value);
• the intervention cost (±50% around the base-case value);
• the costs of neurodevelopment problems (±50% around the base-case value).

In addition to deterministic analysis, a probabilistic analysis was also conducted.

In this case, all model input parameters were assigned probability distributions (rather than being expressed as point estimates), to reflect the uncertainty characterising the available clinical and cost data. Subsequently, 10,000 iterations were performed, each drawing random values out of the distributions fitted onto the model input parameters. This exercise provided more accurate estimates of mean costs and benefits for each intervention assessed (averaging results from the 10,000 iterations), by capturing the non-linearity characterising the economic model structure (Briggs 2006).

The relative risk estimates were given a log-normal distribution. The baseline risk estimates of neurodevelopment problems and utility values were assigned a beta distribution. Costs were assigned a gamma distribution. Where standard error estimate was not available the assumption was made that costs had a standard error of 20% of their mean value.

Results of probabilistic analysis were presented in the form of cost-effectiveness acceptability curves (CEACs), which demonstrated the probability of each treatment option being the most cost effective among the strategies assessed at various cost-effectiveness thresholds.

Sub-group analyses

A recent cohort study in France by Pierrat (2017) looked at neurodevelopment outcomes at 2 years for preterm children born between 22 to 34 weeks’ gestation. The study found that among live births survival at 2 years corrected age without severe or moderate neuromotor and sensory disabilities was 48.5%, 90.0%, and 97.5% at 22-26, 27-31, and 32-34 weeks’ gestation, respectively. Consequently, an exploratory sub-group analysis was undertaken where the base-case probabilities of neurodevelopment problems were reduced by 90.0% and 97.5% to estimate the potential cost-effectiveness of NIDCAP^® in preterm babies who are 27-31 and 32-34 weeks’ gestation, respectively.

In this analysis the duration of initial hospital admission was recalculated based on Seaton (2018) and was estimated to be 50 days for a preterm baby >26 weeks’ gestation. This, in effect, reduced NIDCAP^® professional involvement from 105 days to 50 days and resulted in the NIDCAP^® intervention cost of £1,618 per baby.

Only the deterministic results were calculated for the alternative base-case rates generated using different gestational ages.

The cost-effectiveness of NIDCAP^® in different sub-groups was estimated from both an NHS and PSS perspective and also from a wider public sector perspective that included education costs.

Economic modelling results

Discussion – limitations of the analysis

The results of the economic analysis suggested that NIDCAP^® for parent and carer involvement was likely to be a cost-effective treatment for preterm children <27 weeks’ gestation who are receiving respiratory support. NIDCAP^® resulted in an ICER that was below the threshold of £20,000 per QALY. The probability of NIDCAP^® being cost-effective was 0.673 at a threshold of £20,000 per QALY. The cost effectiveness of NIDCAP^® in preterm children <27 weeks’ gestation was attributed to a number of factors: relatively high baseline risk of neurodevelopment problems in this population, high costs and health related quality of life decrements associated with neurodevelopment problems.

The clinical review searched for evidence on a wider set of neurodevelopmental outcomes, in the economic analysis the clinical data for the effectiveness of NIDCAP^® was based on only 2 studies (n=240) focusing on neurodevelopmental mental delay since this was the only statistically significant finding which was judged by the committee to be noteworthy. The sensitivity analyses indicated that when using the upper confidence interval value for the effectiveness of NIDCAP^® in babies <27 weeks’ gestation from an NHS and PSS perspective, NIDCAP^® resulted in an ICER that was above the threshold of £30,000 per QALY. However, NIDCAP^® remained dominant when considering a wider public sector perspective and the upper confidence interval value for the effectiveness of NIDCAP^®.

Overall, the findings were robust in various scenarios explored in the sensitivity analysis. The estimated cost of NIDCAP^® of £2,887 is substantially higher than that referred to by Westrup (2007). However, even at this much higher estimated intervention cost, NIDCAP^® is a cost-effective intervention in preterm babies of <27 weeks’ gestation requiring respiratory support.

The length of stay in preterm babies of 22-26 weeks’ gestation was approximated using the length of stay reported in Seaton (2018) in preterm babies of 24-26 weeks’ gestation. This could have potentially underestimated the length of stay in preterm babies of 22-26 weeks’ gestation. Although, the median length of stay was simillar for preterm babies of 24, 25, and 26 weeks’ gestation. Also, the deterministic sensitivity analysis indicated that when varying the cost of NIDCAP^® (which, in effect, is equivalent to changing the length of stay) the results for preterm babies of <27 weeks’ gestation were robust to this model input. Moreover, the proportion of babies born at 22-23 weeks’ gestation is small and the impact of this assumption on the cost-effectiveness is likely to be negligible.

The sub-group analysis indicated that the potential for NIDCAP^® is reduced in preterm babies of 27-34 weeks’ gestation, given the relatively low rate of neurodevelopment problems in these babies to start with. The threshold analysis indicated that NIDCAP^® is unlikely to be cost-effective in children born at >27 weeks’ gestation even when considering wider public sector costs. The estimated incremental public sector cost for a child with neurodevelopmental problems would need to be at least £80,000 per annum for NIDCAP^® versus standard care to be cost-effective, which is above what the committee would expect such costs to be. A recent independent review by Schools Week (2018) found that councils spent an average £52,000 per pupil on independent special school places for 2015-16. However, this cost estimate of £52,000 is well below to the estimates obtained from the threshold analysis.

Also, in the sub-group analysis due to the lack of appropriate data, the number of preterm babies developing neurodevelopmental problems (cognitive domain) at various gestational ages were approximated using the percentage of preterm babies with no neuromotor or sensory disabilities at various gestational ages reported in Pierrat 2017. The committee acknowledged that this is not perfect. However, given the lack of more suitable data these estimates provide a reasonable approximation and the resulting rates of neurodevelopmental problems on the cognitive subdomain stratified by the gestational age are in line with the rates observed in their clinical practice.

The cost-effectiveness of NIDCAP^® is likely to have been underestimated since neurodevelopmental problems have significant life-long costs and quality of life consequences. However, due to the lack of suitable data the time horizon of this analysis was limited to 18 years. The committee also noted that NIDCAP^® results in greater parent and carer satisfaction and if NIDCAP^® is made available at such crucial early stages of care the philosophy tends to spread around the nursery. However, to capture such benefits was beyond the scope of this analysis.

Another limitation of the economic analysis was that the costs and utilities were based on postal questionnaires completed by the main parent around the child’s 11th birthday (over the previous year) and may not be fully representative of children in earlier years of life. Nevertheless, the extensive sensitivity analyses indicated that the conclusions were robust to cost and utility estimates and large changes in the base-case values would be required for the conclusions to change.

The committee also discussed implementation challenges, in particular high costs associated with training to set up for NIDCAP^®.