Preparing for Home: a before-and-after study to investigate the effects of a neonatal discharge package aimed at increasing parental knowledge,understanding and confidence in caring for their preterm infant before and after discharge from hospital

Development of the UK Train-to-Home package by a multiple iteration Delphi review process by the Parent Advisory Group and the Clinical Advisory Group

In order to develop a culturally appropriate and acceptable form of the train for use with UK families, nursing and medical staff, we set up two advisory expert panels, a PAG and a Clinical Advisory Group (CAG). The CAG included at least one experienced neonatal nurse and one consultant neonatologist from each of the four participating LNUs. CAG membership also included additional neonatal nurses and consultant neonatologists working in South West England but not in the four study LNUs.

In the first phase we consulted with CAG members individually by e-mail, seeking comments and suggestions on the content and wording of the McMaster ‘Train-to-Home’ documents. In particular, we sought relevant labels for each window, and the criteria to be used to determine window colour: that is, red, yellow or green. The initial responses from each CAG member were not shared with other group members. CAG members were asked open-ended questions on the content and wording of the original McMaster documents, and asked to suggest alternatives to those parts of the documentation that they felt were not appropriate for use in UK neonatal units.

The responses received were collated, categorised and included as a series of unattributed suggestions attached to the original McMaster documentation. These were then forwarded to CAG members individually. Responses from the second round of the Delphi process were collated and a new version of the documentation was constructed using the majority of suggestions received. Where there were significant differences of opinion between CAG members about aspects of the revised documentation (e.g. about how best to categorise types and levels of ventilatory support), these were identified as alternatives and were again forwarded to each CAG member with specific requests for them to consider which approach or wording was the most appropriate. Again, the individual responses from each CAG member were not initially shared with other members of the group. The differences of views were predominantly related to the most appropriate cut-points to use to distinguish between different levels of support that might represent points of change between the different coloured windows, for example the combination of level of inspired oxygen and the use of continuous positive airway pressure to separate ‘red’ from ‘yellow’, and all were easily resolved at this stage.

After the third Delphi round, the documentation was revised in line with the suggestions received, and e-mailed to each member of the PAG for comments and suggestions about content and wording. Delphi methodology was used to involve PAG members individually and seek their comments and suggestions before convening a full PAG meeting to discuss the documentation and suggestions from parents already received. PAG suggestions were incorporated into the third iteration of the documentation. Medical and nursing members of the Project Steering Group reviewed the third iteration, and appropriate comments and suggestions were incorporated into the documentation.

At this stage the proposed documentation was sent, for review, to the senior nurse and a nominated consultant neonatologist in each of the four participating neonatal units, and to ensure that the proposed boundaries defining the red, yellow or green windows on the train were compatible with current management of infants within the defined gestational limits in each of the neonatal units. This process led to a number of relatively minor changes to the definitions of levels of care used in several of the categories, and these changes underwent a further iterative process of refinement to ensure that the wording and definitions used were compatible with current management in all four units.

Finally, a meeting was convened of all members of the CAG, together with the other medical and nursing staff in the participating units who had been involved in the process of developing this documentation who were able to attend – either in person or by telephone conference call. At this meeting the structure, content and presentation of the proposed modified Train-to-Home documentation was discussed and a final version agreed.

This final version (Figure 1; for further details see Appendix 1) was then sent to the lead consultant neonatologist in each of the four participating neonatal units for formal approval. Once this approval had been received the documentation was adopted by the Project Management Group and the Project Steering Group for use in the study.

FIGURE 1.

The ‘Train-to-Home’ package (© University of Bristol and University of the West of England).

Estimating the likely date of discharge for preterm infants from 27 to 33 weeks’ gestation inclusive

We collected data from the routine neonatal database used in South West England (‘BadgerNet’) to estimate the likely discharge date for infants in the study’s gestational age range. Data were collected on all infants in the gestational age range of 27 weeks 0 days to 33 weeks 6 days, who were born in the four study LNUs from 1 April 2011 to 31 March 2013 (a few months before the planned implementation date for the intervention used in the study). We included only those infants who were born in, and discharged to home from, the same LNU. Those infants who were transferred to a regional NICU but transferred back to the LNU in which they had been born, and subsequently discharged to home from the original LNU, were included. Infants who were transferred from the original LNU to another unit from which they were then discharged home and those who did not survive to discharge home were excluded. The ‘Badger’ database provided information on gestation at birth (in weeks and days) and age in days at the time of discharge to home for all infants in the target gestational age range.

Overall, 531 infants met the criteria and were included in the study. In order to achieve large enough groups of infants to assess LOS, we aggregated the data into weeks of gestational age rather than weeks and days. We then calculated the median (50th centile), 75th and 90th centiles for LOS for each week of gestational age within the study range. The numbers of infants at each week of gestational age are given in Table 1, together with the median, 75th and 90th centiles for LOS for infants in each gestational age group.

These data are shown graphically in Figure 2. There was considerable random variation in the 90th centile for LOS because of marked variation in the condition and progress of a small number of infants at each gestation, but for infants of all gestations the 90th centile gave a discharge date considerably sooner than the baby’s EDD, equal to 40 weeks’ gestation. The 75th centile showed a more consistent pattern of change with increasing gestational age, although still showed some week-to-week variation. Overall, the 75th centile for the infant’s postmenstrual age (gestation plus postnatal age) at the time of discharge home was between 36 and 37 weeks for all gestations, except for babies born at 28 weeks, for whom it was 37 weeks and 5 days. In other words, the great majority of preterm infants went home 3–4 weeks before the mother’s original EDD, which is the measure that, until now, has been the one most commonly used by neonatal staff to estimate the likely date of discharge for preterm infants.

FIGURE 2.

Median, 75th centile and 90th centile for LOS and estimate date of delivery for preterm infants, 2011–13.

We needed a consistent and usable chart to allow medical and nursing staff to give parents – within 7–10 days after an infant’s birth – an estimated range of dates for discharge from hospital to home. We decided to use the 75th centile, and to smooth the data to avoid the likely random variations in recorded LOS over only a 2-year period. To make the chart usable and avoid staff making time-consuming calculations for each baby, we included estimated LOS in whole weeks for immature infants of < 31 weeks’ gestation. The more mature infants would mostly be discharged home in < 5 weeks, so we gave the estimated LOS in weeks and days. For simplicity and ease of use, we gave a 2-week range of dates for the less mature infants and 5 days for the more mature infants.

Consultation with the CAG and PAG during the development phase indicated that the range of estimated dates should be no wider than 2 weeks for less mature infants, and should be < 1 week for the more mature infants to be helpful in infant discharge planning. The project aim was to refine the initial range of estimated dates for discharge as the infant progressed through the LNU. As more information became available about each infant’s progress, the estimated dates could be adjusted and the window narrowed.

Figure 3 shows the upper estimate of the LOS, based upon the smoothed 75th centile line (representing the initial estimate of the date by which most infants would have been discharged) together with the earlier limit of the estimated date of discharge, superimposed upon the median and 75th centile lines shown on Figure 2. For all gestations the lower estimate of likely discharge date was just below the median LOS.

FIGURE 3.

Median, 75th centile and estimated LOS for preterm infants based on data from 2011 to 2013.

Using 2011–13 data, this range of estimated dates indicates that 77% of infants born at 27–30 weeks’ gestation inclusive would have been discharged home by the later estimated date based on the smoothed 75th centile, and 75% of infants born at 31–33 weeks’ gestation would have been discharged home by this date. The irregularity in the LOS at each gestation meant that there was a larger variation in the proportion of infants discharged home by the later estimated date for some gestational ages (Figure 4). This is inevitable when using a small database derived from a cohort of infants born in a 2-year period. The proportion of infants still in hospital at the later estimated date of discharge was < 25% for all gestations with the exception of 28 weeks (35%) and 31 weeks (27%). For infants born at 27 and 29 weeks the figures were 19% and 15%, respectively.

FIGURE 4.

Numbers of infants at each gestation who had been discharged or were still in hospital by the later estimated discharge date based on data from 2011 to 2013.

The proportion of infants who had been discharged before the first of the two estimated dates of discharge varied between 50% (at 27 weeks’ gestation) and 27% (at 33 weeks’ gestation).

The final version of the tool that was used in the study is shown in Appendix 2.

Summary

We developed a reliable and simple tool to estimate the likely date of discharge home for infants of 27–33 weeks’ gestation born in the four study LNUs. This was based on the LOS of all infants born in the four units over the preceding 2-year period. We used the smoothed 75th centile for LOS as the later estimate of likely discharge date, with a 2-week window of likely discharge date for less-mature infants (< 31 weeks’ gestation) and a 5-day window for more mature infants.

This easy-to-use tool was applied to all infants within the target gestational age range within 7–10 days of birth, and gave estimated dates of likely discharge that were correct for > 75% of infants. The intention was to refine the initial estimated date of discharge on a 1- to 2-weekly basis as the infant progressed through the LNU (see Appendix 2).

Actual length of stay by gestation for infants included in the study

Figure 5 shows the median (50th centile) and 75th centile of LOS for infants enrolled in the study from the four study LNUs. It shows that the median and 75th centiles for LOS recorded for each gestation for the study period were similar to those in the 2-year period used to develop the tool. However, for infants born at 27 weeks’ gestation, the previously collected data suggested shorter LOS than for infants in the study, although these differences did not reach statistical significance. At all gestations, between 50% and 75% of infants in the study were discharged by the later of the two dates suggested by the tool.

FIGURE 5.

Media and 75th centile for LOS for preterm infants prior to (2011–13), and during, stay period.

Despite considerable variation in infant LOS in the target gestational age range, the tool predicted a likely discharge date for all gestations, by which time close to 75% of the infants had been discharged home. The mean postmenstrual age (i.e. gestation at birth plus postnatal age) at discharge for infants during the study period and during the preceding period from which the above graphs were developed were 36 weeks/4 days and 36 weeks/3 days, respectively – a non-significant difference.

Briefing and training sessions: before and during the intervention

Pre-implementation briefing and training sessions took place in September 2013. The intervention started on 1 October 2013. During each session, the background to the project was introduced, as well as the nature and purpose of the intervention. Time was spent going through and explaining the documentation, and how it had been developed. Information was presented using a PowerPoint^® (2010) (Microsoft Corporation, Redmond, WA, USA) presentation supplemented by paper examples of each of the relevant documents. A project documents pack was given to each LNU attendee. Sessions lasted for between 45 and 60 minutes, with questions taken during the presentations and at the end. LNU attendee feedback was captured by immediate verbal feedback and anonymous written feedback. Feedback was reviewed after each session and used to improve the quality of subsequent presentations.

There was a 15-minute period in between sessions, which allowed for specific detailed discussion with individuals who had particular questions about the project.

The briefing and training sessions were delivered by PJF (a neonatologist) and HB (an experienced neonatal nurse). LNU attendees included a mixture of neonatal nurses, junior medical staff, ancillary staff and consultant neonatologists. Copies of the PowerPoint presentation were made available to each LNU lead clinician to facilitate ‘cascade’ teaching of other LNU medical and nursing staff. A copy of the PowerPoint presentation is given in Appendix 4.

Between 5 and 18 LNU members of staff attended the pre-implementation briefing and training sessions in each unit. At the end of the initial run of sessions we asked LNU lead nurses and lead clinicians to contact us when, and if, further sessions were needed. PJF and HB also visited the LNUs during the first 2 months of implementation to talk informally with LNU staff, answer questions from staff and parents, and arrange further sessions as required. HB delivered a further two to four briefing and training sessions in each LNU during December 2013 and January 2014. PJF delivered two additional sessions each – mostly to LNU junior medical staff in three of the four units. Other planned sessions were cancelled because of heavy LNU workload.

In total, 18 additional sessions were delivered in the four LNUs during the intervention period in addition to the 16 pre-implementation sessions. HB and PJF were available by phone and e-mail to LNU senior staff at each unit to answer queries arising during the intervention process during the period 1 October 2013 to 31 August 2014.

Feedback on the briefing and training session provided for the intervention implementation

A semistructured telephone interview with five of the six LNU lead consultant neonatologists on the briefing and training sessions was carried out at the end of the study. Comments are given below and identities anonymised:

The medical staff found the concepts interesting but the process was ‘not exciting’ to them. Nurses found the initial teaching helpful but it was difficult to cascade the teaching to other nursing staff.

This was really helpful. It was particularly helpful to be able to understand how the estimated discharge dates had been derived.

The teaching was well presented and well explained, but we (i.e. the staff on the unit) were probably too optimistic about how well we would be able to cascade the training to other staff – this was a problem at times.

It was difficult to keep staff up to speed with this because of the rapid turnover of staff – it would probably have been better if we could have made one or possibly two members of the middle grade medical staff responsible for ongoing teaching to ensure all staff understood the project.

The training provided was good and got everyone up and running but cascading training was very difficult – especially for nursing staff. Nurses (or at least some nurses) found it particularly hard to ‘hand over control’ of parts of the process to parents – though this improved as the study progressed.

Parent interviews

Semistructured telephone interviews were carried out with a sample of parents. They explored their experiences of time spent on the LNU caring for their infant, perceptions of communication with staff about their infant’s condition, their infant’s discharge to home, and contact with health services after discharge. Two months after discharge, parents were contacted by phone to complete the final PMPS-E and health resource-use questionnaires. At the end of this call, parents were invited to take part in a semistructured telephone interview at a time of their choice. The researcher explained that they would be contacted once more by phone to answer more in-depth questions about their experiences of having an infant in the LNU, preparation for discharge and contact with health services since discharge.

Parents were purposively selected for interview from those who had agreed to further contact and could be contacted by phone. Maximum variation sampling was used to interview parents from different ethnic and socioeconomic groups, with babies of different gestational ages from single and multiple births. Theoretical sampling was used to develop emerging findings by targeting participants with characteristics of interest (such as younger or first-time mothers).

Interviewers used a topic checklist so that similar areas were covered during each study phase, but with flexibility so that parents could raise issues of importance to them (see Appendix 6 for topic guides). Interviews in phase 2 were also used to explore and understand parents’ views and use of the parent package. The topic checklist was developed with the PAG using relevant literature and insights from study management team meetings.

Nursing and medical staff focus groups and interviews

After family recruitment was completed in phase 2, arrangements were made with the LNUs for researchers to run focus groups with nursing staff in each unit. Visit dates were advertised widely throughout the LNUs several weeks in advance to encourage good attendance. Managers arranged the focus group timings to fit in with staff rotas, usually early afternoons. One-to-one interviews were arranged after the focus groups for staff unable to attend the groups. Short telephone interviews were carried out by a senior researcher with consultant neonatologists from each LNU in early 2015. These interviews were not audio-recorded, but detailed and careful notes were taken and transcribed immediately after the interview.

Rationale for approach to design of economic analysis

Our research study evaluated an intervention that the units had all agreed to adopt as a unit policy. Although we provided (unfunded) assistance and training in achieving this, the aim of the study was not to implement the package but to evaluate its efficacy in neonatal units that had implemented it as a part of routine infant and parent care. Thus analysing the costs of developing the intervention and training the staff using a standard costing approach was not appropriate or relevant information for decision-makers. It is important to recognise that commissioners/sponsors of interventions want to see cost savings demonstrated before they will consider rollout of an intervention that is feasible to implement and acceptable to service users. One of the tensions throughout this project was that some staff in the units were unclear about the nature of the study, and some assumed that they needed to use it only with families who agreed to take part in the research. However, delivery of the Train-to-Home intervention was separate from involving families in answering our questionnaires (the PMPS-E and health-care resource-use tool in the 8 weeks after discharge). The changes associated with use of health-care resources before and after implementation of the Train-to-Home intervention using before-and-after samples, 1 year apart, were costed and did not consider the cost of implementation of the intervention.

Development of the health-care resource-use tool

Resource-use data tools are often designed by developing new questionnaires for each study or modifying questionnaires from a previous study. They are not necessarily validated in their current form, unlike patient-reported measures of health outcome. 46 They need to be appropriate, study specific and easily understood so that participant recall and self-report is accurate. We designed a short data collection tool so that parents could complete it during a busy and stressful time as they were caring for preterm infants during an 8-week post-discharge period. It also needed to be long enough to collect data for estimating cost changes from the perspective of the NHS primary and secondary care, society, and families and carers. Our resource-use data collection tool was piloted with the PAG to check ease of completion and understanding before submission for ethical approval (see Appendix 7). The questions were in four blocks: unplanned use of, and contact with, secondary care services; planned outpatient follow-up contacts; family costs from using those services; and use of primary care and other therapists.

Data preparation and cost estimation

Health resource-use data were entered into the Access database, downloaded into Excel spreadsheets and prepared for cost analyses. The data were reviewed manually to identify low coverage or no-response items, and to clean the data for analysis. There were significant numbers of low-/no-response items in question 2 regarding use of trains, buses, overnight accommodation and number of days off work; these variables were removed from the database. Contact with a health professional – face to face and by telephone, and duration of contact time – was separated out and combined, so that each case had one entry indicating the frequency and duration of contact time by each health professional and contact type. This allowed resource-use volumes and the price of those resources to be treated separately in estimating costs when this was applicable.

Costing the care provided in the LNUs is complex because different tariffs or prices are attached to each care level and recent price information to estimate costs at each level of prematurity is not accessible. This mirrors the costing difficulties reported in the Toolkit for High Quality Neonatal Services. 6 Therefore, we have used proxy prices and recorded their sources, and listed assumptions, unit cost estimates and sources of prices used to calculate costs in Appendix 8.

These data were checked against the health resource-use database and added and/or corrected. Resource-use data in volume units were combined with price and unit cost information from published sources47 to estimate costs per item in pounds sterling (2014 prices). All variables were named and defined (see Appendix 8).

Health-economic data analysis

Cost analysis was performed using SPSS version 21. The reasons for infant inpatient readmission to hospital, outpatient and ED attendance, and hospital centres attended were analysed and compared across the two phases using chi-squared tests. Data were graphed when this was more useful. Continuous data were tested using the Shapiro–Wilk test of normality in distribution (see Appendix 9). These data were not normally distributed, so the Mann–Whitney U-test was applied to test for statistically significant differences in costs during the two phases. We costed the changes associated with use of health-care resources before and after implementation of the Train-to-Home intervention, using before-and-after samples, which were exactly 1 year apart, so we did not adjust for month of discharge or seasonal effects.

Maternal Perceived Maternal Parenting Self-Efficacy distribution

The Shapiro–Wilk test suggests a significant departure from normality at baseline (p = 0.002), at discharge (p < 0.0001), at home (p < 0.0001), between discharge and baseline (p < 0.0001), and was borderline between home and baseline (p = 0.052). Figure 13 shows a typical distribution of PMPS-E scores using maternal scores at discharge for all mothers from both trial phases, therefore a non-parametric analysis approach was adopted.

FIGURE 13.

Distribution of maternal PMPS-E scores at discharge.

Maternal Perceived Maternal Parenting Self-Efficacy scores

Table 12 shows the median PMPS-E score of mothers at baseline, discharge and 8 weeks post discharge. The primary analysis was powered to compare scores for around 100 mothers per group, thus the comparison at home 8 weeks post discharge was slightly underpowered.

Again, there was no significant difference between groups at each time point. The median improvement in maternal PMPS-E score from baseline to discharge taken at an individual level (i.e. subtracting the individual baseline score from the individual discharge score and summing these differences over each group) was slightly higher in phase 2 (+8) than in phase 1 (+7) but this difference was not significant (p = 0.60). The difference in individual improvement was more marked from baseline to home with a median improvement of +14 in phase 2 compared with +11 in phase 1, but this was not quite significant (p = 0.10). Data from Table 12 are illustrated in Figure 14.

FIGURE 14.

Median maternal PMPS-E scores at baseline, discharge and home. The minimum score is 20, indicating poor self-efficacy; the maximum score is 80, representing high self-efficacy.

Regression model

The difference in maternal PMPS-E scores between the two phases at discharge and 8 weeks post discharge were tested using linear regression. The dependent variable for each of the two models was maternal PMPS-E score at discharge and 8 weeks post discharge. The independent variable being tested (Table 13) was the group effect adjusted for the baseline PMPS-E score. Thus the coefficient reported was the mean difference (i.e. phase 2 – phase 1) in PMPS-E score at discharge or home between the two groups.

On average, mothers in phase 2 at discharge scored 0.07 points lower in the PMPS-E score than mothers in phase 1: virtually no difference between the two groups (p = 0.95) with an effect size of 0.005. At 8 weeks post discharge, mothers in phase 2 scored 1.48 points higher in the PMPS-E score than mothers in phase 1, with an effect size of 0.12, although this difference was not significant (p = 0.14).

Imputation of missing values for the primary analyses

Variables predicting missing PMPS-E scores at discharge and 8 weeks post discharge included the number of children in the family, maternal education, gestational age in weeks and maternal age. Table 14 compares the results of the primary analysis given in Table 13 with the results after imputing data for mothers with missing PMPS-E data to include all 245 mothers in the model.

The results are quite similar. At discharge, the group effect still centred around zero [–0.07 points in the initial model (p = 0.95) and +0.11 points in the imputed model (p = 0.91)]. At 8 weeks post discharge, the group effect suggested an increase in scores for those mothers in phase 2 [+1.48 points in the initial model (p = 0.14) and +1.26 points in the imputed model (p = 0.19)], but neither finding suggested significant differences. The mean values and SEs were similar using both approaches.

Adjusting for significant factors

The only significant difference between infants in phase 1 and phase 2 was a higher prevalence of infants in phase 1 having conditions related to the metabolic, endocrine, nutritional, gastroenterological or neurological systems (p = 0.006). Adjusting for this in the models presented in Table 13 made little difference to the comparison of PMPS-E scores between the two groups. At discharge, the association between infant medical conditions and the PMPS-E score was not significant (p = 0.28) and the influence on the group effect was minimal (phase 2 mothers scored 0.15 points higher; p = 0.89). At 8 weeks post discharge, again the association between infant medical condition and PMPS-E score was not significant (p = 0.71), and the influence on the group effect was to slightly reduce the difference to 1.39 points (p = 0.18).

Maternal Perceived Maternal Parenting Self-Efficacy subscales

The 20 questions of the PMPS-E score can be split into four mutually exclusive subscales. These are:

• Caretaking procedures (mother’s perceptions of her ability to perform the activities and tasks related to the baby’s basic needs such as feeding).

• Evoking behaviour(s) (perceptions in her ability to elicit a change in the baby’s behaviour, e.g. soothing the baby).

• Reading behaviour(s) or signalling (perceptions in her ability to understand and identify changes in her baby’s behaviour, e.g. ‘I can tell when my baby is sick’).

• Situational beliefs (mothers’ beliefs about her ability to judge her overall interaction with the baby).

Table 15 compares the median maternal PMPS-E score at baseline, discharge and at home between the two phases, broken down by each of these four subscales.

There was little difference between phases 1 and 2 at the subscale level. Mothers in phase 2 scored slightly less on the evoking behaviour subscale at discharge but slightly more than those in phase 1 at 8 weeks post discharge.

Maternal Perceived Maternal Parenting Self-Efficacy between-centre variation

Figure 15 illustrates the differences in maternal PMPS-E scores at baseline, discharge and home between each LNU. In unit 1, the scores at baseline were lower for mothers in phase 2, although the difference decreased over time. In unit 2, the scores at baseline were exactly the same but increased more among phase 2 mothers over time. In unit 3, the baseline scores were higher among phase 2 mothers but decreased over time, whereas in unit 4 (like unit 1) the baseline scores were lower for phase 2 mothers but were much higher at discharge and at home 8 weeks post discharge.

FIGURE 15.

Median maternal PMPS-E scores at baseline, discharge and home by centre. (a) Unit 1; (b) unit 2; (c) unit 3; and (d) unit 4.

Table 16 shows the median difference in maternal PMPS-E from baseline to home comparing phase 1 with phase 2 for each of the units.

Adjusting the primary analysis models in Table 13 for centres made little difference to the group effect.

Per-protocol analysis

There was evidence when the researchers visited units, and confirmed in subsequent qualitative interviews, that staff from one particular unit were less engaged with the intervention and there were questions about whether or not the intervention was delivered properly. Dropping this unit from the analysis, Table 17 shows the median PMPS-E score of the mother at baseline, discharge and 8 weeks after discharge from just three units.

Again there was no significant difference between groups at each time point when restricted to three units. The median improvement in maternal PMPS-E score from baseline to discharge taken at an individual level (i.e. subtracting the individual baseline score from the individual discharge score and summing these differences over each group) was slightly higher in phase 2 (+9) than in phase 1 (+7) but this difference was not significant (p = 0.27). This difference in individual improvement was more marked from baseline to home, with a median improvement of +16 in phase 2 compared with +11 in phase 1, and this difference was significant (p = 0.02).

Figure 16 illustrates the data from Table 17.

FIGURE 16.

Median maternal PMPS-E scores at baseline, discharge and home (three units).

The difference in maternal PMPS-E scores for the three units between the phases at discharge and 8 weeks post discharge (Table 18) were tested using linear regression similar to Table 13.

On average, mothers in phase 2 at discharge scored 0.21 points higher in the PMPS-E score than mothers in phase 1, a difference that was not significant (p = 0.86). At 8 weeks post discharge, mothers in phase 2 scored 1.78 points higher in the PMPS-E score than mothers in phase 1; again, this difference was not significant (p = 0.13). As Figure 15 shows, there appears to be a slightly larger difference in PMPS-E scores when restricting the analysis to three units but the smaller numbers used yield wider CIs and a similar p-value, suggesting that this study may be underpowered to test the significance of this finding.

Paternal Perceived Maternal Parenting Self-Efficacy score

As anticipated, fewer fathers participated in the study, so, given the lack of power, the assessment of paternal PMPS-E scores will be exploratory in nature.

The data for the days taken from birth to various time points to measure paternal PMPS-E were not normally distributed at baseline (p < 0.0001) or discharge (p = 0.002), thus, although means could be reported for days from birth to home (p = 0.44), medians have been presented for consistency. Table 19 suggests there was no significant difference between the days taken to measure the paternal PMPS-E from birth to baseline, discharge or 8 weeks post discharge between the two groups. The relatively large difference in median days at discharge reflects the small numbers observed and was not significantly different between phases 1 and 2.

The paternal PMPS-E scores were borderline normally distributed at birth and discharge but not the home scores (Shapiro–Wilk test; p = 0.046). For consistency, Table 20 shows the median PMPS-E score of the partner at baseline, discharge and 8 weeks post discharge.

The paternal PMPS-E scores were similar to maternal PMPS-E scores at baseline but tended to be lower at discharge and home for both groups. There was no significant difference between groups at each time point. The median summed individual difference in paternal PMPS-E score from baseline to discharge was slightly higher in phase 2 (+6) compared with phase 1 (+5) but this difference was not significant (p = 0.57). The median improvement from baseline to home was lower in phase 2 (+12) compared with phase 1 (+13), but, again, this difference was not significant (p = 0.75).

The difference in paternal PMPS-E scores was modelled in the same way as the maternal PMPS-E scores.

Table 21 shows that on average fathers in phase 2 at discharge scored 0.92 points higher in the PMPS-E score than fathers in phase 1, a difference that was not significant (p = 0.62). At 8 weeks post discharge, fathers in phase 2 scored 0.54 points higher in the PMPS-E score than fathers in phase 1; again, this difference was not significant. The numbers in both models were very low.

‘Rollercoaster of emotions’

Emotional upheaval marked parents’ experience of the admission of their baby to a neonatal unit. Overall, they recalled mixed feelings about having a baby in the LNU. Their feelings became increasingly positive with the improvements in their baby’s health and with the passage of time.

Well I suppose it was a mixture of like, well it was a worrying time, but also it was in the end a more positive time to be able to watch your baby grow, and know that she was doing well.

Mother #305 phase 1

How would I describe my time? Very emotional, very up and down like a rollercoaster of emotions. And I suppose the biggest thing: every minute felt like an hour, an hour felt like a day, a day felt like a week; it just seemed to go on and on. But um obviously the staff were very supportive and helped us to feel, you know, confident that [baby] would come home and he’d be fine, so yeah.

Mother #315 phase 1

A spectrum of emotions was described by parents across the different LNUs from ‘hell’ to ‘comforting’ and ‘reassuring although still scary’. Even although one unit was described as ‘relaxing’ by parents, they still experienced traumatic and negative feelings.

We found it really, strangely relaxing. Once we got over the initial um shock of seeing her with – with all the wires and everything around, which was always stressful, it was really nice and relaxing.

Mother #514 phase 1

Parents talked about the distress of being separated from their newborn baby, and later the stress of having to leave their baby on the LNU when they went home.

It was just the fact that my baby was in the unit and I wasn’t there with him, and I had to go home every night; that was the worst thing ever. It was fine when I was there in the day um with him, but leaving and going home was awful. But the staff at the unit were very positive and reassuring and helpful and brilliant really.

Mother #719 phase 1

They also talked about the feelings of emotional isolation as they felt that staff didn’t want to give them false or unrealistic hopes about their baby’s condition.

They never say to you ‘Don’t worry it’s all going to be fine,’ because they can’t give you any of that false hope, and they don’t know if everything will be fine, so he might be fine today but then by tomorrow you know as I said they can get infections so quick. So you do feel quite, um, like emotionally isolated with it all.

Mother #213 phase 2

Positive experience was linked to support from LNU staff, primarily the nursing team. The physical environment is challenging – it was very hot, light levels were low and parents spent a long time doing very little other than being with their baby. In phase 2, parents commented that having Wi-Fi would have helped pass the time and relieved the boredom.

They keep the lights very low, and it’s very hot, the environment is quite depressing, they have tried to keep the colours brighter, staff are really friendly as well, even small things like Wi-Fi access would have made a massive difference . . . I think it is quite hard as a parent, or like the mum who has to go and sit there all day every day, and if you don’t you feel bad, then if you do you’re bored.

Mother #213 phase 2

‘You learn a lot in there’

In both phases, parents talked about the LNU as a place where they learnt parenting skills. This opportunity would not be available to parents of term babies as they are in hospital for much shorter periods.

I think in some ways, and this sounds silly, but having a premature baby is almost a bit of a better benefit than actually having a full-term baby. Because having a full-term baby you’re suddenly thrown into the world of, ‘Oh my God, we’ve got a baby: what do we do now?’ With a premature baby, obviously they needed that specialist care for the time till they’re able to go home, and in that time you’re taught so many things that you would have to just learn on the spot at home.

Father #307 phase 1

‘They were approachable and listened to me’: establishing relationships with nurses

Mothers mostly found nurses and midwives easy to talk to, and they described most staff as helpful and supportive. Staff with a friendly approachable manner were valued, and they appreciated having ‘a normal everyday conversation’ with them. Support for parents included ‘being listened to’, ‘staff knowing about their family’ and having continuity of care for them and their baby. Mothers commented on the importance of the shift handover between staff as a way of ensuring that information was passed thoroughly and appropriately. They also commented on the advantages of having the same staff caring for their baby in building relationships with the family. Both of these aspects supported parents in their communication with the staff.

‘Conflicting advice about feeding’

Less positive support was recalled by some mothers trying to express breast milk or learning to breastfeed their baby. A lack of consistent advice, and the attitude of some staff, made them feel apprehensive and lacking in confidence about caring for their baby.

There was one nurse, I think she was a sister, and she basically made my life hell. She did it with other mums who had the same problem, in as far as if I wasn’t expressing enough [breast milk] that was my fault and I’m not feeding or looking after my baby properly. Whereas the other nurses were, ‘Look, don’t worry. You’ve got formula we can mix. You need to relax to be able to do this’ . . . but whenever this other nurse came up it was just hellish.

Mother #320 phase 1

I would have to ask specifically for them to come and help. I was encouraged every time to put him to the breast (when) I tube fed. But if I actually wanted some help with latching on or positioning or anything I would have to sort of say, ‘Can you give me a hand?

Mother #719 phase 1

I was getting different advice . . . because you kind of get moved on to nursery nurses, and health-care workers rather than neonatal nurses, and they all do things differently, because the level of experience or the training they have had, and I had one there say to me, ‘Stop expressing now in between feeds,’ then the next night another nurse said, ‘Are you still expressing?’ And I said no I was told to stop, and she went, ‘Oh that’s strange, I don’t know why because we’re told to tell you to carry on.

Mother #213 phase 2

It was actually, the breast feeding was one of the issues of why we had to wait to go, because [twin 1] was breastfeeding and [twin 2] wasn’t, and we actually had a lot of pressure on me, um, from the nurses to get [twin 2] to breast feed, so much so like they would you know really be kind of forceful about putting her to the breast. Um, and so we never talked about bottles or formula or anything, because they you know were . . . we doing breast feeding . . . Then the day I got home I stopped producing milk, I had no idea what to do with bottles and formula, because no one taught us as a backup. That was hard.

Mother #617 phase 2

Medical staff were mentioned too, but they were less visible on the LNUs. Some mothers linked communication with the consultant to receiving good or bad news about the baby’s condition. In phase 2, particularly at one LNU, parents reported that doctors had been very helpful and supportive.

The baby doctors were always there, the neonatal doctors, there was always a doctor around in either side of the unit. Um so you didn’t – if you wanted to find something out you could always approach one of them, you didn’t have to go searching for someone.

Mother #307 phase 1

I mean obviously the doctors are very busy, um, but they certainly . . . if you had concerns they would definitely come and it might not be straight away but they would definitely come and spend some time talking things over with you, . . ., but they very much wanted to put you at your ease . . . and take you through it slowly . . . you know, he was happy to sit and talk to us again once we’d thought about actually the things that we wanted to ask him.

Mother #819 phase 2

‘The ‘Neonatal Intensive Care Unit family’: support from other parents

These new friendships were valued and they felt part of a NICU family. They talked about developing their ideas with other families and learning from those who were ahead of them in their progress to going home. Some families kept in touch after they had gone home, too.

We were talking to lots of other parents who were going through the stage just ahead of us . . . I think we just bounced off ideas all the time about what we were thinking of . . . the dilemmas we had about whether to continue trying breastfeeding or just to go to bottle feeding.

Mother #315 phase 1

Every day we would have lunch together and sit there together expressing and everything, so like they know exactly, you don’t have to explain anything to them because they know exactly what’s going on.

Mother #206 phase 2

I thought it was a great experience, um, at the end of the nine weeks we felt like we were almost like part of a NICU family, not just with, um, not just with the staff, but also with the other parents that we had met there.

Mother #617 phase 2

Practical preparation: ‘knowledge and skills transfer, but not enough notice’

Parents considered the knowledge they received practically prepared them for going home and most recalled the practical skills they had been taught, including holding their baby, sleeping positions, first aid emergency and resuscitation (cardiopulmonary resuscitation), bathing, bottle-feeding and nappy changing. The nurses trained the parents in readiness for going home with their baby. There was no reported systematic information delivery using checklists or going home packs throughout the LNUs in phase 1. One LNU identified that they had a pre-existing checklist; however, parents reported that processes seemed to be reactive rather than proactive.

In phase 1, many mothers were very concerned about the lack of notice about discharge and not having a date to go home or any structured or meaningful discussion about going home. This made it difficult for them to alert other family members about their imminent departure if they needed to make family arrangements.

It was all very quick in the end . . . we didn’t really talk about going home and then all of a sudden it was, ‘Let’s try demand feeding. You could go home in 2 days,’ um very, very sudden.

Mother #719 phase 1

I did actually, admittedly, think we would have a bit more notice to ‘er – to get some stuff in.

Mother #713 phase 1

We didn’t actually know we were going home until sort of the very last minute.

Mother #221 phase 1

Because my parents were looking after my other children, so it was important that they were highly involved in what was going on. I texted them all – well we had to wait for the paediatrician to do his round on the day that we were discharged, and I texted them all going, ‘Yeah we’re coming home today’.

Mother #112 phase 1

In phase 2, some mothers were irritated about the inconsistency between staff about the timing of messages about discharge, particularly between doctors and nurses. They reported that the decision about whether they would go home at the planned date lay with the doctors. Nevertheless, mothers mostly reported being quite well prepared for going home.

They [the nurses] just made decisions on you know upping the milk and things like that, they didn’t make any major decisions.

Mother #808 phase 2

As far as the standard like training or procedure that the unit goes through for parents I do feel, it was good, and I didn’t get home and think oh my goodness what do I do with this baby.

Mother #213 phase 2

Yeah, it was quite good that I was very hands on, um, from as early on as I could be really, um, so yeah things like changing nappies, getting her changes, giving her a wash, giving her a bath, all those sort of things . . . so after all that I don’t think there was anything else they could have really done.

Mother #807 phase 2

Emotional preparation: ‘uncertainty, feeling rushed, motivation to get home’

Mothers described their uncertainty about going home as a mixture of feeling scared, nervous and excited. Some felt that they were not well supported by staff in this preparation and they needed more reassurance. In phase 1, a few families from two LNUs felt that they were not ready for home for various reasons, including the health of the baby. The scariness of taking home a ‘tiny baby’ was raised by a small number of mothers, but this was overcome by the desire to have the family under one roof. For others, the thought of putting in place the routines learnt at hospital supported their preparation.

It came to the point where we, yeah, we were getting prepared to do it and, yeah, nervous, scared, but excited at the same time . . . mentally kind of preparing how we were going to do things you know, continuing it at home rather than having the hospital around to help.

Mother #521 phase 1

These concerns were also echoed in phase 2. A few mothers felt their family needed to be all together, other mothers felt they would never be ready. However, more parents in phase 2 commented that they felt mentally prepared to take their baby home.

I just got on with it . . . it did help me feel mentally prepared by the train date.

Mother #814 phase 2

I was feeling ready to go home.

Mother #623 phase 2

Me too . . . we couldn’t wait to come home actually to be honest . . . yeah, we could bring the babies home.

Father #623 phase 2

The role of breastfeeding: ‘breastfeeding is the harder way to do it’

Breastfeeding was supported and encouraged by nurses across all four LNUs. In two LNUs mothers identified a nurse who functioned as a breastfeeding advisor. In the other LNUs, mothers said they could approach almost any nurse with varying degrees of ease depending on their relationship. They mentioned having advice about expressing breast milk regularly and frequently, various ways to position a baby for feeding, and how to ensure that the baby was latched on, but there was conflicting advice about on some points, particularly using nipple shields.

Got loads of support with breastfeeding. You just felt able to, you know, ask questions. They were particularly useful, very supportive, yeah.

Mother #305 phase 1

Yeah, sort of positive spin, because I had difficulty, um, producing enough breast milk, um, and they were always sort of really helpful and positive about it, they just sort of you know said, ‘keep trying,’ so really sort of supportive.

Mother #814 phase 2

There was mixed opinion on nipple shields: some people said it might help and other people said ‘no don’t’.

Mother #719 phase 1

She suggested that I try some nipple shields, and another nurse on the day shift helped me use them, and between the two of those nurses, that’s what got the breastfeeding going. Whereas – one nurse was like, ‘Why are you using nipple shields? You haven’t got flat nipples,’ but it wasn’t the point. And because she just came in she didn’t have a clue.

Mother #305 phase 1

All of us who are friends all had to use the nipple shields, we always had to use it, and then we all went home with them, and I just found it too hard, but some of the mums have kept on with it and stuff. But I just found it [breastfeeding] . . . he found it too hard.

Mother #408 phase 2

When mothers encountered issues with breastfeeding, nurses continued to support them. For some nurses, this involved supporting a mother’s decision to switch to formula feeding or suggesting topping up with formula to ensure that the baby could go home sooner.

So that we could go home they said, ‘Why don’t you just do top-ups with formula? Because then we know that she’s got a means of having enough milk when you go home.’

Mother #514 phase 1

I was expressing for about eight weeks, um, so we were using that, and then gradually I think it was probably about a week before we thought she might go home they started introducing a bit of formula with my milk, um, to get her used to the taste, and then I think we were on about 50/50 when we went home, and then when we got home we just used formula straight away which she was fine with, um, because that was always my plan to do formula anyway. So it was like a bonus that she’d had my milk.

Mother #206 phase 2

There was a sense in the interviews that breastfeeding was difficult for this group of babies but it is not clear how nurse to parent conversations or mother to mother conversations contributed or improved this.

I understand breast is best, and I’m completely aware of that; I didn’t realise how difficult it would be though.

Mother #320 phase 1

Mothers were also aware of the delays that could be associated with breastfeeding and weight gain.

Everybody acknowledged that breastfeeding is the harder way of doing it. And a lot of babies that were bottle-fed left sooner than us. Like there was a baby that was only 1 day older, and he went home a week sooner because he went home on a bottle.

Mother #305 phase 1

What ended up stopping us from going home . . . was his weight because we were breastfeeding. He was tube fed with breast milk which is then fortified . . . and then they stopped fortifying it, so his weight took a sudden drop, because all of a sudden it wasn’t being fortified. So then they needed to see that weight go back on, to ensure he could home healthy, so we’re struggling to put the weight back on, and he was losing maybe like a couple of grams each time, so I found that very frustrating, it was very upsetting, because they weigh every other day, and then it was quite disappointing every time we didn’t get to go home.

Mother #213 phase 2

Parent-reported positive comments about the Train-to-Home package

Many parents responded positively to the Train-to-Home package. Mothers, fathers and siblings found the visual representation of their baby’s progress helpful and enjoyable. Positive themes were the ‘value of staff engagement’, ‘giving us hope’, ‘something you can visually see’, ‘fathers and families loved it’, ‘the dates prepare you for home’, ‘feeling in control’ and ‘do it right at the beginning’, as shown in Figure 17.

FIGURE 17.

Themes generated about the Train-to-Home intervention in phase 2.

‘The value of staff engagement’

Where staff engaged with the intervention and delivered care through the intervention, parents responded positively and felt that it helped them understand what was going on around them.

It was one of the nursery nurses sat down with me and explained what it [Train-to-Home] all was, um, and did the first kind of round of stickers and everything, and then the, um, nurse that had been the most kind of trying to get us home she went through it again a bit later, updated it with a different discharge date and everything.

Mother #219

We loved it [Train-to-Home], we felt like it made us feel like a part of it, and we loved getting the little booklet out, and you know we would discuss it between the two of us and discuss it with the nurses, and you know we would get excited, we would take pictures of it, and, um, we enjoyed it.

Mother #617

‘Giving us hope’

An impression of progress, an end in sight, and of moving towards home indicated by the changing of the stickers on the train from red, through yellow to green, was reflected in the parents comments captured in most LNUs.

The train date was earlier obviously, [than due date] and so it made us kind of think well actually there’s no reason why he won’t be out earlier, um, if he achieves all of these things, um, which obviously are on the train. So I think, um, it gave us that sort of encouragement really.

Mother #223

It was such a long process that it made it feel like you know there was an end in sight.

Mother #617

It’s good because it shows you the progress that your baby is having.

Father #623

The train was one of the only things I remember from those first couple of days . . . I was in shock, pain, emotional and very tired . . . I remember the train, remember thinking – OK so he works his way along until he leaves.

Mother #802

Yeah, it was, it was really helpful, because what happened is even when we got there and we were given a day I always looking forward to that day, and I was happy that the date didn’t move as in she went home within that day, I was really, really glad that I had a date to look up to.

Mother #823

The train is meant to give you like a bit of hope for when your baby is coming home. We all knew it’s not definite, but it’s meant to give you that bit of hope.

Mother #808

‘Something you can visually see’

Some parents took more ownership of the process and explained it to their wider families. The system of colours also provided parents with an ‘at-a-glance’ affirmation that their baby was improving in health.

I really liked it, I thought it was a great idea, because it’s a quick easy way of seeing where they are and showing . . . when my parents came, instead of having to explain everything I could show them the little train, and they had their colours on there.

Mother #219

It helped, um, us visually see how sort of much progress he’d made, and it was sort of quite encouraging to see him going from a red to an amber, and then an amber to a green. So that was useful, and also it was quite useful to show relatives and friends, sort of how far he was progressing, and sort of provide pointers on what to go through with them and what to explain and things. It helped provide it in a bit more simplistic terms really.

Mother #223

Well it was good information because they were giving you, um, the colour they would say that’s a red it means the baby is not ready, so you could just see day by day if the baby is . . . the problems with the baby’s health. It was very helpful because when I came the next day I was just like oh she’s got more green than red, it was just very optimistic saying, you were like oh my God, it won’t take long to go home with the baby.

Mother #622

It is a bit more involvement, obviously it’s not a lot, but it’s something, and it’s something you can visually see, because looking at your child in an incubator and that, the only difference you can really see is maybe a few tubes are coming, um, but when you can look at a chart, and look at colours and obviously going from red to green and that it’s all positive. So a little bit helps really.

Mother #807

‘Fathers and families loved it’

Looking at the train, reviewing the criteria, changing the stickers and taking photographs to send to their family were all mentioned by mothers as helpful activities for fathers and siblings when in the LNU.

Visually, um, for my partner especially he could just see how well he was doing. Obviously I had put the stickers, I would talk to people and then I would put the stickers on, um, but it’s also something nice to take home as well. Just say look, you know, this is how well he came on.

Mother #608

My partner he really got into it for the first, um, week or so, he read all through the booklet and he wanted to do stickers and he wanted to update it. I think for him it was like something he could look at, you know, focus on.

Mother #225

Yes, um, one of the nurses sat down and sat through it all with . . . and, um, we went through everything so that myself and my partner completely understood, um, it, and it was great because we’ve got a four year old, so when he came in he could . . . he loved the train, he thought the train was fantastic, and it was great for him to be able to understand a bit more. He used to stand there hoping that it would turn green in front of his eyes, he thought it was magic.

Mother #418

It made us feel good when he got an amber [sticker]. I took pictures of the train and showed it to friends, to the grandparents and colleagues at work. Towards the end I was able to say that he’d got four greens and we were just working on the feeding. I thought it made it visual – it was really excellent.

Father #816

Weirdly enough my husband loved it, um, I liked it, but he wanted to put the new stickers on as soon as physically possible, whereas I wanted to not put the upper stickers until we were absolutely sure that was the level that she was at, because there was a couple of times where it felt like back tracking if you had to change the colours of your stickers. My husband couldn’t wait.

Mother #819

‘The dates prepare you for going home’

Parents confirmed that the package helped them to prepare for taking their baby home by giving them a time window in advance that they could work towards.

It just says the date you will be going home, it’s just a rough idea about how you can be prepared for it.

Mother #622

And you know like roughly when he can go home, so it’s quite helpful as well.

Mother #623

‘Feeling in control’

A few parents spoke of the control that they felt about engaging with the Train-to-Home package.

It was great because it put me even more in control, and I felt like I was doing more, so it was great.

Mother #418

Well actually the doctors came round one point while I was there early and they said that I could do the stickers myself. So, um, that was quite good. Something so small actually makes a big impact, you kind of feel quite good about being able to sort of put a different colour sticker in, because you can see how she is improving really.

Mother #807

‘Do it right at the beginning’

Some parents had not received the train at the beginning of their LNU stay and this impeded their chance to engage with their baby’s progression.

We did think it was a really good idea that you can see him progressing, but it’s just unfortunate that we didn’t do it right at the beginning . . . So we didn’t actually see much of a progress . . . so that’s the only one thing I do suggest . . . is to do it right at the beginning so that you can see a change . . . more consistent with it, and like right from the beginning, because I think it’s fantastic that you can see the change and see the progress.

Mother #408

It might have been [nurse], who I think, um, you know, went through it in quite a lot of depth with us, which was helpful. But that did happen a bit later down the line . . . maybe just, um, I don’t know whether to be done slightly sooner?

Mother #621

So I would definitely say within the first week, you need to know that’s an option, that’s a good idea, but sort of 3/4 days onwards I think.

Mother #814

Using the booklets: ‘questions to ask and getting to know your baby’

Over half of the parents who were interviewed used the booklets. Those who used them found the questions helpful and ‘loved getting the booklet out’ to use. Both mothers and fathers found the booklet useful, it facilitated engagement and understanding, helped them to ask questions and many of parents wrote comments or observations in it.

‘Staff didn’t explain it’

Some parents in three LNUs reported that the train was not always introduced or fully explained to them to enable them to engage with the intervention.

Yeah, nobody ever explained that train to us, like, and then one day I was like I’m going to stick some stickers on there, so I did, because everyone else I knew had put their stickers on, and everyone was like looking at [baby]’s, because it had nothing on, so um yeah nobody went through that like booklet or anything . . . it didn’t really mean as much for us. But it was quite nice, when I put all the green stickers on, I took a photo of it and I put it in her baby book. Um, yeah it is quite nice to have it in there.

Mother #206

The train appeared one day with some dates on it, um, and . . . it must have been that I wasn’t around when the nurse came to do it, and then I didn’t know what it meant for quite some time, I think it was another couple of weeks before it was explained to me.

Mother #213

They didn’t really use the stickers or anything. I didn’t really get explained what to do with that . . . mine didn’t really get used to be honest.

Mother #417

‘It was confusing’

The parents mentioned being confused by nursing staff comments about the materials developed for the study or when they appeared not to use them. This included not explaining the criteria for changing train window colours, challenging the going home dates, and being advised to use ‘due date’ as the date when their baby would be going home. In half of the LNUs, nursing staff told parents to use the due date as the target for going home, and then added a different, earlier, date to the train.

I was talking it through with one of the other nurses, and she was explaining it to me, and she was saying actually some of the, um, the things there that are written down to get each spot could be changed around really.

Mother #221

I think nearly all of them [nurses] said that, it [going home date range] was unrealistic. It seemed very strange, because when it [the train] was first presented to us we were kind of like ‘oh wow that’s really soon’, because the consultant, the first thing they said was you have to go with his due date . . . So that was very confusing.

Mother #225

You always get told to wait until due date.

Mother #418

‘Give me more information to change the colours’

Some parents reported that nursing staff handed over responsibility for maintaining the train to them but did not give them enough information to use it correctly. This meant that stickers were not updated as parents did not understand the basis on which to change them.

Perhaps it didn’t get updated often enough, because we had it done initially, and then it was only updated when they’d got to all green, um, and there were a couple of things that I felt, um, [baby] in particular had . . . she’d kind of moved on, she had achieved it, but it wasn’t updated particularly . . . wasn’t sure if I was supposed to touch the train, um, it kind of seemed like that was something that they did, um, and it wasn’t for me to do, because they would determine whether they had met all the little targets.

Mother #219

I think it just needs a little bit more work around explaining it to parents, and, you know, who is leading on it really, and at what point, you know, what’s the criteria so that you can make judgements about where your baby is. Because things like for example sleeping, I thought well does that mean . . . you know what does that mean? Does that mean she is sleeping lots so that’s a good thing?

Mother #621

‘Health visitor support was either fantastic or a bit vague’

Once at home, parents had a range of services they could turn to for support and advice. Many parents from all the LNUs regarded their HV as most useful, helpful and knowledgeable in discussing a variety of routine problems that they encountered, including breastfeeding and tongue tie.

My health visitor has been fantastic. She’s been coming out weekly and has been really, really good support . . . More with the breastfeeding side because I was having problems, obviously [baby] was tongue-tied and I was having real doubts whether or not I could actually carry on breastfeeding when I first brought her home, because it was just so wearing. And she was really, really supportive in, ‘Well let’s look beyond the problem. Let’s see what’s causing it.’ And, you know, she was absolutely fantastic.

Mother #318 phase 1

The health visitor has been very good. Unfortunately she’s retired now, which is a shame because I would have liked to have had her for a bit longer. She was very good because she noticed the tongue tie. I think just her sheer experience was really helpful, um, and she was like . . . she’s referred us to the cranial osteopath and then she gave me some advice about, um, foods and things to avoid and stuff like that. So she’s been really good.

Mother #213 phase 2

My health visitor is amazing, um, yeah she’s really friendly, she’s really easy to get hold of, um just need to send her a text and she rang me back when I had a problem . . . I mean I haven’t had many problems, um, but I have she’s been really, helpful, really happy to answer any questions.

Mother #219 phase 2

However, there were other parents who had concerns about the lack of direction they received from their HV, the lack of specific understanding about premature babies, and a focus on child protection aspects of the role rather than parental support.

Um the health visitor was good for some basic stuff, but I did find that she was – she’s quite vague. I almost got the impression that she didn’t want to give me too much information because she didn’t want to be wrong, like she didn’t want to give advice and just say, ‘Well you could try this or you could try that,’ but, I don’t know, I found her a bit vague in what she was saying. I don’t know if that’s just her. I don’t know what other health visitors are like.

Mother #109 phase 1

When the health visitor came round she didn’t have a great deal of experience with prem babies, so I found that what she was saying was contradicting special care, what they told us before we came home. So I kind of ignored what she said.

Mother #418 phase 2

‘Neonatal outreach team is a great support’

One unit supported mothers with a hospital neonatal outreach team after leaving the LNU. This support was sometimes preferred over community-based HVs because the outreach team gave advice specific to premature babies.

My health visitor actually used to do NICU herself in the past, which actually helped. But she did more practical – rather than [outreach team] were solely giving knowledge and information regarding NICU and prematurity. Whereas the health visitor was more, I suppose, generalised of how things would be going normally. So any more worries and concerns I had, it was easier to speak to neonatal outreach team members regarding things about her being early, asking them rather than the health visitor.

Mother #521 phase 1

Yeah, the NICU health visitor and my own health visitor . . . found them really helpful . . . all three of them, they’ve been there for me as well as [baby], obviously that’s what they’re there for, but I found them a great support.

Mother #608 phase 2

‘GPs helpful for medical things’

General practitioner (GP) knowledge seemed to be more mixed. Some parents felt supported by their GP in terms of routine illnesses, but as with HVs, most do not have specific preterm knowledge and advise parents to contact the LNU or in one situation go to an ED.

My doctor was brilliant, because obviously she was the one that was saying, you know, ‘Let’s get this [hernia] sorted, and bring him back tomorrow, or bring him back in two days’ time and let me have a look then.’ So . . . she was fantastic.

Mother #315 phase 1

Actually the GP has been very helpful from a baby point of view, the sort of reassurance about just generally having a baby, when I’ve been to see them, but not really about a premature baby.

Mother #719 phase 1

Also my GP . . . it’s literally the medical, illness side, then it’s definitely the GP.

Mother #807 phase 2

‘Staying in contact with other parents met on the unit’

Several parents remained in contact with other parents that they met on the LNU and one unit had started a support group for recently discharged parents.

Well I’ve kept contact with the friends I made on the ward, we meet up every four weeks and we talk. Also the [name] group have got the Facebook page, they also come to the unit, they’ve been really good actually when I’ve had any sort of questions, just posted it on there and then loads of other mums have replied saying, ‘I was in exactly the same situation, I tried this’ so it’s been quite nice.

Mother #206 phase 2

I have joined the support group that runs out of [the unit], I’ve been to one of their community meet ups, and met some mums that way, so yeah we just chat to a few . . . If you’ve got any questions really, all mums from the unit, and I think they let you stay for about 2 years, so if you’re experiencing something they’ve been through they can help out. It’s a big Facebook group as well, so that’s quite nice, so it’s really easy to get hold of somebody.

Mother #219 phase 2

‘Wider family support is most helpful’

Families in phase 2 identified the support they receive from their wider family (grandparents and sisters) as being helpful and having the most useful knowledge. Understanding the baby’s needs while in the unit was also important for the wider family in knowing how to help after discharge.

Family really that has given most of the help . . . it’s two babies feeding is quite difficult isn’t it? . . . this is the most important thing I think, that’s why we get help from family.

Mother #623 twins phase 2

[My sister] she’s got a baby who is, I think just 11 weeks older than [baby], so she’s kind of recently been through the kind of stuff that I’m about a couple of months behind on. So it’s good to have that . . .

Mother #814 phase 2

Yeah, definitely my mum [most helpful knowledge] because I don’t really see my health visitor.

Mother #808 phase 2

‘Using the internet and Facebook to get information’

Mothers described using the internet cautiously to find more information on a range of issues specific to premature infants, which they would have liked to have been given before going home. They mentioned using Facebook, Mumsnet and Netmums, NHS websites for illness and BLISS for support. They looked for information about how their baby would be followed up, how many routine appointments they will have, when to use corrected age, why certain tests/scans are necessary and the long-term implications of prematurity. Mothers described feeling stranded in the community and that other people did not understand.

We’ve got like a chat on Facebook, so we put our things on there as well. I’d probably say those mums have been the most helpful, um, because our babies were all premature and very similar age, and if somebody said, ‘My baby is doing this, what do you think I should do?’ And then another mum will say, ‘I tried this, it seems good,’ and it’s you know we’ve all sort of had each other for support. So I would probably say that’s the best really.

Mother #206 phase 2

Because I’ve read things that if they’re not gaining weight . . . that can have effects on their general development. Yeah you see that was me being naughty and Googling. But I was reading articles, research articles, because I have access to the BMJ and stuff for my own job so . . . yeah as opposed to just junk like . . . think we’ve been quite anxious recently.

Mother #112 phase 1

We avoided it [using the internet] when she was in hospital, um, for obvious reasons, um, but it’s more to you know find out what milestones she should be doing, um, for her current actual age, of her corrected age as well. So just sort of see we’re in the right ballpark area of what she’s supposed to be doing, so just for that sort of thing really.

Mother #819 phase 2

There’s only one main website for premature babies, and that’s the BLISS website, but I think again it’s more for support rather than actual information, and I think it is because the ranges are so large, so they can’t give you . . . so you go on the NHS website about weaning and it will say around average of 6 months, but you can’t do that with premature babies, because also they then develop at different rates as well, so everything is just a bit up in the air with them.

Mother #213 phase 2

I felt actually quite stranded about having a premature baby. All my antenatal friends’ babies were all full-term. They’re younger in chronological age than [my baby] but gestationally they were full-term, so it’s all different, and they don’t quite understand.

Mother #719 phase 1

‘I would have liked to have known more about . . .’

There was practical information that parents thought would be helpful in advance such as how to establish a routine at home, how much medicine a premature baby can have, understanding weight and feeding. In phase 2, several mothers mentioned baby tongue tie not being identified before going home, which had contributed to feeding difficulties.

. . . with the breastfeeding side because I was having problems, obviously [my baby] was tongue-tied and I was having real doubts whether or not I could actually carry on breastfeeding when I first brought her home, because it was just so wearing. And she [HV] was really, really supportive in, ‘Well let’s look beyond the problem. Let’s see what’s causing it’.

Mother #318 phase 1

What I did find quite strange, because also, he had tongue tie, which my health visitor only picked up three weeks ago, and she said, ‘I’m surprised they didn’t pick up on it in hospital’.

Mother #213 phase 2

[twin 1] was showing no interest in breastfeeding, and he was really struggling. We have since found out that he had a tongue tie that wasn’t picked up, so he had that cut last week, so he may have been able to feed had it been picked up earlier, but I don’t know, he didn’t seem interested.

Mother #219 phase 2

Well there’s one medical thing, [twin 1] was tongue tied and it was not caught. [When was it caught?] Two weeks after discharge, and it was, found by the cranial osteopathy that we went to, and it’s caused like a lot of feeding problems. If that would have been found a lot sooner I think that we could have probably prevented some of the issues we’re having today with her feedings.

Mother #617 phase 2

Summary

Parents described their time in the LNU as a ‘roller coaster emotional journey’, where ‘they learnt a lot’, ‘established relationships with nurses’ and received ‘support from the ‘NICU family’. They felt prepared practically for going home but in phase 1 many were very concerned about the lack of notice about the expected discharge date and a lack of structured, meaningful discussion about going home. In phase 2, mothers mostly reported being quite well prepared for going home. Mothers expressed emotional uncertainty about going home and some felt that they needed more reassurance. In phase 2, more parents said they felt mentally prepared to take their baby home. In both phases it was acknowledged that breastfeeding was often difficult and getting it right could delay the timing of their baby’s discharge home.

Set against this background of a lack of notice and feeling rushed just before babies were discharged home, the Train-to-Home intervention was introduced. Many parents responded positively to the Train-to-Home package and found it helpful for showing them in a visual way that their baby was progressing. Mothers, fathers and siblings enjoyed using it. Any negative comments were focused around nursing staff attitudes, confusing messages and an apparent lack of understanding or engagement with the package.

Once home, parents had a range of services they could turn to for support and advice including HVs, hospital neonatal outreach nurses and GPs. Several parents remained in contact with others whom they had met on the LNU, and one unit had started a support group for parents discharged from the unit. Mothers described a cautious use of the internet to find more information on a range of issues specific to premature infants, which they would have liked to have been given before going home. They mentioned using Facebook, Mumsnet and Netmums, NHS websites for illness and BLISS for support.

Preparing parents to take their baby home: ‘gaining weight, feeding and maintaining temperature’

During the focus groups and interviews we initially discussed nursing staff views on preparing parents for discharge home and what they thought were the main issues for parents. They all used basic criteria for discharge home from the neonatal unit highlighting that they told parents that babies had to gain weight, feed well (particularly if breastfed), and maintain their temperature before they could go home. They thought that parents worried about coping and feared taking their baby home.

They discussed the factors that facilitated earlier discharge or delayed discharge home. Although they mentioned using ‘aim for your due date’ as a guide, they said that now they advise parents that around 37 weeks’ gestation is a more realistic expectation.

We used to say that babies go home about their due date and anything before is a bonus. We know that they don’t go home on their due date, so now we tend to say . . . it will probably be between 37 and 38 weeks in most cases. Term is anything from 37 weeks. The focus here would be to get them breastfeeding, establish breastfeeding.

There is a perceived feeling that if they bottle feed they will get home quicker. For a mum who is breastfeeding they get frustrated with that last hurdle and they will switch to bottle if most of the others around them are bottle feeding.

The whole communication with parents and working with them has changed quite dramatically really over the last sort of 4 or 5 years really . . . we’re much more focused on their gestation, where they are in their gestational developmental journey and their nutrition. This is linked in with a Trust-wide tool looking at discharge planning.

Unit #1

It’ll be your due date, but we’re not very specific about setting that date. We expect babies to be going home now at 35 weeks. A few years ago they weren’t going home till 38/39 weeks, so our expectations are that we build the parents up to going a bit earlier than that. If they end up staying a bit longer it becomes harder doesn’t it.

Unit #2

Clinical and other justifications for not setting discharge dates were provided by some staff:

Infection keeps them longer, problems with breastfeeding, reflux, poor feeding.

The thing is though when you’re going home receiving breast milk you will have a longer stay. You do get home more quickly when you’re bottle feeding, um, and you have to look very carefully at the support that parents need in the last few days before they go home if they’re breastfeeding, and it’s actually huge.

Babies need to be 35 weeks, feeding well, keeping warm.

There’s a good guide in our admission booklet.

I think mainly breastfeeding, it can take a lot longer for the feeding.

Parents lose all their confidence initially and we have to build that back up again.

Unit #3

Feeding keeps them in a bit longer, but you know when you need to establish breastfeeding it does take that bit longer than if they were going home bottle feeding, and it’s the baby itself though really.

We get quite a few social issues, so they’re ready to go home on a Thursday or Friday but we can’t send them home until Monday so there’s a full care package available.

Unit #4

‘A visual representation to explain baby’s progress’

We sought staff views on having visual materials to help parents understand that discharge is approaching. Staff agreed that this was a good idea but there were very mixed views about using the package on the LNUs. Using it to help explain a baby’s progress was welcomed by some and most acknowledged that it helped to start a discussion:

I think it definitely helped discussion and it did promote conversation, so as a tool it engaged I think people and actually made you focus more on discharge planning and targets and actually engaging conversation. That visual is great isn’t it? And I know for some parents it’s absolutely fantastic, but for others it’s not been.

Unit #1

It looks good, it looks nice and some parents love it, and then they get really cross because they can’t take it home and they say what’s the point? It starts off a discussion. But to be honest some people love it, some people hate it.

Unit #2

A lot of people talk about their journeys really, so why can’t it be a train? I think if you’re going to get a family that, like that way, that visual way of seeing stuff . . . I think it would work. Anything we can provide for parents that is going to make their journey easier, that’s good. I think there’s a place for it but not for every family though.

Unit #3

Thus, although the value of the intervention package was acknowledged by many staff, for some it was implemented in a less than wholehearted way, and a degree of ambivalence was reflected in their responses.

In contrast, staff in the fourth unit really embraced the materials and appreciated the principles involved.

I think traditionally we’ve always told parents that they will go home around the time they had been due, but actually babies don’t . . . they’re going home around that 35/36 week mark, much sooner. By having a more structured plan to prepare them for home . . . they’ve got a date to work for. It’s the first question they always ask, ‘When can we go home? When can we take our baby home?’ It’s one you can’t properly answer, but this gives them a bit more of an idea when to expect them home. It allows them to prepare at home, so family coming to support them, that sort of thing.

Unit #4

Seeing that visual representation that their baby is getting better; that they are making that journey towards home. It’s fantastic; it’s been a really good way of engaging the parents with their baby’s progress. It’s definitely a good sort of starting point for communication with them. Dads love it. They also appear with the parents talking to the relatives, and they’re explaining it and they feel empowered.

I think it should be used with all our parents. They can all understand it and because it’s quite visual you can explain it to them in a way that they understand. Once they’ve been given it they’ve all been really interested.

‘A chance to sit down with the parents’: using the train on the unit and explaining the dates

The dates on the Train-to-Home package were usually changed following a discussion with doctors. Most nurses struggled with explaining the dates to parents but in one LNU nurses found that talking about the dates and changing the window colours helped them to promote discussion with parents and explain how their baby was progressing.

I think initially the parents really like the train, they see it as something quite positive, and they like the dates. But the problem we get is that as it comes nearer the time and they don’t necessarily meet their dates, I know that you can change the dates, but there’s no guidance on how to change them. So it’s quite hard to know what to change them to, because it’s almost a guess, because feeding is a bit you can’t tell, you can’t predict how long it’s going to take somebody to feed, and when they don’t reach those dates or you change the dates then they get quite disappointed. So initially it’s quite positive, and for the babies that become unwell during their journey and go backwards, that’s really difficult for them. We’ve only had a few of these though.

Unit #1

The first date is always very optimistic for what happens in our unit. I found the dates difficult as there was no guidance on how to extend them rather than narrow the dates.

The uncertainty was emphasised and the big differences between individual babies. Some staff expressed a need for more guidance on changing the dates, with some feeling that this was not their role.

It’s a huge difference between each individual child. So once you start writing the date on there then the parents are like glued to that date, and it’s a bit difficult to then try and move it. The consultant would actually write the dates and explain to the parents.

Unit #2

We tend to do it quite often on the doctor’s round, often the doctors will alter the date or put a date on if there isn’t one and sometimes speak to the parents about it as well.

Unit #3

Others relied on a shared discussion and emphasised the ‘rough guide’ aspect of initial dates, with a more flexible and reflective attitude to changes in the likely timing of discharge home.

We change the dates in discussion with the doctors. Sometimes the frame is just a little bit too narrow. I don’t think the doctors have been so on board. I think we are because we’re face to face and there’s so many things that . . . and for me it’s an excuse to sit down with the parents and discuss the train and feeding plan.

Unit #4

I think it’s been lovely to be able to give them an idea right from the word go. I really liked that, because that is the one thing that they always ask is you know, ‘When will my baby be going home?’ It’s like, ‘Well, this is a rough guide,’ and what I do feel we haven’t done well is to reassess it nearer the time that they’re going home, yeah, keep reassessing it.

‘Intensive care may be too early?’: when to introduce the package

A number of staff on three LNUs thought that it would be better to use the package in the lower-dependency areas after the first weeks of intensive care. However, following the implementation and use of the Train-to-Home package, the fourth LNU was keen to use the package as soon as babies were admitted and wanted to continue with its use.

There’s that anxiety of moving from high dependency to low dependency where there’s less nurses and, you know, babies aren’t monitored as much, but it gives them a focus they’re actually going home. And actually what would be really nice is thinking of when we’re working with parents and actually empowering them beginning to sort of hand over the reins, because that’s actually a really difficult transition for some parents, especially when they have been in for a long time, there’s actually that nervousness of going home, and taking over that sort of care, and we’re obviously trying to empower at an earlier stage with teaching parents to tube feed and take on more sort of in high dependency. But as they come round the unit, having that tool where they could actually start to share some decision-making.

Unit #1

Use of the train as a signal to parents was preferred by some staff once a baby was moving on from intensive care and in some instances from high-dependency care, and they saw the train and the booklet as a way of providing information for parents at, what they perceived as, a more appropriate time, bearing in mind the needs of the baby and staff workload:

I think there’s more time in low dependency. You’ve got also more of an idea of when that baby is likely to go home, because you don’t know if his journey in high dependency is going to be two or three days or a long time. Whereas once they have gone into low dependency normally it’s just growing and just getting . . . so they’re heading towards home. And parents relax more about it don’t they? They start enjoying their baby rather than being worried about it. To start you’d almost give them something different to focus on wouldn’t it? Once the baby went to low dependency, right we can start thinking about discharge now.

Unit #2

They have a lot going on in the first week or couple of weeks and they don’t remember the peripheral stuff, so intensive care is too early.

Unit #3

I think when they get into the special care side it’s probably good . . . to introduce them, only because in ITU and high dependency there’s quite a lot going on, and things changing quite quickly. When they get to special care they can be there for a couple of days, but some could be there for a couple of weeks.’

There is that negativity built up about it, because some staff have felt that, obviously it’s an important thing, but to staff who are trying to save lives actually it comes way down on the list. So it’s getting . . . I think we need to sort of be allowed to use it and introduce it when it’s appropriate, and it’s getting the timing right for the parents and for the workload of the staff to be quite honest. If we did take this forward I think the intensive care side is too early.

However, not all staff saw it that way, particularly those in the fourth unit:

I think overall it’s worked really well, it’s been quite positive for the parents, which is the main thing really. We just do it every day now, you kind of get used to doing it, so it’s part of the daily thing you do with the parents where we’ve figured oh we’ve got to get the train, need to do this, need to do that, we just do it now.

Unit #4

The parent pathway booklets: ‘some parents used it like a diary’

Nursing staff did not report seeing parents using the booklets much. It was also clear that most of them did not offer parents guidance on using them.

It’s interesting looking at the moment at some of the developmental work that’s being looked at and they’ve got the pathway there which is a similar sort of pathway over how you would expect a journey to take place. And I think the whole pathway idea working with parents is a really positive one. But I felt that it sort of almost made it over simplified . . . I think you almost needed to have a more individualised approached, so to use the pathway as an overarching goal, target, but to be able to have individualised it a little bit more.

Unit #1

Some parents would look in the book and literally they would have essays written on every question. Other parents would never write in it. They might have read it but they never wrote in it. So it just was an individual thing really. But some parents literally would write everything down. It got better with the booklets, and the booklets were quite useful I think.

Unit #2

I haven’t actually, no I haven’t actually seen them [parents] use them [booklets]. Um, I’ve seen them, but my knowledge on them is not the best. Is that something that would be kept . . . would usually be kept on the cot side with them?

Unit #3

I think they’ve asked about the temperature more and things like that. They feel they’re able to ask, because they’re involved in this idea of the colour stickers. It’s almost like they need the verification of the book to say to you, ‘In the book it says . . .’ so does that mean that I can? . . . Some of them used them daily and read you know questions out, or underlined things, and wrote how their babies were doing, like a diary really.

Unit #4

Ongoing use of the Train-to-Home materials

Responses to continued use of the materials after phase 2 of the study was mixed: one LNU was negative, another was very positive and the others were ambivalent. There was some discussion about possible formats for future versions including a website or an app for parent use.

I think having a generic tool that you can then adapt and working with your own individual unit culture or own flow, or design is actually really helpful. So I think any electronic system that you’re talking about, or apps or things like . . . are going to be really useful because then actually you can mould it slightly can’t you and make it individualised.

Unit #1

Others were satisfied with what they already had in place:

I personally think we are really good at our discharge planning. I think because of our tick lists we go through all those things, and I think that works really well. We’re good at our planning, I don’t say we’re necessarily good at our dates. We are not using the trains any more, I haven't seen any around.

Unit #2

However, some staff recognised the value of this kind of innovation for new staff and as a way of working with parents:

I think if I was a new member of staff, obviously I’m not, but I would actually almost use this as, well as a learning tool, to have those conversations with people, and you know, if you don’t know the answer, if a parent asks you something then you find out somebody who does know that. So, you know, there’s quite a good learning tool I feel for new members of staff.

Unit #3

I would be quite upset if they were taken away. Because it’s been so positive, and I have seen that, and that’s really nice, you do feel you’re making a difference for the parents.

Unit #4

We’d really like to keep the trains. The visual aspect of it and actually find a way of extending that to all our babies, not just those babies that fall inside the criteria. That would be our ideal. There’s a lot in the NHS at the moment about all patients having this journey card and having the discharge dates, and knowing what’s going to happen next and that sort of thing. We would really like the train to be ours rather than it being an odd bit of like postcard that gets tatty and forgotten. This is just a really good way of us doing that with our parents and giving them a way forward to work towards going home. It’s been brilliant, and they love changing those stickers and going to the next colour.

Unit #4

The NHS initiative on discharge planning was mentioned by two groups of nursing staff and they considered the Train-to-Home package fitted well with this approach because the pathways focused on developmental care. Staff told us that they have started to focus on babies’ stages in their gestational developmental journey and our approach seemed to fit with this. They thought that having more time to embed the materials into each LNU was needed. This normalisation period would allow more individuals to be trained and promote optimal delivery of the package.

Comparison with other studies

Relatively few studies have investigated parent-centred approaches to improving parental involvement in the care of their babies while in neonatal care. We are not aware of any studies that have systematically attempted to assess the impact of such an intervention on parental self-efficacy or babies’ LOS.

In the McMaster University study, Gaal et al. 43 observed that a major barrier to effective implementation of the intervention was a ‘failure of physician endorsement through action’. She noted that ‘the neonatologists supported the committee’s efforts, readily shared personal philosophies and practices, and yet remained disengaged from incorporating the train into daily bedside management rounds’.

In our study we found that the consultant medical staff supported the principles underpinning the intervention and were enthusiastic about the study aims and objectives. However, they identified reluctance from nursing staff to engage fully with the intervention and they were unwilling to promote it when nurses were unenthusiastic.

Nursing staff feedback confirmed this. Some of the more experienced nurses did not consider that the intervention materials were significantly different from what they already did for parent education and preparation for home. This echoes the findings of Gaal et al. 43 – that some ‘staff saw the discharge planning pathway as a barrier because of its perceived redundancy’.

Some of these implementation difficulties were recognised by parents, who reported nursing staff ambivalence about the intervention. Parents found the intervention helpful, informative and reassuring as they became more engaged and involved in understanding their baby’s progress through the LNU.

Family-centred care and planning for discharge has been emphasised for several years in Department of Health Neonatal Toolkit documents,6 NICE guidance25 and in the POPPY systematic review. 21 Despite this rhetoric around family-centred care in neonatal units, discharge processes continue to be reactive and poorly planned. 24,27 A recent study in NICUs in England showed that mothers did not report positive family-centred practices. 27 Mothers experienced a state of liminality (a state of uncertainty or ambiguity about their role) and were acutely sensitive to inconsistencies in care. They formed deferential relationships with health professionals and remained in a state of anxious surveillance. 27 The study concludes that it is important to minimise the consequences of the liminal experience in the NICU, improve staff–mother interactions and facilitate mothers’ opportunities to be primary caregivers. 27 Our study implemented a family-centred care package and, where it was adopted by the staff, changes were seen in parental confidence in caring for their baby by the time they left the neonatal unit and in the weeks following discharge home.

Dunn et al. 48 developed a web-based family care map to support family-based care practices in neonatal units in New England, USA. Their web-based tool was felt to be the ideal medium to assist individual care providers and family advisors to provide comprehensive family-centred care to infants and families. They concluded that implementation of the family-centred care map had the potential to affect positively the quality of newborn intensive care and lead to improved long-term outcomes. 48 Our Train-to-Home package would also be enhanced by being web-based, to enable staff and parents to access the parent pathways to facilitate parent communication with staff. It would help parents to ask pertinent questions and promote discussion to increase their confidence about taking their baby home after their extended stay in hospital.

Maternal self-efficacy studies have explored the associations with future maternal behaviour. 33,36 Our study measured maternal self-efficacy at three time points and, although there were no significant changes in measured maternal self-efficacy using the PMPS-E, parents reported feeling more confident and better informed about how to care for their babies in phase 2. This is supported by the observed reduction in unplanned ED attendances. This is an important finding, as reducing ED use can save money for the NHS.

Working with the Parent Advisory Group and Clinicians Advisory Group

Our PAG initially comprised eight mothers with experience of a baby in neonatal care and they were very helpful in designing the parent-focused materials. Their contribution to the Delphi process in the development of the study documentation was invaluable. We managed to keep contact with five of them throughout the study, but the other three moved away from Bristol and were not able to attend. Their interest and involvement with the study progress helped to keep the parent perspective firmly in mind when discussing recruitment difficulties and interpreting the findings of the interviews. Their contributions to Project Steering Group meetings were also very valuable.

Clinicians Advisory Group meetings provided a lively discussion forum for developing and refining the intervention materials, and contributions made by the members of this group were central to the development of the study documentation. The Delphi process undertaken to develop and refine the study materials was time-consuming and we are particularly grateful to the members of the CAG for their enthusiastic involvement in this process.

Strengths

Despite involving families with very young preterm infants, at a time when they were under considerable stress, there was a high level of involvement and commitment to the study by those families taking part.

The level of patient and public involvement – particularly in the development and implementation of the intervention – was high, with a PAG being actively involved at all stages.

The development of the intervention involved staff input from all of the four neonatal units taking part in the study. This ensured that the intervention materials were compatible with existing protocols and practices in all participating units.

The four neonatal units involved were self-selected, all having approached the research team in response to publicity about the proposed implementation of the intervention.

The holistic approach used ensured that we identified a wide range of effects of the intervention, on both staff and families.

Limitations

Despite the involvement of staff in the participating units from an early stage we experienced difficulty in some of the units in obtaining full engagement with the nursing staff. Many of the nursing staff showed poor understanding of the difference between the intervention and the research. These issues may have been related to poor cascading systems for changes in practice. Cascading of learning about the intervention in some of the participating units was patchy and limited, and a longer period of time to embed the intervention into practice would have been helpful.

The use of ‘nurse champions’ recruited from the LNU staff (and funded via the Clinical Research Network) did not work well, as not all had volunteered for this role, and clinical requirements were a much higher priority for all of the nurses involved.

As a consequence of the limited time availability of the nurse champions, there was commonly some delay in obtaining assent from families, which led to variation in timing of the initial PMPS-E questionnaires. This may have been an important contributory effect in the lack of identifiable changes in this measure. It also highlights a significant limitation of using such a measure without dedicated research staff to ensure prompt and effective data collection.

In common with most studies involving families from wide socioeconomic backgrounds, recruitment of families from the more-deprived groups was lower than from the less-deprived groups.

1. Babies born at 27–30 weeks’ gestation

In the first week after birth, estimate the likely discharge date, as shown in Table 29.

Then adjust the estimated date of discharge according to the baby’s progress and condition as follows:

• second or third week after birth: aim to narrow the range of dates slightly to a 5- to 7-day range.

• by 1 month: aim to narrow the range further to 2–4 days.

2. Babies born at 31–33 weeks’ gestation

For these more mature babies it is may be easier to calculate the estimated date of discharge in days from the date of birth.

In the first week after birth: estimate the likely date of discharge as shown in Table 30.

Adjust the likely date of discharge according to the baby’s progress and condition as follows:

• second week: aim to reduce the range of dates

• third week: aim to narrow the range further to 2–4 days.

Slide 1

Slide 2

Slide 3

Slide 4

Slide 5

Slide 6

Slide 7

Slide 8

Slide 9

Slide 10

Slide 11

Slide 12

Slide 13

Slide 14

Slide 15

Slide 16

Slide 17

Slide 18

Slide 19

Slide 20

Slide 21