Use of selective gut decontamination in critically ill children: PICnIC a pilot RCT and mixed-methods study

Handling of missing data

The proportion of variables included in the analyses that were missing were reported. No measures were taken to replace missing values.

Patient and public involvement

Engagement with patients was vital to the successful conduct of the PICnIC pilot RCT and included the mixed-methods approaches (see Chapter 4), along with involvement of patient stakeholders in the development and oversight of the PICnIC RCT. A parent of a child who was admitted to a PICU with severe infection provided input to the PICnIC pilot RCT design. A former patient on a PICU (IS) is a coinvestigator and provided patient input for development of the pilot RCT, including reviewing the study literature given to patients and their families [e.g. participant information sheets (PISs)]. In addition to this, patients and members of the public were engaged through the PICnIC qualitative study, which has provided invaluable insight that has been incorporated into the recommendations arising from the overall study.

Design and development of the protocol

The PICnIC pilot RCT protocol was designed to inform key parameters and to test the design and possible conduct of the proposed definitive PICnIC RCT.

Amendments to the Infection Control in Paediatric Intensive Care pilot randomised controlled trial protocol

Following receipt of approval from the HRA on 20 November, 12 non-substantial amendments were approved and categorised.

National Health Service support costs

The NHS support costs were agreed prior to the submission of the research grant to include screening to identify eligible patients and obtaining informed consent from parents. These included costs for microbiology swab tests and the administration of SDD at sites allocated to intervention in Period Two of the study.

Site selection

Six sites that had previously expressed an interest in participating in the PICnIC cRCT were selected. Potential sites were asked to complete a site feasibility questionnaire to confirm eligibility by the ICNARC CTU. None of the sites had experience of delivery of SDD in clinical care. Two sites expressed interest as reserve sites but did not need to participate. The reasons for selection included a good research track record, a display of a high level of enthusiasm and ensuring a good diversity of sites.

Site initiation

Site teams from all six participating sites attended a site initiation meeting prior to the commencement of patient screening. These were held online to minimise the risk of infection transmission during the COVID-19 pandemic. During the first meeting, the chief investigator presented the background and rationale for the PICnIC pilot RCT. The meetings included discussion of the protocol, screening and recruitment of patients, data collection and microbiology screening procedures.

Following randomisation, a second site initiation visit was undertaken for the three sites randomised to SDD, where the local PIs and pharmacists attended and were given information on the protocol for prescribing and administering the SDD oral solution and paste. The operational challenges of conducting the PICnIC pilot RCT at sites were discussed, including strategies for communicating the study to all PICU staff. The PI and trial manager maintained close contact with these sites to address any questions that arose as the clinical teams integrated SDD into their infection control procedures.

Investigator site file

An investigator site file was provided to all participating sites. This contained all essential documents for the conduct of the PICnIC pilot RCT and included the approved protocol; all relevant approvals (e.g. local confirmation of capacity and capability); a signed copy of the clinical trial site agreement; the delegation of trial duties log; and copies of the approved PISs, parent/legal representative consent forms and all standard operating procedures (e.g. for screening participants, delivery of the intervention, obtaining informed consent and entering data onto the secure, dedicated eCRF). The site PI was responsible for maintaining the investigator site file.

Pilot cluster-randomised controlled trial eligibility criteria

Ecology period eligibility criteria

Baseline patient characteristics

Baseline data were collected at critical care admission via data linkage with PICANet, and directly via trial case report forms (CRFs). The following baseline demographic and clinical data were summarised overall and by allocated treatment group and study period:

• age (years) – median [interquartile range (IQR)]

• age group (< 1 year, 1 year, 2 – 4 years, 5 – 9 years, 10 – 16 years) – number and %

• sex (male, female) – number and %

• weight (kg) – median (IQR)

• ethnic group

• Paediatric Index of Mortality (PIM) 3 score18 – median (IQR)

• main reason for PICU admission.

Outcome definitions

The outcomes of this study were selected to measure the respective objectives as previously stated and were focused on assessing the feasibility of a larger-scale definitive study. Specifically, these were as follows:

• The ability to randomise PICUs to either control or intervention was assessed by the successful random assignment of three PICUs to the intervention without delay to subsequent phases of the trial.

• The willingness and ability of healthcare professionals to screen and recruit eligible children were assessed by the proportion of eligible children recorded on study screening logs successfully recruited to the pilot cRCT as a percentage of total eligible admissions, and the reported reasons for non-recruitment.

• The potential recruitment rate for a future definitive cRCT trial of SDD-enhanced infection control in eligible children was estimated by combining the proportion of eligible children recruited to the pilot cRCT with the size of the potentially eligible population (estimated from nesting the screening log data from participating PICUs with the national UK PICU data from PICANet).

• Adherence to the SDD protocol was assessed using the proportion of eligible children allocated to the intervention receiving (1) both elements and (2) each individual element of the SDD intervention, the number of days on which these elements were received relative to days eligible for the SDD intervention and the reported reasons for non-adherence.

• Procedures for assessing and collecting selected clinical and ecological outcomes and for AE reporting were assessed using the proportion of children with complete data for these outcomes (as listed later) including, for ecological outcomes, the proportion consenting to additional study specific sample collection.

• Generalisability of the study results to all UK PICUs was assessed by comparing baseline characteristics (as listed earlier) and potential outcome measures (as listed later) for children recruited to the pilot cRCT with data from all potentially eligible children (receiving invasive mechanical ventilation for at least three calendar days) within participating PICUs and within all UK PICUs (from PICANet).

With the aim of understanding potential patient-centred primary and secondary outcome measures for the definitive cRCT, the following potential outcome measures were reported by arm and trial period (during the ecology surveillance weeks, only HCAIs and positive microbiology results were recorded):

• HCAI (confirmed/presumed) – as defined by a local microbiology and clinical team as infections acquired during the PICU stay (presented as number and % of enrolled patients)

• any positive microbiology swab or sample – result, n/N (% of enrolled patients)

• duration of invasive ventilation (days) – mean [standard deviation (SD)] and median (IQR)

• days alive and free of ventilation to day 28 – median (IQR)

• length of hospital stay days from confirming eligibility to hospital discharge – mean (SD) and median (IQR)

• length of stay in PICU: hours from confirming eligibility to PICU discharge – mean (SD) and median (IQR)

• PICU survival: status (alive or dead) before PICU discharge – number and %

• hospital mortality: status (alive or dead) before hospital discharge – number and %.

• mortality within 30 days post enrolment (alive or dead): from the patients’ survival status as of 30 days post enrolment on the CRF – number and %.

Safety monitoring and adverse events

Adverse event reporting followed the HRA guidelines on safety reporting in non-clinical trial investigational medicinal product studies.

The ICNARC CTU monitored data for documented AEs that are not considered to be related to the trial treatment. In the event that any trial procedure appeared to be resulting in AEs, the TMG was contacted for their opinion.

Data management

Participant data were entered onto a secure web-based data entry system. The site PI oversaw and was responsible for data collection, quality and recording. Collection of data could be delegated by the site PI to qualified members of the research team recorded on the Delegation Log. All data were collected and processed in line with General Data Protection Regulation (GDPR). Data entered onto the secure trial database underwent validation checks for completeness, accuracy and consistency of data. Queries on incomplete, inaccurate or inconsistent data were sent to the research team at participating sites for resolution. During the conduct of the trial, all electronic participant data were encrypted, and all trial documents stored securely at the site or the ICNARC CTU, as appropriate. On completion of the trial, all participant data (electronic and paper) and other trial documents were archived securely and will be retained for 10 years at the site, the sponsor or at the ICNARC CTU, as appropriate.

Screening and eligibility

The numbers of participants screened and deemed eligible are reported overall and by site.

Recruitment

The numbers of participants enrolled, consented (for study-specific procedures and use of their clinical data) are reported by site and treatment summarised as a Consolidated Standards of Reporting Trials (CONSORT) flow diagram. 20 The numbers of patients enrolled are reported per site and period of the study. When reported, the reasons for exclusion and non-consent are also summarised. Recruitment to the pilot cRCT was summarised graphically and presented as a rate of patients per week, overall and by site and study period.

Baseline demographic and clinical data were summarised for the trial population overall and by treatment group and study period. In addition, using PICANet data, patients’ characteristics are reported for potentially eligible children within participating PICUs, and within all UK PICUs, by study period. There was no statistical testing for any of the summary measures while comparing the baseline variables.

Adherence and protocol deviations

The number of screened patients, and the number and percentage of patients found to have been ineligible, together with the reasons for ineligibility (inclusion criteria not met or exclusion criteria met), were reported by treatment group and study period.

The number and percentage of patients with at least one protocol deviation in the SDD intervention group was reported, and the total number of such deviations. A deviation was defined as any of the following:

• SDD treatment starts > 6 hours after being identified as eligible

• SDD treatment starts within 6 hours of being identified as eligible but continues for more than 30 days (treatment period) or finishes before the patient is extubated/no longer mechanically ventilated

• dose of one or both SDD treatments were not given, and patient was mechanically ventilated

• dose was wrongly administered (four per day, every 6 hours while intubated).

Analysis principles

Data completeness of clinical and ecological outcomes and for AE reporting are summarised using counts and percentages, by intervention group and treatment period.

Potential outcome measures were summarised using counts and percentages (for binary outcomes), median and IQR (for all continuous outcomes), and means and SDs (for length of stay and duration of treatment), and they were reported by intervention group and study period.

To account for cluster randomisation, multilevel logistic or generalised linear regressions were used to estimate potential treatment effects. The effect estimate was calculated as the interaction between treatment group and time period and reported either as an odds ratio with 95% confidence interval (CI; for binary outcomes) or as difference in means with 95% CIs (for continuous outcomes). These calculations were done only to inform planning of a definitive trial and should not be used to draw any conclusions regarding differences in outcomes between treatment groups.

To determine the most appropriate primary outcome for a definitive trial, for all potential outcome measures the number of patients with complete data in each treatment group are reported. For measures requiring data linkage with routine data sources (PICANet), the proportion of successfully linked records are reported.

Missing data

To assess the follow-up procedures, the number (%) of participants with complete follow-up data for each of the potential outcome measures for the definitive RCT, overall and by treatment group, is reported. All analyses were undertaken in the trial population. There was no imputation of missing data.

Harms

The number and percentage of participants experiencing each prespecified AE (plus any other AEs as reported) while in PICU were collected for each treatment group. Numbers of serious AEs (SAEs), severity and reported relatedness SDD administration in the intervention sites during Period Two are also reported.

Eligibility criteria

Inclusion criteria:

• parents/legal representatives of children involved in the pilot cRCT, including those who withdraw from data collection.

Exclusion criteria:

• parents/legal representatives who do not speak English.

We use the term parents to include legal representatives from this point forward for brevity.

Informed consent

The University of Liverpool researcher (MP) contacted parents/legal representatives to arrange an interview within 1 month of consent (or withdrawn consent) for the pilot cRCT. Parents were offered either a telephone or online video conference interview via Zoom. Informed consent for participation, for audio recording and to receive a copy of the mixed-methods study findings was sought verbally before the interview commenced. This involved the researcher (MP) reading each aspect of the consent form to parents. Each box was initialled by the researcher on the consent form when verbal consent was provided by the parent. Informed consent discussions were audio-recorded for auditing purposes.

Conduct of the interviews

All parent interviews were conducted by MP using the parent/legal representative interview topic guide (see Appendix 2 and 3). Respondent validation was used so that previously unanticipated topics will be added to the topic guide and discussed with participants as interviewing and analyses progress. An example was additional questions to explore parental understanding about whether or not their child would have received the intervention if the pilot cRCT was not being conducted. A distress protocol was available but did not need to be used. After the interview, participants were sent a copy of the consent form and a thank you letter, including a £30 Amazon (Amazon.com, Inc., Bellevue, WA, USA) voucher to thank them for their time.

Informed consent

Informed consent for participation, for audio recording and to receive a copy of the mixed-methods study findings was sought verbally before the focus group commenced. This involved a member of the University of Liverpool team reading each aspect of the consent form to individual practitioners in Zoom breakout rooms. Each box was initialled by the researcher on the consent form when verbal consent was provided. Informed consent discussions were audio-recorded for auditing purposes.

Conduct of the focus groups

Focus groups began with introductions and were semistructured, informed by the topic guide. Unanticipated topics were added as data collection and analysis progressed. Voting took place via Poll Everywhere throughout each focus group on topics as they were discussed. Once the focus group was complete, participants were thanked for their time.

Paediatric intensive care unit staff characteristics

Survey participants included nurses (23/44, 52%), doctors (19/44, 43%) and pharmacists (2/44, 5%). The focus group sample included nurses (17/26, 65%), doctors (8/26, 30%) and one pharmacist (1/26, 5%). Those who took part in the focus group may also have taken part in the survey.

The majority of participants were involved in the clinical care of children. All units were research active and the majority of participants had been actively involved in previous trials. Only six survey participants indicated they were new to research and had no prior clinical trial experience. None of the units were using the SDD intervention before the PICnIC pilot trial.

Focus groups took, on average, 70 minutes (range 46–100 minutes).

Parent characteristics

The majority of parents identified their ethnic group as being white (British n = 13, other n = 3), with the rest identifying as African (n = 1), black British (n = 1) and Pakistani (n = 2; missing = 3). The majority of parents had non-medically related backgrounds (n = 22) and most parents were in employment (n = 18).

Parent interviews related to 23 children (17 boys) between the ages of 4 weeks and 12 years (average 2.6 years). Parents were non-bereaved, while one child had been transferred to a child’s hospice, but was back home at the time of the interview. The main reasons that children were admitted to PICU were viral infections or conditions caused by it (rhinovirus, respiratory syncytial virus, bronchiolitis, COVID-19, meningitis), conditions affecting the airway (laryngomalacia, tracheomalacia) and bacterial pneumonia. Other reasons included trauma, cardiac arrest and neurosurgery.

Interviews took, on average, 54 minutes (range 16–70 minutes).

Acceptable screening process and inclusion/exclusion criteria

Most survey and focus group participants indicated they were satisfied with the screening process and ‘quite happy with the inclusion and exclusion’ (P17, FG5, control, PI/doctor) criteria. However, three nurses and multiple focus group discussions suggested that site staff found aspects of checking patient eligibility challenging, which led to eligible patients being missed or excluded unnecessarily. Discussion centred upon the need for staff to decide whether or not the patient would be mechanically ventilated for at least 48 hours in order for them to be eligible. As the quotations below illustrate, this was difficult to predict with any certainty at the point of PICU admission, which led to potentially eligible patients being excluded:

I think the main issue with recruitment was this wobbly window about they’ve got to be ventilated for 48 hours. If you ask any of them they’d say, ‘We haven’t got a crystal ball, we don’t know’.

P10, FG3, intervention group, research nurse

It’s hard to predict the future. We could have a good idea, but sometimes you think they were going to walk out the door the next day but it changed.

P03, FG2, intervention part 2, research nurse

Although some patients did become eligible the day after admission, they often were missed as screening only took place at admission.

We think, ‘This person came in, they’ll be extubated tomorrow’. Then, the day after, it’s like, ‘Great, they didn’t’. Then, if I’m not here and I’m like, ‘Guys, it’s okay, we can screen them today’, they’ve already forgotten about it. Because if they don’t do it for admission, then it’s gone … I find that the team then struggles to have that click of, ‘Yes, we can include them now’.

P01, FG1, intervention site, doctor

One unit described how screening would take place on multiple occasions for the same patient, which would help identify children who became eligible sometime after admission. For others, ‘re-screening’ was not possible due to staffing constraints. Research nurses also described how some patients were no longer eligible when ventilation was removed fairly soon after admission, or the time window to administer the SDD intervention had passed by the time they had screened and found a clinician to prescribe the SDD.

Not user friendly for clinical staff (i.e. ability to rescreen the patient if unexpectedly remains ventilated).

P20, research nurse, pilot site survey

It was very much a research led, chasing the doctor-screening and chasing doctors to prescribe, sort of thing. It wasn’t the clinical team, and we sort of explained their load and the fact that it was just such a short period of time that we were doing it, that obviously they were probably never going to catch on, because by the time people had started saying, ‘Oh, actually this patient–’ you’re like, ‘Oh, we’re not doing it anymore, sorry’. That sort of thing.

P04, FG2 part 2, intervention group, research nurse

Research nurses used free-text responses in the survey to highlight the need to clarify the screening process. Some suggested widening the inclusion criteria to all patients, or all ventilated patients and re-screen, to help prevent potentially eligible patients being excluded at the point of admission.

Perhaps include all ventilated patients at admission? Patients rescreened daily – if not expected to be ventilated ‘day after tomorrow’ then they were repeatedly excluded.

P21, research nurse, pilot site survey

I guess the only question would be whether or not it would be all patients, irrespective of whether they were invasively ventilated, if you were doing it across the whole unit, would you just give it to everybody who was admitted if you really wanted to look at embedding it in practise. Whenever you’re wanting something to happen, it’s easier if you say you do it for everybody rather than having subsets of patients. Even if they’re a large, you know, let’s say 70% + of all our patients are intubated at some point. It’s still potentially easier if you just say, ‘Let’s do it for everybody’.

P17, FG5, control group, PI/doctor

Support for cluster randomisation in the proposed trial

During focus groups, there was a general consensus that cluster randomisation was acceptable for the proposed PICnIC trial and ‘I think there’s no other, that I can think of, way that you’d do it’ (P05, FG2 part 2, intervention group, PI/doctor). Three survey participants used open text responses to state that ‘Cluster/step wedge RCT [is the] way to go’ (P29, doctor, pilot site survey). When asked to consider whether individual randomisation would be preferable, staff described how this may have caused confusion:

I think the choice to cluster was a good plan. I think that would have caused a lot of confusion if we were randomised to that within each unit, so I think the clustering seemed to work very well. It would be a good idea for this sort of study.

P17, FG5, control group, PI/doctor

Approach to recruitment and consent including samples to assess clinical and ecological data

All but one practitioner involved in recruitment to the pilot trial (22/44 survey participants) stated the PICnIC recruitment process was acceptable (16/22, 70%), or very acceptable (6/22, 26%). The one recruiter who did not find the process acceptable ‘disliked assumed consent’ for the SDD intervention.

Across focus groups, practitioners commented on how parents did not comment or react negatively to their child being given SDD, or to having routine care samples taken, without their prior consent. Many implied that consent did not need to be sought for the intervention or routine sample aspects of the study as parents were not dissatisfied that they had been taken.

I don’t think, as a general feeling, I don’t feel that we had any parents who have been particularly outspoken about it.

P06, FG2 part 1, intervention site, research nurse

I found that they were never unhappy about us collecting routine samples, and particularly if you said that this is our standard practise. They were very happy with that.

P16, FG5, control site, research nurse

Some staff at intervention sites were surprised that parents had not expressed concern about their child being given antimicrobials in a research study without their prior consent. They reflected on how it may have been how they had presented study information to families. As the following quotation highlights, accessing SDD may have been presented and therefore viewed in a positive light, with parents hoping that the intervention may help their child:

I don’t know if it’s the way we explain the medication to them, as well? I make a point of telling them some centres are doing it, others aren’t, because we want to see if it helps them, in a way, to give this as part of infection control, prophylactics. I think maybe they see it in a positive way? Like, ‘We’re getting it, so hopefully we won’t grow any bugs’. I don’t know if that’s the way they see it.

P01, FG1, intervention site, research nurse

Those involved in pilot trial recruitment described how parents were not concerned that prior consent had not been sought as they had been informed that samples were part of standard care in their unit, or that other sites were using SDD.

I don’t think many parents minded because we hadn’t really done anything to their child that was out of standard care anyway. So I think they’d have probably had more – they’d have been more upset about it if we’d have done things to the child, if that makes sense.

P20, FG6, control site, research nurse

However, for some sites, admission samples, referred to as ‘routine or standard samples’ in the study, were not part of standard infection control measures for their unit. Nurses described how they felt uncomfortable collecting these samples, particularly for older children or those who were awake, as they were not routine for their unit:

They were awake patients, often older and they were difficult to approach. Also, it felt a bit wrong that we were asking for all of these admission swabs to be done on them that would not normally be done on them ...

P12, FG4, control site, research nurse

After admission samples, additional samples were taken twice weekly until discharge from the nose, rectum and wound (if present). Practitioners spoke of how parents’ responses to collection of these additional samples varied, which was in contrast to the collection of admission samples, or administration of the SDD intervention. Some parents did not consent for this aspect of the pilot trial as ‘they just did not want to put their child through anything extra’ (P14, FG4, control site, PI/Doctor), or had declined at a later date when parents decided the swabs were upsetting their child.

I’d say I’ve had a few parents who when going through the consent form after reading the information sheet, they have said that they just point blank refuse any rectal swabs unless they’re needed, because … adding more additional tests and things that their child doesn’t clinically need. That distresses the parents.

P15, FG5, control site, research nurse

Not the initial ones, they know, when we explain, say, ‘They’re samples that are taken as part of their routine care. They would have the regardless of the PICnIC study or not’. The second, the extra samples, very varied. Some are very anti it, some were willing.

P04, FG2 part 2, intervention site, research nurse

As much, you know, the PIS sort of states that it’s – they won’t harm your child or they won’t cause any pain. Well, actually, you know, putting a swab up your bottom and a swab up your nostrils, I mean even though you’re a bit sedated, I don’t think that’s the most pleasant thing … we’d go back maybe like the third time, the third set of swabs, or something, they were like, ‘Actually, the child’s quite scared now of people coming up and doing things’. We’re like, ‘That’s absolutely fine, thank you so much for agreeing to other things’.

P06, FG2 part 2, intervention site, research nurse

Practitioners at one control site described how they did not to approach parents and take rectal samples due to how critically ill the patient was, or the ‘parents’ personality’:

If they’re in a really critical situation, I don’t want to then go and speak to them about rectal swabs because in the grand scheme of things, it’s just not that important.

P16, FG5, control site, research nurse

The sample processing increased workload for microbiology labs. At one site, the staffing and pandemic pressures meant additional discussion was undertaken at local level to allow set-up.

You know, the microbes. I do know that there are many. Say, for our unit, staph aureus … Candida was not a part of our routine. And vancomycin resistance was also not part of our routine swabs … With COVID and the lab running to almost 300% of its capacity, it took a lot of negotiation to agree to get them to even do this.

P02, FG1, intervention site, PI/doctor

Such discussions might have been better before rather than after the site was included in the study and this will need to be part of the set-up discussion in any future definitive study. Microbiologist discussion will need to include the specific expectations of the study to allow appropriate ecology screening and the resources to enable this within the local laboratories.

Not an opt-out approach

Focus group discussions and a few open responses in the survey indicated that there was some confusion about the consent process being described as an ‘opt-out’ approach in the protocol as ‘they can’t opt out of having the drug’ (P06, FG2 part 2, intervention site, research nurse). Some staff stated there was a need to clarify what parents were consenting for:

Clarity on what they are actually consenting for. The consent form talks about consent for data collection but that was done anyway. The consent was purely for the follow up.

P21, research nurse, pilot site survey

Some voiced their confusion about studies that use an opt-out approach as its ‘“opt-out consent” is an oxymoron. It doesn’t make sense’ and others described how they did not use the term during recruitment conversations with parents.

Clinician 1:I don’t know that I necessarily used the words opt out.

Clinician 2:No, I’ve never used that.

FG3, intervention group, research nurses

Approaching every parent for consent for specific elements of the study meant that the PICnIC consent process differed to a traditional opt-out consent model, which often does not involve a formal consent form. However, there was support for the same approach to consent used in the PICnIC pilot trial to be used in the proposed PICnIC trial. This support related to not seeking consent for the intervention and samples at the point of admission with consent sought for additional samples, which was referred to as ‘a combination approach’ (P04, FG2 part 1, intervention site, research nurse).

Approaching parents to discuss the pilot trial – the importance of timing

As consent did not need to be sought prior to SDD delivery, and/or samples collected at the point of admission, staff were able to assess when was the most appropriate time to approach parents to discuss the study. Many emphasised how ‘timing is key’ given the PICU context, which is ‘a very intense environment for them [parents] to come into unexpectedly’ (P10, FG3, intervention group, research nurse). Those involved in recruitment discussions with parents described how they observed parents and patients to establish the most appropriate time to approach, which was never the ‘first day that a parent arrives to the PICU because I feel like that’s the day that they’re most stressed out’ (P01, FG1, intervention site, research nurse). Practitioners would commonly wait until the child’s condition had stabilised or improved, which was checked with the bedside nurse, to establish timing. This checking helped to ensure parents would be ready and receptive to research discussions:

I think the difference here is its intensive care, isn’t it? If they’re really unwell, they’re less likely to consent as well. Because the last thing parents want is more things added on to that. We do see that if we go day two, if the child is improving, then they’re quite happy. They’re quite receptive to listen.

P02, FG1, intervention site, research nurse

Some we’ve approached within 24 hours and others we’ve left for a number of weeks, because it’s just not been something to add to their mental load. It’s really varied.

P03, FG2 part 2, intervention site, research nurse

We would generally talk to the bedside nurses, the clinicians, to know what situation they’re in at that point.

P03, FG2 part 2, intervention site, research nurse

Not all parents were not approached about the study, or given the SDD intervention. Some practitioners described how a decision was made by the clinical team due to concerns about how parents would react to finding out that something had changed in their child’s clinical care and something ‘new’ had been given.

There were some patients that we specifically excluded or clinicians wouldn’t necessarily start it on, because we knew how this family can behave and react to new things. If someone is extremely hospitalised, they’re very aware of medicines, and something new, it just doesn’t sit well with them, sort of thing. We kind of excluded a couple of patients before even starting it. Not many. I think it was maybe like three. Yes, the clinicians were sort of like, ‘I don’t really think I want to give this to this patient because of what repercussions – the family are just going to kick off’. That sort of thing. We pre-empted patients that we already knew would have an issue, potentially, to try and avoid that issue.

P04, FG2 part 2, intervention site, research nurse

Patients were missed due to sites not having research cover over weekends, or if the research team were not based in a PICU: ‘I think if the research team were based on PICU, it would be a lot easier … probably would not miss as many as we missed’ (P11, FG4, control site, research nurse). One doctor described how they would provide brief information to parents of patients who arrived on a Friday, as they were aware it would be a few days before the research team would be available to speak to them: ‘The research team will come and speak to you in more detail, but this is just to sew the seed’ (P09, FG3, intervention site, research nurse) as sometimes families were ready to be discharged at the point that the team approached them on the Monday. One site described how some parents were not provided with any information about the study as their child’s condition did not improve and care was withdrawn:

For some PICnIC patients we haven’t approached at all because if they come in and it’s been a very difficult couple of days and then the focus has changed to palliative care and then they’ve moved towards withdrawal of care, then we’ve made that decision that actually it, it doesn’t feel appropriate to – because in theory it should be opt-out, like you say. We haven’t felt that for all patients it is appropriate to burden them with that information when realistically they’re not going to be – it’s not going to change their management.

P04, FG2 part 1, intervention site, research nurse

PICnIC pilot trial recruitment took place at a time when multiple other studies had started, or restarted after the COVID-19 national lockdowns. As a result, research teams were approaching families about multiple studies during their PICU stay. As the following quote illustrates, this required careful consideration so that parents were not burdened with too much paperwork and decisions to make:

Normally it would be the following day, at least, really. I think the problem we had, not with your study, but at the time that PICnIC was running is we’ve got a raft of other studies. The timing became more and more crucial because these patients could have been eligible for four or five different studies and you can’t give the parents masses of paper and just say, ‘If you just read these and I’m going to sign here’. We had to do it gradually.

P10, FG3, intervention group, research nurse

Challenges in communicating the pilot trial to parents

During focus groups, the researcher asked those involved in recruitment about how they explained the study to parents, including cluster randomisation. At control sites, staff described how they often provided a broad overview of the study aims and how their unit was not taking part in the intervention side of the study. Cluster randomisation or the SDD intervention was not always mentioned in the recruitment discussion, with a focus on the sample aspect of the pilot trial that their child would be participating in.

Well, basically say that it is a study, an infection control study that is looking at introducing a new measure, compared to just our standard infection control, to see whether that made a difference to [hospital-wide] infections. Depending on the family, I would obviously go into more or less detail about what those things all mean. Then I would say that they are looking at introducing medication to treat gut bacteria. But our unit are not doing that. Then discuss the sampling.

P14, FG4, control site PI/doctor

I think I would mention the PICU was randomised but I didn’t go much into what SSD was for our patients.

P20, FG6, control site, research nurse

At sites delivering the SDD intervention, recruiters had also simplified their explanations as much as possible for parents and avoided terms such as cluster randomisation.

I have to be honest, I try to simplify it as much as I can when I’m explaining it to parents. Because then, if you use those terms with parents, then it’s just making your life harder.

P01, FG1, intervention site, research nurse

One intervention site clinician described how a cRCT design was easier to explain to parents than individual randomisation as all children would receive the same treatment:

PICnIC, then, is actually probably the easiest one to explain, because we don’t randomise them in two. I think it’s going to be more difficult to say if, in the same unit, children got randomised to treatment and some didn’t. I think it was because the whole unit has the same thing, it’s easier.

P01, FG1, intervention site, research nurse

However, there appeared to be subtle differences in how staff described the nature of the intervention at sites administering SDD, which may have led to parental misconceptions reported under the ‘SDD acceptability’ subtheme. Across examples of recruitment descriptions provided by staff, many had used the term ‘standard practice’ when explaining the trial and the nature of intervention. While some staff at intervention sites gave the example of SDD being used in other hospitals, or ‘in (the) adult world’ (P10, FG3, intervention group, research nurse) as their rationale for using the term ‘standard practice’, they emphasised how they were careful to clarify this was a research study. Others stated that admission swabs were standard practice in their PICU before the pilot trial, which appeared to underpin the use of this term in their discussions with parents.

Because I like to make sure that they know that this is part of our practice anyway, to do the admission swabs anyway. It would be something that we would have done anyway. I don’t think parents question me because of that.

P01, FG1, intervention site, research nurse

Although SDD was not used at any of the intervention sites prior to the pilot trial, there were also examples given of staff informing parents that SDD was their current standard practice, without explaining how SDD was not used in their PICU before the research.

So that everybody – this is now our standard practice as a unit, this is our standard practice.

P04, FG2 part 1, intervention site, research nurse

Yes, it’s just our standard treatment for patients with a breathing tube in. That’s what our PICU is doing for these ten weeks, it is part of our standard care. Every patient with a tube that meets inclusion, sort of thing.

P03, FG2 part 2, intervention site, research nurse

Staff at two intervention sites spoke of the challenge of communicating the study to parents who did not speak the English language. Although interpreters had been used, staff described how they were unsure how the study had been explained and how these parents would often not provide consent for their child’s participation. They felt the challenge was compounded by consent forms only being available in English, which meant that these parents did not know what they were signing. Recommendations were made to have information sheets and consent forms available in multiple languages for the proposed trial.

You never know what the translator’s saying. People always don’t consent there because it’s better not to consent than to commit to something which they’re not entirely sure of … I remember this very well, of one of the doctors translating for me for one of the families from Egypt. And mum was looking at the form like, ‘Yes, you’re telling me, but is it actually what’s written here?’ Because how will she know? I wouldn’t want to sign something without knowing what it says.

P02, FG1, intervention site, PI/doctor

Acceptability of the selective decontamination of the digestive tract intervention – support for the proposed trial

Overall, focus group participants were mainly positive about the SDD intervention and its use in the proposed trial. Many supported the study as they wished to know the answer to the research question in the hope that it would help their patients in the future.

I’m generally pro [use of SDD in the proposed trial]. I’m very aware of healthcare-associated infections, and I think anything that can reduce that has to be good. I would love this pilot to work and this trial to go ahead, to see whether there’s the evidence to back up my hope for it to be of use in reducing healthcare-associated infections. I think it has to be in a way which is acceptable clinically and for families.

P05, FG2 part 2, intervention site, PI/doctor

I was interested in the fact that it can reduce the amount of antimicrobials used in that patient, but also in other patients on the unit. That’s what interested me quite a bit was the effect on the microbiological sort of flora within the unit.

P18, FG5, control group, PIC consultant

Discussions about the potential risks and benefits of SDD suggested many clinicians were in equipoise: ‘We just do not know. We need to know’ (P12, FG4, control site, research nurse).

I don’t have any ethical convictions or anything … VAPs are not good. It prolongs ventilation, which then weakens … It just prolongs recovery for a child. I can see the advantage of it, but I can also see that you’re giving an antimicrobial that they don’t necessarily need. If they never develop a VAP, then they never needed it.

P09, FG3, intervention site, research nurse

There were a few focus group participants at one intervention site who expressed concern about the use of SDD without evidence in children. Their concerns related to the potential impact on the gut flora and whether this would lead to allergies in individual children, or negatively affect the infection profile on the unit. After discussion, two of these clinicians stated they would find it acceptable to deliver SDD in the proposed trial, although one remained uncertain: ‘this is something new that has come in. I don’t have to say no or yes yet’ (P07, FG2 part 1, intervention site, bedside nurse) and had additional concerns about the ability to deliver the trial unless there was sufficient nurse support (see Concerns about delivering the selective decontamination of the digestive tract to paediatric intensive care unit patients).

Others spoke of how they were aware of colleagues who had polarised views on SDD: ‘Some of them are very pro. Some of them are very anti. It is a bit Marmite’ (P14, FG4, control site, PI/doctor).

Is there the risk for altering your unit flora in an adverse way as well?

P17, FG5, control site, PI/doctor

One of my only concerns would be there is another study in one of the neonatal units that I was involved in the recruiting team that was giving prophylactic antimicrobials. And one of the concerns about that trial appeared in neonates was how this was going to affect, as [Doctor/PI] said, the gut flora long term and whether this was going to lead to further down the line increases in allergies and essentially it just wasn’t known. I guess the same concerns would therefore apply to this as well.

P19, FG6, control site, doctor

One clinician was concerned about the sample size needed to answer the research question:

In principle, I don’t have an issue with it. I can see the logic behind it, but I would want to see more evidence which obviously is why we’re doing the trial we’re going to need a huge, huge number of patients because actually in children, VAP is not a huge problem.

P04, FG2 part 2, intervention site, research nurse

The practitioner survey (pilot site and wider PICU versions) included questions to access views on the acceptability of the SDD intervention and its use in the proposed trial. As shown in Table 17, a series of statements were proposed, with participants asked to indicate how strongly they agreed with each statement using a Likert scale (strongly agree to strongly disagree). Survey data reflected findings from focus group discussions, with the majority (31/44, 71%) indicating they were not opposed to SDD (statement 11). A quarter (11/44, 25%) did not agree nor disagree with this statement, while a minority (2/44, 4%) were opposed to SDD. Most (38/44, 86%) agreed that ‘the decision to adopt SDD requires consensus between my colleagues’ (statement 2).

Responses to other statements in Table 17 suggested that pilot site and wider PICU staff were in equipoise, as the majority did not have strong feelings (neither agree nor disagree) about the effectiveness of SDD in increasing hospital antimicrobial resistance (statement 1) or intensive care unit (ICU) Clostridium difficile infections (statement 7). Although many indicated their unit did have other strategies to address hospital-acquired infections (statement 9) and VAP (statement 10), over half (32/44, 53%) disagreed or were uncertain (9/44, 21%) they knew the SDD evidence base well enough to have an informed opinion regarding its use (statement 8).

Most staff (40/44, 90%) were working in hospitals that tried to reduce antimicrobial use (statement 5), yet over half (26/44, 59%) agreed that research to date has not adequately addressed concerns about antimicrobial resistance and SDD and agreed (38/44, 86%) that a review and appraisal of current best practice was needed to inform a decision about whether to adopt SDD (statement 4).

Concerns about delivering the selective decontamination of the digestive tract to paediatric intensive care unit patients – the look and texture of the formulation

In the survey, PICU pilot site and wider PICU staff were asked about how acceptable they did or would find delivering the SDD intervention to patients. Of those who stated this question was applicable to them (33/44, 75%), the majority stated it was acceptable (19/33, 57%) or very acceptable (8/33, 24%) across both groups. Four of the six staff that did not find delivering the SDD acceptable were from pilot sites.

Both survey and focus group findings highlighted some concerns and challenges regarding the delivery of SDD to patients.

Some survey participants had doubts over clinical equipoise, while one stated they ‘can’t see how this has a favourable risk–benefit ratio’ (Pox, 872699-872681-93590253, male wider PICU survey). Another raised concerns about the size of the sample needed and how findings of a trial in adult critical care setting may mean this study is not needed:

This is a waste of time, money and energy. The power of your study to show a difference will need to be massive and it will not affect outcome. This has been shown in adults who have stopped doing SDD.

P10, doctor, male, pilot site

More commonly there were concerns raised about the look and taste of SDD paste, which was yellow in a child’s mouth, which ‘[j]ust looks vile’ (P10, FG3, intervention site, research nurse) and ‘looks like they haven’t had any mouth care’ (P05, FG2 part 2, intervention site, PI/doctor). Intervention site staff described how parents would question whether or not their child had been sick due to the look of the paste and how some children had become unsettled. This became more of a concern when mouth care could not be done for two hours after SDD delivery, as per protocol.

To leave that in a child’s mouth or in any patient’s mouth I think would be quite unsettling. The parents when they see their children, they want to see that they’re comfortable and they’re settled. If they’re thrashing their head around because they clearly have got this stuff in their mouth, I think that was a bit of a problem – vile.

P09, FG3, intervention site, research nurse

The mouth care, the protocol says that mouth care is supposed to be not done for two hours and that really is very difficult, impossible, pretty much. If that’s seen as a deviation from the protocol, then that’s an issue.

P03, FG2 part 2, intervention site, research nurse

As a complex intervention that leaves a gunk in the mouth of kids, will be hard to convince families of its role.

P28, doctor, male, pilot site survey

Those involved in the clinical care of children commented on the texture of the paste and how its thick sticky texture interfered with tapes that were securing tubes on ventilated children and made them look dirty.

This thing [SDD paste] is very nasty. And even talking with my colleagues, I think the form of the paste is too thick and to try and dissipate a little bit, it’s just so sticky that it is just going on one place. I found it very difficult dealing with the paste physically.

P07, FG2 part 1, intervention site, bedside nurse

There is definitely a degree of concern, so concern about the look, and concern – there’s been concern about the tapes, and whether or not it affects the security of the tape that’s securing the ET [endotracheal] tube.

P04, FG2 part 2, intervention site, research nurse

A concern at one control site was that the paste would block nasogastric tubes as ‘we do have that problem with small NG tubes in kids’ (P18, FG5, control site, PICU consultant). Two intervention sites described how the paste blocked the suction tubes during mouth care: ‘The nurses would say when they do the mouth care and the suction it blocks up the suction tubing they use. They had to keep chucking it away, the Yankauer sucker’ (P09, FG3, intervention site, research nurse). This clinician also stated that their unit’s four-hourly mouth care at some sites ‘meant that throughout 24 hours they’re going to be suctioning it pretty much straight out because they’re doing their next mouth care’.

As the following quotation illustrates, having a generic amount of the SDD paste prescribed to children regardless of size or age was also discussed by two intervention sites who felt that the amount was too much for smaller babies.

I think it would be difficult to be standardised for every patient. Because, obviously, in paediatrics, you might get a zero-month one and you can get a 16-year-old. And the syringes being the same size for both of the patients is very different. Is there some way of changing it depending on the weight? Rather than have a generic amount for everybody. Because the technique of giving a 3-kilo baby, as opposed to a 60-kilo child is very different, isn’t it?

P01, FG1, intervention site, research nurse

Across focus group and survey data there were descriptions of how research nurse teams had ‘done the vast majority of the work’ (P06, FG2 part 2, intervention site, research nurse) and would remind doctors about the elements they would need to do. There was agreement that the proposed trial would need to be a nurse-led study. Supporting nurse research staff to run this trial was highlighted as important to ensuring trial success.

This study was very much a nursing-led study from our perspective … it’s about the nurses administering it and trying to fit it into their routine of all the other things they’re having to do in terms of observations and other medications, etc., etc. It just becomes another piece of the jigsaw of what they have to do for the care for their patients.

P17, FG5, control site, PI/doctor

Protocol adherence – issues to consider for the proposed trial

Five research nurses (5/27, 19%) who took part in the survey noted difficulties in adhering to the protocol. During focus groups, there were examples provided of how the clinical teams had decided not to take additional swabs for a variety of reasons, including perceived burden on the patient, and therefore ‘it was not appropriate’, and staff capacity issues, particularly during nights shifts, which meant that swabs became less of a priority or forgotten.

Getting the swabs taken twice weekly is a very big challenge. We did not adhere properly to the protocol for that for a lot of patients … The unit went through a very busy period in October/November time and it is not the priority for them to do that when they have got so much else to do.

P14, FG4, control site, PI/doctor

We’re finding that if something’s happening at night, it’s missed because people just aren’t at their best at those times.

P16, FG5, control site, research nurse

Intervention sites described difficulties accessing drugs, or delayed access, which meant that the administration of the first dose was delayed or missed as not all staff knew where it was stored or the nurse was busy when the intervention should have been given.

We had one child, for example, they had the drug, it was prescribed, but they couldn’t find it. It was in the fridge marked ‘PICnIC study drug’, just in here, this one here. They just couldn’t find it …. I think that delayed the start of the medicine on many occasions, that I personally saw. The doctor would write it up. We’d tap the doctor on the shoulder, ‘Can you just write the SSD up?’ They’d write it up. Then the nurse was busy and because the drug was as is, it was then delayed, they’d missed the first dose and it would be done for the second dose.

P10, FG3, intervention site, research nurse

Yes, sure, we had missed doses and for varying different reasons and it was hard at some points to figure it out. if things weren’t done perfectly, maybe if someone hadn’t maybe written a date on it – obviously no one is going to give a drug that they don’t really know anything about if the date was missing. So then obviously missed dose is until the next. Because it’s obviously something new and people don’t have a lot of knowledge on, if it wasn’t set up perfectly, it just creates questions and it’s safer just not to give it.

P06, FG2 part 1, intervention site, research nurse

In one focus group, staff stated that the relatively short shelf life of the SDD contributed to protocol adherence issues. They suggested some of these issues could be rectified by the addition of a time of drug preparation added to the bottle label.

[B]ecause the drug had a very short shelf life, it was only 5 days once you’d made it up, if they’d made it up on the 5th at 10:00 at night, when it got to the 10th, say, and it had expired use before the 10th they would bin it and not give it. But actually there wasn’t space on the bottle to say what time they’d made it up, so actually you could give it until 10:00 that night. So you’d have missed doses where people perhaps thought that it had expired, hadn’t necessarily gone to make up a new bottle.

P04, FG2 part 1, intervention site, research nurse

Logistics and packaging

Sites described a number of logistical issues such as temperature deviations with storage of the intervention in fridges and the work involved in making up the drug for each individual child, which was described as a burden on nurse time, delaying the intervention being administered. Some commented on how it was wasteful to have one per child rather than one bottle per unit.

… in the fridge before it’s made up and then it sits in the unit. It was all a bit faffy ... I understand because it was a pilot study. The fact that each of these children had to have their own kit, which seems wasteful. A huge amount of work, which also delayed the making up of the drug. Obviously, the nurse that’s looking after the patient has got other things to do, she’s now to got to make this drug up in order to give it to the patient. If it was just a bottle in the cupboard she would have done it, but because of the way it’s been dispensed in the fact that each patient has to have their own kit.

P10, FG3, intervention site, research nurse

Others commented on how the SDD packaging was fragile and difficult to store, which lead to bottles being smashed and wasted:

P03, research nurse: I think the box, the bottle and individual syringes in the box, storage-wise, is really tricky. If the paste could come in like a tube, like Daktarin that squeeze into your syringe, for storage that would be so much better. The box wouldn’t stay shut. It had to be stored upright but kept falling over, so the bottle would fall out–

P06, research nurse: And smash. I smashed four.

FG2 part 2, intervention site

Training and the importance of sufficient set-up time

Of those who had completed the pilot site survey, just over half (15/27, 56%) had been trained by a member of their site team, while the remainder (11/27, 40.7) had been trained by the study team (ICNARC). Only one member of staff had received training from both groups. Training received was rated highly (good or excellent) regardless of the provider.

In free-text responses, staff recommended that more time to prepare in advance of the proposed trial starting would help ensure ‘it will be read by clinical staff and busy team’ (P18, research nurse, female, pilot site survey). Suggestions were made to clarify guidance for sample collection and to conduct training in groups (as opposed to individually) or in an online format for the proposed trial. A number of wider PICU and pilot site staff suggested that a 2-month lead in time would be needed for training and preparation to engage as many staff as possible including microbiologists. Part of this training should include how to access the intervention at each site.

Clearer guidance at times re the swabbing timings at the start of the study and data input of both consented and non-consented patients.

P8, research nurse, female, pilot site survey

The biggest stumbling blocks are the hospital microbiologists ... need to get them on board first.

P01, clinical academic nurse, female, PICU-wide survey

The importance of ‘getting the doctors on board’ (P14, FG4, control site, PI/doctor) was emphasised by two sites. Due to doctor rotations, it became challenging to ensure all staff were familiar with the protocol, which suggests the need for frequent training to help ensure patients, samples or doses of the intervention are not missed.

I think the underlying thing is still there, which is the cohort of doctors on PICU is ever-changing. People are on a rolling rota of people coming in for three or four months at a time and then going off again and then coming in. So, actually, the time to keep people up to date and trained, and enough people so that there is always somebody on shift that has got PICnIC at the front of their mind is just relatively hard.

P17, FG5, control site, PI/doctor

Overall acceptability of the proposed trial

Towards the end of focus groups, staff were asked to consider discussions that had taken place and conclude whether or not they felt the proposed trial was acceptable to conduct. They were also asked to use a link to a voting system (Poll Everywhere) to vote on acceptability using a five-point Likert scale (very acceptable, acceptable, neutral, not acceptable and very unacceptable). Both verbal responses and handset voting suggested that staff involved in the PICnIC pilot trial sites found the proposed trial acceptable. All of those who voted on trial acceptability stated they would be interested in participating in the proposed PICnIC trial and no staff used the voting poll system to indicate the trial was unacceptable; most indicated it was acceptable (17/24, 71%), while four (4/24, 17%) stated the proposed trial was very acceptable. Two (2/24, 8%) provided a neutral response. The wider PICU survey responses also indicated support and trial acceptability. Wider pilot site and wider PICU staff feedback highlighted that the challenges described in this chapter, such as delivering the SDD, collection of additional samples, sufficient support and engagement, or nurse research teams, and time for trial set-up, should be taken into consideration and changes made to help improve trial acceptability and ultimately trial success.

I think acceptable. I would say acceptable with, clearly, the feedback we have given. It needs a bit of finetuning.

P01, FG1, intervention site, research nurse

It depends how the proposal is going to work, especially who is going to randomise, if it’s going to involve the clinical team, like, the doctors, then really there will be some challenges.

P07, FG2 part 1, intervention site, bedside nurse

I think the pilot has shown that, yes, we can definitely – Well, from the control side anyway, it was possible. It was acceptable to parents and, yes, the only issue was getting all of the swabs. I do not know what kind of leeway of how – What percentage of swabs you need to get, to get the information you need. So, that is the only sticking point, I think.

P11, FG4, control site, research nurse

A professional approach to recruitment and consent

Parents described how staff were professional and ‘respectful’ (P15, father, control) of their situation, explaining ‘the key points of the study’ and ‘what the treatment involved’ (P07, father, SDD). Across each of the control, ecology and intervention groups, parents described how they appreciated the time and care taken by staff when explaining the study.

I think also the people coming round are obviously very knowledgeable about it, and they’re talking to the parents up front, ahead of reading the material. So, normally, if there was something that wasn’t covered, that they generally would get a common question about, they normally bring that to your attention at the beginning. So, I think the approach is very good.

P07, father, SDD

Most had been approached about the study by a research nurse. Views on whether it should be research nurses or medical staff who discuss research with families were supportive of a nurse-led approach as ‘these women were highly specialist’ (P11, mother, control) and had knowledge of the trial as well as the clinical care of their child’s condition. Views on whether or not staff broaching the trial should be involved in clinical care were mixed, some suggesting independence was important so that clinical care remains focused on their child, while others highlighted that staff involved in care would have the best knowledge of their child’s condition.

Someone involved would be better because they know, and they have all the background, and they have more information. Rather than someone from outside, who doesn’t have so much knowledge of what’s going on with their care.

P18, mother, SDD

I think it was nice to have somebody separate come and explain it. Because, from my personal view, if you’ve got a nurse that is looking after your child in intensive care, you want their focus to be 100% on just looking after your child.

P08, father, SDD

The importance of appropriate timing

The majority of survey respondents (n = 63, missing = 1) agreed that staff had checked whether it was a convenient time to discuss research before discussing the PICnIC trial (96.9%; Table 18). During the interview, the majority of parents emphasised the importance of appropriate timing when approaching parents, which is when the child is in a stable condition:

I don’t think there’s ever a great time to approach anyone with a child that’s that ill about it. But the best time is I think when the diagnosis has settled down a bit and is not all fresh in the mind.

P22, father, ecology period 3

However, appropriate timing can depend on how sick the child is, and ‘how big a shock things have been, which may affect how willing people are to have those sorts of discussions about research’ (P19, father, control), but usually parents reported that a day or two into the child being admitted to hospital would be acceptable.

Nevertheless, one parent suggested that for the units that are randomised to the treatment arm, it may be more important to tell the parents ‘then and there’ in order to provide an early opportunity to opt out of receiving treatment rather than only a decision about using data:

If it’s a ward where everyone on that ward is being given it, maybe as close to the time that the medicine is given or just before the medicine is given, maybe speak to the parent then and there because that is what you are doing and this is the reason why. There is a choice for them to opt out. Not giving them, ‘Do you mind if we do this to your child?’ Go, ‘This is what we’re doing. This is why we’re doing it but you have an option for us not to use the data and to opt out.’ But obviously that’s a tricky decision to make I suppose.

P16, father, control

PICnIC participant information was positively viewed

The majority of parents (96%) who completed the survey agreed that the information they received about the study was clear and straightforward to understand (see Table 18).

Parents were asked to reflect on the written information they were given about the PICnIC study during their child’s hospital visit. A copy of the intervention or control information leaflet was also sent to parents so they had an opportunity to see the alternative version of PICnIC information. For example, parents whose children were in a control or ecology weeks were sent the intervention PICU leaflet.

Being provided with an information leaflet by a research nurse or doctor was commonly the first-time parents were aware of the study. Only two parents (1 ecology, 1 SDD group) remembered seeing posters about the PICnIC study in the PICU waiting room but did not focus on it too much due to what was happening to their child.

Parent:As you’re waiting in ICU just to … As soon as we got sent into ICU, I wasn’t allowed to go in while they ventilated him and stuff. I was, obviously, sitting in the ICU waiting room. Automatically, your eyes wander to the boards and stuff so I did see something there. […] I wasn’t thinking, if I’m honest. I was more concerned about the fact that I was told we were going to HDU and now we’re being moved to ICU and what they were doing.

P02, mother, ecology

Many recalled how staff providing them with the information leaflet, which was ‘a couple of pages’, was ‘pretty much to the point’ and ‘providing the information that was needed’ (P2, mother, ecology). The leaflet was viewed positively, with parents describing the information as being clear, short and jargon-free.

I think it was really informative, actually. So we did reflect back it on later on when we were discussing it, but I think it was quite informative really. It had the key points on there, and I think it had someone to contact if we needed, had any questions. I’m sure what there was a point of contact on there.

P03, mother, control

It had the main points in, how they were going to do the study. That’s all that I thought, at that time, was needed.

P02, mother, ecology

They were really helpful. They were definitely very helpful. And they worded it in, like, a way that was easy to, kind of, comprehend and understand, without trying to have to understand that medical jargon. It was really – it was well-written, do you know what I mean? It was easy to understand why, and what was happening.

P21, mother, SDD

All confirmed that there was opportunity for any questions, yet few could recall asking any as ‘there was just so much going on. It didn’t seem like a big deal once we’d had time to think about it’ (P05, mother, control). Those that did have questions about potential risks of the SDD paste, or how the samples would be returned, felt their queries had been addressed sufficiently.

Because she came to me and gave me some specimen … There was an envelope with some specimen containers. So, I just asked her to clarify to me as and when they want the specimens, and how I should post them back to them.

P20, mother, SDD

I did ask about the paste, I guess, and she did answer my question about the paste. I said, ‘If there is any kind of risks to her involved?’ and she answered that saying, ‘No, there wasn’t’.

P18, mother, SDD

While most parents stated that they appreciated being given a leaflet and/or a PIS that they could read in their own time, in order to understand more about the study, some mentioned that, as they were being approached at a very difficult time, it was difficult to read the study information and they therefore read the leaflet at a later point. Some had poor recall of the study yet were still satisfied with their consent decision having read the PIS again prior to the interview. They emphasised the importance of keeping the information brief and the benefit of receiving a full explanation from staff to aid decision-making in such stressful circumstances:

I couldn’t read the leaflet and all the information that she gave me because, until now, I’m handling him. I spend a lot of time with my kid because he is still not very well […]. So, I think, if that information could be a bit brief, or if someone can explain it a bit better, rather than the long leaflets, for the information.

P04, mother, SDD

I can’t really remember much of it because it was such an emotional time. But I knew that it was quite simple and it wasn’t going to stop our lives.

P17, mother, control

Parents suggested creating an animation or a video that explains the study or adding diagrams to the PIS to make it more appealing than text. They suggested this may help parents engage with proposed trial information, particularly during such emotional and stressful times.

So I think it was just two sides of fairly full-on text, I seem to remember. I don’t know if there are ways to use diagrams or anything to make it visually as well more appealing.

P19, father, control

Some people cannot do just text, so it is good to see the information on a video or something.

P13, mother, control

Time to decide

The amount of time it took parents to decide to consent or not varied from a very quick decision, taking less than an hour, to decisions that took a few days, which was often due to wanting to discuss the study with family members, or the research nurse not returning for a consent decision until after the weekend. Most suggested that 24–48 hours would be an optimal time frame to allow for decision-making in the proposed PICnIC trial.

I didn’t sign straightaway, I said that I would have to discuss it with dad and then come back. I think, a couple of days later, they came back and we had agreed to be a part of the study.

P2, mother, ecology

I knew in the first 20 minutes but it took my partner a little longer to come to terms with it. I think four hours is a good amount of time.

P15, father, control

Parents’ understanding of the rationale

Parents were then asked to recall how staff had explained the PICnIC study and if they could tell the researcher (MP) what the study was looking at in their own terms. The level of parental understanding of the PICnIC pilot trial and proposed trial varied. Some parents had no recall of what the study was about, which they attributed to the stress they were under when the PICnIC pilot study was discussed. A few confused the study with other trials that were being conducted at the same time: ‘It’s to do with the ventilation, isn’t it, the diseases?’ (P08, father, SDD). For this reason, fairly early in the research process the research team began to send the PICnIC pilot PIS to parents prior to interviews to assist parental recall.

To be honest, I can’t remember word for word what they said, not if I’m totally honest with you.

P03, mother, SDD

I can’t remember, to be honest with you. There has been so much going on.

P08, father, SDD

However, most parents interviewed had a general understanding that the study was exploring infection control measures or those in the intervention group could describe the SDD intervention:

It was the paste that they had to use for antibacterial, the gut

P18, mother, SDD

I cannot remember now, but I think it was to do with babies that are chosen, or babies that are put forward for the trial, to test to see about any infections.

P13, mother, control

Few could recall staff explaining any risks or potential benefits of their child’s participation in the pilot trial, but felt they had received sufficient information through the leaflet provided, and did not recall being concerned about participation posing a risk to their child’s safety:

I don’t remember them talking about any risks or benefits, to be perfectly honest with you, but then I don’t really remember a whole lot of it. Again, I walked away from it feeling as though I knew everything that I needed to know.

P15, father, control

Yes. It was all in the handouts that they gave us at the time. They didn’t actually talk us through it, but they did give us a full leaflet, that explained all of the drawbacks to it as well.

P07, father, SDD

Parents’ perspectives on the acceptability of a definitive trial that includes the selective decontamination of the digestive tract intervention

All survey respondents (n = 65) indicated that they thought it was acceptable to conduct a larger PICnIC study in hospitals across the UK. Despite supporting the proposed cRCT, two parents from the same site decided to opt out of the study describing how ‘I was happy to go ahead at first, I then decided that my child already has plenty of people poking him all day and didn’t want any more if not needed’ (survey, P42, father, control).

Similarly, after the discussion during the interview about their experience of taking part in the study, most parents who had experienced randomisation to control, intervention and ecology aspects of the pilot trial were in favour of the proposed cRCT:

I said to the nurse at the time as well when I read it, ‘You should definitely roll this out into different intensive care units around the UK or further afield’. Like I say, it can only benefit.

P9, father, SDD

I think it’s absolutely necessary. To be honest, I think the pilot study you’re going to get a good idea of the accuracy but you won’t necessarily get as accurate data as you would if you did the whole of the UK. It just makes sense to do the study across the whole of the UK, increase the accuracy of the study. Yes, I think it’s a good idea.

P15, father, control

Reasons for the acceptability of the proposed study were linked to the nature of the intervention, which was viewed as non-invasive, as well as the provision of information and reassurance by staff on site, particularly about the low-risk nature of the trial, which was seen as safe for the child to be involved in.

And the drawbacks were very unlikely and very minimal in their impact anyway. So, I think it was fine, as far as I was concerned, that that had already been started. And the reassurance was given that, if there were any concerns that it would have affected her in any way, they wouldn’t have started her on the treatment.

P7, father, SDD

I think the acceptability depends on the information that is given to mothers or care-givers. If whoever is doing the study explains the same way that they did to me, I think it will be very much acceptable to parents and care-givers. It all depends on the information dissemination. If it’s good, then I think you can have good results out of that.

P20, mother, SDD

On the basis that it is 100% safe to children, then I’m all for it. Medical research is an important part of development and in evolution of protection against viruses and bacteria, so, yes, I’m, in principle, for it.

P10, father, control

Some parents (two control, four intervention) hoped or held the misconception that their child would directly benefit from the study:

Like I say, even if nothing came from it, nothing developed from it, but … You wouldn’t know obviously, if he hadn’t had the medicine, if he then also didn’t get anything. You wouldn’t know if there was a benefit to it. But from a point of having it, being in his system, there, ready, it’s just another thing, isn’t it? It’s like another protective layer almost, for him to have.

P23, mother, SDD

I was fine with that. I mean, you know, if this is going to help keep our child safe, and especially with his injury, if there was a risk of him getting an infection that could cause further damage, or a longer stay in hospital, or further complications, then I would have been happy to take part in that study.

P12, mother, control

While others spoke of how the trial was presented to them as something that may help others in the future rather than having direct benefit for their own child:

Yes, of course. So, when I got to PICU, they gave me a leaflet with what PICnIC is, and they explained, you know, it doesn’t actually benefit or change anything for your child, but it’s for future, like, future researches.

P21, mother, SDD

Acceptability and understanding of randomisation

Overall, parents acknowledged the importance of randomisation in a study such as PICnIC and trusted that the practitioners would ‘treat the study in an ethical manner’ (P15, father, control).

I think it’s acceptable, like I said, because you’re not going to have [any comparison, no outcome]. And if [they] weren’t doing the study, they wouldn’t be giving it to [them] anyway. So, it’s not something they’re removing from them, the standard practice. It’s an additional – as a measure outcome really.

P3, mother, Period One control

Two parents thought the question of randomisation was hard to answer, as it would depend on how sick the child is and whether the benefits outweigh the risks.

For me, personally, if my son was really very ill, which he was at one point, to find out that there could be a medication out there that could help him to not get worse by catching an infection, then I would want him to have it. So, it’s a tricky one, isn’t it?

P5, mother, Period One control

That’s a hard one. I don’t know because it’s [outweighing] the benefits and the risks, at the end of the day, isn’t it? It’s difficult. I think, because it’s something we don’t routinely use here, yes, I wouldn’t ask questions if they weren’t to receive it, to be honest. […] But obviously for those that are using it, hopefully the benefits outweigh the risks, at the end of the day. So yes. I suppose all trials and things have got to start somewhere.

P6, mother, SDD

When asked about staff explanations of how PICnIC sites had been randomised, the majority of parents reported that the process of cluster randomisation was not explained to them:

I don’t think so. I wasn’t under the impression that everybody in [PICU] would be in a control … I think I knew there were other hospitals involved, but I hadn’t grasped that one hospital would be giving all the treatment and we were a control.

P19, father, control

Only one parent reporting remembering having this process explained to them.

Some parents wanted to know that the hospitals were comparing the two methods; however; one parent explained that they were not bothered by the lack of knowledge of the technical word: ‘Yes. It’s more important to know that than the name. The name would have been nice information to know beforehand, but at the same time we still knew what the words meant, just without the words’ (P15, father, control). Parents in both groups wanted to know more about the actual intervention, rather than the trial arm the child had been randomised to:

I don’t think it’s as important as explaining what the actual treatment involves and what that means for your child and what the drawbacks could be to the treatment actually. I think that’s more important than knowing whether your child is actually getting the treatment or not for it.

P7, father, SDD

If we were in the intensive care unit that obviously were obviously then yes, maybe it’s probably worth saying to them, ‘Look, this is why we’re giving this to your child because everyone on this unit we’d like to give it because we’re comparing it to another unit who is not having it to see the results.’ Yes, maybe we could have been told that information.

P16, father, control

Following a discussion in which the researcher described the cluster approach to randomisation used in the PICnIC study, parents were then asked how acceptable they found this method, how they felt when they found out which trial arm their child was in (e.g. control or intervention) and how they would have felt had their child been randomised to the other trial arm.

Some parents were under the impression that children would be chosen to be given the medication for a particular reason; however, one of them did not think it was acceptable when it was explained that units were (and would be in a future trial) randomly assigned to give medication to children.

I think she’d have been better off being in the unit she was because at the end of the day certainly no intervention and just leaving wouldn’t have really … That’s pretty much what’s happened for [Name of child] anyway. We knew that she was probably chosen for what was probably the better one for her.

P9, father, intervention

Researcher: How acceptable do you feel it is to give SDD, to give the medication, to only half of the babies who take part in the study?

Parent: I think, obviously, they must be chosen for a reason.

Parent: I would say it is not acceptable.

P11, mother, Period One control

Two parents (one control, one intervention) indicated that they would prefer a cluster design, as that would be beneficial ‘because you’re not going to accidentally give treatment to the wrong person’ (P15, father, control).

If the babies are not in the same unit, I think it’s fine. But it only raises questions if half of them are given and the other half is not given. But if all of them are given or if all of them are not given, I don’t think that becomes an issue.

P20, mother, intervention

Researcher: Okay. Given that your child was in the control group, how do you think you would feel if you were informed that the intensive care unit that your child was in, if you were informed that all children in that unit were given SDD?

Parent: Again, personally, I wouldn’t be upset. I’d maybe ask some more questions regarding it, but I wouldn’t have any adverse feelings.

P15, father, control

Selective decontamination of the digestive tract acceptability and lack of parental understanding

While parents had a rough understanding of the study aims, most of them were not aware of what the SDD medication was. Parents whose children were in the intervention unit remembered their children being given the SDD paste, ‘Yes, it was like a thick yellowy – it was, like, an off-yellow colour paste that they put in’ (P18, mother, SDD), but they did not know that it was an antimicrobial:

No. I mean, they call it, on the ward, concrete, because it’s quite dense and stuff. But I don’t really know what’s in it, no.

P7, father, SDD

Only one parent was concerned about the child being given an antimicrobial: ‘Oh my God, basically stripping the gut of stuff’ (P11, mother, Period One control). Their aversion to their child receiving the SDD medication was based on past experience in ICU when the child had experienced severe adverse events as a result of being given antimicrobials. However, they were open to persuasion if given the right information: ‘it’s difficult I suppose, it depends on how you pitch it anyway doesn’t it? […] However, if I think my son is at greater risk because he’s more vulnerable or this and that then I’m open to persuasion’ (P11, mother, Period One control).

It’s really a tough call isn’t it because I understand that these infections are absolutely horrendous, however [Name of child] received some very aggressive antibiotics while he was in the ICU which then led to his skin breakdown as he was trying to poo and wee. It’s taken me two months to recover, because he’s incontinent so he wears incontinence products so it’s really difficult.

P11, mother, control

Some parents described the practice of giving SDD to children as ‘standard practice’, believing that it was what children would have gotten anyway outside of the trial:

Because when they explained everything, they said that it was all very low risk anyway, because they’re not doing anything more to him than he would be having, other than just giving him the medicine that he was going to be having.

P8, father, SDD

While most parents trusted the medical team and considered their child to be in safe hands [‘It’ll be fine, she’s in the safest hands and everything.’ (P23, mother, SDD)], parents (both who understood that SDD was new, as well as those who were informed about the medication during the interview) indicated that they would want the research team to explain better what medication they were giving their child, but they would have still agreed for their child to participate:

Yes, I definitely think that’s important information, it should be clarified. Especially because it’s not typical standard of care, do you know what I mean? Like, I think you should have all the knowledge, as a parent, going into it, exactly. Especially what is being put into your child; you should know.

P21, mother, SDD

Yes. My initial thought is, ‘Gosh, is this using children to test on?’ but I understand it’s the hospitals that are carrying on in their usual way aren’t doing anything different, so I guess children aren’t being put at any risk by not using the medication. I guess some parents wouldn’t want their children to have medication, so, yes, I guess it’s an important thing to do, definitely.

P5, mother, control

It could have been just nice to know information, but it wouldn’t make any changes in my decision-making.

P20, mother, SDD

No, I don’t think it really – Well no but I didn’t, sort of, know that, that it wasn’t used widely across … It doesn’t really change my view of it, I don’t think, because she didn’t develop any infections or anything.

P23, mother, SDD

Some parents in the intervention group perceived the study as ‘increasing the amount of care that people would have had’, which was seen as a benefit and a reason why it was acceptable and felt ‘a bit sorry for the ones that, obviously, haven’t’ (P7, father, intervention).

Acceptability of sample collection and the importance of being informed

The majority of parents thought that it is acceptable to collect routine samples as ‘they’re taking them anyway’ (P3, mother, control):

Yes, because that is part of established nationwide routine care. That is part and parcel of a routine admission, so you have presumed consent.

P19, father, control

However, as the following quotes highlight, many parents would have preferred to have been informed about the samples and what they would be used for before samples were taken. Few stated that they would have preferred for their prospective consent to have been sought for routine sample collection as they were part of a research study:

So yes, I would not be happy if it was taken without my consent. Again, because my child’s enrolment, I do not know and now you are taking samples, I do not know what you are doing with it, if I knew more of the information.

P13, mother, control

I think it’s best for the clinicians to explain to the care-givers of the babies, ‘That we are going to collect some specimens for the study, or just as routine investigations’, because sometimes you really wonder, Why they are taking that urine, how are they going to use it? But it’s fair if they can explain to you that, ‘We are going to do this, we are doing it routinely’.

P20, mother, SDD

No, I think parents would want to know and would want to give consent on providing samples. Obviously, I guess, it’s all about the protection of their child. Knowing where the blood samples are going, who are they going to, what the bloods are going to be used for, etc. I think it’s important to get parents’ consent around that for any, I guess, fluids or any other tests that result in something taken from a child. I think parental consent is important.

P2, mother, ecology

Some parents held the misconception that routine samples would be destroyed if they decided to opt out:

And it was explained to us at the time that they had already taken the samples. They hadn’t used them, they hadn’t done anything with them, they just took them because they needed to have them within a time frame. That makes complete sense. So, they’d got it there. And they said, ‘If you agree to the study, we’ve got them, we can get them sent off. If you don’t agree to it, they will be destroyed, that’s the end of it.’ ‘Again, that’s no …’

P08, father, SDD

The majority of the parents did not have an issue with the collection of additional samples. However, some have expressed concern about doing so without consent or without being given the relevant information.

If they’re routine, I think that’s fine. If they’re taking the additional samples without consent, then no, not so good.

P5, mother, Period One control

Researcher: […] and then seek consent for additional samples?

Parent: Well, I think it’s just part of being aware of what was going on. Yes, so no issues.

P19, father, control

Two survey respondents declined consent for samples, saying that their children had been through enough while they were ventilated or that the child ‘already has plenty of people poking him all day and didn’t want any more if not needed’ (survey, P42, father, control). Out those who attended the interview, two (one control, one intervention) declined consent mostly because of the condition that their child was in, saying that ‘if her health condition was stable, then I would have said yes, but because she was going through a phase of seizures, so other health issues. That’s probably what made me say no’ (P18, mother, SDD).

Reasons for participation or opting out.

As discussed previously, parents often stated that they took part in the PICnIC pilot study to help other children in future. As one following quotation illustrates, one father was not aware of his child’s participation in the pilot trial until it was broached, but felt it was important to consent, even if his child did not recover, as he found comfort in the thought of helping other children in the future:

I didn’t have any knowledge about the trial, I didn’t actually even realise it was something that hospitals did. But my initial reactions to it were, if something bad was to happen with him, it would be nice to know that any data that you could take from him would be used to help someone else. And that was almost a consoling kind of feeling.

P15, father, control

This finding was also supported by survey data with parents (50/61 who consented, 81%) indicating that they did not opt out/provided consent for the study ‘to help other children in the future’ with 11/50 (22%) indicating this was their main reason. This was the most common main reason indicated by parents. Other reasons were because they ‘trusted the doctor or nurse who explained PICnIC’ (41/61, 67%), because ‘medical studies like PICnIC are important’ (49/61, 80%), or to ‘help my child’ (36/61, 59%). Less commonly, parents did not opt out/provided consent because ‘the treatment had already been given’ to their child (8/61, 13%), they ‘didn’t feel comfortable saying no to the nurse or doctor who explained the study’ (1/61, 2%), or because their ‘child recovered’ (5/61, 8%).

During interviews, many parents spoke of the trust they had in the clinical and research staff to act in their child’s best interests and how their children were in ‘the safest hands’ (P23, mother, SDD).

The thing is, the research people for the study is not going to harm your kid, it’s just the information that you are giving to them. So, you can think quite quickly. You have to trust the medical team and medical people.

P04, mother, Period One control

Some spoke about how their child’s condition had improved which meant they were therefore willing for their child to take part to help others and reflected on how they could understand if parents of children who were very unwell were hesitant.

I would say because I knew he was okay, I was happy to speak to somebody who would tell me about the programme, and … But I can see if other parents didn’t want to take part, or didn’t want to speak to someone at the time, because of their child’s condition, it’s very understandable.

P12, mother, control

This was supported by parents who did decline consent for additional samples as they felt that their child was too poorly to take part in the study and did not want any ‘additional tests’.

I guess if her health condition was stable, then I would have said yes, but because she was going through a phase of seizures, so other health issues. That’s probably what made me say no. Had she not had that, I probably would have said yes. I guess it depends on the outcome of health, current health, at that point.

P18, mother, SDD

Although I was happy to go ahead at first, I then decided that my child already has plenty of people poking him all day and didn’t want any more if not needed.

Survey P42, father, control, opt out

Importantly, all parents who completed the questionnaire (65/65, 100%) and were interviewed (23/23, 100%) stated that the proposed PICnIC RCT was acceptable to conduct. This included parents who declined consent for some aspect of the trial. Having the option to decline certain aspects of trial involvement, such as additional samples, appeared to make the pilot trial more acceptable to parents.

Outcomes measures of importance to parents.

Parents were asked to review the list of outcomes that had been sent to them prior to the interview. In the few cases in which parents did not have access to the materials, a definition of each outcome was read to them, including an explanation about why it is important to explore parents’ perspectives about important outcomes (Box 1).

MP then asked parents, ‘Thinking about your experience of your child being admitted to the PICU, what would you hope the study would do to help your child?’ Most parents hoped that their child’s quality of life will improve, that is:

… getting back eating; obviously, not having relapses; I think getting back to where she was, developmentally, before she got poorly.

P19, father, control

[That] it would reduce the risk of an infection and a need for further intervention.

P02, mother, ecology

Overall, parents felt that the list of outcomes presented to them was comprehensive:

I think you’ve got the majority of it covered.

P02, mother, ecology

Most prioritised outcomes that were included in the list they were provided with, such as reduced likelihood of complications or infections:

Less complications because they received the SDD, then I think those would be priority.

P12, mother, control

So, from a parent’s perspective, I’m hoping that this study drives an outcome where secondary infections are less common in children receiving intensive care treatment.

P07, father, SDD

Parents were then asked to rank the outcomes listed in order of importance from the list provided. The top prioritised outcomes can be seen in Box 2. Survival (1) and health complication/AEs (2) were ranked joint as most important for the majority of parents.

Although many did not initially mention it, the majority of parents considered survival an important outcome. When MP inquired about survival, the most reported reason for not mentioning it was that the death of their child was not something they wished to consider [‘I don’t know, the outcome of survival … I try not to think about that really’ (P9, father, SDD)], while others thought it was obvious that survival would be a top one [‘Survival is obviously quite a big one. (Laughter)’ (P5, mother, Period One control)]. One parent in the control group found it difficult to answer, as it made them wonder whether they would question which treatment arm their child was in and whether being in the intervention arm could have improved their child’s chance of survival:

If I was taking part in the study and he didn’t survive, would I be asking, ‘Could he have survived if he had received SDD?’ And that would obviously depend on, was it because of complications and an infection that they picked up while they were in hospital? So, that’s a difficult one to call.

P12, mother, control

Next, associated with the health complications outcome were the infections acquired in the ICU, tolerance of intervention, number of specific treatments their child received, and looking and behaving like normal self, that is overall feeling of return to health or normality:

‘Oh, this treatment will help [name of child] be her normal self a week quicker,’ I’d jump at it, wouldn’t I? So, I think that is my overarching thing and I want her to be her normal self as soon as possible.

P19, father, control

Although health complications were one of the top prioritised outcomes, most parents agreed that the length of PICU and hospital stay, associated with this outcome, was not a priority for them as,

there is no time limit on the rehabilitation of your child, so to me, it’s not up there in the priority. You would stay for however long it would be required. I think [their] wellbeing would be first, in terms of their behaviour, and their presentation.

P03, mother, Period One control

The number of, duration spent, and type of treatment for their child’s vital organs (including mechanical support) (3) was generally reported as being third most important outcome, ‘Because if they’re on it for longer, that risk of infection is increased. So, I’d put that as number one’ (P10, father, SDD).

Current health status in terms of development, functioning and/or life quality (4) and re-admission to hospital (5) were seen as the least important outcomes for the majority parents:

The returning to good health, that is difficult, it’s really difficult because you don’t know what a new-born should do or when you have a special needs child the knock-on effect is far greater. That’s not something that I would focus on exclusively in the hospital, in the ICU, that would be something then to focus on normal ward.

P11, mother, Period One control

Overall, however, as previous research has shown,22 parents seemed to find the questions on outcomes difficult to answer and there was general consensus that they were all as important as each other:

They’re all quite important facts. If you put one then there is another one you read and you’re like, ‘Well actually that one is as important.’ I think to put them in an actual first, second, third is pretty hard to do. Sorry. I don’t think I can.

P16, father, control