Routine gastric residual volume measurement to guide enteral feeding in mechanically ventilated infants and children: the GASTRIC feasibility study.

BACKGROUND

The routine measurement of gastric residual volume to guide the initiation and delivery of enteral feeding is widespread in paediatric intensive care and neonatal units, but has little underlying evidence to support it.

OBJECTIVE

To answer the question: is a trial of no gastric residual volume measurement feasible in UK paediatric intensive care units and neonatal units?

DESIGN

A mixed-methods study involving five linked work packages in two parallel arms: neonatal units and paediatric intensive care units. Work package 1: a survey of units to establish current UK practice. Work package 2: qualitative interviews with health-care professionals and caregivers of children admitted to either setting. Work package 3: a modified two-round e-Delphi survey to investigate health-care professionals' opinions on trial design issues and to obtain consensus on outcomes. Work package 4: examination of national databases to determine the potential eligible populations. Work package 5: two consensus meetings of health-care professionals and parents to review the data and agree consensus on outcomes that had not reached consensus in the e-Delphi study.

PARTICIPANTS AND SETTING

Parents of children with experience of ventilation and tube feeding in both neonatal units and paediatric intensive care units, and health-care professionals working in neonatal units and paediatric intensive care units.

RESULTS

Baseline surveys showed that the practice of gastric residual volume measurement was very common (96% in paediatric intensive care units and 65% in neonatal units). Ninety per cent of parents from both neonatal units and paediatric intensive care units supported a future trial, while highlighting concerns around possible delays in detecting complications. Health-care professionals also indicated that a trial was feasible, with 84% of staff willing to participate in a trial. Concerns expressed by junior nurses about the intervention arm of not measuring gastric residual volumes were addressed by developing a simple flow chart and education package. The trial design survey and e-Delphi study gained consensus on 12 paediatric intensive care unit and nine neonatal unit outcome measures, and identified acceptable inclusion and exclusion criteria. Given the differences in physiology, disease processes, environments, staffing and outcomes of interest, two different trials are required in the two settings. Database analyses subsequently showed that trials were feasible in both settings in terms of patient numbers. Of 16,222 children who met the inclusion criteria in paediatric intensive care units, 12,629 stayed for > 3 days. In neonatal units, 15,375 neonates < 32 weeks of age met the inclusion criteria. Finally, the two consensus meetings demonstrated 'buy-in' from the wider UK neonatal communities and paediatric intensive care units, and enabled us to discuss and vote on the outcomes that did not achieve consensus in the e-Delphi study.

CONCLUSIONS AND FUTURE WORK

Two separate UK trials (one in neonatal units and one in paediatric intensive care units) are feasible to conduct, but they cannot be combined as a result of differences in outcome measures and treatment protocols, reflecting the distinctness of the two specialties.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN42110505.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in

Nurses looking after babies and children on intensive care units in the UK usually pass a tube and aspirate whatever food or fluid is in the baby’s stomach before they give a feed. The idea is to ensure that the stomach is not overdistended with food and prevent the baby vomiting or, worse, aspirating food into the lungs. However, there is little justification for this practice. It is rarely done in many other countries. It may not be pleasant for the child and perhaps is unnecessary. Some experts have suggested that the policy should be evaluated in a randomised controlled trial. This would mean allocating a large number of children at random to either have the stomach aspirated before feeds, or not. Such a trial would be a major undertaking and we are unsure if parents or staff would be willing to allow children to participate. The aim of this study was to see if it is possible to conduct such a large trial in the UK. Two surveys (of 119 units) showed us that regularly measuring the stomach contents when starting and increasing feeds is common practice for both newborn and older children in UK intensive care units. However, in some countries, such as France, this practice is rarely done. We asked 31 parents and 51 health-care professionals about a future study. Overall, parents were supportive of a trial if it was explained to them well by a knowledgeable and caring professional, and if they were approached at the right time. Some concerns were expressed about not picking up complications early if gastric residual volume was not measured. Health-care professionals were also mainly positive about a future trial, but mentioned similar concerns about not picking up complications early and the difficulty of changing a long-standing routine practice. Parents suggested study outcomes that were important to them. These, along with other outcomes, were voted on in a further survey of 106 professionals and at face-to-face meetings involving 41 participants. Overall, our findings suggest that a trial is feasible to perform and acceptable to parents. However, because of differences in both treatments and important outcomes between children’s intensive care units and newborn baby intensive care units, two trials would be needed, one in each type of intensive care unit. These two trials will test whether or not the benefits of not measuring gastric residual volume (e.g. improved calorie intake) outweigh the potential harms (e.g. delayed diagnosis of complications).

Déclaration de conflit d'intérêts

Lyvonne N Tume was a National Institute for Health Research (NIHR) Health Technology Assessment (HTA) panel member during the conduct of the study and is the deputy chairperson of the HTA Prioritisation Committee. Chris Gale reports grants from the Medical Research Council during the conduct of the study; and grants from the NIHR (research grant and fellowship for a Doctor of Philosophy student), the Mason Medical Research Foundation (London, UK), Rosetrees Trust (Edgware, UK) and from the Canadian Institute for Health Research (Ottawa, ON, Canada), outside the submitted work. Chris Gale also reports grants and personal fees from Chiesi Pharmaceuticals (Parma, Italy), outside the submitted work (the grant is for a research study and the personal fee was to support attendance at an educational meeting). Chris Gale is vice-chairperson of the NIHR Research for Patient Benefit London Regional Assessment Panel (2016–present). Frederic V Valla reports personal fees from Baxter International (Deerfield, IL, USA) and personal fees from Nutricia (Zoetermeer, the Netherlands), outside the submitted work. Jon Dorling reports grants from the NIHR and from Nutrinia (Nazareth, Israel), outside the submitted work (the grant from Nutrinia in 2018 was for part of his salary to work as an expert advisor on a trial). Jon Dorling was a member of the NIHR HTA General Board (2017–18) and the NIHR HTA Maternity, Neonatal and Child Health Panel (2013–18).