Background Some critically ill children have malnutrition which may worsen while they are in hospital and delay their return home. They can recover faster when they are given tube feedings to improve their nutrition. Unfortunately, in the hospital these feedings are often interrupted and so these children do not get all the nutrition they need. The usual procedure is to set hourly rates for the tube feedings and to accept that they get less when feedings are interrupted. The researchers would like to test if children are fed better if the bedside nurses were to check the volume provided through the day and then ensure the child gets closer to the prescribed volumes.

Aim To determine the feasibility of performing a Randomized Control Trial assessing the use of a Volume-based feeding algorithm in critically ill children admitted to the Alberta Children's Hospital Pediatric Intensive Care Unit (PICU).

Objectives

1. Obtain information to inform sample size calculations for nutrition and clinical outcomes for a larger RCT: energy adequacy and protein adequacy, feed tolerance, infections, changes in anthropometric measurements at transitions of care, 28-day ventilator free days, length of stay, 60-day mortality, and 60-day hospital readmission?
2. Assess adherence of medical staff to the study protocol
3. Evaluate the timing of study enrollment and participant allocation
4. Evaluate the proposed deferred consent strategy.

Methods The researchers will conduct a randomized control feasibility trial of critically ill children admitted to the Alberta Children's Hospital (ACH) Pediatric Intensive Care unit who require tube feedings. Children will be randomly assigned to the intervention arm (Volume-based algorithm) or the comparison arm (rate-based algorithm).

Significance The proposed study will provide evidence of whether a novel approach to feeding critically ill children is feasible during PICU admission. This trial will inform a larger Randomized Control Trial on this topic that will assess if using a Volume-based feeding algorithm will improve outcomes of clinical importance including energy adequacy, protein adequacy, feed tolerance, infections, changes in anthropometric measurements at transitions of care, 28-day ventilator free days, length of stay in PICU and hospital, 60-day mortality, and 60-day hospital readmission.

Malnutrition is prevalent in hospitalized critically ill children, may worsen throughout hospital admission, and is associated with negative outcomes including increased ventilator days and longer hospital stays. Early optimal enteral nutrition (EN) is associated with improved outcomes including reduced mortality and shorter hospital stays. Unfortunately, feed interruptions are common which result in underfeeding and accumulating nutrient deficits. Traditionally in pediatrics, tube feedings are ordered and provided using hourly rate-based EN algorithms, which direct the nurse to run the feeding pump at a prescribed hourly rate. The problem with this approach is that the nurse is not authorized to compensate for feed interruptions. Using a daily volume-based EN goal would circumvent this issue by allowing bedside nurses to adjust feedings to be able to deliver the 24-hour desired goal and compensate for feed disruptions. Research conducted in adult critical care comparing volume based versus rate based EN algorithms have shown superior energy and protein delivery. To the researchers knowledge, no pediatric studies have compared volume to rate-based EN.

Aim To determine the feasibility of performing a randomized controlled trial (RCT) assessing the use of a volume-based versus rate-based feeding algorithm in critically ill children admitted to the pediatric intensive care unit (PICU).

Objectives

The primary objective is to assess the feasibility of the proposed randomized control trial to evaluate a Volume-based EN algorithm in the PICU by:

Assessing participant enrollment and recruitment Assessing adherence of medical and nursing staff to the study protocol Evaluating the acceptability of the proposed deferred consent strategy The secondary objective will be to obtain data to inform sample size calculations for nutrition (energy and protein adequacy) and clinical outcomes (feed tolerance, infections, changes in anthropometric measurements at transitions of care, 28-day ventilator free days, length of stay, 60-day mortality, and 60-day hospital readmission) for the larger RCT.

Methods: The investigaters will conduct a single-center parallel partially-blinded 1:1 randomized feasibility trial of 20 children admitted to the Alberta Children's Hospital PICU. Children/adolescents aged 1 month to 18 years, who the investigators anticipate will be admitted to the PICU for ≥ 48 hours, and who initiate enteral nutrition support will be eligible. The trial will compare a volume-based EN algorithm (intervention) to the standard of care rate-based EN algorithm (control). Randomization will be block-stratified by age and ventilator status (invasive ventilation or other). The clinical team will remain unblinded to group allocation to be able to perform care. The research team will remain blinded to allocation to minimize bias.

Adherence to study protocol will be assessed as the number of times that medical or nursing staff deviate from the allocated feeding algorithm.

Enrollment and recruitment of participants will be evaluated by assessing the proportion of eligible participants who are successfully enrolled into the study. Reasons for non-enrollment will be recorded.

Deferred Consent: There is growing evidence that requesting consent close to admission into a PICU puts undo stress on parents/caregivers which can result in limited study enrollment. Deferred consent allows us to randomize eligible children at the time of the decision to use tube feedings and to start the study intervention prior to obtaining consent. This approach provides time to approach caregivers when they are not dealing with the fact that their child has been admitted to intensive care. As far as the investigators are aware, there is no evidence for using a deferred consent model in an enteral feeding RCT. Therefore, part of this feasibility study will assess this strategy for nutrition care research. Quantitative evaluation will be used to estimate the length of time it takes between randomization of eligible participants to starting the assigned feeding algorithm to obtaining consent/assent, the proportion of eligible subjects who provide consent/assent, and any concerns raised. Qualitative methods using the Theoretical Domains Framework will be used to assess parents' perceptions and experiences related to the deferred consent strategy using semi-structured interviews.

Baseline data collected will include age, sex, PRISM IV score (severity of illness), admission diagnosis, and admission comorbidities. Anthropometrics will be measured as soon as possible after PICU admission and at transfer to another unit and/or to home. Data for sample size estimations will be collected prospectively including daily prescribed calories and protein, daily received calories and protein, feed intolerance, deviations from assigned algorithm, ventilator days, days with inotropic support, and length of stay and mortality at PICU and hospital discharge will be collected.

Inclusion Criteria:

• Children 1-month post-natal age to 18 years of age
• Children anticipated by the intensivist or nurse practitioner to be admitted to the PICU for ≥ 48 hours
• Children who will be initiated on EN support

Exclusion Criteria:

• Children will not be eligible for this study if they are palliative
• Children who have contraindications to EN (i.e., a non-functional GI tract)
• Children who are on parenteral nutrition
• Children who are being fed by a bolus feed regime
• Children who cannot progress past trophic feed volumes within 24 hours of EN initiation
• Children anticipated to be admitted to PICU for <48 hours

A double-blinded strategy is not feasible for this study because the medical team cannot remain adequately blinded to study arms after randomization. Bedside nurses delivering EN support need to know which feeding algorithm to follow. The medical team (including PICU physicians, nurse practitioners, and dietitians) have access to patients' medical records which include current hourly infusion rates of enteral feeds so they will be able to deduce the patients' allocations. The team uses current hourly infusion rates to assess fluid balance and to make clinical decisions.

To minimize bias, the research assistant tasked with data collection will remain blinded to study allocation throughout data collection and will enter the data into the REDCap Database in a blinded fashion. After data collection and prior to data analysis, all identifying information will be saved in a separate file to ensure the Study Coordinator will be re-blinded to participant allocation during data analysis.