The DINE-Normal Proof-of-concept Study

Overarching hypothesis In critically ill adults enteral feeding in a diurnal intermittent pattern improves patient centred outcomes. Research questions for this study Are the same derangements in metabolic and hormonal function observed in healthy volunteers when fed continuously via a nasogastric tube observed in critically ill patients and can those derangements be mitigated by intermittent diurnal feeding? Aim of this study Assess the effect of an enteral nutrition regimen mimicking the usual diurnal meal pattern on hormonal profile and metabolism in critically ill adults. This will generate novel and important proof of concept data and support progression to a clinical trial integrating investigation of physiological responses and patient centred outcomes. Objectives of this study Laboratory: Characterise patterns of hormone, lipid and metabolite response to intermittent diurnal feeding in critically ill adults. Clinical: assess feasibility, tolerability (vomiting and gastric residual volume) and efficacy (calorie delivery) of intermittent diurnal feeding in critically ill adults.

Setting Single adult general intensive care unit in Bristol with 48 beds and \~2500 admissions annually Design Parallel group randomised, open-label trial Population We will recruit from a representative critically ill population and exclude patients with conditions that pose undue risk and/or introduce bias. See eligibility criteria. Screening and recruitment Usual care team will screen daily. Eligibility will be confirmed by health care professional on the delegation log and recorded in the ICU clinical information system. This may be done remotely. A screening log will be kept. Randomisation Via sealed opaque envelopes prepared in advance, allocation 1:1 to intervention or control, stratified by sex. Consent Enteral nutrition is a necessary intervention. Early nutrition (within hours) is the current standard of care. Prolonged continuous feeding prior to enrolment may bias results. An emergency waiver will be used and deferred consent sought from participants. Where participants lack capacity, this will be from a personal or professional consultee as appropriate. Intervention Participants will receive their estimated nutritional requirement in 3 equal "meals" each over 30-60 minutes at 0800, 1300, and 1800 with first meal given after an overnight fast from 1900 on the day of enrolment. Controls Participants will receive their estimated nutritional requirement as continuous feed starting at 0800 after an overnight fast from 1900 on the day of enrolment. The daily nutritional requirement will be estimated by dietetic staff in the Intensive Care Unit as per usual practice. All other care for both groups will follow usual unit practice as directed by treating intensivist. Once blood sampling for the primary outcome is complete the study intervention period ends and feeding will continue according to usual unit practice. Blood sampling Primary outcome assessment requires six one-hourly samples will be hourly around the fourth bolus feed with equivalent timing in control group. Outcomes See outcomes section Adverse events The natural history of critical illness includes many events that might be considered adverse events or serious adverse events including death, organ failure or nosocomial infection. Such expected events do not need to be reported. Events that are not recorded as study outcomes and are considered possibly related to the study will be reported. The reporting period runs for 72 hours from the start of the overnight fast. Follow up Routinely collected clinical audit data will be used for ICU and hospital length of stay and mortality. Sample size Overall sample size is 30. Based on the data from healthy subjects this study is powered to detect smaller (but substantial) effect sizes (d=1.26) with 15 per group while allowing for some drop out. Analysis Analysis will be undertaken blinded to group allocation. Normality will be tested and appropriate parametric / non-parametric tests used for the primary outcome. Additional analysis of the insulin response (e.g. area under the curve) will be undertaken. Repeated measures ANOVA will be used for physiological secondary outcomes. P \< 0.05 will be taken as significant.