Unexpected deaths arising from the rapid reinstitution of nutritional support in people who were severely malnourished were first reported in medical literature during the mid-1940s.1 Nowadays, the so-called refeeding syndrome is known to occur in patients who are apparently well nourished after only a short period of fasting.2
The “hallmark clinical sign” of refeeding syndrome is serum hypophosphataemia,3 but patients often present with additional metabolic disturbances including hypokalaemia, fluid overload, and thiamine deficiency.4 In patients who are accepted to have refeeding syndrome, the clinical sequelae of extreme metabolic disturbances might include respiratory failure, cardiac failure, delirium, rhabdomyolysis, haemolytic anaemia, seizures, coma, excess infections, and death.4, 5
Correction of phosphate and other electrolyte imbalances, plus thiamine supplementation, are accepted as the first steps in the appropriate treatment for refeeding syndrome.2, 4, 5, 6 However, expert recommendations regarding other aspects of medical care for such patients differ. For example, healthy patients' serum phosphate concentrations can decrease after the ingestion or intravenous infusion of glucose.7 Thus, some nutritional experts recommend that nutritional support should be stopped or restricted during phosphate and electrolyte replacement for refeeding syndrome.2, 5 With an absence of evidence from clinical trials5 and outcome studies,8 other experts recommend that “feeding can be continued” during electrolyte replacement.5 This discordance in expert recommendations is mirrored by clinical practice. In a survey undertaken in 33 intensive care units (ICUs) throughout Australia and New Zealand, about half (52%) of the ICU clinicians interviewed reported that they restrict energy intake during the treatment of refeeding syndrome, whereas the other half reported that they continue feeding without energy restriction.9
Research in context
Evidence before this study
We searched MEDLINE and Embase, without language restrictions, between Jan 1, 1945, and Aug 20, 2015, to identify clinical trials that evaluated competing approaches for the treatment of refeeding syndrome in any population of patients. We used the search term “refeeding” in combination with “malnou*” or “starv*”. Although case reports and case series were identified, we could not find any clinical trial that objectively evaluated the effects of competing approaches to the treatment of refeeding syndrome, in any patient population. All randomised and pseudo-randomised interventional studies done in any patient population with refeeding syndrome were eligible for inclusion. Observational studies, opinion papers, animal studies, and case reports were not eligible.
Added value of this study
The Refeeding Syndrome Trial is, to our knowledge, the first reported randomised clinical trial undertaken in any patient population with refeeding syndrome. Our results provide important evidence to guide treatment choices for the management of critically ill patients with this condition.
Implications of all the available evidence
Many health-care professionals, patients, and families might now judge caloric restriction during treatment for refeeding syndrome in critically ill adults preferable to continued standard caloric intake. Future research is needed to objectively assess the effects of caloric restriction in other patient populations and to investigate how appropriate the other aspects of care are for patients with refeeding syndrome (eg, thiamine dose, electrolyte replacement triggers, and doses).
We aimed to address the equipoise surrounding the decision to restrict or continue the provision of nutritional support during electrolyte replacement in adult critically ill patients who developed refeeding syndrome; and assess whether energy restriction affects the duration of critical illness compared with standard care.