ORIGINAL ARTICLE Throw caution to the wind: is refeeding syndrome really a cause of death in acute care? KL Matthews 1 , SM Capra 1 and MA Palmer 2 BACKGROUND/OBJECTIVES: Refeeding syndrome (RFS), a life-threatening medical condition, is commonly associated with acute or chronic starvation. While the prevalence of patients at risk of RFS in hospital reportedly ranges from 0 to 80%, the prevalence and types of patients who die as a result of RFS is unknown. We aimed to measure the prevalence rate and examine the case histories of patients who passed away with RFS listed as a cause of death. SUBJECTS/METHODS: Patients were eligible for inclusion provided their death occurred within a Queensland hospital. Medical charts were reviewed, for medical, clinical and nutrition histories with results presented using descriptive statistics. RESULTS: Across 18 years (1997–2015) and ~ 260 000 hospital deaths, five individuals (4F, 74 (37–87)yrs) were identified. No patient had a past or present diagnosis, such as anorexia nervosa, that would classify them as at high risk for RFS. RFS was not listed as the primary cause of death for any patient. No individual consumed 43400 kJ per day. Limited consensus was observed in the signs and symptoms used to diagnose RFS, although all patients experienced low levels of potassium, phosphate and/or magnesium. Eighty percent of electrolytes improved before death. CONCLUSIONS: RFS was a rare underlying cause of death, despite reported high prevalence rates of risk. Patient groups usually considered to be at high risk were not identi fied, suggesting a level of imprecision with the interpretation of criteria used to identify RFS risk. More detailed research is warranted to assist in the identification of those distinctly at risk of RFS. European Journal of Clinical Nutrition advance online publication, 16 August 2017; doi:10.1038/ejcn.2017.124 INTRODUCTION Refeeding syndrome (RFS), a life-threatening medical condition, is commonly associated with patient groups suffering from acute or chronic starvation, including those with anorexia nervosa, certain types of cancer or alcoholism. 1 RFS was first identified in prisoners of war at the conclusion of World War II following the consumption of a higher calorie intake after months of starvation. 2 After decades of observational research, RFS is now primarily identifiable by electrolyte imbalances, fluid disturbances, and/or life-threatening complications, such as respiratory failure. 3,4 However, limited consensus has been reached as to which signs and symptoms constitute an accurate diagnosis of RFS. 5,6 Rio et al. 4 recommend using a three-facet diagnostic tool, incorporating episodes of severely low-serum electrolyte levels, oedema and organ dysfunction, in an attempt to ensure accuracy in the diagnosis of RFS; however, this process has only been utilised in three other studies thus far, using patients with anorexia nervosa and hunger strikers as population groups of interest. 7–9 With limited consensus, hypophosphataemia is still considered a hallmark sign, with the potential to develop within hours of re-establishing feeding. 6 Development of RFS typically occurs within the initial 72 h of feeding following a period of starvation; 6 however United Kingdom (UK) recommendations suggest monitoring electro- lyte and fluid balances daily for a minimum of 10 days following feeding initiation. 10,11 In addition to daily blood tests, restricted energy intake is recommended in the initial stages of parenteral or enteral feeding to assist in the avoidance of developing RFS. However, studies have found that few individuals with RFS risk factors actually develop RFS symptoms. 4,12 Utilising hypocaloric feeding in an attempt to avoid the development of RFS may then result in unnecessary delays in the provision of adequate nutrition to already malnourished patients. 4,13 Despite reports of RFS developing in individuals consuming an oral intake, 3 it has been considered the lowest threat to those individuals at risk of RFS. 12 While guidelines exist for parenteral and enteral nutrition regimens, no recommendations have been established specifically for patients without an eating disorder diagnosis and consuming an oral intake. Regardless, studies reporting prevalence rates of RFS or RFS risk in hospitals typically include those patients consuming food orally. 6,14,15 The incidence of RFS in acute care reportedly ranges from 0 to 80%, 6 however Rio et al. 4 reported a rate of 2% utilising the three- facet diagnostic criteria. RFS can result in sudden death, 6 yet current RFS research focuses on patient groups at higher risk of the development of RFS rather than the incidence of fatal outcomes. No studies could be located that have examined the prevalence of death either as a primary or secondary outcome of RFS or investigated the groups most likely to die as a result of RFS. We aimed to measure the prevalence rate and describe the types of patients who passed away with RFS listed as a cause of death in Queensland (one of the eight states of Australia) between 1997 and 2015. 1 School of Human Movement and Nutrition Sciences, University of Queensland, St Lucia, Queensland, Australia and 2 Nutrition and Dietetics, Logan Hospital, Meadowbrook, Queensland, Australia. Correspondence: KL Matthews, School of Human Movement and Nutrition Sciences, University of Queensland, St Lucia 4072, Queensland, Australia. E-mail: k.matthews2@uq.edu.au Received 6 February 2017; revised 2 July 2017; accepted 13 July 2017 European Journal of Clinical Nutrition (2017), 1 – 6 © 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved 0954-3007/17 www.nature.com/ejcn