The incidence of the refeeding syndrome. A systematic review and meta-analyses of literature

Background & aims: The refeeding syndrome (RFS) has been recognized as a potentially life-threatening metabolic complication of re-nutrition, but the definition widely varies and, its incidence is unknown. The aim of this systematic review and meta-analyses was to estimate the incidence of RFS in adults by considering the definition used by the authors as well as the recent criteria proposed by the American Society of Parenteral and Enteral Nutrition (ASPEN) consensus. Furthermore, the incidence of refeeding hypophosphatemia (RH) was also assessed. Methods: Four databases were systematically searched until September 2020 for retrieving trials and observational studies. The incidences of RFS and RH were expressed as percentage and reported with 95% confidence intervals (CI). Results: Thirty-five observational studies were included in the analysis. The risk of bias was serious in 16 studies and moderate in the remaining 19. The incidence of RFS varied from 0% to 62% across the studies. No substantial change in the originally reported incidence of RFS was found by applying the ASPEN criteria. Similarly, the incidence of RH ranged between 7% and 62%. In the subgroup analyses, inpatients from Intensive Care Units (ICUs) and those initially fed with >20 kcal/kg/day seemed to have a higher incidence of both RFS (pooled incidence = 44%; 95% CI 36%–52%) and RH (pooled incidence = 27%; 95% CI 21%–34%). However, due to the high heterogeneity of data, summary incidence measures are meaningless. Conclusion: The incidence rate of both RFS and RH greatly varied according to the definition used and the population analyzed, being higher in ICU inpatients and in those with increased initial caloric supply. Therefore, a universally accepted definition for RFS, taking different clinical contexts and groups of patients into account, is still needed to better characterize the syndrome and its approach.