Accepted Manuscript

Revisiting the Refeeding Syndrome: Results of a Systematic Review

Natalie Friedli, Zeno Stanga, Lubos Sobotka, Alison Culkin, Jens Kondrup,
Alessandro Laviano, Beat Mueller, Philipp Schuetz

PII: S0899-9007(16)30090-9
DOI: 10.1016/j.nut.2016.05.016
Reference: NUT 9786

To appear in: Nutrition

Received Date: 11 April 2016

Revised Date: 28 May 2016
Accepted Date: 28 May 2016

Please cite this article as: Friedli N, Stanga Z, Sobotka L, Culkin A, Kondrup J, Laviano A, Mueller B,
Schuetz P, Revisiting the Refeeding Syndrome: Results of a Systematic Review, Nutrition (2016), doi:
10.1016/j.nut.2016.05.016.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Revisiting the Refeeding Syndrome:
Results of a Systematic Review

Natalie Friedli1, Zeno Stanga2, Lubos Sobotka3, Alison Culkin4, Jens Kondrup5,
Alessandro Laviano6, Beat Mueller1 and Philipp Schuetz1

PT
1
Medical University Department, Clinic for Endocrinology/Metabolism/Clinical Nutrition,
Kantonsspital Aarau, Aarau and Medical Faculty of the University of Basel, Switzerland

RI
2
Department of Endocrinology, Diabetes and Clinical Nutrition, Bern University Hospital, and
University of Bern, Switzerland

SC
3
Department of Medicine, Medical Faculty and Faculty Hospital Hradec Kralove, Charles
University, Prague, Czech Republic

U
4
Department of Nutrition and Dietetics, St Mark's Hospital, Harrow, England
AN
5
Clinical Nutrition Unit, Rigshospitalet University Hospital, Copenhagen, Denmark

6
Department of Clinical Medicine, Sapienza University, Rome, Italy
M

Corresponding author:
D

Prof. Philipp Schuetz, MD, MPH

University Department of Medicine, Kantonsspital Aarau
TE

Tellstrasse, CH-5001 Aarau, Switzerland

Phone: +41 (0)62 838 4141
Fax: +41 (0)62 838 4100
EP

E-mail: [email protected]
C

Conflict of interest statement: All authors have completed the Unified Competing
Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the
AC

corresponding author) and confirm that they do not have a conflict of interest
associated with this manuscript.

Funding: This study was supported in part by the Swiss National Science
Foundation (SNSF Professorship, PP00P3_150531 / 1) and the Research Council of
the Kantonsspital Aarau (1410.000.044).

Word Count: 3014

1
 ACCEPTED MANUSCRIPT
Abstract

Introduction:
Although described more than 70 years ago, the refeeding syndrome remains
understudied with lack of standardized definition and treatment recommendations.
The aim of this first systematic review was to gather evidence regarding standardized
definition, incidence rate and time course of occurrence, association with adverse

PT
clinical outcomes, risk factors and therapeutic strategies to prevent or treat this
condition.

RI
Methods:

SC
We searched MEDLINE and EMBASE for interventional and observational clinical
trials focusing on refeeding syndrome, excluding case reports and reviews. We

U
extracted data based on a predefined case report form and assessed bias.
AN
Results:
Out of 2207 potential abstracts, 45 records with a total of 6608 patients were included
M

(3 interventional trials, 16 studies focusing on anorexic patients). Definitions for

refeeding syndrome were highly heterogenous with most studies relying on blood
D

electrolyte disturbances only and others also including clinical symptoms. Incidence
TE

rates varied between 0% and 80% depending on the definition and patient population
studied. Occurrence was mostly within the first 72 hours of start of nutritional therapy.
Most of the risk factors were in accordance with NICE guidelines, with older age and
EP

enteral feeding being additional factors. There was no strong evidence regarding
associations of refeeding syndrome and adverse outcomes, as well as regarding
C

preventive measures and treatment algorithms.

AC

Discussion:
This first systematic review focusing on refeeding syndrome found consensus
regarding risk factors and timing of occurrence, but wide variations regarding
definition, reported incidence rates, preventive measures and treatment
recommendations. Further research to fill this gap is urgently needed.

Keywords: Nutritional therapy, nutritional support, hypophosphatemia

2
 ACCEPTED MANUSCRIPT
Introduction

Disease-related malnutrition is a common condition in medical and surgical inpatients

and is associated with detrimental clinical outcomes including mortality, complications
and prolonged hospital stay [1-3]. Instigating nutritional therapy is commonly
recommended in malnourished patients in order to prevent these adverse outcome,
although large clinical trials demonstrating benefit in the polymorbid inpatient

PT
population are largely lacking [4-6]. A potential risk of nutritional therapy in
malnourished patients is the refeeding syndrome (RFS) with electrolyte disturbances

RI
leading to clinical deterioration and possible sudden cardiac death [7, 8]. Yet, as of
today, the importance of the RFS in the polymorbid inpatient population is

SC
understudied lacking a common definition and treatment recommendation.

U
For the first time, RFS was described in prisoners liberated from concentration camps
at the end of World War II after a long period of starvation. Despite allegedly
AN
adequate nutritional support these prisoners showed an unexpectedly high mortality
rate of up to 20% [9, 10]. To date the pathophysiology of RFS is still incompletely
M

understood. Proposed mechanisms include starvation and catabolic state-induced

reduction in insulin and increased glucagon secretion [11] with resulting activated
D

gluconeogenesis and proteolysis, as well as depletion of intracellular vitamin and

electrolyte supplies [12, 13]. After the initial start of nutritional intake in patients in
TE

such a starved and catabolic state, glucose concentration and insulin secretion rises
suddenly leading to an overwhelming shift in electrolytes with increased sodium and
EP

water retention and extracellular volume expansion [11]. The potential clinical
consequences are volume overload with risk of heart failure and peripheral edema
C

[14], as well as transcellular shift and redistribution of phosphate, potassium and

AC

magnesium[15] with life-threatening complications such as spasm, or cardiac

arrhythmias [12, 15-17]. Additionally the hypophosphatemia also affects the
phosphate-dependent ATP production [18, 19], resulting in possible muscle
weakness, rhabdomyolysis, and impaired hematopoiesis with symptoms of anemia
and reduced oxygen supply [12].
Numerous reviews regarding RFS have been published focusing each on specific
aspects of RFS. Yet, there is lack of a systematic literature search to better
understand the current state of knowledge regarding RFS. The aim of this first
systematic review was to provide evidence in regard to a standardized definition,

3
 ACCEPTED MANUSCRIPT
incidence rate and time course of occurrence, association with adverse clinical
outcome, risk factors and therapeutic strategies to prevent or treat this condition in
adult or adolescent patients.

PT
RI
U SC
AN
M
D
TE
C EP
AC

4
 ACCEPTED MANUSCRIPT
Methods

Study objectives

This systematic review adheres to the PRISMA guidelines [20]. Our aim was to
investigate the importance of the RFS in adult or adolescent patients. Particularly we
focused on the following six research questions:

PT
I. What are the definitions used for RFS?
II. What is the incidence of RFS?

RI
III. When does RFS occur?
IV. Is RFS associated with adverse clinical outcome?

SC
V. What are the risk factors for RFS?
VI. What are therapeutic strategies to prevent or treat RFS?

Eligibility Criteria
U
AN
Randomized clinical and observational studies investigating refeeding and RFS in
anorexic and non-anorexic adult or adolescent patients (medical and surgical), were
M

considered eligible for this review. There was no restriction in regard to publication
date. We did not include case reports, narrative reviews, audits, surveys, animal
D

studies and unpublished records. Also, we excluded studies with pediatric and
TE

neonate patient populations.

EP

Information sources and search strategy

Electronic libraries were searched in order to identify all abstracts written in English.
C

The search was performed with MEDLINE and EMBASE library and in cooperation
AC

with a subject specialist in Medicine & Life Sciences of the University Library of the
University of Basel. The searches were performed using keyword search and also
Mesh Term search, they were run on 20th, 23rd and 27th of March 2015. Two
updated searches were performed in mid-September and mid-December 2015.
The following search terms were used to perform the literature search in the two
databases: “refeeding”, “NOT review”, “NOT child”, “NOT children”, “NOT neonates”
“NOT neonatal”, “NOT animal”, “English”. The full electronic search protocol is
presented in Appendix 1.

5
 ACCEPTED MANUSCRIPT
Study selection and data collection process

Eligibility assessment was performed by one of the authors (NF) in discussion with
the study team. All search results were assessed by short review of abstracts.
Records, which did not meet our inclusion criteria, were excluded. For data extraction
a predefined case report form (Appendix 2) was used after undergoing a testing
period. During the process of data extraction, the individual risk of bias of each study

PT
was assessed. Final extraction results were checked and discussed among the
research team. We also contacted authors from abstracts and posters for

RI
international congresses by email to obtain final reports and further information.

From each included trial we extracted the following information: (1) name, year and

SC
title of the trial, (2) number of included patients and patients characteristics (e.g.
whether patients were anorectic or not, adult or elderly patients, oral, enteral or

U
parenteral refeeding, state of illness, reason for hospital admission, etc.), (3)
AN
definition of RFS if used, (4) methods including inclusion and exclusion criteria, (5)
outcome measures, (6) main results including information concerning our main
investigation questions, (7) conclusions, (8) limitations and (9) information about the
M

nutritional treatment consisting of energy, protein, electrolytes, fluids, micronutrients,

days and rate of nutritional rehabilitation.
D

We reported phosphate, potassium and magnesium in mmol/L (1mg/dl phosphate

TE

equals 0.323mmol/L phosphate, 1mg/dl potassium equals 0.256mmol/L potassium

and 1mg/dl magnesium equals 0.411 mmol/L magnesium).
EP

Risk of Bias Assessment of Individual Studies and across studies

C

In order to assess the risk of bias we conducted an evaluation for each single trial
AC

using the SIGN checklists [21] for bias assessment (Appendix 3/4).

Data Synthesis and Analysis

Due to the lack of quantitative data and wide heterogeneity among trials we decided
post-hoc not to perform a quantitative meta-analysis, but focused on presentation of
data qualitatively corresponding to each of the objectives.

6
 ACCEPTED MANUSCRIPT
Results

Trials identified through the search strategy

Our systematic search identified 2910 titles and abstracts of potentially eligible
studies through database searching. After duplicate removal, 2205 records were
screened and 69 full texts were assessed for eligibility. Through two updated
searches in mid-September and mid-December 2015 two additional studies were

PT
added. A total of 45 studies with a total of 6608 patients were included in the analysis
including 3 interventional trials (Figure 1). A total of 16 studies focused on anorexic

RI
patients and 29 on non-anorexic patients.

SC
Study characteristics and risk of bias

U
Most of the included studies were smaller, single-center studies including
heterogeneous adult or adolescent medical or mixed medical/surgical patients with
AN
disease-related malnutrition. Studies were conducted between 1989 and 2015 in
different countries (Table 1). In regard to bias assessment (Appendix 3), there was a
M

high risk of bias resulting from the observational / retrospective study design used in
most studies, as well as high risk for attrition bias due to incomplete outcome
D

reporting. Also many trials did not deal with RFS as their main research aspect but
TE

reported RFS data as secondary aims.

EP

Definitions and incidence rates for RFS

Among the 45 studies, 38 studies reported a definition for RFS (Table 1). Definitions
C

were highly heterogeneous among studies with some studies only relying on
AC

electrolyte disturbances with again different cut-offs, and others also integrating
clinical parameters into the definition. Most studies had hypophosphatemia, either as
a cut-off or as a relative decrease from baseline, as part of the definition. Cut-offs
used for hypophosphatemia differed among studies (the range varied from phosphate
<1mmol/L and below normal range to <0.32 mmol/L or decreased rate from baseline
>30% or >0.16mmol/L). Of note, the definition used by RIO’s study [22] was later
used by 2 other studies [23, 24]. Also Marik’s [25] definition was reused by another
study [26] and slightly adapted in 2 other studies [27, 28]. These definitions are thus
the most commonly used definitions.

7
 ACCEPTED MANUSCRIPT
Incidence rates for RFS were highly dependent on the definition used and similar in
anorexic and non-anorexic patient populations. As shown in Table 1, 11/32 studies
(34%) did not find any patient with either signs or symptoms of RFS. These studies,
however, used a more rigorous definition with clinical and electrolyte abnormalities.
Incidences of up to 80% [29] were reported in studies using a broader definition
based on electrolyte disturbances only.

PT
Time course and association with adverse outcomes

RI
Only 11 studies reported on the time point when RFS occurred (Table 1). In most of
these studies, it occurred within the first 72 hours after start of nutritional therapy. Out

SC
of these 11 studies providing information regarding the timely occurrence of RFS, 7
used only hypophosphatemia as their definition criteria of RFS. Except for one, all

U
studies reported that RFS occurred within 84 hours. Two studies using clinical
AN
symptoms for their definition reported time of occurrence of RFS within 24 hours and
after 5 days, respectively.
M

A total of 11 studies reported outcomes in their RFS patient population compared to

patients without RFS, mainly all-cause in hospital mortality and length of hospital stay
D

(Table 2). Of 8 studies reporting mortality, 5 reported a tendency to higher mortality

rates with one study finding a mortality rate of 71% in anorexic RFS patients from
TE

critical care [30]. However, these studies did not adjust for confounders and in most
of these studies p-values were not reported or did not show a significant result. No
EP

association of RFS and mortality or a mortality rate of 0% was reported in the 4

remaining studies. A total of 4 studies reported LOS with all studies finding longer
C

LOS in patients showing signs or symptoms of RFS. Again, no adjustment for

confounding was done in these studies.
AC

When separating studies based on the criteria used to define RFS, there was no
differences regarding association of RFS and mortality. In all three definition groups
(hypophosphatemia, electrolyte disturbances or clinical symptoms) some studies
showed a mortality rate of 0% and others an increased risk of mortality in patients
showing signs and symptoms of RFS. Of note, the highest mortality rate was
reported in one study [30] using clinical symptoms as a definition criteria. Studies
reporting on length of hospital stay used either hypophosphatemia or general
electrolyte disturbances as definition criteria.

8
 ACCEPTED MANUSCRIPT
Risk factors for RFS

Several studies focused on risk factors for development of RFS (Table 3 / 3a). Most
risk factors are in accordance with the NICE guidelines [31], i.e. low BMI,
unintentional weight loss, starvation, history of alcohol abuse and low initial
electrolyte concentrations. Most of the studies relying on NICE guidelines as risk
factors for RFS used clinical symptoms as part of their RFS definition. Additionally to

PT
criteria included in the NICE guidelines, higher age was reported to be a risk factor in
one study [32] and showed a non-significant trend in another study [27]. Other

RI
parameters not included in the NICE guidelines were low albumin or pre-albumin,
higher nutritional intake during feeding, low insulin-like growth factor (IGF)-1,

SC
nutritional risk screening (NRS) 2002 [33] ≥3 points and enteral feeding as part of
the nutritional therapy. A total of 3 studies did not validate the above risk factors to be

U
associated with RFS. Of note, in two studies [34, 35] NICE guidelines had a low
sensitivity and only 30% of patients were identified.
AN
M

Therapeutic strategies to prevent or treat RFS

Different studies evaluated preventive strategies for RFS, mainly substitution of

D

electrolytes, thiamin infusions and hypocaloric nutrition. As shown in Table 4, 5

TE

studies showed a preventive effect of hypocaloric feeding whereas 6 studies did not
find such an effect. Regarding substitution of electrolytes and vitamins, 6 studies
found a reduced risk of RFS while 3 studies found no such effect. One study [36]
EP

proposes monitoring of electrolytes whereas another [37] did not show a reduction in
risk for RFS in this measure. There was no correlation between criteria used to define
C

RFS and effect of preventive measurements.

AC

Three studies reported on the treatment of RFS [26, 36, 38]. In two of these the
investigators used phosphate supplementation in order to treat signs and symptoms
of RFS and concluded that these measures were helpful (Table 4). In the most
recent RCT involving critically-ill patients [26], protocolised caloric restriction proofed
to be a suitable option for treatment of patients with improvement in mortality and
overall survival time although the primary outcome (the number of days alive after
ICU discharge) was not significantly improved.

9
 ACCEPTED MANUSCRIPT
Discussion

Summary of evidence

Although described more than 70 years ago, RFS remains incompletely understood.
The main findings of this first systematic and comprehensive literature review are
threefold. First, there is no well-accepted definition of RFS and no consensus
whether the definition should rely only on laboratory parameters or additionally

PT
include clinical symptoms. Most studies, however, had a low or decreasing
phosphate concentration as part of their definition. Depending on the definitions

RI
used, incidence rates showed wide heterogeneity. Second, in regard to risk factors,
the current NICE guidelines seem to adequately reflect findings of the included

SC
studies with some additional factors (i.e. higher age, low albumin or pre-albumin,
higher nutritional intake during feeding, low IGF-1, NRS ≥3 and enteral feeding) being

U
described. Third, only a few studies have investigated the clinical consequences and
AN
therapeutic measures of RFS. Within these studies, there was a lack of association of
this syndrome and adverse clinical outcomes, as well as preventive measures and
improved outcomes. Similarly, there was only one randomized study [26] evaluating
M

therapeutic approaches of RFS as the main outcome. In this recent RCT Doig et al.
reported caloric restriction to be a suitable therapeutic measure for RFS in ICU
D

patients. Further randomized studies confirming these findings are largely lacking.
TE

Interestingly, our results did not find strong differences for the anorexic and non-
EP

anorexic populations. For this reason, we pooled these two populations in our
analysis. Still, from a pathophysiological standpoint, it can be argued that anorexic
patients without an acute illness would be different compared to acutely ill
C

malnourished patients. The lack of differences found in our review may also be the
AC

result of the low study quality and it is thus possible that anorexic patients need
particular attention in regard to the RFS.

The lack of common definitions for RFS is problematic and makes a comprehensive
comparison of different study populations challenging or even literally impossible.
Most studies used electrolyte disturbances as part of their definition – with, however,
differences in cut-offs and types of electrolytes included. In acute illness electrolyte
disturbances are common and may be the result of different nutritional and non-
nutritional factors lowering their specificity for RFS. There is no general consensus
10
 ACCEPTED MANUSCRIPT
whether electrolyte disturbances seen in RFS are part of the normal metabolic
response to disease, or a maladaption which need therapeutic action. From a
biological perspective, concentration gradients of electrolytes between intracellular
and extracellular compartments are basic conditions seen in most eukaryotes and
promote different metabolic functions. Phosphate is essential for different metabolic
and regulatory processes like protein phosphorylation/de-phosphorylation and energy

PT
production (generation of adenosine triphosphate [ATP]) [39]. Similarly, magnesium
is an essential component for ATP stabilization and various enzymatic processes

RI
[40]. Also, sodium - potassium membrane gradient is important for membrane
excitation, signal transduction and for transport processes [41]. During catabolism,

SC
the intracellular ion storage is being depleted while intracellular sodium increases in
exchange. This is reversed after refeeding [12, 13]. Yet, the change in ion
composition of currently-used processes food (i.e., higher amount of sodium, but

U
lower concentrations of potassium, phosphate and magnesium) may interfere with
AN
this natural reaction leading to a more pronounced electrolyte disturbances with
adverse clinical effects. Thus, the observation of low plasma electrolytes during
M

illness may in fact be a consequence of lower nutritional intake of these ions rather
than RFS. These speculations would favor using clinical symptoms in addition to
electrolyte disturbances as a goldstandard diagnosis of RFS.
D

A rational approach to devise up with a standard definition of RFS may also be to

TE

focus on factors associated with adverse clinical outcome, as targets of preventive

and therapeutic measures. Only a few studies, however, have focused on such
EP

premises. For example Heimburger et al. [42] reported a lower survival probability in
patients with hypophosphatemia independent of other metabolic parameters. This
study, however, was conducted in a HIV infected study population limiting
C

generalizability of their findings. There is need for further high quality research in
AC

regard to better understanding the importance of RFS in the medical patient

population.

There is some consensus among studies in regard to the time point of RFS
occurrence, namely within the first 72 hours. Thus, particular attention may be drawn
to at risk patients within this initial time period. Still, due to the limited study quality,
we cannot exclude that in certain patients a later development of the syndrome can
occur.

11
 ACCEPTED MANUSCRIPT
Even though most studies agree with the NICE criteria as risk factors for RFS, the
use of this guidelines by itself has low sensitivity and specificity as Goyale et al [34]
and Zeki et al [35] stated in their studies.

This review has several limitations, mostly associated with the limited quality of
included studies. First, we found only three interventional studies, two of them not

PT
having RFS as their main trial focus. Second, most included studies had small
sample sizes (26 studies included <100 patients) and low quality according to the

RI
SIGN methodology [20], with high heterogeneity in regard to the type of interventions
and patient populations. The risk of bias analysis revealed high risks for performance

SC
bias, attrition bias, and reporting bias. We can therefore not draw firm conclusions
regarding the definition, prevention, risk factors, incidence and treatment of RFS.
There is currently a large, multi-center randomized controlled intervention study

U
(“EFFORT”-study) ongoing focusing on nutritional therapy in inpatients which will help
AN
to fill this gap (https://clinicaltrials.gov/ct2/show/NCT02517476) [4, 5].
M

Conclusion

Although there is consensus in regard to risk factors and timely occurrence of RFS,
D

there is still wide variation in definition, reported incidence rates and preventive
TE

measures and treatment recommendations. This first systematic review is an

important step towards a better understanding of the current state of knowledge
regarding RFS and it’s clinical consequences. Further prospective high-quality
EP

research is warranted to fill this gap.

C
AC

12
 ACCEPTED MANUSCRIPT

Contributors:

NF and PS wrote the initial protocol. NF performed the literature search and the data
extraction. PS and ZS reviewed and commented on the final data extraction. NF
drafted the manuscript. All authors amended and commented on the manuscript and
approved the final version. PS and ZS oversaw the study and act as guarantors.

PT
RI
Acknowledgement:

We thank Heidrun Janka (Universitätsbibliothek Medizin Basel, Switzerland) for

SC
helping with the search. She was not given any compensation for her help with this
study.

U
AN
M
D
TE
C EP
AC

13
 ACCEPTED MANUSCRIPT
Tables and Figures

Figure 1 – Flow Chart

1253 records found through 1657 records found through

PubMed searching Embase searching

PT
Excluded duplicates (n=705)

RI
Review of Titles and Abstracts
(n=2205)

SC
Excluded records:

- Case reports (n=110)

U
- Survey or Audits (n=19)
AN
- Other: Reviews, Animal
studies, records not
meeting our inclusion Eligible for full text analysis
criteria (n= 2007) (n=69)
M
D

Observational studies (n=65) Interventional studies (n=4)

- Anorexia nervosa 19 - Anorexia nervosa 0
TE

- Not AN 46 - Not AN 4

Excluded records:
EP

- Unavailable full text (only

congress abstract or poster
presentation) n= 15
2 additional studies found
C

through two updated searches - Not providing any relevant

information n=11
AC

Records included in systematic

literature review (n=45)

- RCT (n=3)
o Anorexia nervosa 0
o Not AN 3
- Observational studies (n = 42)
o Anorexia nervosa 16
o Not AN 26

14
Table 1 – Definition, incidence and timely occurrence of RFS

Firstauthor Y N= Population ACCEPTED MANUSCRIPT

Definition of RFS Incid.
When did it
occur?
Goyale, A. [34] 2015 52 patients receiving PN PO4 drop of >30% during the first 36h of PN administration. 62% NR
Coskun, R. [43] 2014 117 ICU patients receiving EN or PN PO4 ≤ 0.78 mmol/L 52.14% NR
Gaudiani, L. [44] 2012 25 AN patients PO4 < 0.87mmol/L 45% 3.4 days
Gaudiani, L. [45] 2014 132 AN patients hypophosphatemia 37% NR
ICU patients starved for at least drop of PO4 by > 0.16mmol/L to below 0.65mmol/L accompanied by possible other
Marik, P. E. [25] 1996 62 34% 1.9 +/- 1.1 days
HYPOPHOSPHATEMIA as definition criteria

48h electrolyte disturbances

PO4 drop of ≥ 0.15mmol/L to < 0.80mmol/L from baseline in the first 7days of PN in
Marvin, V. [28] 2007 250 patients receiving PN 34% 3 days
patients with a period of > 48h without food
Brown, CA. [46] 2015 123 female AN patients PO4 < 0.87mmol/L 33% 2 days
frail elderly patients with feeding
Lubart, E. [47] 2009 40 PO4 < 0.52mmol/L 25% 2 - 3 days
problems >72h
Gonzalez Avila, G. cancer patients receiving EN or
1996 106 PO4 < 0.81mmol/L 24.5% Within 72h
[32] PN

PT
Zeki, S. [35] 2011 321 patients receiving PN or EN PO4 < 0.6mmol/L 15% NR
Garber, A.K. [48] 2012 56 AN patients shifts in electrolytes 0% NR
patients with cancer of the upper
Grasso, S. [49] 2013 34 aerodigestive tract, doing an 12 HP and/or other shifts in electrolytes during RF NR NR

RI
hours over night fast
Terlevich, A. [38] 2003 30 patients with refeeding syndrome PO4 < 0.50 mmol/L NR NR
Nagata, J.M. [50] 2015 356 AN patients PO4 < 0.97 mmol/L NR NR
patients >65yrs with at least one 10.9+/- 21.5 days

SC
Kagansky, N. [51] 2005 651 episode of HP compared with PO4 < 0.77 mmol/L NR after
patients without HP hospitalisation
PO4 drop of ≥ 0.15 mmol/L to < 0.80 mmol/L from baseline in the first 7days of PN in
Marvin, V. [27] 2008 140 patients receiving PN NR NR
patients with a period of > 48h without food
Doig, G. [26] 2015 339 ICU critically ill adult patients PO4 drop >0.16mmol/L to < 0.65 mmol/L within 72h after starting nutritional support NR Within 72h

U
patients with mild and severe
SHIFT IN ELECTROLYTE

Hernandez-Aranda,
1997 50 malnutrition receiving EN or PN drop of any electrolyte value under the reference range 48% 3 days
J.C. [52]
for >7d
AN
PO4 < 0.87 mmol/L, for patients ≤ 16yrs < 1mmol/l, hypomagnesaemia, hypokalemia
Redgrave, G.W. [53] 2015 461 AN patients 18.5% NR
and hypoglycemia
Manning, S. [37] 2014 36 alcoholic patients changes in fluid and electrolyte balance in malnourished receiving intensive feeding 0% NR
malnourished patients receiving severe shifts in electrolytes in patients that are underweight, severely malnourished or
Luque, S. [54] 2007 11 0% NR
PN starved and started on refeeding
M

Whitelaw, M. [36] 2010 46 AN patients electrolyte and fluid shifts during refeeding NR NR
malnourished patients receiving
Flesher, M.E. [29] 2005 51 shifts in electrolytes associated with complications during early EN 80% NR
EN
patients having a high risk for
Chen, L. [55] 2014 56 RFS according to NICE shifts in electrolyte and clinical symptoms 20% NR
D

guidelines admitted for RF

all adverse events occurring during nutritional rehabilitation of malnourished patients
Vignaud, M. [30] 2010 68 AN patients 10% NR
or having undergone a prolonged fast
TE

patients with gastrointestinal

Fan, C.G. [56] 2004 15 shifts in electrolytes and clinical symptoms 9.4% within 24h
fistula
Saito, S. [57] 2014 111 AN patients cardiovascular, respiratory, neurological or psychological changes with HP during RF 3.6% NR

electrolyte disturbances (K <2.5mmol/l, PO4 <0.32mmol/l, Mg <0.5 mmol/L) AND

Eichelberger, M. [24] 2014 37 hungerstrike patients clinical symptoms (peripheral edema or acute circulatory fluid overload) AND 3% NR
EP

disturbance to organ function (respiratory failure, cardiac failure or pulmonary edema)

CLINICAL SYMPTOMS

electrolyte disturbances (K <2.5mmol/l, PO4 <0.32mmol/l, Mg <0.5 mmol/L) AND

patients on artificial nutrition
Rio, A. [22] 2013 243 clinical symptoms (peripheral edema or acute circulatory fluid overload) AND 2% NR
support
disturbance to organ function (respiratory failure, cardiac failure or pulmonary edema)
C

Heimburger, D.C.
2010 142 HIV infected patients HP and clinical symptoms 0.7% NR
[42]
shifts in electrolytes (PO4, K, Mg) AND edema AND impaired carbohydrate
AC

Herranz, A. [58] 2013 312 patients receiving PN 0% NR

metabolism during PN
Faintuch, J. [59] 2001 8 hungerstrike patients relevant shifts in electrolytes and clinical symptoms 0% NR
Elnenaei, M.O. [60] 2011 35 patients admitted for PN clinical features 0% NR
Leclerc, A. [61] 2013 29 AN patients signs and symptoms of shifts in fluid and electrolytes (particularly phosphate) 0% NR
clinical signs and symptoms of water retention, HP or of depletion of other electrolytes
Gentile, M.G. [62] 2010 33 Very severe AN patients 0% NR
or the thiamine co-factor of glycolysis

electrolyte disturbances (K <2.5mmol/l, PO4 <0.32mmol/l, Mg <0.5 mmol/L) AND

Hofer, M. [23] 2014 65 AN patients clinical symptoms (peripheral edema or acute circulatory fluid overload) AND 0% NR
disturbance to organ function (respiratory failure, cardiac failure or pulmonary edema)

Golden, N.H. [63] 2013 310 AN patients HP associated with possible hypomagnesaemia and hypokalemia and clinical features 0% NR
Raj, K.S. [64] 2012 541 AN patients metabolic and clinical changes under RF NR 4.9 +/- 5.5 days
Garber, A.K. [65] 2011 40 AN patients NR 0% NR
Patients with moderate or severe
Hoffmann, M. [66] 2008 621 NR 10% NR
HP
NR = not reported AN = anorexic Not-AN = not anorexic HP = hypophosphatemia PO4 = phosphate K = potassium Na = sodium Mg = magnesium
EN = enteral nutrition PN = parenteral nutrition RF = refeeding

15
 Table 2 – Adverse outcome of RFS
ACCEPTED MANUSCRIPT Other
Mortality Association with
Firstauthor Y N= Population Definition of RFS RFS
Adverse
rate Outcome
death within 7 days and
Zeki, S. [35] 2011 321 patients receiving PN or EN PO4 < 0.6mmol/L NR NR
HP were not associated
HYPOPHOSPHATEMIA

frail elderly patients with No correlation in 1st

Lubart, E. [47] 2009 40 PO4 < 0.52mmol/L 6% NR
feeding problems >72h week and 1st month

PO4 drop of >30% during the first 36h of PN

Goyale, A. [34] 2015 52 patients receiving PN or EN 0% NR NR
administration
patients >65yrs with at least threefold
not an independent
Kagansky, N. [51] 2005 325 one episode of HP compared PO4 < 0.77mmol/L increased Longer LOS
predictor for mortality
with patients without HP mortality rates
ICU patients receiving EN or ICU LOS
Coskun, R. [43] 2014 117 PO4 ≤ 0.78 mmol/L NR NR
PN

PT
longer
patients with mild and severe
ELECTROLYTE

Hernandez-Aranda,
1997 50 malnutrition receiving EN or PN drop of any electrolyte value under the reference range 29% NR Longer LOS
SHIFT IN

J.C. [52]
>7 days
ICU patients starved for at least drop of PO4 by > 0.16mmol/L to below 0.65mmol/L

RI
Marik, P. E. [25] 1996 62 NR NR Longer LOS
48h accompanied by possible other electrolyte disturbances

electrolyte disturbances (K <2.5, PO4 <0.32, Mg <0.5

mmol/L) AND clinical symptoms (peripheral edema or
Eichelberger, M. [24] 2014 37 hungerstrike patients acute circulatory fluid overload) AND disturbance to 0% NR NR

SC
organ function (respiratory failure, cardiac failure or
CLINICAL SYMPTOMS

pulmonary edema)

electrolyte disturbances (K <2.5, PO4 <0.32, Mg <0.5

mmol/L) AND clinical symptoms (peripheral edema or
patients on artificial nutrition

U
Rio, A. [22] 2013 243 acute circulatory fluid overload) AND disturbance to 0% NR NR
support
organ function (respiratory failure, cardiac failure or
pulmonary edema)
AN
all adverse events occurring during nutritional
Vignaud, M. [30] 2010 68 AN patients rehabilitation of malnourished patients or having 71% NR NR
undergone a prolonged fast
lower survival
Heimburger, D.C. [42] 2010 142 HIV infected patients HP and clinical symptoms probability at 12 NR NR
weeks
M

NR = not reported AN = anorexic Not-AN = not anorexic HP = hypophosphatemia PO4 = phosphate K = potassium Na = sodium Mg = magnesium
EN = enteral nutrition PN = parenteral nutrition RF = refeeding LOS = length of stay
D
TE
C EP
AC

16
 Table 3 – Risk factors of RFS
ACCEPTED MANUSCRIPT
Firstauthor Y N= Definition of RFS Risk Factors

alcoholism, malignancy, anorexia nervosa, Crohn's disease, recurrent small

Terlevich, A. [38] 2003 30 PO4 < 0.50 mmol/L
HYPOPHOSPHATEMIA as definition criteria

bowel obstruction, pneumonia, dysphagia

PO4 drop of ≥ 0.15mmol/L to < 0.80mmol/L from baseline in the

Marvin, V. [28] 2007 250 cancer patients
first 7 days of PN in patients with a period of > 48h without food

PO4 drop of ≥ 0.15mmol/L to < 0.80mmol/L from baseline in the

Marvin, V. [27] 2008 140 high NRS ≥3, lower Mg, age, <12mmol PO4 in first day's PN regimen
first 7 days of PN in patients with a period of > 48h without food
Brown, CA. [46] 2015 123 PO4 < 0.87mmol/L low BMI, low K, low prealbumin and high hemoglobin concentration
Gaudiani, L. [45] 2014 132 hypophosphatemia low prealbumin concentration
Coskun, R. [43] 2014 117 PO4 ≤ 0.78 mmol/L EN with energy provided in a short time
Gonzalez Avila, G. [32] 1996 106 PO4 < 0.81mmol/L age > 60yrs and moderate or severe malnutrition

PT
Goyale, A. [34] 2015 52 PO4 drop of >30% during the first 36h of PN administration IGF-1 <63.7 ug/l
infusion of glucose solutions, EN, weight loss, fever, high WBC count and
Kagansky, N. [51] 2005 651 PO4 < 0.77mmol/L
creatinine kinase, low baseline concentration of PO4 and hypoalbuminaemia
Zeki, S. [35] 2011 321 PO4 < 0.6mmol/L enteral tube feeding

RI
Gaudiani, L. [44] 2012 25 PO4 < 0.87mmol/L no risk factors found
Redgrave, G.W. [53] 2015 461 shifts in electrolytes lower admission BMI
ELECTROLYTE

Whitelaw, M. [36] 2010 46 electrolyte and fluid shifts during RF malnourished with % IBW <68
SHIFT IN

SC
Grasso, S. [49] 2013 34 HP and/or other shifts in electrolytes during RF hypokalemia
drop of PO4 by > 0.16mmol/L to below 0.65mmol/L accompanied
Marik, P. E. [25] 1996 62 poor caloric intake for as short as 48hours and low prealbumin concentration
by possible other electrolyte disturbances

electrolyte disturbances (K <2.5, PO4 <0.32, Mg <0.5 mmol/L)

U
AND clinical symptoms (peripheral edema or acute circulatory
Hofer, M. [23] 2014 65 NICE
fluid overload) AND disturbance to organ function (respiratory
failure, cardiac failure or pulmonary edema)
AN
Raj, K.S. [64] 2012 541 metabolic and clinical changes under RF NICE
all adverse events occurring during nutritional rehabilitation of
Vignaud, M. [30] 2010 68 poor caloric intake, NICE and higher nutritional intake during RF
malnourished patients or having undergone a prolonged fast

electrolyte disturbances (K <2.5, PO4 <0.32, Mg <0.5 mmol/L)

M

AND clinical symptoms (peripheral edema or acute circulatory

Eichelberger, M. [24] 2014 37 NICE
fluid overload) AND disturbance to organ function (respiratory
CLINICAL SYMPTOMS

failure, cardiac failure or pulmonary edema)

Chen, L. [55] 2014 56 shifts in electrolyte and clinical symptoms NICE and EN in nursing homes
D

electrolyte disturbances (K <2.5, PO4 <0.32, Mg <0.5 mmol/L)

AND clinical symptoms (peripheral edema or acute circulatory
Rio, A. [22] 2013 243 weight loss, poor caloric intake, NICE and low serum Mg concentration
fluid overload) AND disturbance to organ function (respiratory
TE

failure, cardiac failure or pulmonary edema)

clinical signs and symptoms of water retention, HP or of depletion

Gentile, M.G. [62] 2010 33 weight loss
of other electrolytes or the thiamine co-factor of glycolysis
EP

HP associated with possible hypomagnesaemia and hypokalemia

Golden, N.H. [63] 2013 310 poor caloric intake and degree of malnutrition
and clinical features
Flesher, M.E. [29] 2005 51 shifts in electrolytes associated with complications during early EN malnourished patients
nutrition support suddenly introduced at full strength and patients with GI-
Fan, C.G. [56] 2004 15 shifts in electrolytes and clinical symptoms
C

fistula
Elnenaei, M.O. [60] 2011 35 Clinical features no risk factors found
AC

history of significant alcohol abuse, 5% weight loss in the preceding 30 d or

Charles, E.J. [67] 2014 83 NR
10% loss in the preceding 6 months, or poor caloric intake for at least 7 days
Henderson, S. [68] 2010 43 NR no risk factors found

NR = not reported AN = anorexic Not-AN = not anorexic HP = hypophosphatemia PO4 = phosphate K = potassium Na = sodium Mg = magnesium
EN = enteral nutrition PN = parenteral nutrition RF = refeeding

17
Table 3a – Risk factors of RFS
ACCEPTED MANUSCRIPT
No of patients
Risk Factor No of Studies Studies and Firstauthors included

Vignaud, M. [30], Rio, A. [22], Charles, E.J. [67], Golden, N.H. [63], Flesher, M.E. [29], Gonzalez Avila,
poor caloric intake / malnutrition 8 969
G. [32], Marik, P. E. [25], Whitelaw, M. [36]
NICE 6 Hofer, M. [23], Raj, K.S. [64], Vignaud, M. [30], Eichelberger, M. [24], Chen, L. [55], Rio, A. [22] 1000
low electrolyte concentrations 5 Rio, A. [22], Marvin, V. [27], Brown, CA. [46], Grasso, S. [49], Kagansky, N. [51] 1191
weight loss 4 Rio, A. [22], Charles, E.J. [67], Gentile, M.G. [62], Kagansky, N. [51] 1010
low prealbumin / albumin concentration 4 Gaudiani, L. [45], Brown, CA. [46], Marik, P. E. [25], Kagansky, N. [51] 977
enteral nutrition 3 Chen, L. [55], Zeki, S. [35], Kagansky, N. [51] 707
no risk factors found 3 Elnenaei, M.O. [60], Henderson, S. [68], Gaudiani, L. [45] 103
higher or faster nutritional intake during RF 3 Vignaud, M. [30], Coskun, R. [43], Fan, C.G. [56] 200

PT
low admission weight / BMI / AN patients 3 Terlevich, A. [38], Redgrave, G.W. [53], Brown, CA. [46] 614
alcohol abuse 2 Charles, E.J. [67], Terlevich, A. [38] 113
age 2 Marvin, V. [27], Gonzalez Avila, G. [32] 246
malignancy 2 Terlevich, A. [38], Marvin, V. [28] 280

RI
IGF-1 <63.7 ug/l 1 Goyale, A. [34] 52
high NRS ≥3 1 Marvin, V. [27] 140
high hemoglobin concentration 1 Brown, CA. [46] 123

SC
AN = anorexic EN = enteral nutrition RF = refeeding

U
Table 4 – Preventive and therapeutic measurements

Therapeutic
AN
Firstauthor Y N= Preventive Medication Effective?
Medication
Terlevich, A. [38] 2003 30 NR 50mmol PO4 over 24 h yes
HYPOPHOSPHATEMIA

Gonzalez Avila, G. [32] 1996 106 PO4 supplementation NR yes

as definition criteria

during the first 24h slow PN regimen providing <70% of protein and calories but
Marvin, V. [27] 2008 140 NR yes
M

> 12mmol PO4

Garber, A.K. [65] 2011 40 no effective preventive measures found NR no
Coskun, R. [43] 2014 117 lower energy intake NR no
Doig, G. [26] 2015 339 NR lower caloric intake yes
D

prophylactic administration of PO4 no

ELECTROLYTE

Whitelaw, M. [36] 2010 46 supplementation of

SHIFT IN

lower initial energy intake, monitoring of PO4 yes

TE

phosphate
Luque, S. [54] 2007 11 PO4 supplementation, thiamine 3.51mg/day NR yes
Manning, S. [37] 2014 36 repeated electrolyte testing NR no
Fan, C.G. [56] 2004 158 PO4 supplementation NR yes if phosphate < 0.30
EP

Gentile, M.G. [62] 2010 33 prophylactic administration of PO4 and K, cautious nutritional rehabilitation NR yes
CLINICAL SYMPTOMS

for patients at risk initial nutritional support 10kcal/kg/d falling to as low as

Vignaud, M. [30] 2010 68 NR yes
5kcal/kg/d
Chen, L. [55] 2014 56 thiamine and multivitamin supplementation, 15kcal/kg/day NR yes
C

Golden, N.H. [63] 2013 310 lower caloric intake NR no

Leclerc, A. [61] 2013 29 hypocaloric feeding NR no
Flesher, M.E. [29] 2005 51 thiamine supplementation, cautious feeding NR no
AC

Rio, A. [22] 2013 243 hypocaloric feeding NR no

NR = not reported AN = anorexic Not-AN = not anorexic HP = hypophosphatemia PO4 = phosphate K = potassium Na = sodium Mg = magnesium
EN = enteral nutrition PN = parenteral nutrition RF = refeeding

18
 ACCEPTED MANUSCRIPT
References
1. Pirlich M, Schutz T, Norman K, Gastell S, Lubke HJ, Bischoff SC, Bolder U, Frieling T,
Guldenzoph H, Hahn K, et al: The German hospital malnutrition study. Clin Nutr 2006,
25:563-572.
2. Katona P, Katona-Apte J: The interaction between nutrition and infection. Clin Infect Dis
2008, 46:1582-1588.
3. Felder S, Lechtenboehmer C, Bally M, Fehr R, Deiss M, Faessler L, Kutz A, Steiner D, Rast AC,
Laukemann S, et al: Association of nutritional risk and adverse medical outcomes across
different medical inpatient populations. Nutrition 2015, 31:1385-1393.

PT
4. Schuetz P: "Eat your lunch!" - controversies in the nutrition of the acutely, non-critically ill
medical inpatient. Swiss Med Wkly 2015, 145:w14132.
5. Schutz P, Bally M, Stanga Z, Keller U: Loss of appetite in acutely ill medical inpatients:

RI
physiological response or therapeutic target? Swiss Med Wkly 2014, 144:w13957.
6. Bally MR, Blaser Yildirim PZ, Bounoure L, Gloy VL, Mueller B, Briel M, Schuetz P: Nutritional
Support and Outcomes in Malnourished Medical Inpatients: A Systematic Review and

SC
Meta-analysis. JAMA Intern Med 2016, 176:43-53.
7. Preiser JC, van Zanten AR, Berger MM, Biolo G, Casaer MP, Doig GS, Griffiths RD, Heyland DK,
Hiesmayr M, Iapichino G, et al: Metabolic and nutritional support of critically ill patients:
consensus and controversies. Crit Care 2015, 19:35.

U
8. Felder S, Braun N, Stanga Z, Kulkarni P, Faessler L, Kutz A, Steiner D, Laukemann S, Haubitz S,
Huber A, et al: Unraveling the Link between Malnutrition and Adverse Clinical Outcomes:
AN
Association of Acute and Chronic Malnutrition Measures with Blood Biomarkers from
Different Pathophysiological States. Ann Nutr Metab 2016, 68:164-172.
9. Schnitker MA, Mattman PE, Bliss TL: A clinical study of malnutrition in Japanese prisoners of
M

war. Ann Intern Med 1951, 35:69-96.

10. Burger GC DJ, Sandstead HR.: Appendices to malnutrition and starvation in western
Netherlands September 1944–July 1945 (part II). he Hague General State Printing Office,
1948.
D

11. McCray S, Walker S, Parrish CR: Much ado about refeeding. Practical Gastroenterology 2005,
29:26+31-33+37-40+43-44.
TE

12. Boateng AA, Sriram K, Meguid MM, Crook M: Refeeding syndrome: treatment
considerations based on collective analysis of literature case reports. Nutrition 2010,
26:156-167.
EP

13. Mehanna HM, Moledina J, Travis J: Refeeding syndrome: what it is, and how to prevent and
treat it. Bmj 2008, 336:1495-1498.
14. Mehler PS, Winkelman AB, Andersen DM, Gaudiani JL: Nutritional rehabilitation: practical
guidelines for refeeding the anorectic patient. J Nutr Metab 2010, 2010.
C

15. Crook MA, Hally V, Panteli JV: The importance of the refeeding syndrome. Nutrition 2001,
17:632-637.
AC

16. Solomon SM, Kirby DF: The refeeding syndrome: a review. JPEN J Parenter Enteral Nutr
1990, 14:90-97.
17. Brooks MJ, Melnik G: The refeeding syndrome: an approach to understanding its
complications and preventing its occurrence. Pharmacotherapy 1995, 15:713-726.
18. Knochel JP: Hypophosphatemia. Clin Nephrol 1977, 7:131-137.
19. Stanga Z, Brunner A, Leuenberger M, Grimble RF, Shenkin A, Allison SP, Lobo DN: Nutrition in
clinical practice-the refeeding syndrome: illustrative cases and guidelines for prevention
and treatment. Eur J Clin Nutr 2008, 62:687-694.
20. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P: Preferred reporting items for systematic
reviews and meta-analyses: the PRISMA statement. J Clin Epidemiol 2009, 62:1006-1012.
21. Becker KL: Pedigree analysis. 2. Clinical discrepancies and limitations. 1966, 40:171-174.
22. Rio A, Whelan K, Goff L, Reidlinger DP, Smeeton N: Occurrence of refeeding syndrome in
adults started on artificial nutrition support: prospective cohort study. BMJ Open 2013, 3.

19
 ACCEPTED MANUSCRIPT
23. Hofer M, Pozzi A, Joray M, Ott R, Hahni F, Leuenberger M, von Kanel R, Stanga Z: Safe
refeeding management of anorexia nervosa inpatients: an evidence-based protocol.
Nutrition 2014, 30:524-530.
24. Eichelberger M, Joray ML, Perrig M, Bodmer M, Stanga Z: Management of patients during
hunger strike and refeeding phase. Nutrition 2014, 30:1372-1378.
25. Marik PE, Bedigian MK: Refeeding hypophosphatemia in critically ill patients in an intensive
care unit. A prospective study. Arch Surg 1996, 131:1043-1047.
26. Doig GS, Simpson F, Heighes PT, Bellomo R, Chesher D, Caterson ID, Reade MC, Harrigan PW,
Refeeding Syndrome Trial Investigators G: Restricted versus continued standard caloric

PT
intake during the management of refeeding syndrome in critically ill adults: a randomised,
parallel-group, multicentre, single-blind controlled trial. Lancet Respir Med 2015.
27. Marvin VA, Brown D, Portlock J, Livingstone C: Factors contributing to the development of
hypophosphataemia when refeeding using parenteral nutrition. Pharm World Sci 2008,

RI
30:329-335.
28. Marvin V, May C, Livingstone C, Davis J: Incidence of hypophosphataemia in patients on
parenteral nutrition. Hospital Pharmacist 2007, 14:166-169.

SC
29. Flesher ME, Archer KA, Leslie BD, McCollom RA, Martinka GP: Assessing the metabolic and
clinical consequences of early enteral feeding in the malnourished patient. JPEN J Parenter
Enteral Nutr 2005, 29:108-117.

U
30. Vignaud M, Constantin JM, Ruivard M, Villemeyre-Plane M, Futier E, Bazin JE, Annane D,
group A: Refeeding syndrome influences outcome of anorexia nervosa patients in intensive
AN
care unit: an observational study. Crit Care 2010, 14:R172.
31. Care NCCfA: Nutrition support in adults Oral nutrition support, enteral tube feeding and
parenteral nutrition. National Collaborating Centre for Acute Care, London 2006.
32. Gonzalez Avila G, Fajardo Rodriguez A, Gonzalez Figueroa E: [The incidence of the refeeding
M

syndrome in cancer patients who receive artificial nutritional treatment]. Nutr Hosp 1996,
11:98-101.
33. Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc EWG: Nutritional risk screening
D

(NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 2003,
22:321-336.
TE

34. Goyale A, Ashley SL, Taylor DR, Elnenaei MO, Alaghband-Zadeh J, Sherwood RA, le Roux CW,
Vincent RP: Predicting refeeding hypophosphataemia: insulin growth factor 1 (IGF-1) as a
diagnostic biochemical marker for clinical practice. Ann Clin Biochem 2015, 52:82-87.
35. Zeki S, Culkin A, Gabe SM, Nightingale JM: Refeeding hypophosphataemia is more common
EP

in enteral than parenteral feeding in adult in patients. Clin Nutr 2011, 30:365-368.
36. Whitelaw M, Gilbertson H, Lam PY, Sawyer SM: Does aggressive refeeding in hospitalized
adolescents with anorexia nervosa result in increased hypophosphatemia? J Adolesc Health
2010, 46:577-582.
C

37. Manning S, Gilmour M, Weatherall M, Robinson GM: Refeeding syndrome is uncommon in

alcoholics admitted to a hospital detoxification unit. Intern Med J 2014, 44:512-514.
AC

38. Terlevich A, Hearing SD, Woltersdorf WW, Smyth C, Reid D, McCullagh E, Day A, Probert CS:
Refeeding syndrome: effective and safe treatment with Phosphates Polyfusor. Aliment
Pharmacol Ther 2003, 17:1325-1329.
39. Suki WR, D.: Renal transport of calcium, magnesium, and phosphorus. Philadelphia1991.
40. Jahnen-Dechent W, Ketteler M: Magnesium basics. Clin Kidney J 2012, 5:i3-i14.
41. Robinson JF, MS: The (Na+ + K+)-activated ATPase.1979.
42. Heimburger DC, Koethe JR, Nyirenda C, Bosire C, Chiasera JM, Blevins M, Munoz AJ,
Shepherd BE, Potter D, Zulu I, et al: Serum phosphate predicts early mortality in adults
starting antiretroviral therapy in Lusaka, Zambia: a prospective cohort study. PLoS One
2010, 5:e10687.
43. Coskun R, Gundogan K, Baldane S, Guven M, Sungur M: Refeeding hypophosphatemia: a
potentially fatal danger in the intensive care unit. Turk J Med Sci 2014, 44:369-374.

20
 ACCEPTED MANUSCRIPT
44. Gaudiani JL, Sabel AL, Mascolo M, Mehler PS: Severe anorexia nervosa: outcomes from a
medical stabilization unit. Int J Eat Disord 2012, 45:85-92.
45. Gaudiani JL, Sabel AL, Mehler PS: Low prealbumin is a significant predictor of medical
complications in severe anorexia nervosa. Int J Eat Disord 2014, 47:148-156.
46. Brown CA, Sabel AL, Gaudiani JL, Mehler PS: Predictors of hypophosphatemia during
refeeding of patients with severe anorexia nervosa. Int J Eat Disord 2015.
47. Lubart E, Leibovitz A, Dror Y, Katz E, Segal R: Mortality after nasogastric tube feeding
initiation in long-term care elderly with oropharyngeal dysphagia--the contribution of
refeeding syndrome. Gerontology 2009, 55:393-397.

PT
48. Garber AK, Mauldin K, Michihata N, Buckelew SM, Shafer MA, Moscicki AB: Higher calorie
diets increase rate of weight gain and shorten hospital stay in hospitalized adolescents
with anorexia nervosa. J Adolesc Health 2013, 53:579-584.
49. Grasso S, Ferro Y, Migliaccio V, Mazza E, Rotundo S, Pujia A, Montalcini T: Hypokalemia

RI
during the early phase of refeeding in patients with cancer. Clinics (Sao Paulo) 2013,
68:1413-1415.
50. Nagata JM, Park KT, Colditz K, Golden NH: Associations of elevated liver enzymes among

SC
hospitalized adolescents with anorexia nervosa. J Pediatr 2015, 166:439-443.e431.
51. Kagansky N, Levy S, Koren-Morag N, Berger D, Knobler H: Hypophosphataemia in old
patients is associated with the refeeding syndrome and reduced survival. J Intern Med

U
2005, 257:461-468.
52. Hernandez-Aranda JC, Gallo-Chico B, Luna-Cruz ML, Rayon-Gonzalez MI, Flores-Ramirez LA,
AN
Ramos Munoz R, Ramirez-Barba EJ: [Malnutrition and total parenteral nutrition: a cohort
study to determine the incidence of refeeding syndrome]. Rev Gastroenterol Mex 1997,
62:260-265.
53. Redgrave GW, Coughlin JW, Schreyer CC, Martin LM, Leonpacher AK, Seide M, Verdi AM,
M

Pletch A, Guarda AS: Refeeding and weight restoration outcomes in anorexia nervosa:
Challenging current guidelines. Int J Eat Disord 2015.
54. Luque S, Berenguer N, Mateu de Antonio J, Grau S, Morales-Molina JA: [Patients at risk of
D

malnutrition: assessment of 11 cases of severe malnutrition with individualised total

parenteral nutrition]. Farm Hosp 2007, 31:238-242.
TE

55. Chen HL, Bair MJ, Lin IT, Wu CH, Lee YK: Clinical experience of refeeding syndrome in
southeastern Taiwan. Journal of Gastroenterology and Hepatology 2012, 27:308.
56. Fan CG, Ren JA, Wang XB, Li JS: Refeeding syndrome in patients with gastrointestinal fistula.
Nutrition 2004, 20:346-350.
EP

57. Saito S, Kobayashi T, Kato S: Management and treatment of eating disorders with severe
medical complications on a psychiatric ward: a study of 9 inpatients in Japan. Gen Hosp
Psychiatry 2014, 36:291-295.
58. Herranz Antolin S, Alvarez De Frutos V, Blasco Guerrero M, Garcia Martinez Mdel C, Gimeno
C

Fernandez Mdel C: [Nutritional support with parenteral nutrition. Course and associated
complications]. Endocrinol Nutr 2013, 60:287-293.
AC

59. Faintuch J, Soriano FG, Ladeira JP, Janiszewski M, Velasco IT, Gama-Rodrigues JJ: Refeeding
procedures after 43 days of total fasting. Nutrition 2001, 17:100-104.
60. Elnenaei MO, Alaghband-Zadeh J, Sherwood R, Awara MA, Moniz C, le Roux CW: Leptin and
insulin growth factor 1: diagnostic markers of the refeeding syndrome and mortality. Br J
Nutr 2011, 106:906-912.
61. Leclerc A, Turrini T, Sherwood K, Katzman DK: Evaluation of a nutrition rehabilitation
protocol in hospitalized adolescents with restrictive eating disorders. J Adolesc Health 2013,
53:585-589.
62. Gentile MG, Pastorelli P, Ciceri R, Manna GM, Collimedaglia S: Specialized refeeding
treatment for anorexia nervosa patients suffering from extreme undernutrition. Clin Nutr
2010, 29:627-632.

21
 ACCEPTED MANUSCRIPT
63. Golden NH, Keane-Miller C, Sainani KL, Kapphahn CJ: Higher caloric intake in hospitalized
adolescents with anorexia nervosa is associated with reduced length of stay and no
increased rate of refeeding syndrome. J Adolesc Health 2013, 53:573-578.
64. Raj KS, Keane-Miller C, Golden NH: Hypomagnesemia in adolescents with eating disorders
hospitalized for medical instability. Nutr Clin Pract 2012, 27:689-694.
65. Garber AK, Michihata N, Hetnal K, Shafer MA, Moscicki AB: A prospective examination of
weight gain in hospitalized adolescents with anorexia nervosa on a recommended
refeeding protocol. J Adolesc Health 2012, 50:24-29.
66. Hoffmann M, Zemlin AE, Meyer WP, Erasmus RT: Hypophosphataemia at a large academic

PT
hospital in South Africa. J Clin Pathol 2008, 61:1104-1107.
67. Charles EJ, Petroze RT, Metzger R, Hranjec T, Rosenberger LH, Riccio LM, McLeod MD, Guidry
CA, Stukenborg GJ, Swenson BR, et al: Hypocaloric compared with eucaloric nutritional
support and its effect on infection rates in a surgical intensive care unit: A randomized

RI
controlled trial. American Journal of Clinical Nutrition 2014, 100:1337-1343.
68. Henderson S, Boyce F, Sumukadas D, Witham MD: Changes in serum magnesium and
phosphate in older hospitalised patients--correlation with muscle strength and risk factors

SC
for refeeding syndrome. J Nutr Health Aging 2010, 14:872-876.
69. Birmingham CL, Puddicombe D, Hlynsky J: Hypomagnesemia during refeeding in anorexia
nervosa. Eat Weight Disord 2004, 9:236-237.

U
70. Hayek ME, Eisenberg PG: Severe hypophosphatemia following the institution of enteral
feedings. Arch Surg 1989, 124:1325-1328.
AN
71. Koethe JR, Blevins M, Nyirenda CK, Kabagambe EK, Chiasera JM, Shepherd BE, Zulu I,
Heimburger DC: Serum Phosphate Predicts Early Mortality among Underweight Adults
Starting ART in Zambia: A Novel Context for Refeeding Syndrome? J Nutr Metab 2013,
2013:545439.
M
D
TE
C EP
AC

22
 ACCEPTED MANUSCRIPT
Appendices:

Appendix 1 - SEARCH PROTOCOL:

1. Pubmed
I. Search ((Refeeding) NOT child*) NOT neonat*) NOT Review[Publication Type] NOT
(animal*)
II. Search ((("Deficiency diseases"[Mesh] OR "Refeeding syndrome"[Mesh] AND
"refeeding"[Text Word] OR "refeeding syndrome"[Text Word])) OR ((("refeeding

PT
syndrome"[Text Word] OR "refeeding"[Text Word] NOT medline[sb])) OR ("Refeeding
Syndrome"[Mesh] AND "humans"[MeSH Terms] AND "adult"[MeSH Terms] NOT
"Review"[ptyp])))

RI
2. Embase
I. 'refeeding' NOT 'review':it NOT 'child' NOT 'children' NOT 'neonates' NOT 'neonatal'
NOT 'animal' AND [english]/lim
II. 'refeeding syndrome'/exp OR 'refeeding syndrome'/de OR 'refeeding syndrome' OR

SC
'refeeding'/exp OR 'refeeding'/de OR 'refeeding' AND ([article]/lim OR [article in
press]/lim OR [conference abstract]/lim OR [conference paper]/lim OR [conference
review]/lim) AND [humans]/lim AND [english]/lim AND [embase]/limAND ([adult]/lim
OR [aged]/lim OR [middle aged]/lim OR [very elderly]/lim) AND ('clinical article'/de OR

U
'cohort analysis'/de OR 'comparative study'/de OR 'controlled clinical trial'/de OR
'controlled study'/de OR 'double blind procedure'/de OR 'major clinical study'/de OR
AN
'medical audit'/de OR 'normal human'/de OR 'observational study'/de OR 'prospective
study'/de OR 'randomized controlled trial'/de OR 'retrospective study'/de)
M

Appendix 2 - EXTRACTION FILE:

Length of
Kind of Definition Treatment
Author Y Title n= Methods Outcome Mainresults Conclusion Limitations Energy Proteins Electrolytes Liquid Micronutritients nutrition
study RFS RFS
therapy
D
TE

Appendix 3 – SIGN QUALITY ASSESSMENT:

EP

SIGN Quality Assessment

Y = YES N = NO NA = Does not apply CL No = Checklist Number (2 = Controlled Trials, 3 = Cohort studies, 4 = Case-control studies)
C

INTERVENTIONAL STUDIES
Study
CL
AC

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 2.1 2.2 2.3 2.4 additional comments and remarks
No

More patients at risk for RFS in eucaloric group (25 vs 13), RFS is no main
Charles E.J. 12% / outcome and not main subject of study. Very small study group, due to slow
2 Y Y Y N Y N Y Y NA +/++ N N
[67] 26% enrollment study was closed before full planned enrollment was reached. Single
centre study, designed as pilot study.
Terlevich A. Uncontrolled, not randomized study. Unprecise definition for RFS, no precise
2 Y NA NA NA NA NA Y X NA NA - N N
[38] endpoints, small study group, lack of detailed patients characteristics.
All patients had standard nutritional support in the beginning. Definition of RFS
4% /
Doig, G. [26] 2 Y Y Y N Y Y Y Y NA ++ Y N was hyposphophataemia only. Only ICU patients, no double blinding. No
3%
standard calorie amount for standard care group.
OBSERVATIONAL STUDIES
CL
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 1.14 2.1 2.2 2.3 2.4 additional comments and remarks
No
Hofer, M.
3 Y NA N X 10.5% Y Y NA X Y Y Y Y N + Y Y Retrospective study design.
[23]
Birmingham RFS is not the main subject, small study
3 Y NA NA Y 0% NA Y NA NA NA Y Y N N + Y N
[69] size, no control group, single centre study.
Groups not enrolled at the same time (10
Garber, A.K. yrs difference; secular trends may have
3 Y Y NA NA 0% NA Y NA N Y Y Y Y Y + Y Y
[48] influenced patient characteristics), small
study size, single centre trial.

23
 ACCEPTED MANUSCRIPT
Garber, A.K. No control group, RFS not the main subject,
3 Y Y N NA 12.5% N Y NA NA Y Y Y N N + Y N
[65] small study size, single centre study.
Gaudiani, L. Single centre study, very small study group,
3 N NA N Y NA NA Y NA NA Y Y Y Y Y + Y Y
[44] no control group.
Single centre study, no control group, low
prealbumin levels cannot be said to cause
any medical complications per se. Nadir
Gaudiani, L.
3 Y NA N Y NA NA Y NA NA Y Y Y Y Y + Y Y weight was used in statistical analyses. No
[45]
adolescents have been treated nor less
acutely ill patients such as might be found in
outpatient settings.
Small cohort group, no strict refeeding
protocol, patients were encouraged - not
Gentile,
3 Y Y N N 12 % NA Y NA NA Y Y Y N Y + Y N forced - to eat enough to restore their
M.G. [62]
weight, no control group, single centre
study.
Groups not in all respects comparable

PT
(group sizes, duration of illness). Initial
caloric prescritpion was made by individual
providers based on their clinical
Golden, N.H.
3 Y N N NA 0% NA Y NA Y Y Y Y Y Y + Y Y assessments. Some providers elected to
[63]
supplement phosphorus prophylactically
before levels dropped below 3mg/dl. No
large difference in the amount of calories

RI
prescribed between the two groups.
Extremely malnourished patients were
excluded (below 70% of their ideal body
Leclerc, A. weight). Any patients needing nastogastric
3 Y Y NA Y NA NA Y NA NA Y Y N N Y + Y Y
[61] supplementation were removed from the
study. Small number of patients. No control

SC
group, single centre.
Retrospective study. Inability to prove
causality given its noncontrolled setting of
Nagata, J.M. the study. Selection bias as data were
3 Y Y NA N NA NA Y NA NA Y Y N Y Y + Y N
[50] gathered at admission for all patients but not
necessarily repeated through the
hospitalization if initially normal.

U
RFS is not the central investigated subject,
single centre, retrospective, no control
Raj, K.S. group, important missing data (urine pH was
3 Y Y NA N NA NA Y NA NA Y Y N Y Y + Y N
[64] available only for a subgroup of patients),
AN
lack of uniformity of supplementation of Mg
and PO4.
Specialty behavioral eating disorders unit on
a psychiatry service. Difficult to implement
the protocol in non-hospital based
Redgrave, residential treatment centers. Study is
3 Y Y NA Y NA NA Y NA NA Y N N N Y + Y N
G.W. [53] unable to assess potential effects of
M

medication on weight gain. Study is

observational and does not randomize
participants.
No information about calories used for
Saito, S. [57] 3 Y NA Y N 0% NA Y NA X Y N N N Y + N N refeeding. Very small study cohort. Single
centre and retrospective design.
D

Only ICU patients, small cohort.

Vignaud, M.
3 Y NA Y N 0% NA Y NA X Y Y Y N N + N Y Retrospective study, no precise definition of
[30]
RFS.
Small number of patients included,
TE

calculations for energy intake were based

Whitelaw, M.
3 Y NA Y N 0% NA Y NA N Y Y Y N Y + N Y on what was prescribed without taking into
[36]
account variation in serving sizes or what
was actually eaten.
Data accuracy relies on accuracy of
Eichelberger recorded medical charts. May have missed
3 Y Y NA Y 0% NA Y NA NA Y Y Y Y Y + Y Y
, M. [24] some hunter strike episodes, missing data
EP

did not allow for a systematic analysis.

Underestimated number of RFS due to
incomplete data collection and insufficient
knowledge with regards to RFS. Only the
Chen, L. [55] 3 Y NA N Y NA NA Y NA Y Y Y Y Y Y + Y Y
high risk groups were recognized but there
were no definitive criteria for diagnosis or
confirmation in a real world setting.
C

Retrospective design, unavailability of the

scoring systems regarding malnutrition,
Coskun, R. missing data on the daily dosage of
3 Y NA Y Y NA NA Y NA Y Y Y Y N Y + Y Y
AC

[43] intravenous phosphorus, magnesium, and

potassium replacements at the time of the
treatment.
Relatively small study population. Inability of
IGF1 assay to accurately quantify
concentrations below 25 ug/l and the fact
Elnenaei,
3 Y NA NA Y NA NA Y NA NA Y Y N N Y + N N that age and sex-unified values of IGF1 and
M.O. [60]
leptin were employed to obtain a single
factor to calculate the RI. Only patients
receiving PN were included into this study.
No prospective investigation or systematic
Faintuch, J. clinical protocol could be applied, all findings
3 Y NA Y NA 0% NA Y NA N Y N Y N Y + Y N
[59] were analyzed retrospectively. Not a
controlled trial. Very small study cohort.
Small cohort group, only patients with GI-
fistula included. No refeeding protocol, only
Fan, C.G.
3 Y NA NA Y 0% NA Y NA N Y Y Y N Y + Y N patients presenting with RFS included,
[56]
retrospective study, lack of detailed
definition of RFS.
Flesher, Missing precise definition of RFS. Small
3 Y NA Y Y NA N Y NA NA Y Y Y Y N + Y Y
M.E. [29] study cohort. Retrospective study.
Gonzalez Small study cohort. Single centre sutdy. No
3 Y Y N N NA NA Y NA NA Y Y Y N Y + Y Y
Avila, G. [32] RCT.
Age and gender specific IGF-1
Goyale, A.
3 Y NA NA Y NA NA Y NA N Y Y Y N Y + N N concentrations were not used due to the
[34]
small cohort. Single center study. Small

24
 ACCEPTED MANUSCRIPT
cohort.

Study was not designed to assess the

mechanisms of RFS. Accurate nutritional
Grasso, S.
3 Y Y Y N 0% NA Y NA N Y Y Y N Y + Y N intake investigation was not performed to
[49]
evaluate micronutrient intake. Small sample
size.
Very small study group, retrospective study
Hayek, M.E.
3 Y NA NA Y NA NA Y NA N Y Y Y N N + N N design, inclusion and ecxlusion criterias are
[70]
not clear, only intensive care patients.
Because all patients had to be provided
with supplementation of PO4, it cannot be
concluded that phophate supplementation
had no effect (lack of appropriate
Heimburger,
3 Y Y N Y 23 % Y Y NA N Y Y Y Y Y + Y N comparison group). Observational design
D.C. [42]
which prevents from concluding a causal
relationship between low serum phosphate

PT
and mortality. Participant attrition. Small
study cohort, 23% lost to follow up.
Small study sample. Very frail and ill
Henderson,
3 Y NA Y Y 0% NA Y NA N Y Y Y N Y + Y N patients are likely to be underrepresented.
S. [68]
Designed as a pilot study
Hernandez- Small study group, no multi centre study.
Aranda, J.C. 3 Y NA N Y NA NA Y NA N Y Y Y N Y + N N Observational study. RFS defined as any

RI
[52] electrolyte imbalance.
Most patients received PN because feeding
by the gastrointestinal route was not
Herranz, A. possible, rather than because of the
3 Y NA N Y NA NA Y NA N Y Y Y N Y + Y Y
[58] presence of associated malnutrition (no
starved patients). Composition of PN was

SC
not analyzed. Only PN analyzed.
Retrospective study design, records were
often incomplete, single centre study.
Hoffmann,
3 Y Y N Y NA NA Y NA N Y Y Y N N + N Y Diagnosis of possible refeeding was based
M. [66]
on history and clinical information only and
could not be confirmed.
Case control study, not randomized. Single

U
Kagansky,
4 Y Y Y X Y Y Y N Y N Y NA NA NA + Y Y Centre study. No defined nutritional
N. [51]
intervention.
HIV infected population, inability to
AN
determine actual causes of death and to
Koethe, J.R. make definitive diagnoses of opportunistic
3 Y Y N Y 12.8% N Y N N Y Y Y Y Y + Y N
[71] infections. Single centre study, no RFS
definition, no nutritional intervention took
place.
No patients with severe acute medical
Lubart, E.
3 Y NA N Y NA NA Y NA N Y Y Y N Y + N N problems included, small number of subjects
[47]
included.
M

Luque, S. Small study cohort, single centre study, no

3 Y NA Y N 0% NA Y NA N Y Y Y N Y + N N
[54] randomized trial, no exact definition of RFS.
Manning, S. Small study cohort, no refeeding algorithm,
3 Y NA Y Y NA N Y NA N Y Y Y N Y + N N
[37] single centre study, no randomised trial.
Small study cohort, single centre study, no
D

Marik, P. E.
3 Y Y Y N NA N Y NA N Y Y Y N Y + N N randomized trial, RFS definition as HP. No
[25]
information about exact feeding algorithm.
Heterogenous group of adult patients most
receiving PN after surgery.. PO4
TE

concentrations do not necessarily correlate

Marvin, V.
3 Y Y Y N 0% NA Y NA N Y Y Y N Y + Y N with development of symptoms as only total
[28]
plasma inorganic PO4 was measured (0.1%
of total body plasma). No information about
PN (amount, electrolyte substitution).
Recall bias on previous consumption of
alcohol and diet. Only patients receiving PN
EP

Marvin, V.
4 Y Y Y - Y Y Y N Y N Y NA NA NA + Y Y analyzed. No information about PN regime
[27]
(amount of calories, duration, electrolyte
supplementation, etc.).
Selection bias: The main source of data loss
was the excluded group which may
potentially have contained participants who
went on to develop the refeeding
C

Rio, A. [22] 3 Y Y Y Y 0% NA Y N N Y Y Y N Y + Y Y syndrome.The diagnosis of only three

participants limited the statistical analyses.
Lack of consensus on the electrolyte
AC

concentration values to diagnose the

syndrome. Single centre study.
Low PO4 as the sole surrogate marker for
RFS. No analysis of K and Mg shifts or
morbidity. Single centre study, retrospective
design. Patients were likely to have
Zeki, S. [35] 3 Y Y Y Y 0% NA Y NA N Y Y Y N Y + Y N
electrolyte disturbances corrected before
clinical manifestations became apparent
due to close monitoring. No details about
EN/PN amout of calories, etc.
Brown, CA.
4 Y Y Y 50% N Y Y N Y Y Y + Y Y Onyl female, small sample size.
[46]

25
 ACCEPTED MANUSCRIPT
Appendix 4 - SIGN CHECKLISTS:

Methodology Checklist 2: Controlled Trials

SIGN

Study identification (Include author, title, year of publication, journal title, pages)
Guideline topic: Key Question No: Reviewer:

Before completing this checklist, consider:

1. Is the paper a randomised controlled trial or a controlled clinical trial? If in doubt, check the study design algorithm
available from SIGN and make sure you have the correct checklist. If it is a controlled clinical trial questions 1.2, 1.3, and 1.4

PT
are not relevant, and the study cannot be rated higher than 1+

2. Is the paper relevant to key question? Analyse using PICO (Patient or Population Intervention Comparison Outcome). IF NO
REJECT (give reason below). IF YES complete the checklist.

RI
Reason for rejection: 1. Paper not relevant to key question
2. Other reason
(please specify):

SECTION 1: INTERNAL VALIDITY

SC
In a well conducted RCT study… Does this study do it?

1.1 The study addresses an appropriate and clearly focused question. Yes
No

Can’t say

1.2 The assignment of subjects to treatment groups is randomised. Yes
No

U
Can’t say

1.3 An adequate concealment method is used. Yes
No

AN
Can’t say

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation. Yes
No

Can’t say

1.5 The treatment and control groups are similar at the start of the trial. Yes
No

M

Can’t say □

1.6 The only difference between groups is the treatment under investigation. Yes
No

Can’t say

D

1.7 All relevant outcomes are measured in a standard, valid and reliable way. Yes
No

Can’t say

TE

1.8 What percentage of the individuals or clusters recruited into each treatment
arm of the study dropped out before the study was completed?

1.9 All the subjects are analysed in the groups to which they were randomly Yes
No

allocated (often referred to as intention to treat analysis). Can’t say
Does not apply

1.10 Where the study is carried out at more than one site, results are comparable Yes
No

EP

for all sites. Can’t say
Does not apply

SECTION 2: OVERALL ASSESSMENT OF THE STUDY

2.1
C

How well was the study done to minimise bias? High quality (++)

Code as follows:
Acceptable (+)

AC

Unacceptable – reject 0

2.2 Taking into account clinical considerations, your evaluation of the

methodology used, and the statistical power of the study, are you certain
that the overall effect is due to the study intervention?

2.3 Are the results of this study directly applicable to the patient group targeted
by this guideline?

2.4 Notes. Summarise the authors’ conclusions. Add any comments on your own assessment of the study, and the extent to which
it answers your question and mention any areas of uncertainty raised above.

Methodology Checklist 3: Cohort studies

SIGN

Study identification (Include author, title, year of publication, journal title, pages)

26
 ACCEPTED MANUSCRIPT
Guideline topic: Key Question No: Reviewer:

Before completing this checklist, consider:

3. Is the paper really a cohort study? If in doubt, check the study design algorithm available from SIGN and make sure you have
the correct checklist.

4. Is the paper relevant to key question? Analyse using PICO (Patient or Population Intervention Comparison Outcome). IF NO
REJECT (give reason below). IF YES complete the checklist..

Reason for rejection: 1. Paper not relevant to key question □ 2. Other reason □ (please specify):
Please note that a retrospective study (ie a database or chart study) cannot be rated higher than +.

SECTION 1: INTERNAL VALIDITY

PT
In a well conducted cohort study: Does this study do it?

1.1 The study addresses an appropriate and clearly focused question. Yes □ No □

RI
Can’t say □

SELECTION OF SUBJECTS

SC
1.2 The two groups being studied are selected from source populations that are Yes □ No □
comparable in all respects other than the factor under investigation.
Can’t say □ Does not apply □

1.3 The study indicates how many of the people asked to take part did so, in

U
Yes □ No □
each of the groups being studied.
Does not apply □
AN
1.4 The likelihood that some eligible subjects might have the outcome at the Yes □ No □
time of enrolment is assessed and taken into account in the analysis.
Can’t say □ Does not apply □
M

1.5 What percentage of individuals or clusters recruited into each arm of the
study dropped out before the study was completed.

1.6 Comparison is made between full participants and those lost to follow up, by Yes □ No □
D

exposure status.
Can’t say □ Does not apply □
TE

ASSESSMENT

1.7 The outcomes are clearly defined. Yes □ No □

Can’t say □
EP

1.8 The assessment of outcome is made blind to exposure status. If the study is Yes □ No □
retrospective this may not be applicable.
Can’t say □ Does not apply □
C

1.9 Where blinding was not possible, there is some recognition that knowledge Yes □ No □
of exposure status could have influenced the assessment of outcome.
Can’t say □ □
AC

1.10 The method of assessment of exposure is reliable. Yes □ No □

Can’t say □

1.11 Evidence from other sources is used to demonstrate that the method of Yes □ No □
outcome assessment is valid and reliable.
Can’t say □ Does not apply□

1.12 Exposure level or prognostic factor is assessed more than once. Yes □ No □

Can’t say □ Does not apply □

CONFOUNDING

1.13 The main potential confounders are identified and taken into account in the Yes □ No □
design and analysis.
Can’t say □

27
 ACCEPTED MANUSCRIPT
STATISTICAL ANALYSIS

1.14 Have confidence intervals been provided? Yes □ No □

SECTION 2: OVERALL ASSESSMENT OF THE STUDY

2.1 How well was the study done to minimise the risk of bias or confounding? High quality (++) □

Acceptable (+) □

Unacceptable – reject 0

2.2 Taking into account clinical considerations, your evaluation of the

Yes
No

methodology used, and the statistical power of the study, do you think there

PT
is clear evidence of an association between exposure and outcome?
Can’t say

2.3 Are the results of this study directly applicable to the patient group targeted Yes □ No □
in this guideline?

RI
2.4 Notes. Summarise the authors conclusions. Add any comments on your own assessment of the study, and the extent to which
it answers your question and mention any areas of uncertainty raised above.

SC
Methodology Checklist 4: Case-control studies
SIGN

U
Study identification (Include author, title, year of publication, journal title, pages)
AN
Guideline topic: Key Question No: Reviewer:

Before completing this checklist, consider:

M

1. Is the paper really a case-control study? If in doubt, check the study design algorithm available from SIGN and make sure you
have the correct checklist.

2. Is the paper relevant to key question? Analyse using PICO (Patient or Population Intervention Comparison Outcome). IF NO
REJECT (give reason below). IF YES complete the checklist.
D

Reason for rejection: Reason for rejection: 1. Paper not relevant to key question □ 2. Other reason □ (please specify):

SECTION 1: INTERNAL VALIDITY

TE

In an well conducted case control study: Does this study do it?

1.1 The study addresses an appropriate and clearly focused question. Yes
No

EP

Can’t say

SELECTION OF SUBJECTS

1.2 The cases and controls are taken from comparable populations. Yes
No

C

Can’t say

1.3 The same exclusion criteria are used for both cases and controls.
AC

Yes
No

Can’t say

1.4 What percentage of each group (cases and controls) participated in the Cases:
study?
Controls:

1.5 Comparison is made between participants and non-participants to establish Yes
No

their similarities or differences.
Can’t say

1.6 Cases are clearly defined and differentiated from controls. Yes
No

Can’t say

1.7 It is clearly established that controls are non-cases. Yes
No

Can’t say

28
 ACCEPTED MANUSCRIPT
ASSESSMENT

1.8 Measures will have been taken to prevent knowledge of primary exposure Yes
No

influencing case ascertainment.
Can’t say
Does not apply

1.9 Exposure status is measured in a standard, valid and reliable way. Yes
No

Can’t say

CONFOUNDING

PT
1.10 The main potential confounders are identified and taken into account in the Yes
No

design and analysis.
Can’t say

STATISTICAL ANALYSIS

RI
1.11 Confidence intervals are provided. Yes
No

SC
SECTION 2: OVERALL ASSESSMENT OF THE STUDY

How well was the study done to minimise the risk of bias or confounding?
2.1 High quality (++) □

U
Acceptable (+) □

Unacceptable – reject 0 □
AN
2.2 Taking into account clinical considerations, your evaluation of the
methodology used, and the statistical power of the study, do you think there Yes
No

is clear evidence of an association between exposure and outcome?
Can’t say

M

2.3 Are the results of this study directly applicable to the patient group targeted Yes
No

by this guideline?

2.4 Notes. Summarise the authors conclusions. Add any comments on your own assessment of the study, and the extent to which
it answers your question and mention any areas of uncertainty raised above..
D
TE
C EP
AC

29
 ACCEPTED MANUSCRIPT
Highlights

- This is the first systematic review focusing on RFS

- Definitions for RFS rely on electrolyte disturbances with or without clinical
symptoms
- Incidence rates for RFS highly depend on the definition used
- Most of risk factors for RFS are in accordance with the NICE guidelines
- No strong evidence for adverse outcomes and preventive measures in patients
with RFS was found

PT
RI
U SC
AN
M
D
TE
C EP
AC