- Email: [email protected]
Propofol sedation substantially increases the caloric and lipid intake in critically ill patients
Accepted Manuscript Propofol sedation substantially increases the caloric and lipid intake in critically ill patients. Mélanie Charrière, Emma Ridley, Jennifer Hastings, Oliver Bianchet, Carlos Scheinkestel, Mette M. Berger PII:
S0899-9007(17)30105-3
DOI:
10.1016/j.nut.2017.05.009
Reference:
NUT 9965
To appear in:
Nutrition
Received Date: 23 February 2017 Revised Date:
11 April 2017
Accepted Date: 15 May 2017
Please cite this article as: Charrière M, Ridley E, Hastings J, Bianchet O, Scheinkestel C, Berger MM, Propofol sedation substantially increases the caloric and lipid intake in critically ill patients., Nutrition (2017), doi: 10.1016/j.nut.2017.05.009. 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 1
Propofol sedation substantially increases the caloric and lipid
2
intake in critically ill patients.
3 4
Mélanie Charrière1*, Emma Ridley 2, 3*, Jennifer Hastings 4, Oliver Bianchet 4, Carlos
5
Scheinkestel 4, Mette M. Berger1
7
RI PT
6
*: M. Charrière AND E. Ridley should be considered both as first co-authors.
8 9
1
Service of intensive care medicine, University Hospital (CHUV), Lausanne,
Switzerland
11
2
12
University, Melbourne, Australia
13
3
Nutrition Department, The Alfred Hospital, Melbourne, Australia
14
4
Intensive Care Unit, The Alfred Hospital, Melbourne Australia
M AN U
ANZIC RC, Department of Epidemiology and Preventive Medicine, Monash
15
26 27 28 29 30 31 32 33
TE D
Prof. Mette M. Berger Service of intensive care medicine, University Hospital (CHUV) Rue du Bugnon 46 1011 Lausanne, Switzerland, E-mail: [email protected]
Coauthor mails: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
EP
24 25
Correspondence:
AC C
16 17 18 19 20 21 22 23
SC
10
Conflict of interest: none of the authors has any conflict of interest to disclose
Word count Text n= 2870 Abstract: n= 250
10APR17- 1
ACCEPTED MANUSCRIPT 34
Abstract:
35
Objective:
The amount of lipid delivered to patients varies considerably depending on the non-nutritional intake from sedation, and on the
37
feeding solution. Our study aimed at quantifying the magnitude and
38
proportion of lipids and energy provided from propofol sedation in
39
intensive care (ICU) patients.
40
Methods
RI PT
36
Retrospective analysis of prospectively collected data in consecutive patients admitted in the ICUs of 2 university hospitals. Inclusion
42
criterion: ICU stay >5 days. Data were collected for maximum 10 days.
43
Propofol sedation using 1% or 2% propofol solutions was defined as
44
>100 mg/d. Nutritional management was per protocol in both centres,
45
recommending enteral feeding. Data as means±SD. Results
M AN U
46
SC
41
Altogether 701 admissions (687 patients aged 59±16 years, SAPS2 51±17) and 6485 days including 3484 propofol sedation days were
48
analysed. Energy targets were 1987±411 kcal/day; mean energy
49
delivery was 1362±811 kcal/day (70±38% of prescription) including
50
propofol and dextrose. Enteral feeding dominated (75% of days) and
51
progressed similarly in both ICUs. Mean propofol sedation dose was
52
2045±1650 mg/day, resulting in 146±117 kcal/day. Fat from propofol
53
constituted 17% of total energy (up to 100% during the first days). Fat
54
delivery (40±23 g/day: maximum 310 g/day) was significantly
55
increased by the combination of propofol sedation, the 2% solution,
56
and high fat containing feeds. In survivors, high fat proportion was
57
associated with prolonged ventilation time (p<0.0001)
AC C
EP
TE D
47
58
Conclusion Propofol sedation resulted in large doses of lipids being delivered to
59
patients, some receiving pure lipids during the first days. As the
60 61 62 63 64 65
metabolic impact of high proportions of fat are unknown, further research is prompted.
Key words: critical care, nutrition, energy delivery, fat, nutrient composition, sedation
10APR17- 2
ACCEPTED MANUSCRIPT Introduction
67
Over the past 3 decades, sedation has considerably evolved in critical care with the
68
appearance of short acting agents such as propofol. Conversely, there has been
69
little evolution in commercial enteral nutrition (EN) solutions in regards to nutrient
70
composition. The focus of such EN solutions is mainly on total energy, glucose in
71
the context of glucose control, and recently on protein intakes, but with little concern
72
regarding lipid 1. In the critically ill, the commercial enteral nutrition (EN) and
73
parenteral nutrition (PN) solutions may deliver up to 55% of total energy as lipids.
74
This contrast with recommendations in healthy individuals to avoid fat intakes above
75
35% of total energy, aiming at reducing cardiovascular risk 2: however, such
76
recommendations do not exist for critically ill patients. There are only general
77
guidelines that recommend that fat should not exceed 1 to 1.5 g/kg/day 3, while data
78
supporting this range are few. The knowledge about the effects of higher amounts
79
of lipid over prolonged periods come from studies investigating the impact of
80
omega-3 fatty acids on intensive care unit (ICU) outcome: the trials used a high
81
dose (55% of energy) long chain triglyceride (LCT) solutions as comparative, and
82
showed poorer outcomes with the higher lipid containing feeds 4. Further, studies in
83
critically ill burn patients, show that feeds with fat content reduced to 15% of total
84
energy result in less infectious complications compared to feeds with 30% fat 5.
85
There has been increasing awareness regarding additional energy provision from
86
non-nutritional sources, such as dextrose from drug dilution and hydration, citrate
87
from dialysate solutions, and lipid from the sedative agent propofol
88
non-nutrition energy sources are added to artificial nutrition delivery overfeeding
89
may result. Propofol comes as lipid emulsion either composed of pure LCT soya
90
lipid emulsion or a balanced mixture of medium chain triglycerides (MCT) and LCT.
91
It is generally solubilized as a 1% or 2% solution: its use, particularly in high doses,
92
increases the overall proportion of fat provided to the patient 8, 9.
93
Alterations of lipid metabolism are frequent during critically illness
94
hypertriglyceridemia is observed in nearly 45% of patients who require more than 3
95
days of ICU treatment. Hypertriglyceridemia is considered to reflect a global liver
96
dysfunction that has several causes, including overfeeding. Propofol and its
97
accompanying lipid emulsion are the strongest risk factors for hypertriglyceridemia,
98
stronger than total energy, glucose and lipids intakes from feeding 8. Moreover,
6-8
. When these
AC C
EP
TE D
M AN U
SC
RI PT
66
10
, and
10APR17- 3
ACCEPTED MANUSCRIPT 99
hypertriglyceridemia, may also occur during the rare but life-threatening propofol
100
infusion syndrome 11.
101
The present study aimed to quantify the quantity of lipid and the proportion of both
102
energy and lipids that resulted from propofol sedation in critically ill patients
103
requiring artificial nutrition in two distinct ICU settings.
RI PT
104 105 Methods:
107
The study was designed as a bi-centric retrospective analysis of prospectively
108
collected data of consecutive patients admitted in the ICUs of two teaching hospitals
109
(Alfred Hospital [=AH], Melbourne, Australia and the Centre Hospitalier Universitaire
110
Vaudois [=CHUV], Lausanne, Switzerland). Inclusion criterion was a ventilation time
111
>5 days and an ICU stay longer than 5 days, with no exclusion criteria. The patients
112
were identified through hospital databases. The study was conducted from August
113
2011 to March 2012 at CHUV and from January 2012 to December 2012 in AH.
114
Data was collected from ICU admission until ICU stay day 10 or ICU discharge
115
(whichever occurred first). Data obtained included demographic and admission data,
116
nutrition assessment information (including energy and protein targets) and daily
117
nutrition therapy data including the mode of nutrition, delivery of energy, and lipid
118
amounts from propofol and feeds. Length of mechanical ventilation, of ICU stay and
119
outcome were recorded, but not infectious complications.
120
The protocol was reviewed by the Commission Cantonale d’Ethique pour la
121
Recherche sur l’être humain (CHUV) and by Human Research Ethics Committee at
122
The Alfred Hospital (AH). In both institutions a low risk ethics approval was obtained
123
and the requirement for consent was waived due to the absence of intervention and
124
low risk nature of the project.
M AN U
TE D
EP
AC C
125
SC
106
126
Propofol sedation days were defined as any 24-hour period with ≥ 100mg of
127
propofol any day with less being aggregated with the “non-propofol days”. The
128
centers used different formulations of propofol: a 1% solution at AH (Fresofol 1%,
129
Fresenius Kabi, Australia: 1.1 kcal/ml, 100mg propofol deliver of 11 kcal as LCT)
130
and a 2% solution at CHUV (Propofol Lipuro, BBraun, Crissier, Switzerland: 0.5
10APR17- 4
ACCEPTED MANUSCRIPT 131
kcal/ml, 100mg propofol deliver 5 kcal). This difference meant that for the same
132
dose of propofol, AH patients received twice the amount of fat.
133 134
Nutrition therapy was as per evidence based nutrition guidelines in both ICUs and
135
EN is systematically favoured. Energy targets were set differently: at AH the
136
Schofield equation with added stress factors was used
137
a weight based target (25kcal/kg/day), or in patients >70 years the Harris-Benedict
138
predicted value times 1.2. Both centers use indirect calorimetry on occasions in
139
patients requiring specific nutrition therapy (burns, trauma, transplantation, obesity,
140
malnutrition). The choice of the EN formula was as per standard practice at each
141
site: the composition of the 3 most frequently used EN solutions per center is
142
depicted in Table 1. At CHUV blood triglycerides are determined twice weekly as
143
part of routine care, but not at AH.
144
For each patient, the mean fat intake and proportion of energy intake from fat was
145
calculated. High fat delivery was defined to occur when more than 45% of total
146
energy was supplied as fat, and excessive fat delivery when >55% of total energy
147
was supplied as fat.
12
M AN U
SC
RI PT
, while CHUV used mainly
TE D
148 Statistical analysis
150
Variables are reported as numbers or percent; normally distributed variables are
151
reported as mean± standard deviation and non-normally distributed variables are
152
reported as median [inter-quartile range (IQR)]. Comparisons between sites were
153
carried out using Chi2 tests for discrete variables, and 2-way ANOVAs for
154
continuous variables repeated over time. Single regressions were calculated
155
between fat doses and outcome variables. Significance was considered at the level
156
of p<0.05. Statistical package: JMP® Version 10.0, SAS Institute Inc. Cary, NC,
157
USA.
AC C
158
EP
149
159
Results
160
Altogether 701 admissions of 687 patients met the inclusion criteria, resulting in
161
6485 study days. The demographics of the population is reported in Table 2. The
162
mean age was 59 ±16 years and Body Mass Index (BMI) was 27.2 ±6.9 kg/m2:
163
patients were significantly younger and heavier at AH: the 2 largest diagnostic
10APR17- 5
ACCEPTED MANUSCRIPT 164
categories were cardiovascular (19% and 23.6% respectively per site) and trauma &
165
musculoskeletal pathologies (higher at AH with 45%, versus 2.5%).
166 Propofol sedation: Of the 6485 study days, 3484 (53.7%) were with propofol
168
sedation (1623 and 1861 propofol days from CHUV and AH, respectively). Overall,
169
2045 ±1650 mg/d propofol were provided during the 10 study days, corresponding
170
to 85 mg/hour propofol with a large inter-ICU and inter-patient variability (Figure 1):
171
the median propofol dose was 1290 mg/day at AH and 2400 mg/day at CHUV
172
(p<0.0001). Propofol was most intensively used during the first 3 days: the
173
proportion of overall energy provided by propofol sedation is shown in Figure 2. As
174
the result of the 1% or 2% propofol emulsion, despite significantly higher propofol
175
dose at CHUV, the median amount of energy resulting from the sedation was lower
176
at CHUV (at AH 136 kcal/day (IQR 61-253) versus 108 kcal/day (IQR 40-178) at
177
CHUV: p<0.001). Energy from propofol was the unique source of kcal in several
178
patients for a few days. Overall, propofol sedation contributed 17 % as a mean of
179
total energy delivery per day over the study period.
M AN U
SC
RI PT
167
180
Nutrition therapy: Overall 75% of days were on EN: 17.4% of days were without
182
nutrition for multiple reasons. Indirect calorimetry was used to adjust the energy
183
target in 79 patients (21 AH, 51 CHUV) and repeated in some patients (total: 109
184
studies). The mean energy target for the population was 1987±411 kcal/day. The
185
energy target was higher at AH (p=0.0001) as result of the use of the Schofield
186
equation and a younger cohort. The mean daily amount of energy delivered from all
187
sources at both sites was 1362±811 kcal corresponding to in 81% of prescription.
188
Other nutrition information is shown in Table 3.
EP
AC C
189
TE D
181
190
Fat Delivery: Median fat delivery was 39 g/day (IQR 31-56) increasing to 43 g/day
191
(IQR 35-59) on propofol days. Maximal daily fat intake was 310 g/day. Total lipid
192
delivery by day is shown in Figure 3. Expressed per kg body weight/day, the
193
cumulated nutrition-propofol intake represents a median of 0.53 g/kg/day with (IQR
194
0.27-0.76) with maximal values in both centres of 2.11 and 2.31 g/kg. This
195
represents 31% (IQR 20-40) of total energy. Considering all days, fat delivery was
196
similar in both sites, but in absence of propofol sedation, the mean fat delivery was 10APR17- 6
ACCEPTED MANUSCRIPT higher at AH (37±17 g/d) compared to CHUV (28±27 g/d: p<0.0001) from the higher
198
fat-containing enteral feeds.
199
The predefined high fat doses (>45% of energy) were observed in 888 days
200
(13.7%) (Figure 2): 705 days at AH (19.6% of AH days), and 183 at CHUV (10.4%
201
of CHUV days: p<0.0001). Fat doses exceeding 55% of energy intake were found in
202
678 days (10.5%): 585 days at AH and 93 at CHUV (p<0.0001). Fat represented
203
100% of energy intake during 358 days (5.5%), constituting a highly unbalanced
204
substrate composition.
205
Mean blood triglyceride value was 1.7±1.0 mmol/l at CHUV (544 values) versus
206
2.5±2.2 mmol/l at AH (p= 0.023: only 9 values).
SC
RI PT
197
207
No association was observed between the propofol dose and mortality. When the
209
relationship between fat intake (mean and maximal in g/day, fat as % of total energy
210
intake, or maximal fat %) were examined, we found no relevant association with
211
either length of mechanical ventilation, or ICU stay or mortality (R2: 0.02 to 0.06 with
212
p<0.001). On the other hand, in survivors (n=547) we observed an association
213
between both Mean fat %/day, Maximal fat %/day and length of mechanical
214
ventilation (R2=0.108 and R2=0.081 respectively: p<0.0001).
215 216
TE D
M AN U
208
Discussion
218
This bi-centric study shows that propofol sedation may result in the delivery of large
219
amounts of lipids, especially during the first days of ICU treatment. While the
220
median fat dose per day was not excessive with 39 g/day (i.e. about 0.5 g/kg) an
221
important proportion of the patients received much more (up to 310 g/day). This
222
proportion was highest with the combination of the 1% propofol solution and high fat
223
containing enteral feeds. Fat from propofol sedation accounted for 17% of daily
224
energy as a mean and added about 10% to the “nutritional” energy.
225
Individual data show a substantial number of days (13.7%) with a fat proportion
226
exceeding 45% of total energy delivery, and even exceeding 55% in 10.5% of days.
227
There are no data to our knowledge that report on the clinical consequences of high
228
fat intakes resulting from propofol sedation since an abstract published in 1995 9:
229
most studies have concentrated on propofol’s impact on plasma triglycerides 8, or
AC C
EP
217
10APR17- 7
ACCEPTED MANUSCRIPT 11
on the propofol infusion syndrome
. Here we report a weak but significant
231
association of high fat proportions with prolonged ventilation time, which should be
232
further explored in prospective studies.
233
The impact of high fat delivery on organ function in critical illness is largely unknown.
234
Data on high fat diets come from studies investigating the impact of the ketogenic
235
diet on brain metabolism in refractory epilepsy, and from tracer metabolic studies.
236
Ketone bodies are an important source of energy for the brain
237
are high fat (but not 100% fat), low carbohydrate and normo-protein diets, used for
238
prolonged periods in epileptic children with apparently no harm
239
frequently in adults. The main side effect is acidosis which is known to stimulate
240
protein catabolism. In 2001 Hart et al 15 showed that patients on high fat diets (44%
241
fat with 42% carbohydrates) remained more catabolic compared to patients on high
242
carbohydrates feeds (82% with 3% fat) with similar 15% amount of proteins. If the
243
high fat delivery is combined with an insufficient protein delivery as is frequently the
244
case in ICUs
245
patients. Other studies in critical care show that whatever the route of feeding, an
246
increase of either fat or carbohydrates above the “normal” proportions of 30% lipids
247
and 55% carbohydrates stimulates hepatic de novo lipogenenesis
248
development of fatty liver.
249
patients testing the impact of a 5 day isocaloric isonitrogenous PN containing 75%
250
(PN-glucose) or 15% (PN-lipid) glucose 18.
251
Sedation is a recurrent worry in the ICU, as it has deleterious effects of its own
252
The present quality control study shows that there is room for improvement of
253
continuous sedation in several patients: delivering close to 9000 mg propofol in
254
24hours is not desirable
255
purpose, being by definition rapidly reversible, but the pharmacological properties
256
change during continuous prolonged sedation and reveal different side effects
257
Even the volatile agent sevoflurane has recently been shown to induce nephrogenic
258
diabetes insipidus
259
propofol has few side effects (except for the dangerous PRIS
260
solubilising lipid emulsion which is part of PN has been widely used, it is perceived
261
as a safe alternative: but its prolonged use out of the PN context has not been given
262
much attention until now.
RI PT
230
13
. Ketogenic diets 14
M AN U
SC
, and less
16
, this may aggravate the lean body mass loss observed in critically ill
and the
TE D
17, 18,
17
AC C
EP
. These data confirmed the results of a study in 16
19
.
19
. Anaesthetic agents appear suitable for sedation 11
.
20
. Other agents, such as dexmedetomidine, are expensive. As 11
), and as the
10APR17- 8
ACCEPTED MANUSCRIPT 263 264
Underfeeding associated with difficult EN is the most frequent observation in the
265
ICU
266
delivery is potentially beneficial for the energy balance, but it increases substrate
267
imbalance: this requires further investigations considering that high fat diets have
268
been shown to favor catabolism
269
than 110% of measured energy expenditure
270
propofol sedation adds 10% energy as a mean, while energy targets are frequently
271
estimated and hence inexact, there is a real risk of overfeeding with this agent
272
particularly when PN is required. The last decade has seen the publication of
273
studies of PN or combined EN and PN with negative outcome results
274
being possibly related to overfeeding as energy targets are frequently equation
275
based which has been shown to inexact in up to 70% of patents
276
hand there are positive outcome studies using either indirect calorimetry based
277
targets
278
deleterious during the early phase of acute disease, via several mechanisms
279
Of note the number of calorimetry studies was also insufficient in this cohort. As the
280
energy contribution from propofol becomes acknowledged with the help of
281
computerized systems
282
recognised, clinicians may want to adjust provision of nutrition therapy to include
283
these non-nutritional calories. If the energy from propofol is integrated in the a priori
284
calculation of the feed target, the consequence may be a reduction in the nutrient
285
delivery, at the expense of proteins
16, 21
: delivering large amounts of fat with sedation, while increasing energy
15
22
RI PT
. Overfeeding defined as the delivery of more
or low targets
25
23
, the latter
6, 24
). On the other
. Exceeding the energy needs may be particularly 26-28
.
TE D
24
M AN U
SC
, is less frequent. As the use of
7
EP
, and the deleterious consequences of overfeeding
29
if their delivery is not monitored, which may
30
adversely affect outcomes
. Monitoring should be implemented including real daily
287
deliveries and triglyceride determination once or twice weekly.
288
Significant energy delivery and imbalance of substrate composition (i.e. >55% of
289
energy as fat) was observed in 10.5% of days and occurred for >2 days in 19.8% of
290
patients. The clinical relevance of these results is still uncertain, but our data
291
suggest an association of high fat proportions with prolonged mechanical ventilation.
292
The provision of large amounts of fat to individual patients on individual days did
293
exceed quantities recommended for cardiovascular prevention (i.e. 35% of total
294
energy 2). Careful monitoring of the overall energy provision including all sources to
295
minimize the risk of substrate imbalance and/or overfeeding should be encouraged.
AC C
286
10APR17- 9
ACCEPTED MANUSCRIPT If nutrition therapy is adjusted to account for non-nutritional energy clinicians should
297
be mindful to monitor protein delivery to maintain a sufficient intake.
298
There are limitations to the present quality control study. The ICU populations in the
299
2 centers differed as did the sedation protocol and the choice of enteral feeds, but
300
this has the advantage of showing the widespread use of propofol across ages,
301
conditions and countries. A cohort study has the advantage of reflecting real life.
302
Both ICUs have guidelines-based feeding protocols, and achieve an elevated feed
303
delivery (80% as median) showing the relevance of the findings in very different
304
settings. Some differences in practice attenuated the strength of our observations,
305
such as the absence of a systematic monitoring of plasma triglycerides at AH: the
306
higher values observed at AH suggest that this monitoring might be useful. And
307
finally as our retrospective study was not designed to test the impact of high fat
308
delivery on clinical outcome (length of mechanical ventilation and stay, infections)
309
this issue should be verified prospectively.
310
Conclusion
311
Propofol sedation may result in a significant fat delivery that is frequently ignored.
312
Whenever propofol is used, it significantly adds to the amount of energy and
313
proportion of fat. In the early phase of ICU stay, this sedation regimen may result in
314
an imbalance in the substrate provision, dominated by fat. The more concentrated
315
2% propofol solution seems to reduce efficiently the risk of fat overload. The
316
consequences of
317
consequences on substrate metabolism, clinical outcomes and possibly on long-
318
term recovery.
SC
M AN U
TE D
EP
the latter require prospective investigation of potential
AC C
319
RI PT
296
10APR17- 10
ACCEPTED MANUSCRIPT Abbreviations
321
AH
Alfred Hospital
322
CHUV
Centre Hospitalier Universitaire Vaudois
323
BMI
Body Mass Index
324
EN
Enteral nutrition
325
PN
Parenteral Nutrition
326
ICU
Intensive Care unit
327
IQR
Interquartile range
328
LCT
Long chain triglycerides
330
M AN U
329
SC
RI PT
320
Ethical Approval and Consent to participate: study was approved by both institutions ethics committees: Commission Cantonale d’Ethique pour la Recherche
332
sur l’être humain (CHUV) and by Human Research Ethics Committee at The Alfred
333
Hospital (AH). In both institutions a low risk ethics approval was obtained and the
334
requirement for consent was waived due to the absence of intervention and low
335
risk nature of the project.
TE D
331
Consent for publication: all authors approved the final version
337
Funding: None. The present study was supported by internal resources of the 2
338
services
EP
336
Authors contribution: MMB, ML, and ER designed the study; All authors
340
participated to the acquisition of the data and interpretation of the results; MC, ER,
341
CS and MBB drafted and revised the manuscript which was approved by all the
342
authors,
343
Acknowledgments: The authors would like to thank Bianca Levkovich, and Owen
344
Roodenburg residents from the Alfred Hospital’s Intensive Care Unit, Melbourne, for
345
assistance in data collection
AC C
339
346
10APR17- 11
ACCEPTED MANUSCRIPT 347
Table 1: Enteral feed composition of the 3 principal solutions used in the respective ICUs.
348 349
CHUV
Nutrison
Nutrison Nutren
Isosource Isosource Promote
Protein-Plus
Energy
Pulmonary
Protein
Energie
Fiber+®
Multifibre®
®
®
Fibre®
®
Abbott
Nutricia
Nutricia
Nestlé
Nestlé
Nestlé
1.33
1.57
1.3
30
35
30
Energy density
1.25
1.5
1.5
35
35
56
% Energy from
% Energy from
45
carbohydrate % Energy from
20
protein
49
26
48
49
43
16
18
20
16
25
AC C
EP
TE D
350
M AN U
fat
SC
kcal/ml
RI PT
Feed
Alfred Hospital
10APR17- 12
ACCEPTED MANUSCRIPT Table 2: Patient characteristics AH
CHUV
Patients
687
373
314
(admissions)
(701)
(373)
(328)
Men (%)
66.8
68.9
64.3
Age (years)
56±17
53±18
SAPSII
50±16
50±14
BMI (kg/m2)
27.2±6.9
27.8±7.3
Weight (kg)
80±21
82±24
Length of Ventilation (d)
10.9±10.5
LICU(d) ICU mortality (%)
p
ns
RI PT
All patients
<.0001
51±18
ns
26.4±5.9
0.0069
77±17
0.0009
12.9±10.8
8.7±9.8
<0.001
15.5±12.4
17±13
13 ±12
<.0001
86 (12.5%)
48 (12.9)
38 (12.1)
ns
22.9
18.2
ns
Data as mean ±SD
M AN U
Total Hospital mortality (%) 20.4
SC
61±15
AC C
EP
TE D
*: includes any internal medicine, neurology and infectious diseases
10APR17- 13
ACCEPTED MANUSCRIPT Table 3: Detail of energy, fat and propofol delivery by site AH
CHUV
p
6485
3607
2878
-
Energy target (kcal)
1957 ±461
2172±384
1743±303
<.0001
Energy delivery (kcal/d)
1362 ±811
1488 ±865
1202 ±707
<.0001
% of prescription Lipid delivery (total g/d) Total g/kg/day % Fat in total daily energy
69.7 ±37.9 [80.7] 40±23
41 ±16
39 ±29
0.0203
0.53±0.30
0.55±0.25
0.52±0.41
0.0127
29.5%
31.4 ± 28.3
27.1 ± 16.1
<.0001
35.8±29.5
35.3±13.5
ns ns
Propofol sedation (mg/d) Energy from propofol sedation (kcal/d)
3214
1778 (49.3%)
1436 (49.9%)
2045±1651
1574 ±1196
2627 ±1927
<.0001
146±117 [119]
169 ±132 [136]
118 ±87 [108]
<.0001
16.6 ±21.4
18.5±24.1
14.2±17.2
[11]
[9]
TE D
% Energy propofol sedation
M AN U
- Propofol sedation days Days on propofol >100 mg
68.6±37.5 [80.2] 70.9±38.1 [81.5] <0.016
SC
Days
RI PT
All
[10]
AC C
EP
Data as mean ±SD, [median]
10APR17- 14
<.0001
ACCEPTED MANUSCRIPT
Legends to the figures: Fig 1 Propofol and fat dose by day during the first 10 days in both institutions
RI PT
(horizontal dotted lines show the medians) during the propofol sedation days (i.e. >100 mg/d): this figure shows the impact of the sedative concentration on the total fat delivery.
SC
Fig 2 Proportion of energy delivery by day from propofol sedation (i.e. >100 mg/d) in both sites with horizontal lines showing the 35%, 45% and 55% of total
M AN U
energy (Max = maximal value, 90% CI = upper 90% confidence interval) Fig 3 Total lipid delivery from feeding solutions plus propofol by day in all patients (box plots with interquartile ranges; doted horizontal line = mean value).
AC C
EP
TE D
Two way Anova: Effect of time p<0.0001, time*site p<0.0001
10APR17- 15
ACCEPTED MANUSCRIPT References Weijs P, Cynober L, DeLegge M, Kreymann G, Wernerman J, Wolfe RR. Proteins and amino acids are fundamental to optimal nutrition support in critically ill patients. Crit care 2014; 18:591.
2.
Krauss RM, Eckel RH, Howard B, Appel LJ, Daniels SR, Deckelbaum RJ et al. AHA Dietary Guidelines: revision 2000: A statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation 2000; 102:2284-99.
3.
Singer P, Berger MM, Van den Berghe G, Biolo G, Calder P, Forbes A et al. ESPEN Guidelines on Parenteral Nutrition: Intensive care. Clin nutr 2009; 28:387-400.
4.
Singer P, Shapiro H. Enteral omega-3 in acute respiratory distress syndrome. Curr opin clin nutri metab 2009; 12:123-28.
5.
Garrel D, Patenaude J, Nedelec B, Samson L, Dorais J, Champoux J et al. Decreased mortality and infectious morbidity in adult burn patients given enteral glutamine supplements: a prospective, controlled, randomized clinical trial. Crit care med 2003; 31:2444-9.
6.
Singer P, Anbar R, Cohen J, Shapiro H, Shalita-Chesner M, Lev S et al. The tight calorie control study (TICACOS): a prospective, randomized, controlled pilot study of nutritional support in critically ill patients. Intensive care med 2011; 37:601-09.
7.
Berger MM, Revelly JP, Wasserfallen JB, Schmid A, Bouvry S, Cayeux MC et al. Impact of a computerized information system on quality of nutritional support in the ICU. Nutrition 2006; 22:221-29.
8.
Devaud JC, Berger MM, Pannatier A, Marques-Vidal P, Tappy L, Rodondi N et al. Hypertriglyceridemia: a potential side effect of propofol sedation in critical illness. Intensive care med 2012; 38:1990-98.
9.
DeChicco R, Matarese L, Hummell AC, Speerhas R, Seidner D, Steiger E. Contribution of calories from propofol to total energy intake. J Am Diet Assoc 1995; 95:A25.
10.
Marik PE. Dyslipidemia in the critically ill. Crit care clin 2006; 22:151-9, viii.
11.
Diaz JH, Prabhakar A, Urman RD, Kaye AD. Propofol infusion syndrome: a retrospective analysis at a level 1 trauma center. Crit care res pract 2014; 2014:346968.
12.
Subramaniam A, McPhee M, Nagappan R. Predicting energy expenditure in sepsis: HarrisBenedict and Schofield equations versus the Weir derivation. Crit Care Resusc 2012; 14:202-10.
AC C
EP
TE D
M AN U
SC
RI PT
1.
13.
Morris AAM. Cerebral ketone body metabolism. J Inherit Metab Dis. 2005; 28:109-21.
14.
Thakur KT, Probasco JC, Hocker SE, Roehl K, Henry B, Kossoff EH et al. Ketogenic diet for adults in super-refractory status epilepticus. Neurology 2014; 82:665-70.
15.
Hart DW, Wolf SE, Zhang XJ, Chinkes DL, Buffalo MC, Matin SI et al. Efficacy of a highcarbohydrate diet in catabolic illness. Crit care med 2001; 29:1318-24.
16.
Alberda C, Gramlich L, Jones N, Jeejeebhoy K, Day AG, Dhaliwal R et al. The relationship between nutritional intake and clinical outcomes in critically ill patients: results of an international multicenter observational study. Intensive care med 2009; 35:1728-37.
17.
Schwarz JM, Chioléro R, Revelly JP, Cayeux C, Schneiter P, Jéquier E et al. Effects of enteral carbohydrates on de novo lipogenesis in critically ill patients. Am J Clin Nutr 2000; 72:940-45. 10APR17- 16
ACCEPTED MANUSCRIPT Tappy L, Schwarz JM, Schneiter P, Cayeux C, Revelly JP, Fagerquist CK et al. Effects of isoenergetic glucose-based or lipid-based parenteral nutrition on glucose metabolism, de novo lipogenesis, and respiratory gas exchanges in critically ill patients. Crit care med 1998; 26:860-67.
19.
Taskforce D. A. S., Baron R, Binder A, Biniek R, Braune S, Buerkle H et al. Evidence and consensus based guideline for the management of delirium, analgesia, and sedation in intensive care medicine. Revision 2015 (DAS-Guideline 2015) - short version. Ger Med Sci 2015; 13:Doc19.
20.
Muyldermans M, Jennes S, Morrison S, Soete O, Francois PM, Keersebilck E et al. Partial Nephrogenic Diabetes Insipidus in a burned patient receiving Sevoflurane sedation with an anesthetic conserving device-A case report. Crit care med 2016; 44:e1246-e50.
21.
Bendavid I, Singer P, Theilla M, Themessl-Huber M, Sulz I, Mouhieddine M et al. NutritionDay ICU: A 7 year worldwide prevalence study of nutrition practice in intensive care. Clin nutr 2016
22.
McClave SA, Lowen CC, Kleber MJ, Nicholson JF, Jimmerson SC, McConnel JW et al. Are patients fed appropriately according to their caloric requirements? JPEN 1998; 22:375-81.
23.
Dissanaike S, Shelton M, Warner K, O'Keefe GE. The risk for bloodstream infections is associated with increased parenteral caloric intake in patients receiving parenteral nutrition. Crit care 2007; 11:R114.
24.
Heidegger CP, Berger MM, Graf S, Zingg W, Darmon P, Costanza MC et al. Optimisation of energy provision with supplemental parenteral nutrition in critically ill patients: a randomised controlled clinical trial. Lancet 2013; 381:385-93.
25.
Doig GS, Simpson F, Sweetman EA, Finfer SR, Cooper DJ, Heighes PT et al. Early parenteral nutrition in critically ill patients with short-term relative contraindications to early enteral nutrition: a randomized controlled trial. JAMA 2013; 309:2130-38.
26.
Grau T, Bonet A, Rubio M, Mateo D, Farre M, Acosta JA et al. Liver dysfunction associated with artificial nutrition in critically ill patients. Crit care 2007; 11:R10.
27.
Berger MM. The 2013 Arvid Wretlind lecture: Evolving concepts in parenteral nutrition. Clin nutr 2014; 33:563-70.
28.
Braunschweig CA, Sheean PM, Peterson SJ, Gomez Perez S, Freels S, Lateef O et al. Intensive Nutrition in Acute Lung Injury: A Clinical Trial (INTACT). JPEN 2015; 39:13-20.
29.
Berger MM, Soguel L, Charrière M, Theriault B, Pralong F, Schaller MD. Impact of the reduction of the recommended energy target in the ICU on protein delivery and clinical outcomes. Clin Nutr 2017; 36
30.
Weijs PJM, Stapel SN, de Groot SDW, Driessen RH, de Jong E, Girbes ARJ et al. Optimal protein and energy nutrition decreases mortality in mechanically ventilated, critically ill patients: A prospective observational cohort study. JPEN 2012; 36:60-68
AC C
EP
TE D
M AN U
SC
RI PT
18.
10APR17- 17
ACCEPTED MANUSCRIPT
200
p<0.001
RI PT
150
SC
100 50
M AN U
Lipids total (g/day)
Fig 1
0
TE D
p<0.001
EP
9000 8000 7000 6000 5000 4000 3000 2000 1000
AC C
Propofol (mg/day)
0
1 2 3 4
5 6
7 8 9 10
1 2 3 4 5 6
Alfred Hospital
CHUV Days in ICU
7 8 9 10
ACCEPTED MANUSCRIPT
RI PT
100 90
SC
80 70
M AN U
60 50 40
TE D
30 20 10 0
Max % 100 90% CI 100 Median% 36.7
2
3
AC C
1
EP
% of total energy as propofol
Fig.2
100 54.5 14.2
100 34.1 9.4
4
5 6 Days in ICU 100 100 100 20.3 22.5 21.2 7.6 8.4 8.0
7
8
9
10
70.1 100 59.8 89.0 18.9 18.3 20.3 20.1 8.2 7.3 8.6 7.9
ACCEPTED MANUSCRIPT
300
SC
p<0.001
M AN U
200 150
TE D
100
EP
50 0 1
2
AC C
Lipids total (g/day)
RI PT
Fig 3
3 4
5
6
7
8
9 10
1
2
3 4
Alfred Hospital
5
6
CHUV Days of ICU stay
7
8
9 10
ACCEPTED MANUSCRIPT
Propofol sedation substantially increases the caloric and lipid intake in critically ill patients NUT-D-17-00105R1
RI PT
Highlights
Continuous propofol sedation results in a significant fat and energy delivery
•
A concentrated 2% solution reduces the unwanted intake of fat
•
Monitoring real daily intakes of fat enables adjusting nutrition therapy and
•
Triglyceride monitoring once to twice weekly assists the detection of impending fat overload
Propofol sedation over several days may have unwanted metabolic
EP
TE D
consequences on catabolism
AC C
•
M AN U
prevents unbalanced diet.
SC
•
S0899-9007(17)30105-3
DOI:
10.1016/j.nut.2017.05.009
Reference:
NUT 9965
To appear in:
Nutrition
Received Date: 23 February 2017 Revised Date:
11 April 2017
Accepted Date: 15 May 2017
Please cite this article as: Charrière M, Ridley E, Hastings J, Bianchet O, Scheinkestel C, Berger MM, Propofol sedation substantially increases the caloric and lipid intake in critically ill patients., Nutrition (2017), doi: 10.1016/j.nut.2017.05.009. 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 1
Propofol sedation substantially increases the caloric and lipid
2
intake in critically ill patients.
3 4
Mélanie Charrière1*, Emma Ridley 2, 3*, Jennifer Hastings 4, Oliver Bianchet 4, Carlos
5
Scheinkestel 4, Mette M. Berger1
7
RI PT
6
*: M. Charrière AND E. Ridley should be considered both as first co-authors.
8 9
1
Service of intensive care medicine, University Hospital (CHUV), Lausanne,
Switzerland
11
2
12
University, Melbourne, Australia
13
3
Nutrition Department, The Alfred Hospital, Melbourne, Australia
14
4
Intensive Care Unit, The Alfred Hospital, Melbourne Australia
M AN U
ANZIC RC, Department of Epidemiology and Preventive Medicine, Monash
15
26 27 28 29 30 31 32 33
TE D
Prof. Mette M. Berger Service of intensive care medicine, University Hospital (CHUV) Rue du Bugnon 46 1011 Lausanne, Switzerland, E-mail: [email protected]
Coauthor mails: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
EP
24 25
Correspondence:
AC C
16 17 18 19 20 21 22 23
SC
10
Conflict of interest: none of the authors has any conflict of interest to disclose
Word count Text n= 2870 Abstract: n= 250
10APR17- 1
ACCEPTED MANUSCRIPT 34
Abstract:
35
Objective:
The amount of lipid delivered to patients varies considerably depending on the non-nutritional intake from sedation, and on the
37
feeding solution. Our study aimed at quantifying the magnitude and
38
proportion of lipids and energy provided from propofol sedation in
39
intensive care (ICU) patients.
40
Methods
RI PT
36
Retrospective analysis of prospectively collected data in consecutive patients admitted in the ICUs of 2 university hospitals. Inclusion
42
criterion: ICU stay >5 days. Data were collected for maximum 10 days.
43
Propofol sedation using 1% or 2% propofol solutions was defined as
44
>100 mg/d. Nutritional management was per protocol in both centres,
45
recommending enteral feeding. Data as means±SD. Results
M AN U
46
SC
41
Altogether 701 admissions (687 patients aged 59±16 years, SAPS2 51±17) and 6485 days including 3484 propofol sedation days were
48
analysed. Energy targets were 1987±411 kcal/day; mean energy
49
delivery was 1362±811 kcal/day (70±38% of prescription) including
50
propofol and dextrose. Enteral feeding dominated (75% of days) and
51
progressed similarly in both ICUs. Mean propofol sedation dose was
52
2045±1650 mg/day, resulting in 146±117 kcal/day. Fat from propofol
53
constituted 17% of total energy (up to 100% during the first days). Fat
54
delivery (40±23 g/day: maximum 310 g/day) was significantly
55
increased by the combination of propofol sedation, the 2% solution,
56
and high fat containing feeds. In survivors, high fat proportion was
57
associated with prolonged ventilation time (p<0.0001)
AC C
EP
TE D
47
58
Conclusion Propofol sedation resulted in large doses of lipids being delivered to
59
patients, some receiving pure lipids during the first days. As the
60 61 62 63 64 65
metabolic impact of high proportions of fat are unknown, further research is prompted.
Key words: critical care, nutrition, energy delivery, fat, nutrient composition, sedation
10APR17- 2
ACCEPTED MANUSCRIPT Introduction
67
Over the past 3 decades, sedation has considerably evolved in critical care with the
68
appearance of short acting agents such as propofol. Conversely, there has been
69
little evolution in commercial enteral nutrition (EN) solutions in regards to nutrient
70
composition. The focus of such EN solutions is mainly on total energy, glucose in
71
the context of glucose control, and recently on protein intakes, but with little concern
72
regarding lipid 1. In the critically ill, the commercial enteral nutrition (EN) and
73
parenteral nutrition (PN) solutions may deliver up to 55% of total energy as lipids.
74
This contrast with recommendations in healthy individuals to avoid fat intakes above
75
35% of total energy, aiming at reducing cardiovascular risk 2: however, such
76
recommendations do not exist for critically ill patients. There are only general
77
guidelines that recommend that fat should not exceed 1 to 1.5 g/kg/day 3, while data
78
supporting this range are few. The knowledge about the effects of higher amounts
79
of lipid over prolonged periods come from studies investigating the impact of
80
omega-3 fatty acids on intensive care unit (ICU) outcome: the trials used a high
81
dose (55% of energy) long chain triglyceride (LCT) solutions as comparative, and
82
showed poorer outcomes with the higher lipid containing feeds 4. Further, studies in
83
critically ill burn patients, show that feeds with fat content reduced to 15% of total
84
energy result in less infectious complications compared to feeds with 30% fat 5.
85
There has been increasing awareness regarding additional energy provision from
86
non-nutritional sources, such as dextrose from drug dilution and hydration, citrate
87
from dialysate solutions, and lipid from the sedative agent propofol
88
non-nutrition energy sources are added to artificial nutrition delivery overfeeding
89
may result. Propofol comes as lipid emulsion either composed of pure LCT soya
90
lipid emulsion or a balanced mixture of medium chain triglycerides (MCT) and LCT.
91
It is generally solubilized as a 1% or 2% solution: its use, particularly in high doses,
92
increases the overall proportion of fat provided to the patient 8, 9.
93
Alterations of lipid metabolism are frequent during critically illness
94
hypertriglyceridemia is observed in nearly 45% of patients who require more than 3
95
days of ICU treatment. Hypertriglyceridemia is considered to reflect a global liver
96
dysfunction that has several causes, including overfeeding. Propofol and its
97
accompanying lipid emulsion are the strongest risk factors for hypertriglyceridemia,
98
stronger than total energy, glucose and lipids intakes from feeding 8. Moreover,
6-8
. When these
AC C
EP
TE D
M AN U
SC
RI PT
66
10
, and
10APR17- 3
ACCEPTED MANUSCRIPT 99
hypertriglyceridemia, may also occur during the rare but life-threatening propofol
100
infusion syndrome 11.
101
The present study aimed to quantify the quantity of lipid and the proportion of both
102
energy and lipids that resulted from propofol sedation in critically ill patients
103
requiring artificial nutrition in two distinct ICU settings.
RI PT
104 105 Methods:
107
The study was designed as a bi-centric retrospective analysis of prospectively
108
collected data of consecutive patients admitted in the ICUs of two teaching hospitals
109
(Alfred Hospital [=AH], Melbourne, Australia and the Centre Hospitalier Universitaire
110
Vaudois [=CHUV], Lausanne, Switzerland). Inclusion criterion was a ventilation time
111
>5 days and an ICU stay longer than 5 days, with no exclusion criteria. The patients
112
were identified through hospital databases. The study was conducted from August
113
2011 to March 2012 at CHUV and from January 2012 to December 2012 in AH.
114
Data was collected from ICU admission until ICU stay day 10 or ICU discharge
115
(whichever occurred first). Data obtained included demographic and admission data,
116
nutrition assessment information (including energy and protein targets) and daily
117
nutrition therapy data including the mode of nutrition, delivery of energy, and lipid
118
amounts from propofol and feeds. Length of mechanical ventilation, of ICU stay and
119
outcome were recorded, but not infectious complications.
120
The protocol was reviewed by the Commission Cantonale d’Ethique pour la
121
Recherche sur l’être humain (CHUV) and by Human Research Ethics Committee at
122
The Alfred Hospital (AH). In both institutions a low risk ethics approval was obtained
123
and the requirement for consent was waived due to the absence of intervention and
124
low risk nature of the project.
M AN U
TE D
EP
AC C
125
SC
106
126
Propofol sedation days were defined as any 24-hour period with ≥ 100mg of
127
propofol any day with less being aggregated with the “non-propofol days”. The
128
centers used different formulations of propofol: a 1% solution at AH (Fresofol 1%,
129
Fresenius Kabi, Australia: 1.1 kcal/ml, 100mg propofol deliver of 11 kcal as LCT)
130
and a 2% solution at CHUV (Propofol Lipuro, BBraun, Crissier, Switzerland: 0.5
10APR17- 4
ACCEPTED MANUSCRIPT 131
kcal/ml, 100mg propofol deliver 5 kcal). This difference meant that for the same
132
dose of propofol, AH patients received twice the amount of fat.
133 134
Nutrition therapy was as per evidence based nutrition guidelines in both ICUs and
135
EN is systematically favoured. Energy targets were set differently: at AH the
136
Schofield equation with added stress factors was used
137
a weight based target (25kcal/kg/day), or in patients >70 years the Harris-Benedict
138
predicted value times 1.2. Both centers use indirect calorimetry on occasions in
139
patients requiring specific nutrition therapy (burns, trauma, transplantation, obesity,
140
malnutrition). The choice of the EN formula was as per standard practice at each
141
site: the composition of the 3 most frequently used EN solutions per center is
142
depicted in Table 1. At CHUV blood triglycerides are determined twice weekly as
143
part of routine care, but not at AH.
144
For each patient, the mean fat intake and proportion of energy intake from fat was
145
calculated. High fat delivery was defined to occur when more than 45% of total
146
energy was supplied as fat, and excessive fat delivery when >55% of total energy
147
was supplied as fat.
12
M AN U
SC
RI PT
, while CHUV used mainly
TE D
148 Statistical analysis
150
Variables are reported as numbers or percent; normally distributed variables are
151
reported as mean± standard deviation and non-normally distributed variables are
152
reported as median [inter-quartile range (IQR)]. Comparisons between sites were
153
carried out using Chi2 tests for discrete variables, and 2-way ANOVAs for
154
continuous variables repeated over time. Single regressions were calculated
155
between fat doses and outcome variables. Significance was considered at the level
156
of p<0.05. Statistical package: JMP® Version 10.0, SAS Institute Inc. Cary, NC,
157
USA.
AC C
158
EP
149
159
Results
160
Altogether 701 admissions of 687 patients met the inclusion criteria, resulting in
161
6485 study days. The demographics of the population is reported in Table 2. The
162
mean age was 59 ±16 years and Body Mass Index (BMI) was 27.2 ±6.9 kg/m2:
163
patients were significantly younger and heavier at AH: the 2 largest diagnostic
10APR17- 5
ACCEPTED MANUSCRIPT 164
categories were cardiovascular (19% and 23.6% respectively per site) and trauma &
165
musculoskeletal pathologies (higher at AH with 45%, versus 2.5%).
166 Propofol sedation: Of the 6485 study days, 3484 (53.7%) were with propofol
168
sedation (1623 and 1861 propofol days from CHUV and AH, respectively). Overall,
169
2045 ±1650 mg/d propofol were provided during the 10 study days, corresponding
170
to 85 mg/hour propofol with a large inter-ICU and inter-patient variability (Figure 1):
171
the median propofol dose was 1290 mg/day at AH and 2400 mg/day at CHUV
172
(p<0.0001). Propofol was most intensively used during the first 3 days: the
173
proportion of overall energy provided by propofol sedation is shown in Figure 2. As
174
the result of the 1% or 2% propofol emulsion, despite significantly higher propofol
175
dose at CHUV, the median amount of energy resulting from the sedation was lower
176
at CHUV (at AH 136 kcal/day (IQR 61-253) versus 108 kcal/day (IQR 40-178) at
177
CHUV: p<0.001). Energy from propofol was the unique source of kcal in several
178
patients for a few days. Overall, propofol sedation contributed 17 % as a mean of
179
total energy delivery per day over the study period.
M AN U
SC
RI PT
167
180
Nutrition therapy: Overall 75% of days were on EN: 17.4% of days were without
182
nutrition for multiple reasons. Indirect calorimetry was used to adjust the energy
183
target in 79 patients (21 AH, 51 CHUV) and repeated in some patients (total: 109
184
studies). The mean energy target for the population was 1987±411 kcal/day. The
185
energy target was higher at AH (p=0.0001) as result of the use of the Schofield
186
equation and a younger cohort. The mean daily amount of energy delivered from all
187
sources at both sites was 1362±811 kcal corresponding to in 81% of prescription.
188
Other nutrition information is shown in Table 3.
EP
AC C
189
TE D
181
190
Fat Delivery: Median fat delivery was 39 g/day (IQR 31-56) increasing to 43 g/day
191
(IQR 35-59) on propofol days. Maximal daily fat intake was 310 g/day. Total lipid
192
delivery by day is shown in Figure 3. Expressed per kg body weight/day, the
193
cumulated nutrition-propofol intake represents a median of 0.53 g/kg/day with (IQR
194
0.27-0.76) with maximal values in both centres of 2.11 and 2.31 g/kg. This
195
represents 31% (IQR 20-40) of total energy. Considering all days, fat delivery was
196
similar in both sites, but in absence of propofol sedation, the mean fat delivery was 10APR17- 6
ACCEPTED MANUSCRIPT higher at AH (37±17 g/d) compared to CHUV (28±27 g/d: p<0.0001) from the higher
198
fat-containing enteral feeds.
199
The predefined high fat doses (>45% of energy) were observed in 888 days
200
(13.7%) (Figure 2): 705 days at AH (19.6% of AH days), and 183 at CHUV (10.4%
201
of CHUV days: p<0.0001). Fat doses exceeding 55% of energy intake were found in
202
678 days (10.5%): 585 days at AH and 93 at CHUV (p<0.0001). Fat represented
203
100% of energy intake during 358 days (5.5%), constituting a highly unbalanced
204
substrate composition.
205
Mean blood triglyceride value was 1.7±1.0 mmol/l at CHUV (544 values) versus
206
2.5±2.2 mmol/l at AH (p= 0.023: only 9 values).
SC
RI PT
197
207
No association was observed between the propofol dose and mortality. When the
209
relationship between fat intake (mean and maximal in g/day, fat as % of total energy
210
intake, or maximal fat %) were examined, we found no relevant association with
211
either length of mechanical ventilation, or ICU stay or mortality (R2: 0.02 to 0.06 with
212
p<0.001). On the other hand, in survivors (n=547) we observed an association
213
between both Mean fat %/day, Maximal fat %/day and length of mechanical
214
ventilation (R2=0.108 and R2=0.081 respectively: p<0.0001).
215 216
TE D
M AN U
208
Discussion
218
This bi-centric study shows that propofol sedation may result in the delivery of large
219
amounts of lipids, especially during the first days of ICU treatment. While the
220
median fat dose per day was not excessive with 39 g/day (i.e. about 0.5 g/kg) an
221
important proportion of the patients received much more (up to 310 g/day). This
222
proportion was highest with the combination of the 1% propofol solution and high fat
223
containing enteral feeds. Fat from propofol sedation accounted for 17% of daily
224
energy as a mean and added about 10% to the “nutritional” energy.
225
Individual data show a substantial number of days (13.7%) with a fat proportion
226
exceeding 45% of total energy delivery, and even exceeding 55% in 10.5% of days.
227
There are no data to our knowledge that report on the clinical consequences of high
228
fat intakes resulting from propofol sedation since an abstract published in 1995 9:
229
most studies have concentrated on propofol’s impact on plasma triglycerides 8, or
AC C
EP
217
10APR17- 7
ACCEPTED MANUSCRIPT 11
on the propofol infusion syndrome
. Here we report a weak but significant
231
association of high fat proportions with prolonged ventilation time, which should be
232
further explored in prospective studies.
233
The impact of high fat delivery on organ function in critical illness is largely unknown.
234
Data on high fat diets come from studies investigating the impact of the ketogenic
235
diet on brain metabolism in refractory epilepsy, and from tracer metabolic studies.
236
Ketone bodies are an important source of energy for the brain
237
are high fat (but not 100% fat), low carbohydrate and normo-protein diets, used for
238
prolonged periods in epileptic children with apparently no harm
239
frequently in adults. The main side effect is acidosis which is known to stimulate
240
protein catabolism. In 2001 Hart et al 15 showed that patients on high fat diets (44%
241
fat with 42% carbohydrates) remained more catabolic compared to patients on high
242
carbohydrates feeds (82% with 3% fat) with similar 15% amount of proteins. If the
243
high fat delivery is combined with an insufficient protein delivery as is frequently the
244
case in ICUs
245
patients. Other studies in critical care show that whatever the route of feeding, an
246
increase of either fat or carbohydrates above the “normal” proportions of 30% lipids
247
and 55% carbohydrates stimulates hepatic de novo lipogenenesis
248
development of fatty liver.
249
patients testing the impact of a 5 day isocaloric isonitrogenous PN containing 75%
250
(PN-glucose) or 15% (PN-lipid) glucose 18.
251
Sedation is a recurrent worry in the ICU, as it has deleterious effects of its own
252
The present quality control study shows that there is room for improvement of
253
continuous sedation in several patients: delivering close to 9000 mg propofol in
254
24hours is not desirable
255
purpose, being by definition rapidly reversible, but the pharmacological properties
256
change during continuous prolonged sedation and reveal different side effects
257
Even the volatile agent sevoflurane has recently been shown to induce nephrogenic
258
diabetes insipidus
259
propofol has few side effects (except for the dangerous PRIS
260
solubilising lipid emulsion which is part of PN has been widely used, it is perceived
261
as a safe alternative: but its prolonged use out of the PN context has not been given
262
much attention until now.
RI PT
230
13
. Ketogenic diets 14
M AN U
SC
, and less
16
, this may aggravate the lean body mass loss observed in critically ill
and the
TE D
17, 18,
17
AC C
EP
. These data confirmed the results of a study in 16
19
.
19
. Anaesthetic agents appear suitable for sedation 11
.
20
. Other agents, such as dexmedetomidine, are expensive. As 11
), and as the
10APR17- 8
ACCEPTED MANUSCRIPT 263 264
Underfeeding associated with difficult EN is the most frequent observation in the
265
ICU
266
delivery is potentially beneficial for the energy balance, but it increases substrate
267
imbalance: this requires further investigations considering that high fat diets have
268
been shown to favor catabolism
269
than 110% of measured energy expenditure
270
propofol sedation adds 10% energy as a mean, while energy targets are frequently
271
estimated and hence inexact, there is a real risk of overfeeding with this agent
272
particularly when PN is required. The last decade has seen the publication of
273
studies of PN or combined EN and PN with negative outcome results
274
being possibly related to overfeeding as energy targets are frequently equation
275
based which has been shown to inexact in up to 70% of patents
276
hand there are positive outcome studies using either indirect calorimetry based
277
targets
278
deleterious during the early phase of acute disease, via several mechanisms
279
Of note the number of calorimetry studies was also insufficient in this cohort. As the
280
energy contribution from propofol becomes acknowledged with the help of
281
computerized systems
282
recognised, clinicians may want to adjust provision of nutrition therapy to include
283
these non-nutritional calories. If the energy from propofol is integrated in the a priori
284
calculation of the feed target, the consequence may be a reduction in the nutrient
285
delivery, at the expense of proteins
16, 21
: delivering large amounts of fat with sedation, while increasing energy
15
22
RI PT
. Overfeeding defined as the delivery of more
or low targets
25
23
, the latter
6, 24
). On the other
. Exceeding the energy needs may be particularly 26-28
.
TE D
24
M AN U
SC
, is less frequent. As the use of
7
EP
, and the deleterious consequences of overfeeding
29
if their delivery is not monitored, which may
30
adversely affect outcomes
. Monitoring should be implemented including real daily
287
deliveries and triglyceride determination once or twice weekly.
288
Significant energy delivery and imbalance of substrate composition (i.e. >55% of
289
energy as fat) was observed in 10.5% of days and occurred for >2 days in 19.8% of
290
patients. The clinical relevance of these results is still uncertain, but our data
291
suggest an association of high fat proportions with prolonged mechanical ventilation.
292
The provision of large amounts of fat to individual patients on individual days did
293
exceed quantities recommended for cardiovascular prevention (i.e. 35% of total
294
energy 2). Careful monitoring of the overall energy provision including all sources to
295
minimize the risk of substrate imbalance and/or overfeeding should be encouraged.
AC C
286
10APR17- 9
ACCEPTED MANUSCRIPT If nutrition therapy is adjusted to account for non-nutritional energy clinicians should
297
be mindful to monitor protein delivery to maintain a sufficient intake.
298
There are limitations to the present quality control study. The ICU populations in the
299
2 centers differed as did the sedation protocol and the choice of enteral feeds, but
300
this has the advantage of showing the widespread use of propofol across ages,
301
conditions and countries. A cohort study has the advantage of reflecting real life.
302
Both ICUs have guidelines-based feeding protocols, and achieve an elevated feed
303
delivery (80% as median) showing the relevance of the findings in very different
304
settings. Some differences in practice attenuated the strength of our observations,
305
such as the absence of a systematic monitoring of plasma triglycerides at AH: the
306
higher values observed at AH suggest that this monitoring might be useful. And
307
finally as our retrospective study was not designed to test the impact of high fat
308
delivery on clinical outcome (length of mechanical ventilation and stay, infections)
309
this issue should be verified prospectively.
310
Conclusion
311
Propofol sedation may result in a significant fat delivery that is frequently ignored.
312
Whenever propofol is used, it significantly adds to the amount of energy and
313
proportion of fat. In the early phase of ICU stay, this sedation regimen may result in
314
an imbalance in the substrate provision, dominated by fat. The more concentrated
315
2% propofol solution seems to reduce efficiently the risk of fat overload. The
316
consequences of
317
consequences on substrate metabolism, clinical outcomes and possibly on long-
318
term recovery.
SC
M AN U
TE D
EP
the latter require prospective investigation of potential
AC C
319
RI PT
296
10APR17- 10
ACCEPTED MANUSCRIPT Abbreviations
321
AH
Alfred Hospital
322
CHUV
Centre Hospitalier Universitaire Vaudois
323
BMI
Body Mass Index
324
EN
Enteral nutrition
325
PN
Parenteral Nutrition
326
ICU
Intensive Care unit
327
IQR
Interquartile range
328
LCT
Long chain triglycerides
330
M AN U
329
SC
RI PT
320
Ethical Approval and Consent to participate: study was approved by both institutions ethics committees: Commission Cantonale d’Ethique pour la Recherche
332
sur l’être humain (CHUV) and by Human Research Ethics Committee at The Alfred
333
Hospital (AH). In both institutions a low risk ethics approval was obtained and the
334
requirement for consent was waived due to the absence of intervention and low
335
risk nature of the project.
TE D
331
Consent for publication: all authors approved the final version
337
Funding: None. The present study was supported by internal resources of the 2
338
services
EP
336
Authors contribution: MMB, ML, and ER designed the study; All authors
340
participated to the acquisition of the data and interpretation of the results; MC, ER,
341
CS and MBB drafted and revised the manuscript which was approved by all the
342
authors,
343
Acknowledgments: The authors would like to thank Bianca Levkovich, and Owen
344
Roodenburg residents from the Alfred Hospital’s Intensive Care Unit, Melbourne, for
345
assistance in data collection
AC C
339
346
10APR17- 11
ACCEPTED MANUSCRIPT 347
Table 1: Enteral feed composition of the 3 principal solutions used in the respective ICUs.
348 349
CHUV
Nutrison
Nutrison Nutren
Isosource Isosource Promote
Protein-Plus
Energy
Pulmonary
Protein
Energie
Fiber+®
Multifibre®
®
®
Fibre®
®
Abbott
Nutricia
Nutricia
Nestlé
Nestlé
Nestlé
1.33
1.57
1.3
30
35
30
Energy density
1.25
1.5
1.5
35
35
56
% Energy from
% Energy from
45
carbohydrate % Energy from
20
protein
49
26
48
49
43
16
18
20
16
25
AC C
EP
TE D
350
M AN U
fat
SC
kcal/ml
RI PT
Feed
Alfred Hospital
10APR17- 12
ACCEPTED MANUSCRIPT Table 2: Patient characteristics AH
CHUV
Patients
687
373
314
(admissions)
(701)
(373)
(328)
Men (%)
66.8
68.9
64.3
Age (years)
56±17
53±18
SAPSII
50±16
50±14
BMI (kg/m2)
27.2±6.9
27.8±7.3
Weight (kg)
80±21
82±24
Length of Ventilation (d)
10.9±10.5
LICU(d) ICU mortality (%)
p
ns
RI PT
All patients
<.0001
51±18
ns
26.4±5.9
0.0069
77±17
0.0009
12.9±10.8
8.7±9.8
<0.001
15.5±12.4
17±13
13 ±12
<.0001
86 (12.5%)
48 (12.9)
38 (12.1)
ns
22.9
18.2
ns
Data as mean ±SD
M AN U
Total Hospital mortality (%) 20.4
SC
61±15
AC C
EP
TE D
*: includes any internal medicine, neurology and infectious diseases
10APR17- 13
ACCEPTED MANUSCRIPT Table 3: Detail of energy, fat and propofol delivery by site AH
CHUV
p
6485
3607
2878
-
Energy target (kcal)
1957 ±461
2172±384
1743±303
<.0001
Energy delivery (kcal/d)
1362 ±811
1488 ±865
1202 ±707
<.0001
% of prescription Lipid delivery (total g/d) Total g/kg/day % Fat in total daily energy
69.7 ±37.9 [80.7] 40±23
41 ±16
39 ±29
0.0203
0.53±0.30
0.55±0.25
0.52±0.41
0.0127
29.5%
31.4 ± 28.3
27.1 ± 16.1
<.0001
35.8±29.5
35.3±13.5
ns ns
Propofol sedation (mg/d) Energy from propofol sedation (kcal/d)
3214
1778 (49.3%)
1436 (49.9%)
2045±1651
1574 ±1196
2627 ±1927
<.0001
146±117 [119]
169 ±132 [136]
118 ±87 [108]
<.0001
16.6 ±21.4
18.5±24.1
14.2±17.2
[11]
[9]
TE D
% Energy propofol sedation
M AN U
- Propofol sedation days Days on propofol >100 mg
68.6±37.5 [80.2] 70.9±38.1 [81.5] <0.016
SC
Days
RI PT
All
[10]
AC C
EP
Data as mean ±SD, [median]
10APR17- 14
<.0001
ACCEPTED MANUSCRIPT
Legends to the figures: Fig 1 Propofol and fat dose by day during the first 10 days in both institutions
RI PT
(horizontal dotted lines show the medians) during the propofol sedation days (i.e. >100 mg/d): this figure shows the impact of the sedative concentration on the total fat delivery.
SC
Fig 2 Proportion of energy delivery by day from propofol sedation (i.e. >100 mg/d) in both sites with horizontal lines showing the 35%, 45% and 55% of total
M AN U
energy (Max = maximal value, 90% CI = upper 90% confidence interval) Fig 3 Total lipid delivery from feeding solutions plus propofol by day in all patients (box plots with interquartile ranges; doted horizontal line = mean value).
AC C
EP
TE D
Two way Anova: Effect of time p<0.0001, time*site p<0.0001
10APR17- 15
ACCEPTED MANUSCRIPT References Weijs P, Cynober L, DeLegge M, Kreymann G, Wernerman J, Wolfe RR. Proteins and amino acids are fundamental to optimal nutrition support in critically ill patients. Crit care 2014; 18:591.
2.
Krauss RM, Eckel RH, Howard B, Appel LJ, Daniels SR, Deckelbaum RJ et al. AHA Dietary Guidelines: revision 2000: A statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation 2000; 102:2284-99.
3.
Singer P, Berger MM, Van den Berghe G, Biolo G, Calder P, Forbes A et al. ESPEN Guidelines on Parenteral Nutrition: Intensive care. Clin nutr 2009; 28:387-400.
4.
Singer P, Shapiro H. Enteral omega-3 in acute respiratory distress syndrome. Curr opin clin nutri metab 2009; 12:123-28.
5.
Garrel D, Patenaude J, Nedelec B, Samson L, Dorais J, Champoux J et al. Decreased mortality and infectious morbidity in adult burn patients given enteral glutamine supplements: a prospective, controlled, randomized clinical trial. Crit care med 2003; 31:2444-9.
6.
Singer P, Anbar R, Cohen J, Shapiro H, Shalita-Chesner M, Lev S et al. The tight calorie control study (TICACOS): a prospective, randomized, controlled pilot study of nutritional support in critically ill patients. Intensive care med 2011; 37:601-09.
7.
Berger MM, Revelly JP, Wasserfallen JB, Schmid A, Bouvry S, Cayeux MC et al. Impact of a computerized information system on quality of nutritional support in the ICU. Nutrition 2006; 22:221-29.
8.
Devaud JC, Berger MM, Pannatier A, Marques-Vidal P, Tappy L, Rodondi N et al. Hypertriglyceridemia: a potential side effect of propofol sedation in critical illness. Intensive care med 2012; 38:1990-98.
9.
DeChicco R, Matarese L, Hummell AC, Speerhas R, Seidner D, Steiger E. Contribution of calories from propofol to total energy intake. J Am Diet Assoc 1995; 95:A25.
10.
Marik PE. Dyslipidemia in the critically ill. Crit care clin 2006; 22:151-9, viii.
11.
Diaz JH, Prabhakar A, Urman RD, Kaye AD. Propofol infusion syndrome: a retrospective analysis at a level 1 trauma center. Crit care res pract 2014; 2014:346968.
12.
Subramaniam A, McPhee M, Nagappan R. Predicting energy expenditure in sepsis: HarrisBenedict and Schofield equations versus the Weir derivation. Crit Care Resusc 2012; 14:202-10.
AC C
EP
TE D
M AN U
SC
RI PT
1.
13.
Morris AAM. Cerebral ketone body metabolism. J Inherit Metab Dis. 2005; 28:109-21.
14.
Thakur KT, Probasco JC, Hocker SE, Roehl K, Henry B, Kossoff EH et al. Ketogenic diet for adults in super-refractory status epilepticus. Neurology 2014; 82:665-70.
15.
Hart DW, Wolf SE, Zhang XJ, Chinkes DL, Buffalo MC, Matin SI et al. Efficacy of a highcarbohydrate diet in catabolic illness. Crit care med 2001; 29:1318-24.
16.
Alberda C, Gramlich L, Jones N, Jeejeebhoy K, Day AG, Dhaliwal R et al. The relationship between nutritional intake and clinical outcomes in critically ill patients: results of an international multicenter observational study. Intensive care med 2009; 35:1728-37.
17.
Schwarz JM, Chioléro R, Revelly JP, Cayeux C, Schneiter P, Jéquier E et al. Effects of enteral carbohydrates on de novo lipogenesis in critically ill patients. Am J Clin Nutr 2000; 72:940-45. 10APR17- 16
ACCEPTED MANUSCRIPT Tappy L, Schwarz JM, Schneiter P, Cayeux C, Revelly JP, Fagerquist CK et al. Effects of isoenergetic glucose-based or lipid-based parenteral nutrition on glucose metabolism, de novo lipogenesis, and respiratory gas exchanges in critically ill patients. Crit care med 1998; 26:860-67.
19.
Taskforce D. A. S., Baron R, Binder A, Biniek R, Braune S, Buerkle H et al. Evidence and consensus based guideline for the management of delirium, analgesia, and sedation in intensive care medicine. Revision 2015 (DAS-Guideline 2015) - short version. Ger Med Sci 2015; 13:Doc19.
20.
Muyldermans M, Jennes S, Morrison S, Soete O, Francois PM, Keersebilck E et al. Partial Nephrogenic Diabetes Insipidus in a burned patient receiving Sevoflurane sedation with an anesthetic conserving device-A case report. Crit care med 2016; 44:e1246-e50.
21.
Bendavid I, Singer P, Theilla M, Themessl-Huber M, Sulz I, Mouhieddine M et al. NutritionDay ICU: A 7 year worldwide prevalence study of nutrition practice in intensive care. Clin nutr 2016
22.
McClave SA, Lowen CC, Kleber MJ, Nicholson JF, Jimmerson SC, McConnel JW et al. Are patients fed appropriately according to their caloric requirements? JPEN 1998; 22:375-81.
23.
Dissanaike S, Shelton M, Warner K, O'Keefe GE. The risk for bloodstream infections is associated with increased parenteral caloric intake in patients receiving parenteral nutrition. Crit care 2007; 11:R114.
24.
Heidegger CP, Berger MM, Graf S, Zingg W, Darmon P, Costanza MC et al. Optimisation of energy provision with supplemental parenteral nutrition in critically ill patients: a randomised controlled clinical trial. Lancet 2013; 381:385-93.
25.
Doig GS, Simpson F, Sweetman EA, Finfer SR, Cooper DJ, Heighes PT et al. Early parenteral nutrition in critically ill patients with short-term relative contraindications to early enteral nutrition: a randomized controlled trial. JAMA 2013; 309:2130-38.
26.
Grau T, Bonet A, Rubio M, Mateo D, Farre M, Acosta JA et al. Liver dysfunction associated with artificial nutrition in critically ill patients. Crit care 2007; 11:R10.
27.
Berger MM. The 2013 Arvid Wretlind lecture: Evolving concepts in parenteral nutrition. Clin nutr 2014; 33:563-70.
28.
Braunschweig CA, Sheean PM, Peterson SJ, Gomez Perez S, Freels S, Lateef O et al. Intensive Nutrition in Acute Lung Injury: A Clinical Trial (INTACT). JPEN 2015; 39:13-20.
29.
Berger MM, Soguel L, Charrière M, Theriault B, Pralong F, Schaller MD. Impact of the reduction of the recommended energy target in the ICU on protein delivery and clinical outcomes. Clin Nutr 2017; 36
30.
Weijs PJM, Stapel SN, de Groot SDW, Driessen RH, de Jong E, Girbes ARJ et al. Optimal protein and energy nutrition decreases mortality in mechanically ventilated, critically ill patients: A prospective observational cohort study. JPEN 2012; 36:60-68
AC C
EP
TE D
M AN U
SC
RI PT
18.
10APR17- 17
ACCEPTED MANUSCRIPT
200
p<0.001
RI PT
150
SC
100 50
M AN U
Lipids total (g/day)
Fig 1
0
TE D
p<0.001
EP
9000 8000 7000 6000 5000 4000 3000 2000 1000
AC C
Propofol (mg/day)
0
1 2 3 4
5 6
7 8 9 10
1 2 3 4 5 6
Alfred Hospital
CHUV Days in ICU
7 8 9 10
ACCEPTED MANUSCRIPT
RI PT
100 90
SC
80 70
M AN U
60 50 40
TE D
30 20 10 0
Max % 100 90% CI 100 Median% 36.7
2
3
AC C
1
EP
% of total energy as propofol
Fig.2
100 54.5 14.2
100 34.1 9.4
4
5 6 Days in ICU 100 100 100 20.3 22.5 21.2 7.6 8.4 8.0
7
8
9
10
70.1 100 59.8 89.0 18.9 18.3 20.3 20.1 8.2 7.3 8.6 7.9
ACCEPTED MANUSCRIPT
300
SC
p<0.001
M AN U
200 150
TE D
100
EP
50 0 1
2
AC C
Lipids total (g/day)
RI PT
Fig 3
3 4
5
6
7
8
9 10
1
2
3 4
Alfred Hospital
5
6
CHUV Days of ICU stay
7
8
9 10
ACCEPTED MANUSCRIPT
Propofol sedation substantially increases the caloric and lipid intake in critically ill patients NUT-D-17-00105R1
RI PT
Highlights
Continuous propofol sedation results in a significant fat and energy delivery
•
A concentrated 2% solution reduces the unwanted intake of fat
•
Monitoring real daily intakes of fat enables adjusting nutrition therapy and
•
Triglyceride monitoring once to twice weekly assists the detection of impending fat overload
Propofol sedation over several days may have unwanted metabolic
EP
TE D
consequences on catabolism
AC C
•
M AN U
prevents unbalanced diet.
SC
•