Supplement 5. Research Questions with Evidence Tables

Supplement 5. Research questions with Evidence tables

Systematic reviews and randomized controlled trials (RCTs) identified in literature search are described in evidence tables below under respective research questions. The risk of bias in RCTs was assessed with Cochrane Tool for Risk of Bias Assessment as follows: (1) for sequence generation; (2) allocation concealment; (3) blinding; (4) assessment of incomplete data; (5) no selective reporting; and (6) other bias.+ is given for low risk; - for high risk; and +- for unclear risk of bias. Where appropriate, other, non-randomized studies are described.

* Significant difference reported (p<0.05); nsp = not specified

Note: Number of randomised patients ≠ number of analysed patients in many studies.

References are provided at the end of this Supplement.

Research questions:

Question 1. Should we use early enteral nutrition (EEN) in critically ill adult patients?

Question 2. Should we delay EN in adult patients with shock receiving vasopressors or inotropes?

Question 3. Should we delay EN in critically ill adult patients with hypoxemia, hypercapnia or acidosis?

Question 4. Should we delay EN in critically ill adult patients receiving neuromuscular blocking agents?

Question 5. Should we delay EN in critically ill adult patients receiving therapeutic hypothermia?

Question 6. Should we delay EN in critically ill adult patients receiving extracorporeal membrane oxygenation?

Question 7. Should we delay EN in critically ill adult patients managed in prone position?

Question 8. Should we delay EN in critically ill adult patients with traumatic brain injury?

Question 9. Should we delay EN in critically ill adult patients with stroke (hemorrhagic or ischemic)?

Question 10. Should we delay EN in critically ill adult patients with spinal cord injury?

Question 11. Should we delay EN in critically ill adult patients with severe acute pancreatitis?

Question 12. Should we delay EN in critically ill adult patients after gastrointestinal surgery?

Question 13. Should we delay EN in critically ill adults after abdominal aortic surgery?

Question 14. Should we delay EN in critically ill adult patients with abdominal trauma?

Question 15. Should we delay EN in critically ill adult patients with bowel ischemia?

Question 16. Should we delay EN in critically ill adult in patients with intestinal fistula?

Question 17. Should we delay EN in critically ill adult patients with an open abdomen?

Question 18. Should we delay EN in critically ill adult patients with intra-abdominal hypertension?

Question 19. Should we delay EN in critically ill adult patients with upper GI bleeding?

Question 20. Should we delay EN in critically ill adult patients with acute liver failure?

Question 21. Should we delay EN in critically ill adult patients with large gastric (residual) volumes?

Question 22. Should we delay EN in critically ill adult patients with absent bowel sounds?

Question 23. Should we delay EN in critically ill adult patients with diarrhoea?

Question 1. Should we useearly enteral nutrition (EN) in critically ill adult patients?

Question 1A. Should we use early EN rather than early parenteral nutrition (PN)?

Question 1B. Should we use early EN rather than delay nutritional intake?

Table 1A.RCTs comparing early EN vs early PN (8 studies identified during primary search using key word block “critical illness”).

Author
Year / Population
No of pt randomized(pt groups analysed) / Definition
Early EN / Definition
Control / Outcome / Outcome
Early EN / Outcome
Control / Risk of Bias Assessment
1 / 2 / 3 / 4 / 5 / 6
Moore
19891 / Laparotomy for abdominal trauma
N=59 (29 vs 30) / EN within 12 h postop
jejunostomy / PN <12 h postop / Major infections* / 1/29 / 6/30 / +- / +- / - / - / + / +
Kompan
19992 / Trauma in shock
N=28 (14 vs 14) / Immediate gastric EN (mean 4.4 h) / PN within 24h
EN started 24h
(mean 36.5 h) / Mortality (hospital)
MOF day 4* / 0/14
1.84±0.7 / 1/14
2.81±0.8 / + / + / - / + / + / +
Kompan
20043 / Trauma
N=42 (27 vs 25) / Immediate gastric EN (mean 10.6h) / PN within 24h
EN started >24h (mean 38.5h) / Mortality (ICU)
Infections* / 0/27
9/27 / 1/25
16/25 / +- / +- / - / + / +- / +
Lam
20084 / Crit ill burn pt.
N=82 (41 vs 41) / EN within 24 h / PN < 24 h / Mortality (nsp)*
Septic shock* / 6/41
10/41 / 15/41
25/41 / +- / +- / - / + / + / +
Altintas
20115 / Crit ill ventilated
N=71 (30 vs 41) / EN within 48 h after intubation / PN <48 h after intubation / Mortality (hospital)
Sepsis
LOS ICU
MV days* / 13/30
7/30
15 (9-22)
7 (5-9) / 20/41
13/41
14 (10-27)
9 (5-14) / +- / - / - / + / + / +
Justo Meirelles 20116 / TBI with GCS 9-12
N=22 (12 vs10) / EN as soon as hemodynamically stable ( 48h) / PN as soon as hemodynamically stable ( 48h) / Mortality (nsp)
Any infection
Pneumonia
LOS ICU / 1/12
2/12
2/12
14 (5-26) / 1/10
4/10
2/10
14 (6-24) / +- / +- / - / - / + / +
Sun
20137 / SAP (Atlanta criteria: local complications (pseudocyst, necrosis or abscess) or/and organ failure and APACHE II >8)
N=60 (30 vs 30) / EN within 48 hrs admission / PN during the first week / Mortality (hospital)
MODS*
Pancreatic infection*
LOS ICU (IQR)* / 2/30
5/30
3/30
9 (5-14) / 1/30
13/30
10/30
12 (8-21) / +- / +- / - / +- / + / +
Harvey
20148 / Unplanned ICU admission. EN/PN both possible.
N=2388 (1197/1192) / EN within 36 h / PN within 36 h,
for five days / Mortality (ICU) Mortality (hospital) Mortality (30-days)
Any infection / 352/1197
450/1186
409/1195
251/1195 / 317/1190
431/1185
393/1188
261/1188 / + / + / - / + / + / +

Table 1B. RCTs on Early EN vs. delayed nutritional intake (14 studies identified during primary search using key word block “critical illness”).

Author
Year / Population
No of pt randomized
(pt groups analysed) / Definition
Early EN / Definition
Control / Outcome / Outcome
Early EN / Outcome
Control / Risk of Bias
1 / 2 / 3 / 4 / 5 / 6
Early Enteral Nutrition (EN) vs. Delayed Enteral Nutrition (EN) 7 studies
Chiarelli
19909 / Burns, 25-60% TBSA
N=20 (10 vs 10) / Immediate EN
Mean 4.4 ± 0.49 h / EN after 48 h
Mean 57.7 ± 2.6 h / Mortality (nsp)
Infections / 0/10
3/10 / 0/10
7/10 / +- / +- / - / - / +- / +-
Eyer
199310 / Blunt trauma admitted to ICU
N=52 (19 vs 19) / EN within 24 h
Mean 31 ± 13 h
Duodenal / EN after 72 h
Mean 82 ± 11 h
Duodenal / Mortality (nsp)
Pneumonia
LOS / 2/19
8/19
11.8±7.9 / 2/19
4/19
9.9±6.7 / +- / +- / - / - / + / +
Minard
200011 / TBI with GCS 4-11
N=30 (12 vs 15) / EN within 60 h
Mean 33±15h
Nasojejunal / Gastric EN after resolving gastroparesis
Mean 84±41h / Mortality (nsp)
Infections
LOS hospital
LOS ICU / 1/12
6/12
30.0±14.7
19.3±8.8 / 4/15
7/15
21.3±13.7
11.7±6.7 / +- / +- / - / + / + / +
Peck
200412 / Burns >20% of TBSA
N=32 (14 vs 13) / EN within 24 h
Gastric, oral allowed / EN from day 7
Oral diet when tolerated / Mortality (nsp)
Infectiousevents / 4/14
12 / 5/13
11 / +- / +- / - / - / + / +
Nguyen
200813 / Mixed ICU
BMI 27-28
N=28 (14 vs. 14) / EN within 24 h / EN from day 4 / Mortality (hospital)
Pneumonia
LOS ICU *Ventilation days* / 6/14
3/14
11.3±0.8
9.2±0.9 / 6/14
6/14
15.9±1.9
13.7±1.9 / - / - / - / - / + / +
Moses
200914 / Poisoning + MV
N=59 (29 vs. 30) / EN within 48 h
Hypocaloric / Iv fluids / Mortality (hospital) / 3/29 / 3/30 / +- / +- / - / +- / + / +
Chourda-kis
201215 / Brain trauma
N=59 (34 vs. 25)
Restricted randomization / EN (gastric) within 48 hafter admission / EN (gastric) after resolving gastro-paresis (>48h) / Mortality (ICU)
Pneumonia
Infectiousevents
LOS ICU / 3/34
13/34
28
24.8±7.6 / 2/25
12/25
29
28.5±8.9 / - / - / - / + / + / +
Early Enteral Nutrition vs. iv fluids to oral diet (timing of oral diet mostly not predefined, but not within 48 h) (7 studies)
Moore
198616 / Post laparotomy abd. Trauma N=63 (31 vs. 32) / EN within 24h
jejunostomy
PN if needed / Glucose 5%
PN after 5 d if no oral / Mortality (nsp)
Sepsis* / 1/32
3/32 / 2/31
9/31 / +- / +- / - / - / +- / +
Hasse
199517 / Liver transplant
N=31 (14 vs. 17) / Within 12 h
Nasoduodenal / IV fuids to oral diet / Infections / 3/14 / 8/17 / - / - / - / + / + / +
Chuntrasakul
199618 / Trauma
ISS 20-40
N=38 (21 vs.17) / Immediate EN,
PN added, if EN insufficient / IV fluids to oral diet / Mortality (nsp)
LOS ICU *
Ventilaton * / 1/21
8.1±1.4
5.3±1.4 / 3/17
8.4±1.2
6.1±1.3 / +- / +- / - / - / - / +
Watters 199719 / Oesophagectomy, pancreatico-duodenectomy.
N=28 (13 vs.15) / Immediate
jejunostomy / IV fluids to oral diet / Anastomotic leaks
LOS hospital
LOS ICU / 1/13
17 ± 9
2.9 ± 1.7 / 4/15
16 ± 7
2.3 ± 1.2 / + / + / - / + / + / -
Singh
199820 / Intestinal perforation
N=43 (21 vs. 22) / 12-24 h
jenunostomy / IV fluids to oral diet / Mortality (nsp)
Pneumonia
Infectious events* / 4/21
3/21
8 / 4/22
8/22
22 / - / +- / - / - / + / +
Pupelis
200121 / Postop peritonitis, severe pancreatitis
n=60 (30 vs. 30) / <24 h postop
jejunostomy / IV fluids to oral diet / Mortality (nsp)
Unresolved peritonitis*
GI fistulas / 1/30
1/30
0/30 / 7/30
8/30
4/30 / - / - / - / - / + / +
Malhotra 200422 / GI perforation
N=200 (100 vs.100) / <48 h
gastric / IV fluids
Oral from day 5 / Mortality (nsp)
Pneumonia / 12/100
21/100 / 16/100
30/100 / +- / + / - / + / + / +

Table 1C. All RCTs from other questions added to our primary search results. Early EN vs early PN. (8 + 11 studies).

Author
Year / Population
No of pt randomized
(pt groups analysed) / Definition
Early EN / Definition
Control / Outcome / Outcome
Early EN / Outcome
Control / Risk of Bias
1 / 2 / 3 / 4 / 5 / 6
Seri
198423 / Abdominal trauma, after laparotomy, N=18 (10 vs 8) / EN started 12 hpostop; jejunostomy / Glucose + amino acids i/v for 7 days / Sepsis / 1/10 / 2/8 / - / - / - / - / +- / +
Adams 198624 / Abdominal trauma with laparotomy; N=46 (23 vs 23) / EN within 12 hpostop, jejunostomy / PN within 12 hrs postop / Mortality (hospital)
Pneumonia / 1/23
11/23 / 3/23
8/23 / - / - / - / +- / +- / +-
Hadley 198625 / TBI with GCS 10 N=45 (21 vs 24) / EN within 48 h / PN within 48 h / Mortality (hospital)
Pneumonia
Infection events / 3/21
10/21
15 / 2/24
9/24
17 / + / - / - / - / + / +
Moore
19891 / Laparotomy for abd. trauma
N=59 (29 / 30) / EN within 12 h,
jejunostomy / PN within 12 h postop / Major infections* / 1/29 / 6/30 / +- / +- / - / - / + / +
Kudsk 199226 / Abdominal trauma with laparotomy, ATI >15
N=98 (51 vs 45) / EN 24 ± 1.7 hours after surgery; jejunostomy / PN 22.9 ± 1.6 hours after surgery / Mortality (ICU)
Infections (any) Pneumonia
LOS hospital / 1/51
9/51
6/51
20.5±2.8 / 1/45
18/45
14/45
19.6±2.8 / +- / + / - / + / +- / +-
Borzotta199427 / TBI with GCS 8 and coma persisting >24h
N=49 (28 vs 21) / EN within 48 h / PN within 48 h / Mortality (nsp)
Pneumonia
Infectious events / 5/28
15/28
51 / 1/21
9/21
39 / + / +- / - / +- / + / +
Kalfarentzos 1997 28 / SAP (Imrie score ≥3 or APACHE II ≥8, CRP >120 mg/l and Balthazar D or E), N=38 (18 vs 20) / EN within 48h
Nasoenteral / PN within 48 h / Mortality (hospital)
Any infection*
Pancreatic infection
LOS hospital / 1/18
5/18
2/18
40 (25-83) / 2/20
10/20
4/20
39 (22-73) / +- / +- / - / + / + / +
Kompan
19992 / Trauma in shock
N=28 (14 vs. 14) / Immediate gastric EN (mean 4.4 h) / PN within 24h,
EN started at mean 36.5 h) / Mortality (hospital)
MOF day 4* / 0/14
1.84±0.7 / 1/14
2.81±0.8 / + / + / - / + / + / +
Bozzetti
200129 / GI cancer surgery, malnourished
N=317 (159 vs 158)
Only 20 pt in ICU / EN within 24h
Jejunstomy or nasojejunal / PN within 24h / Mortality (nsp)
Pneumonia
Anastomotic leak
LOS hospital / 2/159
9/159
7/159
13.4 ± 4.1 / 5/158
14/158
10/158
15.0 ± 5.6 / + / + / - / + / + / +
Gupta
200330 / Predicted SAP (APACHE >5)
N=17 (8 vs 9)
Only 1 pt in ICU / EN within 24h / PN within 48 h / Mortality (nsp)
MOF
Any infection
LOS hospital / 0/8
0/8
1/8
7 (4-14) / 0/9
6/9
2/9
10 (7-26) / + / + / - / +- / + / +
Kompan
20043 / Trauma
N=42 (27 vs. 25) / Immediate gastric EN (mean 10.6h) / PN within 24h
EN started later (mean 38.5h) / Mortality (ICU)
Infections* / 0/27
9/27 / 1/25
16/25 / +- / +- / - / + / +- / +
Eckerwall 200631 / Predicted SAP (APACHE II >8 and/or CRP >150 mg/Land/or peripancreatic fluidon CT). N=48 (23 vs 25)
6 pt in ICU / EN within 24h / PN within 24 h / Mortality (nsp)
MOF
Any infection / 1/23
1/23
3/23 / 0/25
1/25
0/25 / + / + / - / +- / + / +
Petrov
200632 / SAP (APACHE II ≥8 and/or CRP level ≥150 mg/L)
N=69 (35 vs 34) / EN within 24h / PN within 24 hrs of admission / Mortality (hospital)*
MOF*
Pancreatic infection*
Other infections* / 2/35
7/35
7/35
4/35 / 12/34
17/34
16/34
11/34 / +- / +- / - / - / + / +
Lam
20084 / Crit ill burn pt.
N=82 (41 vs 41) / EN within 24h / PN within 24 h / Mortality (nsp)*
Septic shock* / 6/41
10/41 / 15/41
25/41 / +- / +- / - / + / + / +
Altintas
20115 / Crit ill ventilated
N=71 (30vs 41) / EN within 48h after intubation / PN within 48 h after intubation / Mortality (hospital)
Sepsis
LOS ICU
MV days* / 13/30
7/30
15 (9-22)
7 (5-9) / 20/41
13/41
14 (10-27)
9 (5-14) / +- / - / - / + / + / +
Justo Meirelles 20116 / TBI with GCS 9-12
N=22 (12 vs 10) / EN as soon as hemodynamically stable (within 48h) / PN as soon as hemodynamically stable (within 48h) / Mortality (nsp)
Any infection
Pneumonia
LOS ICU / 1/12
2/12
2/12
14 (5-26) / 1/10
4/10
2/10
14 (6-24) / +- / +- / - / - / + / +
Sun
20137 / SAP (Atlanta criteria: local complications (pseudocyst, necrosis or abscess) or/and organ failure and APACHE II score > 8)
N=60 (30 vs 30) / EN within 48 hrs admission / EN delayed, PN within 1st week / Mortality (hospital)
MODS*
Pancreatic infection*
LOS ICU (IQR)* / 2/30
5/30
3/30
9 (5-14) / 1/30
13/30
10/30
12 (8-21) / +- / +- / - / +- / + / +
Boelens 201433 / Colorectal surgery
N=123 (61 vs 62) / EN 8h postop, nasojejunal tube, oral allowed / PN 8h postop, oral allowed / Pneumonia
Anastomotic leak*
LOS ICU
LOS hospital / 4/61
1/61
2.4 ± 0.8
13.5 ± 2.2 / 8/62
9/62
3.3 ± 1.5
16.7 ± 2.3 / + / + / - / + / - / -
Harvey
20148 / Unplanned ICU admission. EN/PN both possible.
N=2388 (1197/1192) / EN within 36 h / PN within 36 h
for five days / Mortality (ICU)
Hospital mortality
30-d mortality
Infections / 352/1197
450/1186
409/1195
251/1195 / 317/1190
431/1185
393/1188
261/1188 / + / + / - / + / + / +

Table 1D. All RCTs from other questions added to our primary search results. Early EN vs. delayed nutritional intake (14+ 8 studies).

Risk of bias assessed with Cochrane Tool for Risk of Bias Assessment.

Author
Year / Population
No of pt randomized / Definition
Early EN / Definition
Control / Outcome / Outcome
Early EN / Outcome
Control / Risk of Bias
1 / 2 / 3 / 4 / 5 / 6
Early Enteral Nutrition (EN) vs. Delayed Enteral Nutrition (EN) 7 +1 studies
Grahm 198934 / TBI with GCS 10
N=32 (17 vs 15) / EN within 36 hrs admission, jejunal / Gastric EN if bowel sounds (>48 h) / Pneumonia
LOS ICU / 2/17
7 (4-19) / 3/15
10 (range NA) / - / - / - / - / +- / +
Chiarelli
19909 / Burns, 25-60% TBSA
N=20 (10 vs. 10) / Immediate EN
4.4 ± 0.49 h / EN after 48 h
57.7 ± 2.6 h / Mortality (nsp)
Infections / 0/10
3/10 / 0/10
7/10 / +- / +- / - / - / +- / +-
Eyer
199310 / Blunt trauma admitted to ICU
N=52 (26 vs. 26) / EN within 24 h
(31 ± 13 h)
Duodenal / EN after 72 h
(82 ± 11 h)
Duodenal / Mortality (nsp)
Pneumonia
LOS / 2/19
8/19
11.8±7.9 / 2/19
4/19
9.9±6.7 / +- / +- / - / - / + / +
Minard
200011 / TBI with GCS 4-11
N=27 (12 vs. 15) / EN within 60 h
Mean 33±15h
Naso-enteral / EN after 60 h
Mean 84±41h
Gastric / Mortality (nsp)
Infections
LOS hospital / 1/12
6/12
30.0±14.7 / 4/15
7/15
21.3±13.7 / +- / +- / - / + / + / +
Peck
200412 / Burns
N=32 (16 vs. 16) / EN within 24 h
Gastric + oral allowed / EN from day 7 Oral diet when tolerated / Mortality (nsp)
Infectious events / 4/14
12 / 5/13
11 / +- / +- / - / - / + / +
Nguyen
200813 / Mixed ICU
BMI 27-28
N=28 (14 vs. 14) / EN within 24 h / EN from day 4 / Mortality (hospital)
Pneumonia
LOS ICU *
Ventilation * / 6/14
3/14
11.3±0.8
9.2±0.9 / 6/14
6/14
15.9±1.9
13.7±1.9 / - / - / - / - / + / +
Moses
200914 / Poisoning + MV
N=59 (29 vs. 30) / EN within 48 h
Hypocaloric / Iv fluids / Mortality (hospital) / 3/29 / 3/30 / +- / +- / - / +- / + / +
Chourda kis
201215 / Severe TBI, GCS mean 5.81± 1.94 vs 5.22 ± 1.95
N=59 (34 vs. 25) / EN (gastric) within 48 hrs admission / EN (gastric) after resolving gastroparesis (>48 h) / Mortality (nsp)
Pneumonia
Infectiousevents
LOS ICU / 3/34
13/34
28
24.8±7.6 / 2/25
12/25
29
28.5±8.9 / - / - / - / - / + / +
Early Enteral Nutrition vs. iv fluids to oral diet (timing of oral diet mostly not predefined, but never early (within 48 h)) 7 + 7 studies
Sagar 197935 / Major resections esophago-GI tract
N=30 (15 vs 15) / EN on the first postop. day, nasojejunal) / nil per mouth for at least 2 days / Any infections
Anastomotic leak
LOS hospital* / 3/15
0/15
14 (10-26) / 5/15
1/15
19 (10-46) / + / +- / - / +- / +- / +
Moore
198616 / Post laparotomy abd. trauma N=63 (31 vs. 32) / <24h
jejunostomy
PN if needed / Glucose 5%
PN after 5 d if no oral / Mortality (nsp)
Sepsis* / 1/32
3/32 / 2/31
9/31 / +- / +- / - / - / +- / +
Hasse
199517 / Liver transplant
N=31 (14 vs. 17) / 12 h
Naso-duodenal / IV fuids to oral diet / Infections* / 3/14 / 8/17 / - / - / - / + / + / +
Beier-Holgersen 199636 / Major resection esophago- GI tract
N=30 (15 vs 15) / EN day of OP, nasoduodenal, / placebo (n=15) / Mortality (nsp)
Any infection*
Anastomotic leak
Total complications / 2/15
2/15
2/15
8 / 4/15
14/15
4/15
19 / +- / +- / + / +- / + / +
Carr
199637 / GI resection
N=30 (14 vs 14 analyzed) / Immediate EN (after return from OP) nasojejunal / nil by mouth until passage or flatus / Any infections
Nausea/vomiting
Abd. distension
LOS hospital / 0/14
1/14
2/14
9.8 (6.6) / 3/14
7/14
4/14
9.3 (2.8) / + / + / - / + / + / +
Chuntrasakul
199618 / Trauma
ISS 20-40
N=38 (21 vs.17) / Immediate
+PN if needed / IV fluids to oral diet / Mortality (nsp)
LOS ICU *
Ventilaton * / 1/21
8.1±1.4
5.3±1.4 / 3/17
8.4±1.2
6.1±1.3 / +- / +- / - / - / - / +
Heslin 199738 / Resection of esophageal, gastric, peripancreatic or bile duct cancer
N=195 (97 vs 98) / <24h immunoenhancing diet through jejunostomy; / IV fluids to oral diet (up to 10 days, then EN or PN). / Mortality (hospital)
Any infection
Anastomotic leak
Any complication
Ileus / 2/97
20/97
3/97
44/97
3/97 / 3/98
23/98
4/98
33/98
4/98 / + / + / - / + / + / +
Watters 199719 / Oesophagectomy, pancreatico-duodenectomy.
N=28 (13 vs.15) / Immediate
jejunostomy / IV fluids to oral diet / Anastomotic leaks
LOS ICU
LOS hospital / 1/13
2.9 ± 1.7
17 ± 9 / 4/15
2.3 ± 1.2
16 ± 7 / + / + / - / + / + / -
Singh
199820 / Intestinal perforation
N=37 (21 vs. 22) / 12-24 h
jenunostomy / IV fluids to oral diet / Mortality (nsp)
Pneumonia
Infectious events* / 4/21
3/21
8 / 4/22
8/22
22 / - / +- / - / - / + / +
Pupelis
200121 / Postoperative peritonitis or severe pancreatitis, N=60 (30 vs 30) / EN within 24 hrs postop, jejunostomy / IV fluids until oral diet / Mortality*
Unresolvedperitonitis*
GI fistulas
LOS ICU
LOS hospital / 1/30
1/30
0/30
13.9 (14.6)
35.3 (22.9) / 7/30
8/30
4/30
16.0(20.5)
35.8(32.5) / - / - / - / - / + / +
Malhotra 200422 / GI perforation
N=200 (100 vs.100) / <48 h
gastric / IV fluids
Oral diet day 5 / Mortality (nsp)
Pneumonia / 12/100
21/100 / 16/100
30/100 / +- / + / - / + / + / +
Kaur
200539 / Non-traumatic intestinal perforation in malnourished pts
N=100 (50 vs 50) / EN within 24 h, nasoenteric tubes / No nutrition until oral diet (commonly day 4-5) / Mortality (nsp)
Sepsis
Major complication*
Tube intolerance / 3/50
3/50
6/50
11/50 / 4/50
8/50
12/50 / +- / +- / - / +- / + / +
Barlow 201140 / Upper GI cancer
N=121 (64 vs 57)
Block randomization. / EN within 12 h after surgery
Needle catheter jejunostomy / IV fluids until oral diet / Mortality (30 days)
Wound infection
Anastomotic leaks
LOS hospital / 3/64
7/64
2/64
16 (IQ=9) / 0/57
16/57
7/57
19 (11) / + / +- / - / + / + / +
Bakker 201441 / Acute pancreatitis (APACHE ≥8)
N=205 (101 vs 104) / EN within 24 hrs admission / Oral diet 72h after admission, if tolerated. / Mortality (6 months)
Any infection
(6 months) / 11/101
25/101 / 7/104
27/104 / + / + / - / +- / +- / +

Table 1E. Early EN vs delayed nutritional intake. Systematic reviews in unselected ICU population

Author year / Patients
Comparison / RCTs / Included studies / Early EN defined / Main results, EEN associated with:
Marik 200142 / Acutely ill / 15 / <36h / Reduced risk of infections (0.45; 95% CI 0.30-0.66; reduced hospital LOS (-2.2 days; 95% CI 0.81-3.63 days; ND in mortality or non-infectious complications.
Heyland 2003 43 / Critically ill / 8 / Moore 1986, Chiarelli 1990; Eyer 1993; Chuntrasakul 1996, Singh 1998, Kompan 1999, Minard 2000, Pupelis 2001, / <24-48h / No significant difference, a trend toward a reduction in infectious complications (RR, 0.66; 95% CI 0.36-1.22).
Doig 2009 44 / Critically ill / 6 / Chiarelli 1990; Chuntrasakul 1996, Kompan 1999, Pupelis 2001, Kompan 2004, Nguyen 2008 / <24 / Reduction in mortality [OR = 0.34, 95% CI 0.14-0.85] and pneumonia (OR = 0.31, 95% CI 0.12-0.78). ND in other outcomes.
Heyland 2009 45 (Canadian Guideline update) / Critically ill
Early EN vs delayed nutritional intake / 14 / Moore 1986, Chiarelli 1990; Eyer 1993; Chuntrasakul 1996, Singh 1998, Kompan 1999, Minard 2000, Pupelis 2000, Pupelis 2001, Kompan 2004, Malhotra 2004; Peck 2004, Dvorak 2004, Nguyen 2008 / <24-48 except 1 study / A trend towards a reduction in mortality (RR 0.68 95% CI 0.46,1.01, p =0.06; reduction in infectious complications (RR 0.76, 95 % confidence intervals 0.59, 0.98, p = 0.04). ND in LOS.
Dhaliwal 201446 (2013 Canadian Guideline update) / Critically ill
Early EN vs delayed nutritional intake / 16 / Moore 1986, Chiarelli 1990; Eyer 1993; Chuntrasakul 1996, Singh 1998, Kompan 1999, Minard 2000, Pupelis 2000, Pupelis 2001, Kompan 2004, Malhotra 2004; Peck 2004, Dvorak 2004, Nguyen 2008, Moses 2009, Chourdakis 2012 / <24-48 except 1 study / A trend towards a reduction in mortality (RR 0.72, 95% CI 0.50, 1.04, p=0.08; a significant reduction in infectious complications (RR 0.81, 95% CI 0.68, 0.97, p=0.02.
Taylor 201647 / Critically ill
Early EN vs delayed EN / 21 / Sagar 1979, Moore 1986, Chiarelli 1990; Schroeder 1991, Eyer 1993; Beier-Holgersen 1996, Carr 1996, Chuntrasakul 1996, Watters 1997, Singh 1998, Kompan 1999, Minard 2000, Pupelis 2000, Pupelis 2001, Dvorak 2004, Kompan 2004, Malhotra 2004; Peck 2004, Nguyen 2008, Moses 2009, Chourdakis 2012 / Not provided in the text. In at least one study >48h / Provision of early EN was associated with a significant reduction in mortality (RR=0.70;95%CI 0.49-1.00) and infectious morbidity (RR=0.74; 95%CI0.58-0.93) compared with withholding EN (delayed EN or STD).

Question 2. Should we delay EN in adult patients with shock receiving vasopressors or inotropes?

Table 2. Studies addressing early EN in patients with shock receiving vasopressors

Author
Year / Population
No of pt studied / Study design / Feeding details / Groups in comparison / Main outcomes
META-ANALYSES (0)
RANDOMIZED CONTROLLED TRIALS (0)
OTHER, NON-RANDOMIZED STUDIES (8)
Mentec
200148 / ICU patients,
N=143 / Prospective observational case control / Nasogastric EN as soon as possible / Normal gastric aspirate volume (GAV <150 ml) vs increased GAV (2x150-500 or once >500 ml) / High GAV (32%) was related to catecholamines (OR, 1.81; 95% CI, 1.21–2.70) and sedation (OR, 1.78; 95% CI, 1.17–2.71)
If upper GI intolerance (high GAV or vomiting) occurred: increased pneumonia, LOS and mortality
Revelly
2001 49 / Cardiac surgery on
hemodynamic support, N=9 / Prospective / Continuous EN in all pt. Baseline period 120 min, followed by EN for 180 min. / None / EN increased CI and metabolic responses compared to baseline. No evidence of GI ischemia (gastric tonometry and ICG clearance)
Berger
2005 50 / Acute heart failure after cardiac surgery, ICU stay ≥5d, N=70 (40 with EN) / Prospective, observational / EN from day 2 with stepwise increase (d2 25%, d3 50%, d4 75%, d5 100%) / Short stay (5-7 days) vs long stay (>7 days) and different feeding routes / Time to initiation was 72h in 68% of patients with EN
70% of energy target reached during the first 2 weeks.
No patient experienced anyserious gastrointestinal complication.
Rai
2010 51 / Septic (shock) patients with EN, MV >3 d, N=43 (33 in shock) / Retrospective / EN started at a mean of 1.4 days after admission / Shock vs no shock / 40% received <60% of goal over 7 days
Mean time to start was the same in shock vs no shock 1.3 (SD, 1.7) days v 1.7 (SD, 1.3) days (P=0.16)
69% in shock reached goals vs 77% without shock (NS).
Khalid
201052 / MV 2 days or longer and vasopressors,
N=1174 / Prospective data collection / Retrospective groups:
EN <48h from start of MV vs EN >48 h after MV / ICU mortality: Early EN 22.5% vs late EN 28.3%; p = 0.03.
Hospital mortality: Early EN 34.0% vs delayed EN 44.0%; P < .001.
After matching (propensity score): lower hospital mortality for early EN (33.9 vs 42.6%, p=0.01).
Mancl
2013 53 / Concomitant EN and vasopressor for >1 h, N=259 (346 episodes) / Retrospective / Start of EN not specified, episodes of concomitant EN and vasopressor / Episodes: EN tolerated vs EN not tolerated / EN tolerability 74.9%.
EN tolerated: 13.6±7.6 kcal/kg/day;
EN not tolerated: 10.9±6.8 kcal(kg/day
An inverse relationship between max norepinephrine dose and EN tolerability (12.5 mcg/min for patients who tolerated EN vs 19.4 mcg/min, P = .0009)
Patel
2014 54 / MV and septic shock, N=66 / Retrospective / EN within 48 h / Groups (within 48h):
no EN vs.
<600 kcal/d vs.
≥600 kcal/d / No EN: 15 pt; EN <600 kcal/d: 37 pt; EN ≥600 kcal/day: 14 pt.
Pt receiving <600 kcal/day had lower LOS and duration of MV compared to 0 or ≥600 kcal/day*
Flordelis Lasierra
201555 / Hemodynamic failure and at least 24h MV, N=37 (33 only EN; 4 pt EN + SPN) / Prospective observational / Feeding protocol. Nasogastric with gradually increasing (daily 25% steps) to achieve 100% in 4d. / Short stay (≤7 days) vs long stay (>7 days).
Pt with vs without EN-related complications. / In 36 pt EN was started within 48h.
33 pt had only EN, 4 supplemental PN.
Complications of EN in 23, none severe.
No significant differences in illness severity scores, mortality, or hemodynamic variables were detected in patients with or without EN-related complications.

Question 3. Should we delay EN in critically ill adult patients with hypoxemia, hypercapnia or acidosis?

Question 3A: Should we delay EN in critically ill adult patients with hypoxemia?

No evidence table.

Question 3B: Should we delay EN in critically ill adult patients with hypercapnia?

No evidence table.

Question 3C: Should we delay EN in critically ill adult patients with acidosis?

No evidence table.

Question 4. Should we delay EN in critically ill adult patients receiving neuromuscular blocking agents?

Table 4. Studies addressing (early) EN in patients receiving neuromuscular blocking agents (NMBAs)

Author
Year / Population
No of pt studied / Study design / Feeding details / Groups in comparison / Main outcomes
META-ANALYSES (0)
RANDOMIZED CONTROLLED TRIALS (0)
NON-RANDOMIZED STUDIES (1)
Tamion 200356 / Sedation and muscle relaxant for adequate MV
N=20 / Prospective observational / All patients:
First phase: analgo-sedation
Second phase: NMBA added. Paracetamol absorption test and gastric residual volume to measure gastric emptying / No difference in gastric emptying before and after NMBA.

Question 5. Should we delay EN in critically ill adult patients receiving therapeutic hypothermia?

Table 5. Studies addressing early EN in patients receiving therapeutic hypothermia

Author
Year / Population
No of pt studied / Study design / Feeding details / Groups in comparison / Main outcomes
META-ANALYSES (0)
RANDOMIZED CONTROLLED TRIALS (0)
NON-RANDOMIZED STUDIES (2)
Oshima
2015 57 / Therapeutic hypothermia
N=7 / Prospective observational / Nutrition started in the rewarming phase,
preferably through the enteral route. Goal 80% of BMR. / No groups. Energy expenditure (EE) assessed in hypothermia and normothermia. / Median EE for the hypothermic phase (day 1) was 1557 kcal/d, in normothermia 2375kcal/d. There was significant association between cumulative energy deficit and the length of ICU stay.
Williams
2014 58 / Therapeutic hypothermia
N=55 / Retrospective
observational / EN started in hypothermia / No group assessment. Feeding intolerance assessed in cooling, rewarming and normothermia / EN was tolerated in 72%, 95% and 100% in respective phases.

Question 6. Should we delay EN in critically ill adult patients receiving extracorporeal membrane oxygenation?

Table 6. Studies addressing (early) EN in patients receiving extracorporeal membrane oxygenation (ECMO)