Background
Methods
Study identification
Study selection criteria
Data analysis
A priori hypothesis testing
Results
Study identification and selection
Study | Population | Methodology (score) | Intervention |
---|---|---|---|
Kuklinski 1991 [38] | Patients with acute pancreatic necrosis | C.Random: not sure | PN + selenium supplementation (500 μg/day) vs. PN without selenium supplementation |
n = 17 | ITT: no | ||
Blinding: no | |||
(4) | |||
Zimmerman 1997 [39] | Patients with SIRS, sepsis, APACHE II score >15, and multiorgan failure score >6 | C.Random: no | IV selenium as sodium selenite 1000 μg as a bolus and then 1000 μg sodium selenite 24 h as a continuous infusion over 28 days vs. standard |
n = 40 | ITT: yes | ||
Blinding: no | |||
(6) | |||
Berger 1998 [40] | Burns >30 % TBSA | C.Random: yes | IV copper (40.4 μmol), selenium (159 μg), zinc (406 μmol) + standard trace elements vs. standard trace elements (copper 20 μmol, selenium 32 μg, zinc 100 μmol) from days 0 to 8, all received early EN |
n = 20 | ITT: yes | ||
Blinding: double blind | |||
(12) | |||
Angstwurm 1999 [41] | Patients with SIRS and sepsis at 11 ICUs | C.Random: not sure | PN with high-dose selenium (535 μg × 3 days, 285 μg × 3 days, 155 μg × 3 days, and 35 μg thereafter) vs. low-dose selenium (35 μg/day for duration of study) |
n = 42 | ITT: yes | ||
Blinding: no | |||
(10) | |||
Porter 1999 [42] | Surgical ICU penetrating trauma patients with Injury Severity Score ≥25 | C.Random: yes | 50 μg selenium IV every 6 h + 400 IU vitamin E, 100 mg vitamin C every 8 h, and 8 g of N-acetylcysteine every 6 h via nasogastric or oral route from days 0 to 7 vs. none |
n = 18 | ITT: yes | ||
Blinding: no | |||
(9) | |||
Berger 2001 [43] | Trauma patients, surgical ICU | C.Random: yes | IV selenium supplementation (500 μg/day) vs. placebo (selenium group randomized further to two groups: 500 μg selenium alone vs. 500 μg selenium + 150 mg α-tocopherol + 13 mg zinc) given slowly for first 5 days after injury (all groups received EN) |
n = 31 | ITT: no | ||
Blinding: double | |||
(9) | |||
Lindner 2004 [44] | Patients with acute pancreatitis admitted to the ICU | C.Random: not sure | IV sodium selenite dose of 2000 μg on day 1, 1000 μg on days 2–5, and 300 μg from day 6 until discharge vs. placebo (isotonic 0.9 % IV NaCl solution) |
n = 70 | ITT: no | ||
Blinding: single | |||
(9) | |||
Angstwurm 2007 [45] | Multicenter mixed ICUs | C.Random: not sure | 1000 μg selenium IV within 1 h followed by 1000 μg selenium for 14 days vs. NaCl (0.9 %) (all patients received EN or PN) |
n = 249 | ITT: no | ||
Blinding: double | |||
(8) | |||
Berger 2007 [46] | Burns >20 % TBSA | C.Random: not sure | IV 100 ml of copper (59 μmol) + selenium (375 μg + zinc (574 μmol) vs. NaCl (0.9 %) from admission for 5–15 days |
n = 21 | ITT: yes | Both groups were on EN | |
Blinding: no | |||
(8) | |||
Forceville 2007 [47] | Septic shock patients | C.Random: not sure | 4000 μg selenium IV on day 1 followed by 1000 μg selenium for 9 days vs. NaCl (0.9 %) (all patients received EN or PN) |
n = 60 | ITT: no | ||
Blinding: double | |||
(8) | |||
Mishra 2007 [48] | Septic ICU patients | C.Random: not sure | 474 μg selenium IV × 3 days followed by 316 μg × 3 days, 158 μg × 3 days, and 31.6 μg thereafter vs. 31.6 μg selenium (all patients received EN or PN) |
n = 40 | ITT: yes | ||
Blinding: double | |||
(9) | |||
Berger 2008 [49] | Mixed ICU | C.Random: not sure | IV selenium supplementation loading dose 540 μg/day + zinc (60 mg) + vitamin C 2700 mg + vitamin B 305 mg + vitamin E enteral 600 mg + vitamin E 12.8 mg IV for 2 days followed by half the dose of all vs. standard vitamins |
n = 200 | ITT: yes | ||
Blinding: no | All groups received EN or PN | ||
(10) | |||
El-Attar 2009 [50] | Patients with COPD | C.Random: yes | IV selenium as sodium selenite 100 μg/day, zinc 2 mg/day, and manganese 0.4 mg/day vs. none |
n = 80 | ITT: yes | Trace elements were administered during the period on mechanical ventilation | |
Blinding: yes | |||
(12) | |||
Montoya 2009 [51] | Medical/surgical septic | C.Random: yes | Day 1 IV sodium selenite 1000 μg , day 2 sodium selenite 500 μg, and thereafter 200 μg during 7 additional days vs. selenite 100 μg/day |
ICU patients | ITT: yes | ||
n = 68 | Blinding: double | ||
(7) | |||
Andrews 2011 [52] | Mixed ICU, multicenter | C.Random: yes | 500 μg selenium supplemented PN (12.5 g nitrogen, 2000 kcal) vs. standard PN (12.5 g nitrogen, 2000 kcal) initiated after ICU admission (actual median 2.6 days) for 7 days (actual duration, mean 4.1 days). |
n = 502 | ITT: yes | ||
Blinding: double blind | |||
(13) | |||
Manzanares 2011 [53] | Septic or trauma patients | C.Random: not sure | IV selenium supplementation loading dose 2000 μg (2 h) on day 1 followed by 1600 μg/day for 10 days vs. NaCl as placebo |
n = 35 | ITT: no (except mortality) | ||
Blinding: single blind | |||
(9) | |||
Valenta 2011 [54] | Patients with sepsis or SIRS | C.Random: not sure | IV selenium supplementation loading dose 1000 μg on day 1 followed by 500 μg/day for 5–14 days + <75 μg/day of sodium selenite added to PN vs. NaCl + <75 μg/day of sodium selenite added to PN |
n = 150 | ITT: yes | ||
Blinding: no | |||
(8) | |||
Heyland 2013 [16] | Multicenter mixed ICUs | C.Random: yes | 500 μg selenium via PN + 300 μg selenium, 20 mg zinc, 10 mg β-carotene, 500 mg vitamin E, 1500 mg vitamin C via EN vs. placebo via PN and EN |
n = 1218 | ITT: yes | ||
Blinding: double | |||
(12) | |||
Woth 2014 [14] | Mixed ICU, severe septic patients with multiorgan failure | C.Random: not sure | 1000-μg/30 minutes loading dose of sodium selenite and 1000-μg/day treatment for a maximum of 14 days vs. control group (not described) |
n = 40 | ITT: yes | ||
Blinding: no | |||
(6) | |||
Bloos 2016 [17] | Multicenter mixed ICU patients with severe sepsis or septic shock in last 24 h | C.Random: yes | IV loading dose of 1000 μg sodium selenite followed by continuous IV of 1000 μg sodium selenite daily until ICU discharge or for 21 days, whichever comes first vs. placebo (NaCl) |
ITT: yes | |||
n = 1089 | Blinding: double | ||
(12) | |||
Chelkeba 2015 [15] | Mixed ICU patients with severe sepsis and septic shock | C.Random: yes | IV loading dose of 2000 μg sodium selenite followed by continuous IV of 1500 μg sodium selenite daily until day 14 vs. standard therapy without selenium |
n = 54 | ITT: no | ||
Blinding: single | |||
(11) |
Study | Mortality (%) | Infections (%) | LOS days | |||
---|---|---|---|---|---|---|
Experimental | Control | Experimental | Control | Experimental | Control | |
Kuklinski 1991 [38] | ICU 0/8 (0) | ICU 8/9 ( 89) | NR | NR | NR | NR |
Zimmerman 1997 [39] | 3/20 (15) | 8/20 (40) | NR | NR | NR | NR |
Berger 1998 [40] | 1/10 (10) | 0/10 (0) | 1.9 ± 0.9 (1–4) | 3.1 ± 1.1 (2–5) | ICU | ICU |
per patient | per patient | 30 ± 12 (10) | 39 ± 13 (10) | |||
Hospital | Hospital | |||||
54 ± 27 (10) | 66 ± 31 (10) | |||||
Angstwurm 1999 [41] | Hospital | Hospital | NR | NR | NR | NR |
7/21 (33) | 11/21 (52) | |||||
Porter 1999 [42] | 0/9 (0) | 0/9 (0) | 5/9 (56) | 8/9 (89) | ICU | ICU |
22 ± 25.2 | 35.8 ± 21.9 | |||||
Hospital | Hospital | |||||
31.3 ± 23.4 | 49 ± 30 | |||||
Berger 2001 [43] | (a) Selenium alone | 1/11 (9) | (a) Selenium alone | 5/11 (42) | (a) | ICU |
2/9 (22) | 5/9 (56) | ICU | 8.6 ± 8.1 (12) | |||
(b) Selenium + zinc + α-tocopherol | (b) Selenium + zinc + α-tocopherol | 8.0 ± 4.0 (9) | Hospital | |||
Hospital | 64 ± 39 (12) | |||||
0/11 (0) | 3/11 (27) | 82 ± 78 (9) | ||||
(b) | ||||||
ICU | ||||||
5.8 ± 4.4 (11) | ||||||
Hospital | ||||||
60 ± 48 (11) | ||||||
Linder 2004 [44] | Not specified | Not specified | NA | NA | Hospital | Hospital |
5/32 (15.6) | 3/35 (8.6) | 24 (9–44) | 26 (11–46) | |||
Angstwurm 2007 [45] | 28-day | 28-day | New infections (HAP) | New infections (HAP) | ICU | ICU |
46/116 (40) | 61/122 (50) | 10/116 (9) | 10/122 (8) | 15.1 ± 10 (116) | 12.7 ± 9 (122) | |
Berger 2007 [46] | 1/11 (9) | 1/10 (10) | 2.1 ± 1.0 | 3.6 ± 1.3 | ICU | ICU |
per patient | per patient | 35 ± 27 (11) | 47 ± 37 (10) | |||
Forceville 2007 [47] | 28-day | 28-day | Superinfection | Superinfection | ICU | ICU |
14/31 (45) | 13/29 (45) | 1/31 (3) | 2/29 (7) | 21 (7–40) | 18 (10–31) | |
6-month | 6-month | Hospital | Hospital | |||
18/31 (59) | 20/29 (68) | 25 (7–68) | 33 (11–51) | |||
1-year | 1-year | |||||
66 % | 71 % | |||||
Mishra 2007 [48] | ICU 8/18 (44) | ICU 11/22 (61) | 1.5 ± 1.9 | 1.8 ± 1.6 | ICU | ICU |
Hospital | Hospital | per patient | per patient | 21.3 ± 16.2 (18) | 20.8 ± 21.8 (18) | |
11/18 (61) | 15/22 (68) | |||||
28-day | 28-day | |||||
8/18 (44) | 11/22 (50) | |||||
Berger 2008 [49] | ICU | ICU | 36/102 (35) | 34/98 (35) | ICU | ICU |
8/102 (8) | 5/98 (5) | 5.8 ± 5.4 (102) | 5.4 ± 5.7 (98) | |||
Hospital | Hospital | Hospital | Hospital | |||
14/102 (14) | 9/98 (11) | 23 ± 20 (102) | 26 ± 20 (98) | |||
3-month | 3-month | |||||
14/102 (14) | 11/98 (11) | |||||
El-Attar 2009 [50] | ICU | ICU | VAP | VAP | NR | NR |
2/40 (5.6) | 1/40 (2.9) | 5/36 (14) | 7/34 (21) | |||
Montoya 2009 [51] | Hospital | Hospital | NR | NR | Hospital | Hospital |
6/34 (18) | 8/34 (24) | 12 (12–14) | 17 (14–20) | |||
Andrews 2011 [52] | ICU | ICU | Confirmed | Confirmed | ICU | ICU |
84/251 (33) | 84/251 (33) | 104/251 (41) | 121/251 (48) | 13.2 (IQR 7.8–23.7) | 15.1 (IQR 8.3–28.4) | |
6-month | 6-month | Hospital | Hospital | |||
107/251 (43) | 114/251 (45) | 29.8 (IQR 14.7–52.4) | 31.2 (IQR 15.1–57.8) | |||
Manzanares 2011 [53] | ICU | ICU | VAP | VAP | ICU | ICU |
3/15 (20) | 5/16 (31) | 3/15 (20) | 7/16 (44) | 14 ± 11 (15) | 13 ± 6 (16) | |
Hospital | Hospital | |||||
5/15 (33) | 7/16 (44) | |||||
Valenta 2011 [54] | 28-day | 28-day | NR | NR | NR | NR |
19/75 (25) | 24/75 (32) | |||||
Heyland 2013 [16] | Hospital | Hospital | All | All | ICU | ICU |
216/617 (35) | 199/601 (33) | 168/617 (27) | 181/601 (30) | 14.2 ± 22.7 (617) | 13.8 ± 23.1 (601) | |
14-day | 14-day | VAP | VAP | Hospital | Hospital | |
154/617 (25) | 132/601 (22) | 71/617 (12) | 95/601 (16) | 31.2 ± 50.2 (617) | 29.5 ± 44.8 (601) | |
28-day | 28-day | |||||
190/617 (31) | 173/601 (29) | |||||
3-month | 3-month | |||||
239 | 222 | |||||
6-month | 6-month | |||||
250 | 235 | |||||
Woth 2014 [14] | In 14-day study period | In 14-day study period | Gram-negative | Gram-negative | NR | NR |
9/21 (43) | 11/19 (58) | 8/21 (38) | 3/19 (16) | |||
Gram-positive | Gram-positive | |||||
3/21 (14) | 2/19 (11) | |||||
Fungal | Fungal | |||||
1/21 (5) | 0/19 (0) | |||||
Bloos 2016 [17] | 28-day | 28-day | Secondary infections, day 14 | Secondary infections, day 14 | ICU | ICU |
152/543 (28) | 137/546 (25) | 11 (5–22) | 12 (6–24) | |||
90-day | 90-day | 243/543 (44.7 %) | 269/546 (49.3 %) | Hospital | Hospital | |
198/543 (38) | 201/546 (38) | Secondary infections, day 21 | Secondary infections, day 21 | 26 (16–42) | 29 (17–50) | |
319/543 (58.8 %) | 323/546 (59.2 %) | |||||
Chelkeba 2015 [15] | 28-day | 28-day | VAP | VAP | ICU | ICU |
9/29 (31) | 10/25 (40) | 16/26 (61.5) | 21/25 (84) | 19.7 ± 11 (26) | 23.8 ± 13 (25) | |
Hospital | Hospital | |||||
25.2 ± 10 (26) | 24.5 ± 9 (25) |
Primary outcome: mortality
Secondary outcomes
Overall effect on new infectious complications
Overall effect on ICU and hospital length of stay and ventilator days
Overall effect on new renal dysfunction
Subgroup analyses
PN selenium monotherapy vs. combined therapy
PN selenium loading dose vs. no loading dose
PN selenium high dose vs. low dose
Higher vs. lower mortality
Study quality on outcomes
Sepsis vs. nonsepsis studies
Effect of geographic representation of study patients on outcomes
Publication bias
Discussion
Conclusions
Key messages
-
Se is an essential trace element with antioxidant, immunomodulatory, and anti-inflammatory effects that has been considered the cornerstone of the antioxidant defense system.
-
Recently, in the largest trial on IV Se monotherapy, investigators were unable to find any clinical benefit of high-dose sodium selenite in patients with sepsis and septic shock.
-
According to our findings, there is no evidence for a beneficial effect on mortality, infections, and other relevant clinical outcomes of high-dose IV Se as single or combined therapy (antioxidant cocktails) in critically ill patients.
-
IV Se may be able to significantly reduce infections in those studies performed with nonseptic patients and when high-dose Se as an initial IV bolus is not administered.