Table 3.
Summary of Robust (Multi-Center, Low Risk of Bias, n > 100) Acute TBI RCTs (Multiple RCTs for an Intervention Highlighted)
| Study Characteristics | Study findings | Risk of bias | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Study | Country (N countries) | Number (Number of centers >1) | Comparison | GCS (mean): Md = Median, R = Range) | Mortality: early (≤1 month) | Mortality: late (> 1 month) | Mortality: NR | ICP* | Length of stay: ICU | Length of stay: hosp. | GOS (E = GOSE)** 6 month unless stated | Random seq. | Allocation concealment | Covariate adjustment | Ordinal approach |
| Fluid management [1] | |||||||||||||||
| Myburgh 2007 | 2 | 460 (16) | Saline vs. albumin | 7 (Md) | S | S | S (E, 24) | L | L | Y | N | ||||
| Hypothermia [1] | |||||||||||||||
| Andrews 2015 | 18 | 387 (47) | Hypothermia vs. normothermia | 3 – 15 (R) | I | ND | ND | I (E) | L | L | Y | Y | |||
| Management guided by ICP [1] | |||||||||||||||
| Chesnut 2012 | 2 | 324 (6) | ICP-guided Mx vs. clinical-imaging guided Mx | 3 – 8 (R) | ND | ND | ND (E) | L | U | Y | Y | ||||
| Pharmacological therapies not elsewhere defined [13] | |||||||||||||||
| Skolnick 2014 | 21 | 1195 (140) | Progesterone vs. placebo | 3 – 8 (R) | ND | ND, ND (E) | L | L | Y | Y | |||||
| Wright 2014 | USA | 882 (49) | Progesterone vs. placebo | 4 – 12 (R) | ND | ND (E) | L | L | Y | Y | |||||
| Shakur 2009 | 8 | 228 (15) | Low-, medium-, or high-dose Anatibant vs. placebo | 7·5 / 7·6 | ND | L | L | Y | |||||||
| Marmarou 1999 | USA | 139 (31) | Bradycor vs. placebo | 6·0 / 6·1 | ND | ND | ND | L | L | Y | N | ||||
| Euro. Study Group 1994 | 13 | 852 (21) | Nimodipine vs. placebo | NR | ND | ND | U | L | Y | N | |||||
| Teasdale 1992 | 3 | 352 (7) | Nimodipine vs. placebo | NR | ND | ND | ND | U | L | N | N | ||||
| Perel 2012 | 2 | 170 (10) | Tranexamic acid vs. placebo | 10·5 / 10·5 | ND | U | L | Y | |||||||
| Robertson 2014 | USA | 200 (2) | Erythropoetin vs. placebo | NR | ND | ND | U | L | L | ||||||
| Maas 2006 | 15 | 861 (15) | Dexanabinol vs. placebo | NR | ND | ND | ND | L | L | L | |||||
| Yurkewicz 2005 | USA | 404 (40) | Traxopodil vs. placebo | 4 – 8 (R) | ND | ND | L | U | L | ||||||
| Morris 1999 | 6 | 693 (99) | Selfotel vs. control [ceased; mortality / adverse events] | 6·1 / 6·1 | ND | ND | U | L | L | ||||||
| Marshall 1998 | 15 | 1120 (15) | Tirilazad vs. placebo | 4 – 12 (R) | ND | ND | L | U | L | ||||||
| Young 1996 | USA | 463 (29) | High-dose pegorgotein vs. placebo | 5·6 / 5·3 | ND | ND | U | L | L | ||||||
| Low-dose pegorgotein vs. placebo | 5·5 / 5·3 | ND | ND | ||||||||||||
| Pre-hospital cazre [3] | |||||||||||||||
| Bernard 2010 | Australia | 312 (4) | RSI vs non-invasive ventilation | 5·0 | ND | ND | ND | ND | S (E) | L | L | N | N | ||
| Morrison 2011 | Canada | 113 (2) | HTS + dextran vs. normal saline | ≤ 8 | ND | ND | ND (E, 4) | U | L | N | N | ||||
| Bulger 2010 | 2 | 1331 (11) | HTS + dextran vs. HTS vs. normal saline | 4·9 / 5·0 / 5·0 | ND | ND | ND | ND | ND | L | L | N | N | ||
| Seizure Prophylaxis [1] | |||||||||||||||
| McQueen 1983 | UK | 164 (2) | Phenytoin vs. placebo | NR | ND | U | L | N | |||||||
| Steroids [3] | |||||||||||||||
| Asehnoune 2014 | France | 336 (19) | Hydrocortisone + fludrocortisone vs. placebo | ≤ 8 | ND | ND | ND | L | L | Y | |||||
| Edwards 2005 | 49 | 10,008 (239) | Methylprednisolone vs. placebo | 3 – 14 (R) | I | ND | U | L | N | N | |||||
| Grumme 1995 | 2 | 396 (9) | Triamcinolone vs. placebo | Most <8 | ND | ND | ND (12) | U | L | N | N | ||||
| Surgery [3] | |||||||||||||||
| Jiang 2005 | China | 486 (5) | Standard large craniectomy vs. limited craniectomy | 5·2 / 5·3 | S | S | S | L | L | N | N | ||||
| Mendelow 2015 | 13 | 170 (31) | Early surgery (hematoma) vs. standard care | <8 – 15 (R) | S | ND, ND (E) | L | H | Y | N | |||||
| Cooper 2011 | 3 | 155 (15) | DC vs. standard care | 5 / 6 (Md) | ND | S | S | ND | I | L | L | Y | Y | ||
There is variability in how RCTs measure and report intracranial pressure (ICP). For this table, statistical significance was judged according to the overall ICP results as reported
GOS (Glasgow Outcome Scale)/GOSE (Glasgow Outcome Scale-Extended) = proportion of favorable outcome (e.g., GOS 4 or 5) between groups unless otherwise stated.
DC, decompressive craniectomy; Random Seq., random sequence generation; H, high risk of bias; Hosp., hospital; HTS, hypertonic saline; ICU, intensive care unit; I, intervention inferior (statistically significant difference favoring control, non-bold in comparison column); L, low; Mx, management; N, no; ND, no statistically significant difference between groups; NR, not reported; RSI, rapid sequence intubation; S, intervention superior (statistically significant difference favoring intervention, bold in comparison column); SA, South Africa; U, unclear; UK, United Kingdom; USA, United States of America; Y, yes.