Table 2 Randomised clinical trials of adjunctive therapies in severe malaria in adults and children

Warrell et al. [114]

Thailand

6 to 60 years

Cerebral malaria N = 100

RCT (allocation in pairs) Dexamethasone (DMO) over 48 hours adults 0.5 mg/kg initially then 7 doses of 10 mg each; children 0.6 mg/kg initially then 7 doses (total dose 2 mg/kg) (50) vs placebo (50) plus quinine

1⁰ Death in hospital

Deaths in DMO 8/50 (16%) vs 9/50 (18%) placebo

No benefit. Prolonged coma among the survivors: the interval between the start of treatment and the full recovery of consciousness was 63.2 +/-5.9 hours (mean +/S.E.M.) in the dexamethasone group, as compared with 47.4 +/-3.2 hours in the placebo group (P = 0.02). Complications, including pneumonia and gastrointestinal bleeding, occurred in 26 patients given dexamethasone and 11 given placebo (P = 0.004)

2⁰ Neurological sequelae at discharge; time to become rousable and to regain full consciousness; complications

Hoffman et al. [115]

Indonesia

1.5 to 42 years

Cerebral malaria N = 43

Dexamethasone (DMO) (3 mg doses 8 hourly for 48 hours) vs placebo plus quinine

1⁰ Death in hospital

Death DMO: 6/21 (28%) vs placebo 7/22 (31.8%) mean time to coma resolution DMO 83.4 hours (SD = 49.3) vs 80.0 hours (SD 59.4) in placebo

No benefit from addition of DMO in cerebral malaria

2⁰ Time to become rousable; time to regain consciousness; duration of fever; complications

Coma or hyperparasitaemia (>5%) and hypoglycaemia at any time during hospitalisation were significantly correlated with a fatal outcome, but were not improved by using dexamethasone. DMO arm increased risk of GI bleeding.

Di Perri et al. [117]

Burundi

<14 years

Cerebral malaria N = 100

RCT

Only coma resolution pre-specified (sample size calculation)

Deaths in Pt 0/26 vs 5/30 (17%) control

Stopped after enrolling 56 patients due to civil conflict

Pentoxifyllin (Pt) (10 mg/kg/day in saline as a continuous infusion over 72 hours) vs control plus quinine

Coma resolution 6 hours vs 4 hours (no SD given). Neurological sequelae Pt arm 2/26 (7.7) vs 2/25 (8%)

Looareesuwan, et al. [118]

Bangkok Hospital for Tropical Diseases

>16 years old

Severe malaria N = 45 (15 per arm)

Three arm placebo controlled trial

Outcomes included mortality, resolution of coma, renal failure, respiration failure, fever and parasite clearance time

No deaths in any group; no difference in coma resolution, duration of intubation, haemodialysis treatments and use of transfusion. Mean parasite clearance times were 59.2, 60.8, and 57.6 hours in Groups I, II, and III

No significant differences among the three treatment groups were found for any of the outcomes. Pentoxifylline as an adjunctive treatment produced no clinically evident benefit.

I/ high dose of Pt (1.67 mg/ kg/hr)

II/ Low dose Pt (0.83 mg/kg/hr).

III/ placebo (0.9% NaCl, 1 ml/kg/hr) over 72 hours

Hemmer et al. [119]

Hamburg, Germany

Adults

Falciparum malaria N = 100 projected

Placebo controlled trial Pentoxyfylin (20 mg/kg/ daily infused over 24 hours) vs placebo (saline infusion). One patient in Pt arm did not receive the drug. Only six patients in each group had severe disease (received quinine)

Only endpoint specified was TNFα

No difference in TNFα in Pt arm vs placebo. Subgroup patients with mild disease TNFα levels on day 4 vs placebo (P <0.01)

Trial stopped early for futility and concern that there were excess adverse events in the Pt arm.

52 randomised (27 Pt vs 24 placebo)

Non severe patients mefloquine (Pt 20 mg/kg x 3 doses taken 6 hours apart) or halofantrine (20 mg/kg x 3 doses 6 hours apart)

No effect of time to fever defervescence, parasite clearance or hospital stay

Pt arm AEs included nausea (8), vomiting (7), general uneasiness (2), palpitations (1), and discomfort with intravenous cannulae plastic needle (1) vs 3 in the placebo group: discomfort with intravenous cannulae plastic needle (3) and general uneasiness (1). (P <0.05)

Das et al. [120]

Orissa, India

Adults

Cerebral malaria N = 52 (30 control vs 22 Pt)

RCT

Not pre-specified

Coma resolution time better in Pt than control 21.6 ± 13.9 vs 63.5 ± 19.7 hours (P <0.001)

Pt arm TNFα levels decreased on day 3 (TNF. 47.92 pg/ml SD . ± 27.9; P = 0.0029), compared to admission values but in control there was a rise in TNFα levels (TNF . 589 pg/ml SD . ± 602.3; P >0.05).

Group 1 quinine only, Group 2: Pt support (10 mg/kg/day) for the initial 3 days plus iv quinine Method of allocation not specified

Mortality Pt 10% (n = 2) vs 27% (8) in control (P >0.05)

Lell et al. [121]

Kilifi, Kenya

9 months to 8 years

Cerebral malaria Estimated N = 20/15

Step-dose escalation

1⁰ Mortality and 3 months neurological sequelae (NS), SAE

Recruited 10 Pt and 5 control

Stopped early for safety concerns, mortality rate was unexpectedly high in the PTX group but sample size too small for definitive conclusions.

10 mg/kg up to 40 ml/kg Pentoxifyllin vs control

2⁰ coma resolution; fever and parasite clearance

1⁰ Death Pt 4 (40%) vs control: 1 (10%) 2 other SAE in Pt arm 72 hour neurological impairment Pt = 1(17%) vs 2 (50%) NS at 3 months Pt = 0 (0%) vs 1 (25%)

2⁰ Coma resolution time (hours) Pt = 8 (4 to 36) vs 8 (4 to 12)

Hemmer et al. (1997) [119]

Hamburg, Germany

>14 years

Severe and non severe malaria (25 had altered or abnormal renal, liver or coagulation tests) N = 97

Prospective RCT

Not pre-specified

Fever defervescence (days)

There were no significant differences in any of the parameters (parasite clearance, defervescence time, or length of hospital stay) between patients receiving heparin, ASA, or the controls.

3 arms: i/33 low-dose heparin (70 units/kg tds for 5 days subcut, ii/31 iv aspirin (500 mg on day 0, 2 and 4) iv, iii/ 33 control.

5 (range 2 to 9) heparin vs 4 (2 to 9) aspirin vs 3 (2 to 6) control

Length of hospital stay 9 (4 to 54) vs 8 (4 to 27) vs 8 (4 to 31) days respectively

van Hensbroek et al. [122]

Banjul, The Gambia

1 to 9 years

Cerebral malaria N = 610 (of 624: 14 either died before or unable to receive medication)

RCT Placebo controlled (in a 2 X 2 Factorial trial Monoclonal Ab (Mab: against TNFα) in human serum albumin (0.1%) vs placebo. A single i.v. infusion over 15-mins (other randomization: im artemether vs quinine)

1⁰ mortality in hospital and residual neurologic sequelae (NS) (6 months).

1⁰ Death Mab (60/302; 19.9%) vs (64/308; 20.8%) P = 0.9, NS at 6 month 6.8% (15/221) vs 2.2 (5/225) P = 0.04

Faster fever clearance on Mab arm. After adjustment for severity features and antimalarial strategy there was no significant survival benefit in children treated with Mab but significantly increased NS rate at 6 month in survivors (15/221 (6.8%) in the Mab arm compared with 5/225 (2.2%) 3.35 (1.08 to 10.4) P = 0.02

2⁰ parasite and fever clearance rates, coma recovery time, neurologic sequelae (NS) at discharge and 1 month

2⁰ NS at discharge 24.4% (59/242) and 22.1% (54/244); P = 0 .6 and NS at 1 month 11% (25/228) vs 6.4% (15/234) P = 0.1

Taylor et al. [123]

Blantyre, Malawi

1 to 12 years

Cerebral malaria (BCS < = 1) n = 31

Placebo controlled trial i.v. immune globulin (IVIG) (pooled from local blood donations) during the first 3 hours treatment plus quinine

1⁰ Composite of mortality or neurological sequelae

16 received IVIG; 15 placebo

Trial was stopped (based on preplanned stopping rules) – no benefit of IVIG. Placebo vs IVIG: Odds ratio death or sequelae = 0.24 (95% CI 0 to 05, 1 to 26 that is, odds of failing on placebo were about 1/4 of the odds of these events on IVIG

2⁰ Hours to regain full consciousness; Fever resolution; parasite clearance

Died or sequelae

IVIG = 5 + 5 = 10/16 (62.5%) Placebo 1 + 2 = 3/15 (20%)

Looareesuwan, et al. [118]

Bangkok, Thailand

>14 years

Severe malaria N = 23 (12 vs 11)

Placebo controlled trial Polyclonal anti-TNF Fab (n = 12) received 250, 500, or 1,000 units/kg (4 pt each dose). 5 received 2,000 units/kg. Controls (n = 11) saline 100 ml. Treatment allocation not stated.

Clinical endpoints included duration of coma (CM only) development of severe complications. Mechanistic and PK data generated

Coma recovery only reported in 5 patients (3 Fab vs 2 control) Adverse effects: weakness was present longer in controls 5 days (range 3 to 21) vs 4.0 (2 to 14) in Fab arms

Too few patients in the trial with CM (n = 7) to assess efficacy on coma recovery. Only one patient died (control) arm.

Maude et al. [126]

Chittagong, Bangladesh

>or = 16 years

P. falciparum (>2%) plus modified WHO severe malaria criteria N = 60

Phase II Controlled RCT Levamisole 150 mg po or ng stat dose immediately vs control plus artesunate

Composite: Clinical Death or coma recovery Parasite clearance and lactate Pharmacodynamics Microvascular flow

Death 5/29 (17%) Levamisole vs 9/27 (33%) P = 0.22

Levamisole inhibits cytoadherence in vitro and reduces sequestration of late-stage parasites.

No differences in proportions of trophozoites, measures of parasite clearance in blood over 30 hours or effects sequestration

Speculated whether rapid clearance of malaria parasites by artesunate may obscure beneficial effects of levamisole.

White et al. [127]

Thailand

Children >6 years and adults

Cerebral malaria N = 46

Double blind RCT: 3.5 mg/kg single dose given IM phenobarbitone (Pb) n = 24 Control (n. saline) n = 24

1⁰ Seizure prevention

1⁰ Convulsions in Pb arm 3/24 (12-5%), vs 13/24 (54%) placebo (P = 0-006).

Seizures prevention superior

2⁰ Death

2⁰ Deaths 8 (33%) in Pb arm 5 (20.8) placebo arm (P = 0.5)

Small numbers in trial not able to assess effect on mortality.

Kochar et al. [130]

Rajasthan, India

Adults age 14 to 74 years

Cerebral malaria N = 185

Randomly assigned:

Not specified

3/102 (3%) developed seizures after admission in Pb arm vs 19/83 (22.9%) control

Reported as a correspondence not formally reported as a peer reviewed manuscript.

10 mg/kg im one dose of phenobabitone (Pb) (n = 102) Control (n = 83)

Deaths Pb: 29/102 (28.4% vs control 33/83 (39.8%)Pb

Crawley et al. [128]

Kilifi, Kenya

Children 9 months to 13 years

Cerebral malaria N = 340

Placebo controlled trial

1⁰ Seizure prevention

1⁰ Seizure frequency lower in the Pb arm vs placebo group (18 (11%) vs 46 (27%) children had 3 or more seizures OR: 0.32 (95% CI 0.18 to 0.58)

Frequency of respiratory arrest was higher in the phenobarbital arm vs placebo arm

170 Pb arm/170 placebo arm

A single intramuscular dose of phenobarbital (Pb) (20 mg/kg) or identical placebo plus quinine

2⁰ Death

2⁰ Mortality higher in phenobarbitone arm (30 (18%) vs 14 (8%) deaths; OR 2.39 (1.28–4.64)).

Mortality substantially increased in children who received phenobarbital plus three or more doses of diazepam (OR 31.7 (1.2 to 814))

Not recommended in CM.

Gwer et al. [129]

two centres Kilifi and Kisumu hospitals, Kenya

Children 9 months to 13 years

Blantyre Coma Score < =2 (included CM and non-traumatic encephalopathy)

Placebo controlled trial a single dose of Fosphenytoin or placebo plus quinine (N = 173)

1⁰ Seizure prevention and neurologic sequelae (3 months)

CM sub group (n = 110), Seizure prevalence fosphenytoin (n = 20/54; 37%) vs placebo (n = 21/56; 38%) (P = 0.233) Neurological sequelae 5 children (11%) vs 9 (19%) in the placebo arm (P = 0.98)

Stopped early (low enrolment/futility). No difference in clinical or EEG evidence of seizures (P = .980); gross neurological sequelae (P = .283).

85 received Fosphenytion

Deaths not specifically reported for CM subgroup but overall fosphenytoin 18 (21%) vs placebo 15 (17%) (P = .489)

88 placebo

2⁰ Death

Mohanty et al. [133]

India

Adults

CM with evidence of brain swelling on admission CT scan

Controlled trial

1⁰ Death

1⁰ Death mannitol 9/30(30%) vs 4/31 (13%) control

Mannitol led to increased mortality (P = .11) and prolonged coma duration (P = .02)

1.5 g/kg mannitol followed by 0.5 g/kg every 8 hours vs control (no mannitol)

2⁰ Time to regain consciousness

Time to coma recovery 90 hours (IQR 22 to 380) vs 32 hours (IQR 5 to 168 hours)

Namutangula et al. (2007) [132]

Kampala, Uganda

Children 6 to 60 months

Cerebral malaria N = 156/156

Placebo controlled trial single dose mannitol (1 g/kg) vs placebo plus quinine

1⁰ Time to regain consciousness

1⁰ Time to conscious was18.9 hours (10 to 38) (mannitol) v 20.5 hours (14 to 53) (control)

No difference in any endpoint; no adverse events.

2⁰ Death

2⁰ Death mannitol 9/75 (12%) vs control 13/76 (16%)

Gordeuk et al. [135]

Zambia

Children <6 years

Cerebral malaria; unrousable coma N = 78

Placebo controlled Desferrioxamine (DFO) 100 mg/kg/day intravenously for 72 hours vs placebo

1⁰ Mortality

Mortality DFO 7/42 (16.7%) vs placebo 9/41 (22%). Coma recovery DFO: 20.2 hours (N = 41) placebo: 43.1 hours (N = 42) P = 0.38

Relative risk of mortality in DFO arm 0.76 (95% CI 0.31, 1.85)

2⁰ Time to recovery of full consciousness; parasite and fever clearance times

Coma recovery 1.3 times (95% CI 0.7 to 2.3) faster in DFO group than in placebo

Thuma et al. [136]

Zambia; two centres

Children <6 years

Cerebral malaria; unrousable coma; clear CSF

Placebo controlled Desferrioxamine (DFO) 100 mg/kg/day intravenously for 72 hours vs placebo N = 352

1⁰ Mortality

Trial stopped early by DSMB after mortality 32/175 (18.1%) DFO arm vs 19/177 (10.7) placebo arm

Increased risk of death in DFO arm (RR 1.70, 95% CI 1.00 to 2.89)

2⁰ Coma recovery (time to Blantyre Coma Scale: 5); parasite and fever clearance times; parasite clearance day 3

Coma recovery time 1.2 times faster in DFO group vs placebo (P = 0.21) DFO: 18.1 hours (N = 143) placebo: 19.0 hours (N = 158) 95% CI 0.97 to 1.6

Persistent seizures >3 lower in DFO 93/168 vs 115/166 (RR 0.8; 0.67 to 0.95) but recurrent hypoglycaemia higher 43/172 vs 29/172 (RR 1.48 (0.97 to 2.26))

Mohanty et al. [137]

Mumbai India

13 to 84 years

Severe malaria N = 45

Blinded placebo controlled trial 1. Deferiprone (75 mg/kg/day divided in 2 daily doses) 2. Placebo capsules plus antimalarials (10 days)

1⁰ Mortality

No deaths

2⁰ Coma recovery; parasite clearance time

Coma recovery time better in DFO arm 29 hours (SD10) vs 56 (14).

Watt et al. [139]

Bangkok, Thailand

Males 18 to 50 years

Severe malaria N = 30 (15 in each arm)

Phase II placebo controlled trial 1/ NAC dose 150 mg/kg over 15 mins, followed by 50 mg/kg over 4 hours, then 100 mg/kg over 16 hours (in 5% dextrose infusion): vs 5% dextrose (placebo) plus quinine

Lactate acidosis resolution

NAC lactate resolution at 24 hours 10/15 better than placebo (3/15) P = 0.011

Small phase II trial

Median coma resolution faster NAC 24 hours (6 to 60) vs 36 hours (18 to 120) P = 0.19.

Overall low mortality and few received renal dialysis.

Treeprasertsuk et al. [140]

Thailand

13 years or older

WHO (2000) definition of severe malaria Patient had to agree to stay in-hospital until Day 28 (N = 108; 54 NAC regimens; 54 placebo)

Placebo controlled trial of 3 NAC dosage regimens plus artesunate

Not specified; method of randomization not specified either

54 received NAC Gp 1 (n = 31), Gp 2 (n = 5)^b and 18 in Gp 3 (N = 18) Gp 4 (n = 54)

^bGroup 2 oral administration not tolerated as given too early and in a large amount of liquid volume)

Group (Gp 1) iv: 140 mg/kg loading dose then 70 mg/kg 4 hourly x 18 doses; Gp 2) a single iv loading dose followed by oral NAC (see doses Gp 1) Gp 3) iv: loading dose (140 mg/kg) then 70 mg/kg 4hourly for 24 hours then oral NAC Gp 4) placebo

Deaths-only 2 in Group 1. Fever clearance and parasite clearance times – no major differences

Data on fever and parasite clearance time were summarised in table by arm but not compared statistically.

Withdrawal rate

Inconclusive study; incomplete reporting.

Gp 1 (4; 13%), Gp 2 (3; 60%), Gp 4 (10; 18.5%)

Charunwatthana et al. [141]

Mae Sot General Hospital, Thailand and Chittagong, Bangladesh

>16 years

Modified WHO severe malaria (N = 108: 56 NAC,52 placebo)

Placebo controlled RCT NAC: 150 mg/kg over 15 mins, followed by 50 mg/kg over 4 hours, then 100 mg/kg over 16 hours (in 5% dextrose) vs saline/5% dextrose (placebo) plus iv artesunate

1⁰ Lactate clearance time, coma recovery time and parasite clearance time. Others: fever clearance time, mortality, red cell deformability

No difference in lactate clearance (adjusted for admission parasitaemia and bilirubin) (hazard ratio 0.98 (0.60, 1.62) or median coma recovery time (72 hours vs 96 hours) adjusted for censored deaths

No difference in case fatality rate: 21 (38%) in the NAC group vs 17 (33%) in the placebo group.

Treatment with Nacetylcysteine had no effect on outcome in patients with severe falciparum malaria

Yeo et al. [142]

Timika, Papua, Indonesia

18 to 60 years

Modified WHO severe malaria criteria N = 8

Placebo controlled RCT 12 g L-arginine hydrochloride vs saline over 8 hours plus iv astersunate

Measures of haemodynamic function, endothelial function and nitric oxide bioavailability, lactate clearance

No deaths; no adverse events; arginine did not improve lactate clearance nor endothelial nitric oxide (NO) bioavailability

L-arginine was given to an increase endothelial nitric oxide: since low NO and hypo-arginaemia associated with severe malaria. Study stopped early due to local political instability.

(6 L-‐arginine; 2 placebo)

L-Arginine was found to be safe but no benefit shown in small number studied. Further studies may require higher dosing.

Day et al. [68]

Ho Chi Minh City, Viet Nam

Adults

23 patients critically-ill with severe sepsis (n = 10) or severe malaria (n = 13).

Open, randomised, crossover study comparing increasing iv adrenaline doses of 0.1 to 0.5 g/kg per min vs dopamine 2.5 to 10 g/kg per min.

Comparing adrenaline and dopamine effect on acid-base balance and haemodynamics: incidence of drug- associated hyperlactataemia (rise in plasma lactate of >3 mmol/L)

9 received adrenaline first, and 14 received dopamine. Overall, 4 patients required both dopamine and adrenaline to normalise SBP. The full dopamine dose-profile protocol was effected in 19 patients. Development of lactic acidosis curtailed the adrenaline dose-profile at some stage in 16 patients (84%) P = 0.0002.

Infusion of inotropic doses of adrenaline in severe infections resulted in the development of lactic acidosis.

No significant differences were found between sepsis and malaria subgroups with respect to disease effects or responses to treatments.

Maitland et al. [60]

Kenya

Children >6 months

Severe malaria plus deep breathing and base deficit >8 N = 53

Dose finding study 0.9% saline (N = 20); 4.5% human albumin solution (HAS (N = 32). Boluses of 10 to 40 ml/kg given over 1st hour after admission

Aliquots of 10 ml/kg given to achieve CVP 5 to 8 cm and improvement in haemodynamic indices. Resolution of acidosis/base deficit reduction at 8 hours

Mean central venous pressure (SE) at admission was 2.9 cm H2O (0.5 cm H2O); in those with base deficit >15 (445 had hypotension); by 8 hours mean CVP = 7.5 mm HG and evidence of resolution of shock and respiratory distress.

Inadequate allocation concealment and inadequate sequence generation.

No evidence of adverse effects of fluid overload; there were only 4 deaths (case fatality 7.5%) and in the survivors, there were no apparent neurologic deficits at discharge.

Maitland et al. [145]

Kenya

Children >6 months

Severe malaria (Hb > 5 g/dl) plus deep breathing and base deficit >8 N = 150

RCT fluid resuscitation (20 ml/kg = bolus) 1. 4.5% albumin N = 61; 2. 0.9% saline N = 56 3. Control^c N = 33.

1⁰ Acidosis correction: mean percentage reduction in base excess admission to 8 hours.

1⁰ no difference in the resolution of acidosis between the groups;

^control only eligible if base deficit >8 but <15:

2⁰ Mortality; SAE of fluid overload; neurological sequelae (NS)

2⁰ mortality lower in albumin arm (N = 2 3.6%) vs saline (N = 11, 18%). Relative risk of mortality saline vs albumin 5.5 (95% CI 1.2 to 24.8; P = .013)

Subgroup analysis Base deficit >15: Deaths: albumin 2/23 (9%) vs saline 8/26 (33%)

NS: albumin 3/21 (14%) vs saline 1/18; (6%)

Base deficit 8 to 15: Deaths: Albumin 0/33, Saline 3/35 (9%) vs Control 2/33 (6%)

Maitland et al. [150]

Kenya

Children >6 months

Severe malarial anaemia (Hb < = 5 g/dl) plus deep breathing and base deficit >8; N = 61

Pre-transfusion bolus management RCT

1⁰ Acidosis correction (as above)

1⁰ no difference: albumin group 44% (95% CI 32 to 57%); saline group 36% (16 to 57%); control group 42% (19 to 66%) P = 0.7

Tolerability of protocol control 4 (22%) developed decompensated shock: albumin 4 (17%) required emergency interventions; two had salicylate toxicity and 1 had sickle cell anaemia. No need for alternative treatments in the saline group.

1. 4.5% albumin N = 23

2⁰ Mortality; SAE of fluid overload: pulmonary oedema, neurological events

2⁰ Deaths albumin 4 (17%); saline3 (15%); control3 (17%); No adverse events

2. 0.9% saline N = 20

Neurological sequelae

3. control N = 18

Akech et al. [151]

Kenya

Children >6 months

Severe malaria, deep breathing and base deficit >8 N = 88

RCT 20 ml/kg over the first hour, repeat if shock persists

1⁰ Acidosis and shock correction

1⁰ No difference in the resolution of shock or acidosis

Per protocol analysis of mortality albumin: 1/40 (2.5%) vs gelofusine 4/40 (10%), (P = 0.36)

1. 4.5% albumin (N = 44)

2⁰ Mortality; SAE related to fluid and neurological sequelae

2⁰ By ITT mortality albumin (1/44; 2.3%) vs Gelofusine (7/44; 16%) P = 0.06). No pulmonary oedema/fluid overload events. Fatal neurological events more common in gelatin arm.

2. Gelofusin (N = 44)

Akech et al. [152]

Kenya

Children >6 months

Severe malaria (Hb > 5 g/dl) plus deep breathing and base deficit >8 N = 79

RCT 20 ml/kg over the first hour, repeat x1 if shock persists

1⁰ resolution of shock over 8 hours.

1⁰ no difference in 8 hour shock resolution (D70: 23/37 (62%) vs HES: 25/39 (64%), respectively (P = .99).

Fluid boluses with either Dextran or 6% HES lead shock and acidosis resolution without eividence of adverse outcome.

Specifically, there was no evidence of a renal impairment with HES over 24

hrs of observation, its use was associated with falling creatinine levels and good urine output

1. Dextran 70 (N = 39)

2⁰ resolution of acidosis, in-hospital mortality, SAEs (allergic reaction, pulmonary oedema, and neurologic sequelae)

2⁰ Acidosis and respiratory distress resolved better in HES: 3/39 (8%) remained acidotic at 8 hours vs D70 10/37 (27%) (P = .05). 4 deaths (5%): two per arm. No SAEs

2. 6% hydroxyethyl starch (N = 40)

Maitland et al. [89]

Multi- centre 6 sites East Africa

Children

Critically-ill with severe sepsis (N = 1330) or severe malaria (N = 1793) plus shock N = 3121

FEAST trial: an open RCT comparing: 1) 5% albumin bolus g; 2) 0.9% saline bolus (saline- bolus group) 20 to 40 ml/kg over one hour; 3) no bolus (control)

1⁰ 48-hour mortality;

10 48 hour mortality

DMC stopped trial early (after enrollment of 3141 of projected 3600)

2 months to 12 years

Plus iv antibiotics and/or quinine

2⁰ pulmonary oedema, increased intracranial pressure, mortality or neurologic sequelae at 4 weeks

Sepsis: Bolus 108/884 (12.2%) vs 38/446 (8.5%) RR 1.43 (1.01 to 2.04).

Neurologic sequelae occurred in 2.2%, 1.9%, and 2.0%, respectively groups 1, 2 and 3 (P = 0.92), Pulmonary oedema or increased intracranial pressure occurred in 2.6%, 2.2%, and 1.7% (P = 0.17), respectively.

Malaria: 110/1202 (9.2%) vs control 34/591 (5.7%) RR 1.59 (1.10 to 2.31). 20: 28 days mortality 12.2%, 12%, and 8.7% respectively (P = 0.004)

Bojang et al. [159]

Banjul, The Gambia

6 months to 9 years

Severe malarial anaemia N = 114 (PCV <15%) Exclusion: immediate transfusion or recent iron treatment

Allocated ‘at random’ to transfusion (15 ml/kg whole blood N = 58) or iron supplementation for month N = 56 (infants = 2.5 mL tds; <20 kg = 5 mL or if > 20 kg = 7.5 mL) 3 times a day plus oral antimalarials

Not specified

Day 7 mean PCV was significantly less in children who received iron than transfused group (P = 0.001) Day 28: iron treatment arm had a significantly higher mean PCV than transfused arm

Day 90 not reported as at Day 28, children were allocated randomly to receive weekly chemoprophylaxis with Maloprim (pyrimethamine and dapsone) or placebo

Clinical reviews (plus malaria slide and haematocrit) on days 7, 28 and 90 after admission.

Olupot Olupot et al. [160]

2 centres, Mbale and Soroti hospitals, Eastern Uganda

2 months to 12 years

Severe anaemia SA (Hb < 6 g/dl) N = 160

Phase II RCT comparing whole blood (30 ml /kg; Tx30: N = 78) vs standard volume (20 ml/kg; Tx20: N = 82)

1⁰ correction of severe anaemia (to haemoglobin >6 g/dl) at 24 hours.

1⁰ Tx30 70 (90%) corrected SA vs Tx20 61 (74%) hazard ratio = 1.54 (95% CI 1.1 to 2.2) P = 0.01

Higher volume of blood than currently recommended was safe and resulted in an accelerated haematological recovery in Ugandan children with SA.

(95 (59%)) had P falciparum malaria (by RDT or slide)

2⁰ Re-transfusion; serious adverse events; mortality (48- hour and 28 days; redevelopment of severe anaemia

2⁰ Global Hb increment admission- Day 28 superior in Tx30 (P <0.0001); SAE and Death Tx30 = one non-fatal allergic reaction and one death (Tx30) vs 6 deaths in the Tx20 arm (P = 0.12)