Single-centre, randomized, controlled trial without the use of a placebo.
I. Blinding of randomization - can't tell
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement(s) - no

76 infants with GA < 34 wk.
Single centre study, Turkey.
May 15, 1993 - June 15, 1994

40 infants with mean GA (SD) 31.4 +/- 2.9 wk, mean BW (SD) 1623 +/- 468 g received 0.5 g/kg of IVIG (Sandoglobulin, Sandoz) within 24 hours of birth.
36 infants mean GA (SD) 32.3 +/-2.4 wk, mean BW (SD) 1684 +/- 519 g served as controls (no placebo was given)

Proved infection (clinical findings and blood and/or cerebrospinal fluid culture positive for a pathogen).
Total mortality, infectious mortality, days in hospital.

Proved infection, mortality from any cause, infectious mortality and days in hospital could be ascertained from this study.
No adverse effects were noted.

Multicentre, randomized, double-blind, placebo-controlled trial.
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes

588 infants with a BW of 500 - 1750 g. Age 3 - 7 days.
Six centres in the U.S.
July 16, 1987 - December 12, 1988

287 infants received 500 mg/kg of IVIG (Gammagard, Baxter Healthcare, Hyland Division, Glendale, Calif.) at enrolment (age - 3 to 7 days), 1 wk later, and then every 14 days until a total of five infusions had been given or until hospital discharge, whichever came first.
297 infants received an equal volume of a sterile solution of 5 % albumin and 0.9 % sodium chloride.

Proved infection [clinical findings of sepsis and at least one of the following: a positive blood culture (bacteria or fungi), the isolation of a pathogen from a normally sterile body site (CSF, pleural, peritoneal, or joint fluid; bone; soft-tissue; or urine obtained by suprapubic or bladder catheterization), or the isolation of virus from an infant with clinical deterioration].
NEC (stage II or III)
IVH (grade I - IV)
BPD (definition not provided)
Total days in hospital
Eleven neonates were excluded from the study but were included in the intention-to-treat analysis.

The following outcomes could be ascertained from this study: any serious infection (bacterial + fungal), IVH, NEC, deaths from all causes. Total episodes for sepsis were reported. There were 50 episodes of sepsis among 287 infants in the IVIG group and 75 episodes of sepsis among 197 infants in the placebo group. The outcome "sepsis, one or more episodes", could not be ascertained from this study.
Adverse reactions were noted during 10 infusions (5 in each study group, or < 1%). Mild increases or decreases in blood pressure, heart rate, or temperature that were reversed when the rate of infusion was slowed. Two infants in each group had fluid overload after an infusion and were treated with a single dose of furosemide.

Randomized, double-blind, placebo-controlled trial.
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - no
IV. Blinding of outcome measurement - yes

240 infants with BW < 1300 g.
Data for 172 patients are presented in this preliminary analysis; of these 46 were excluded from the statistical analysis (29 because they died during the first 5 days of life, 4 because of protocol violations, and 13 because of inadequate follow-up - usually because of their return to the referring hospital. 126 infants remained).
Single U.S. centre.
September 1984 - October 1987.

61 neonates (mean BW 977 g) received a dose of 1 g of a 6% solution of IVIG (Sandoglobulin, Sandoz Pharmaceuticals, East Hanover, N.J.) on 4 of the first 5 days of life, and a fifth dose was administered on day 15 or as close to that day as possible (the dose could be given as late as day 21).
65 neonates (mean BW 1043 g) received an albumin placebo at equal oncotic load at the same times.

Sepsis (signs and symptoms compatible with sepsis and a positive blood or CSF culture)
IVH diagnosed by ultrasonographic examination at 3-7 days of age.
1 neonate excluded because of severe anomalies incompatible with life.
For additional exclusions see "Participants".

This is an interim analysis of a larger study, the results of which have not been reported to date.
Data are available on the outcome of sepsis but not on IVH
No adverse effects were reported

Randomized, controlled trial without the use of a placebo
I. Blinding of randomization - yes
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement - no

In this study a subgroup with a BW</= 1500 g (n = 86) of the total population of 133 infants (BW range 550 - 3340 g; GA range 24 - 40 wk) fulfilled the inclusion criteria for this systematic review.
Single centre, Italy.
Dates not given.

43 infants received 0.5 g/kg of IVIG (Sandoglobulin) weekly for 1 month.
43 infants received no placebo or other intervention.

Criteria for diagnosis of sepsis, meningitis, arthritis, pneumonia, urinary tract infection, and surface infection included both a positive culture of blood, cerebrospinal fluid, tracheal aspirate, urine or pus, respectively, and the presence of clinical and non microbiological laboratory features. For the diagnosis of pneumonia, the appearance of a new infiltrate on a chest roentgenogram was also required.
NEC was diagnosed when typical clinical and radiologic symptoms were present.
3 infants in the control group who died within 24 hours after birth were excluded from the analysis by the authors.

The outcomes of sepsis, any serious infection, NEC, length of hospital stay, death from all causes and deaths from infections could be ascertained from this study. The 3 infants in the control group who died within 3 days of life are included in our analyses as per intention to treat.
No side effects were observed after IVIG administration.

Randomized placebo controlled trial with the use of a non-identical looking placebo (saline).
I. Blinding of randomization - can't tell
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement(s) - no

61 infants with a BW < 1500 g were enrolled. Single centre study, Taiwan.
July 1993 - June 1994

31 infants, mean BW (SD) 1210 +/- 340 g, mean GA (SD) 29.7 +/- 2.2 wk received Gammumine-N (Miles Inc. Cutter Biological, USA). IVIG was infused for 30 minutes to 2 hours within the first 12 hours of birth, and every 2 wk until the patient weighed 1800 g or was discharged. The dose of IVIG was 750-1000 mg/kg/dose if the infant's BW was < 1000 g and 500-750 mg/kg/dose if the infant's BW was between 1001-1500 g.
30 neonates, mean BW (SD) 1320 +/- 250 g and mean GA (SD) 30.6 +/- 1.7 wk received saline infusion

Proved infection was defined as bacteremia (positive blood culture or CSF culture) with clinical deterioration or haemodynamic change.
Blood samples for serum IgG levels were collected prior to the first dose of IVIG and on days 1, 7, 14, 21, 28, 35, 42, 49, and 56 after birth.

Any serious infection, total mortality, IVH, NEC, BPD, days in hospital and serum IgG levels were reported.

Randomized, double-blind, placebo-controlled study
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes

20 preterm neonates, weight < 2000 g at entry to study and < 7 days of age.
Single centre, U.S.
Dates not given.

10 neonates received IVIG (Gamimmune-N, Cutter Biologicals, Berkeley, Calif.) 5% IgG in 10% maltose at 15 ml/kg BW as a single infusion.
10 neonates received equal volume of 0.1% albumin in 10% maltose.

Nosocomial infection (not defined)
Survival

This study provides information on deaths from infections and deaths from all causes.
There were no differences in heart rate, respiratory rate, rectal temperature, and urine output before, during and after the infusions of IVIG or placebo (no differences between the groups)

Randomized, double-blind, placebo-controlled trial
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes

115 infants with BW of 600 to 2000 g and < 48 hours of age.
Single centre, U.S.
November 1, 1986 - August 31, 1987.

56 neonates (GA = 30 wk; Mean BW (SD) 1.3 +/- 0.7 kg) received IVIG (Sandoglobulin). Initial infusions of IVIG were 500 mg/kg for infants weighing > 1000 g at birth and 700 mg/kg for infants weighing < 1000 g. If serum IgG levels were < 700 mg/dl on day 2 or 6 after transfusion in the IVIG group, an additional dose of IVIG was administered at that time and subsequent doses were increased by 200 mg/kg. The objective was to maintain IgG serum levels at >700 mg/dl.
59 neonates (GA 31 wk; mean BW (SD) 1.3 +/- 0.4 kg) received placebo (equal volume of 6% or 10% sucrose solution). When an infant receiving IVIG required an extra dose, the paired patient in the placebo group also received an additional infusion.

Sepsis (systemic clinical deterioration with a positive blood culture, cerebrospinal fluid, or aspirate of another normally sterile body cavity).
NEC (abdominal distension with gastric retention, abdominal erythema, or bloody stools, with radiographic evidence of pneumatosis intestinalis, portal venous gas, or pneumoperitoneum and staged by the modified Bell's criteria)
Length of hospital stay.
Deaths from all causes.
Deaths from infection.
BPD (requiring O2 at 28 days for BPD)
IVH (Papile classification)

From the data presented the outcomes of sepsis, any serious infection, NEC, BPD, IVH, length of hospital stay, deaths from all causes and deaths from infection could be ascertained. Three episodes of sepsis/proved infection occurred in (an) infant(s) born at 24 weeks GA and BW of 600 g. We assumed that this was only one infant and assigned only one outcome in the meta-analyses.
Transient tachycardia and a decrease in blood pressure was noted in one infant who received IVIG. Transient rise in the alanine aminotransferase level was noted in one infant who received IVIG and one who received placebo.

Randomized controlled trial without the use of a placebo
I. Blinding of randomization - yes
II. Blinding of intervention - no
III. Complete follow-up - no
IV. Blinding of outcome measurement - no

66 neonates of < 30 wk GA.
2 centres in the U.K.
Dates not given.

34 infants received 200 mg/kg IVIG (Intraglobin F, Biotest Pharma, FRG) within 48 hours of birth and at 3-weekly intervals until discharge from the neonatal unit. On clinical suspicion of infection, neonates in the treatment group only were given a supplementary dose of IVIG 100 mg/kg. A further 100 mg/kg was given within the next 48 hours if infection was confirmed.
32 infants received routine intensive care.

Sepsis (blood-culture-proven infection).
NEC (clinical findings and pneumatosis intestinalis on abdominal X-ray, or confirmed at autopsy).
IVH (no definition given)
BPD (no definition given)
Length of stay in NICU (median and range)

The outcomes of sepsis and NEC could by ascertained. We used as denominators all randomized patients. We included the infants that were withdrawn because of early death in the outcome of mortality (all causes). One infant with two episodes of NEC was counted as one outcome.
Side effects were not reported

Randomized controlled trial without the use of a placebo
I. Blinding of randomization - yes
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement - no

80 infants with a BW of 2000 g or less.
Single centre, Italy.
June 1985 - December 1986.

40 infants received 0.5 g/kg/week of IVIG (IgVena, Sclavo; Siena, Italy) until they reached the GA of 36 wk and during the entire period of intensive care.
40 infants received no placebo or other intervention

Sepsis defined as clinical manifestations, microbiologic findings (positive blood culture or CSF culture) and non microbiologic laboratory findings (total and differential white blood cell count, erythrocyte sedimentation rate, C-reactive protein, platelet count, tests of haemostatic function).

Any serious infection, deaths from all causes and deaths from infection could be ascertained in this study. As sepsis included neonates with positive CSF cultures the results were included in the any serious infection category only. Data could not be separated between sepsis and meningitis.
No side effects or adverse reactions were observed following IVIG administration.

Multicentre, two-phase controlled trial. Phase I was placebo controlled and double-blinded; phase II was not placebo controlled.
I. Blinding of randomization - yes
II. Blinding of intervention - yes/no*
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes/no*
* This study had two phases; in phase 1a placebo was used but not in phase 2.

2,416 infants with BW 501-1500 g and randomized at a mean age of 44 +/- 25 hours after birth.
8 centres in the U.S.
January 1, 1988 - March 31, 1991 (or through April, 1991).

In phase I
595 infants received IVIG (Sandoglobulin, Sandoz Pharmaceuticals, East Hanover, N.J.)
623 infants received placebo - equal volume of 5 % albumin solution in the same vehicle prepared by the manufacturer of the immune globulin. The infants received their first dose of study drug within 24 hours of randomization. To achieve a target level of 700 mg of immune globulin/dl, infants weighing 501 to 1000 g were given 900 mg of immune globulin per kg of body weight and infants weighing 1001 to 1500 g were given 700 mg per kg. The infusions were repeated every two weeks until the infants weighed 1800 g, were transferred to another hospital, died or were sent home.
In phase II
609 infants received IVIG as per above (Phase I)
589 received no intervention

Sepsis (symptoms compatible with infection and a positive blood culture for bacteria or fungi obtained at least 96 hours after birth and before 120 days of life; for commensals the diagnosis required two positive blood cultures obtained no more than 4 days apart).
The diagnosis of meningitis required a positive culture of CSF. The diagnosis of urinary tract infection required a pure culture from urine obtained by catheterization or suprapubic puncture.
Proved infection (including septicaemia, meningitis, or urinary tract infection) during the first 120 days of life.
NEC (Bell's modified classification)
BPD (Not defined)
Days in hospital

The following outcomes could be ascertained from this study; sepsis, any serious infection, NEC, death from all causes, death from infection, days in hospital.
The infusions were discontinued in < 1% of infants (10 in the IVIG group and 11 in the placebo group) because of tachycardia or acute changes in blood pressure.

Phase I of Fanaroff 1994, placebo-controlled

1218 infants with BW 501-1500 g and randomized at a mean age of 44 +/- 25 hours after birth.
8 centres in the U.S.
January 1, 1988 - March 31, 1991 (or through April, 1991).

In phase I
595 infants received IVIG (Sandoglobulin, Sandoz Pharmaceuticals, East Hanover, N.J.)
623 infants received placebo - equal volume of 5 % albumin solution in the same vehicle prepared by the manufacturer of the immune globulin. The infants received their first dose of study drug within 24 hours of randomization. To achieve a target level of 700 mg of immune globulin/dl, infants weighing 501 to 1000 g were given 900 mg of immune globulin per kg of body weight and infants weighing 1001 to 1500 g were given 700 mg per kg. The infusions were repeated every two weeks until the infants weighed 1800 g, were transferred to another hospital, died or were sent home.

Sepsis (symptoms compatible with infection and a positive blood culture for bacteria or fungi obtained at least 96 hours after birth and before 120 days of life; for commensals the diagnosis required two positive blood cultures obtained no more than 4 days apart).
The diagnosis of meningitis required a positive culture of CSF. The diagnosis of urinary tract infection required a pure culture from urine obtained by catheterization or suprapubic puncture.
Proved infection (including septicaemia, meningitis, or urinary tract infection) during the first 120 days of life.
NEC (Bell's modified classification)
BPD (Not defined)
Days in hospital

Randomized controlled trial, without the use of a placebo
I. Blinding of randomization - yes
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement - no

150 neonates of 28 to 37 wk GA and less than 4 hours of age.
Single centre, Saudi Arabia.
Dates not given.

50 neonates received IVIG (Intraglobulin, Biotest Pharma, West Germany) 120 mg/kg within 2-4 hours of birth
50 neonates received IVIG (Intraglobulin) 120 mg/kg on day 1 and 8 of life
50 neonates received no intervention

Sepsis was defined as presence of clinical features and a positive culture of blood or cerebrospinal fluid.

Sepsis, any serious infection, death from all causes and death from infection could be ascertained in this study. One infant developed pneumonia in the control group.
The mean age at onset of infection was 46.3 hrs (range 8 to 76 hrs), suggesting that some infants had infection acquired in utero and were infected at the time of enrolment.
No adverse effect of the therapy was noted during the study and at 6-month follow-up.

Multicentre, randomized controlled, double-blind study.
I. Blinding of randomization - can't tell
II. Blinding of intervention - yes
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes

235 neonates of less than or equal to 32 wk gestation, hospitalised before 25 hours of life and having endotracheal tube and/or umbilical catheter on admission.
4 centres in Paris, France.
1987 - 1989.

120 neonates received 500 mg (10 ml) of polyvalent Ig (Biotransfusion, France) on days 0, 1, 2, 3, 17, and 31 of life.
In the placebo group 115 neonates received 10 ml of 0.2% albumin in the same fashion.

Deaths from infection
Certain nosocomial infection [clinical signs of infection, positive cultures (blood, urine, cerebrospinal fluid, tracheal aspirate, stools, gastric aspirate), at least two biological signs of infection (abnormal number of leukocytes, immature leukocytes > 5%, thrombocytopenia < 150 000/mm3, rise in fibrinogen levels > 4.5 g/L, rise in C-reactive protein levels > 20 mg/L.]
NEC was diagnosed when bloody stools were associated with radiologic pneumatosis.
Neonatal infection and infection occurring within the first 4 days of life, potentially of maternal origin were not counted as evaluation criteria.

Deaths from infection could be ascertained in this study. The definition of nosocomial infection did not meet our criteria for sepsis or any serious infection. The number of infants with one or more episodes of NEC could not be ascertained. The 46 infants in which the protocol was broken were maintained in the statistical analyses.
"There were neither clinical nor biologic side effects in any of the patients after Ig infusion".

Randomized controlled trial, without the use of a placebo group
I. Blinding of randomisation - can't tell
II. Blinding of intervention - no
III. Complete follow-up - can't tell
IV. Blinding of outcome measurement - no

68 infants with a BW of 1000 - 1500 g.
Single centre, Bangkok, Thailand.
February 1988 - March 1990.

34 neonates received 250 mg/kg of IVIG (Biotest Pharma, West Germany) within 4 hours of birth
34 neonates received 500 mg/kg of IVIG within 4 hours of birth
34 neonates received no intervention

Sepsis (presence of clinical findings of sepsis plus positive blood cultures).

The outcomes of sepsis and deaths from all causes could be ascertained in this study.
"No adverse effects were observed during the period of study".

Randomized double-blind placebo controlled trial.
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - no
IV. Blinding of outcome measurement(s) - yes.

105 infants were randomized into the study. 24 infants (12 in each group) were excluded because of initial serum IgG level > 4 g/L, violation of the study protocol, withdrawal of consent, or intrauterine infection.

40 infants mean GA (SD) 27.5 +/- 2.2 wk and mean BW (SD) 1.06 +/- 0.39 kg received 1g/kg (20 ml/kg) of IVIG (Baxter) on study day 0 (< 48 hours of age), day 3, 7, 14, 21. 41 infants, mean GA (SD) 27.7 +/- 2.5 wk, mean BW (SD) 1.13 +/- 0.38 kg received an equal volume of placebo (human albumin 5%).

Sepsis (symptoms and positive blood culture).
Days on ventilator.
Total mortality from any cause. Infectious mortality.

The outcomes of sepsis, deaths from all causes, and deaths from infections could be ascertained from this study.

Randomized double-blind trial
Published in abstract form only - full quality assessment not possible

111 VLBW infants

54 infants were given 300 mg/kg of IVIG (name of product not given) as 5% solution, once between 24-72 hours of age and again 72 hours later.
57 infants were given the same volumes as 5% dextrose

The outcome of sepsis was not defined in this abstract, but according to the authors sepsis occurred in 17 infants in the IVIG group and in 15 in the control group.
Hospital stay.

The length of hospital stay could be ascertained from this study published in abstract form only.
Respiratory rate increased in the IVIG group following the first infusion. No other side effects occurred.

Single centre, randomized controlled trial without the use of a placebo group.
I Blinding of randomization - can't tell
II Blinding of intervention - no
III Complete follow-up - no
IV Blinding of outcome measurement - no

94 neonates, GA </= 34 wk gestation, or BW </= 1500 g.
Single centre, Rome, Italy.
May 1984 - June 1986.

0.5 g/kg IVIG (Venogamma Polivalente, Ismunit, Pomezia, Italy) on the 1st, 2nd, 3rd, 7th, 14th, 21st, and 28th day of life (treatment group) or no intervention (control group)

Sepsis was defined as clinical signs of systemic infection and positive blood or CSF culture for a pathogen.
14 neonates were excluded from the analysis. 6 neonates in the treatment group were excluded; 3 underwent exchange transfusion, 2 died from severe respiratory distress syndrome and one had suspected prenatal infection. 8 control neonates were excluded; 2 underwent exchange transfusion, 3 died of severe respiratory distress syndrome, one died of respiratory distress syndrome and IVH, and two had suspected prenatal infection.

Sepsis, any serious infection, deaths from all causes and deaths from infection could be ascertained from this study. Infants that died and were excluded by the authors are included in our analysis.
"..no newborn infant showed any local or general reaction either during or after infusions. Three infants showed a temporary increase in IgE from 10-18 to 28-38 IU/ml without concurrent side effects".

Single centre, quasi-randomized trial without the use of a placebo.
I. Blinding of randomization - no.
II. Blinding of intervention - no.
III. Complete follow-up - yes
IV. Blinding of outcome measure(s) - no.

80 preterm neonates. Single centre, Turkey. Dates for the study period not provided.

40 infants with a mean (SE) GA of 36.18 (0.17) wk, mean (SE) BW of 1.85 (0.07) kg were given 500 mg/kg of IVIG (Sandoglobulin R) if they were weighing greater than 1500 g, and 700 mg/kg if they were weighing < 1500 at birth on days one, two and eight of life.
40 infants with a mean (SE) GA of 35.58 (0.19) wk and mean (SE) BW of 1.67 (0.07) kg got no placebo.

Blood culture proven sepsis, mortality from any cause, sepsis related mortality, days in hospital. Serum IgG levels were measured on days 1, 2, 8 and 12.

Sepsis, deaths from all causes could be ascertained from this study. No adverse effects were noted. The outcome of days in hospital was only reported for the control group. Treatment with IVIG resulted in a statistically significant increase in the serum IgG concentrations.

Randomized controlled trial without the use of a placebo group.
I. Blinding of randomization - yes
II. Blinding of intervention - no
III. Complete follow-up - yes
IV. Blinding of outcome measurement - no

116 neonates of < 32 wk GA and <1500 g BW.
Single centre, Antwerp, Belgium.
Dates not given.

56 neonates received 500 mg IVIG (Sandoglobulin) in 10 ml of saline over a period of 30 min within the first 12 h of life. This infusion was repeated every 24 h until the 7th day of life, then administered weekly for another 3 weeks. 60 neonates received no placebo or other intervention

The diagnosis of any serious infection was made when the clinical diagnosis, in association with suggestive laboratory data, was confirmed by a positive blood or CSF culture.

Any serious infection, hospital stay and death from all causes could be ascertained in this study.
Short lasting hypotension was observed shortly after IVIG administration in one patient.

Multicentre, randomized, double-blind, placebo-controlled trial
I. Blinding of randomization - yes
II. Blinding of intervention - yes
III. Complete follow-up - yes
IV. Blinding of outcome measurement - yes

753 neonates with a BW of 500 to 2000 g, GA </= 34 wk, postnatal age </= 12 hours
9 centres in the U.S.
June 1985 - April 1989.

372 neonates received a single intravenous infusion of 10 ml/kg of IVIG (500 mg/kg) (Sandoglobulin)
381 neonates received a single infusion of albumin 5 mg/kg

All outcomes were recorded during the first 8 weeks (56 days) of life
Sepsis was defined as clinical symptoms and signs consistent with sepsis in association with isolation of a causative organism from a blood culture specimen. For the diagnosis of sepsis caused by S. epidermidis two positive cultures were required.
Infection was defined as clinical signs and symptoms and isolation of a causative organism from either blood culture, CSF culture, urine culture by bladder tap or catheterization, or culture of a specimen from a normally sterile site during hospitalisation or at autopsy.
BPD was defined as need for continued oxygen or ventilatory support at 28 days of age for reasons other than apnea and either a chest radiograph or histopathologic criteria compatible with the diagnosis. IVH was graded on a scale of 1-4 according to Papile. Bell's classification of stage II or greater was used to define NEC.

Sepsis, any serious infection, NEC, and death from infection could be ascertained from this study. Although the outcomes of IVH and BPD were well defined, figures for these outcomes were not reported by group.
The IVIG-treated infants had slower heart rate during the infusion than before the infusion and higher systolic blood pressure for 2 hours after the infusion than before the infusion. Both groups tolerated the infusions similarly.

This is an RCT comparing the effect of fresh frozen plasma or gammaglobulin on humoral immunity in neonatal sepsis. This study did not meet the inclusion criteria as a randomized untreated control group was not included. A non-randomized concurrent group of infants without suspicion of infection and matched for age, birth weight, and gestational age served as a control group.

This is a double-blind placebo-controlled RCT assessing the efficacy of prophylactic use of IVIG in 21 pairs of ventilated neonates weighing more than 1500 g. The mean weight in the IVIG group was 2702 g and in the placebo group 2679 g; thus most neonates did not full fill the entry criterion of a weight < 2500 g. In this study IVIG did not significantly reduce the rate of infection, the duration of ventilation or the time to clinical recovery.

The authors do not provide enough information regarding definitions of outcomes for inclusion in this systematic review.

This is a double-blind RCT designed to determine whether IVIG administration modifies the incidence of infections in high-risk neonates. 170 infants were enrolled. The study population included neonates of > 1500 g birth weight with no upper limit stated by the authors. This study did thus not meet our inclusion criteria. The authors "found no evidence that the administration of IVIG affected parameters that might be related to the occurrence of systematic or localized infectious processes".

This is a randomized controlled trial but the infants enrolled were > 38 weeks gestation and weighed > 2500 g and therefore the study population did not fulfil our inclusion criteria

This study has been published in abstract form only, and the authors do not provide enough information regarding definitions of outcomes for inclusion in this systematic review.

Immunoglobulin was given intra-muscularly. It is unclear whether this is an RCT. Sepsis was not clearly defined.