This article has Open Peer Review reports available.
The INIS Study. International Neonatal Immunotherapy Study: non-specific intravenous immunoglobulin therapy for suspected or proven neonatal sepsis: an international, placebo controlled, multicentre randomised trial
© The INIS Study Collaborative Group; licensee BioMed Central Ltd. 2008
Received: 10 September 2008
Accepted: 08 December 2008
Published: 08 December 2008
Sepsis is an important cause of neonatal death and perinatal brain damage, particularly in preterm infants. While effective antibiotic treatment is essential treatment for sepsis, resistance to antibiotics is increasing. Adjuvant therapies, such as intravenous immunoglobulin, therefore offer an important additional strategy. Three Cochrane systematic reviews of randomised controlled trials in nearly 6,000 patients suggest that non-specific, polyclonal intravenous immunoglobulin is safe and reduces sepsis by about 15% when used as prophylaxis but does not reduce mortality in this situation. When intravenous immunoglobulin is used in the acute treatment of neonatal sepsis, however, there is a suggestion that it may reduce mortality by 45%. However, the existing trials of treatment were small and lacked long-term follow-up data.
This study will assess reliably whether treatment of neonatal sepsis with intravenous immunoglobulin reduces mortality and adverse neuro-developmental outcome.
Methods and design
A randomised, placebo controlled, double blind trial. Babies with suspected or proven neonatal sepsis will be randomised to receive intravenous immunoglobulin therapy or placebo.
Babies must be receiving antibiotics and have proven or suspected serious infection AND have at least one of the following: birthweight less than 1500 g OR evidence of infection in blood culture, cerebrospinal fluid or usually sterile body fluid OR be receiving respiratory support via an endotracheal tube AND there is substantial uncertainty that intravenous immunoglobulin is indicated.
Babies are excluded if intravenous immunoglobulin has already been given OR intravenous immunoglobulin is thought to be needed OR contra-indicated.
Babies will be given either 10 ml/kg of intravenous immunoglobulin or identical placebo solution over 4–6 hours, repeated 48 hours later.
Mortality or major disability at two years, corrected for gestational age.
Data will be collected at discharge from hospital and at 2 years of age (corrected for gestation) using a parental questionnaire and a health status questionnaire completed during a face-to-face follow-up appointment with the child's paediatrician.
Current Controlled Trials ISCRTN94984750.
This protocol is for a large, simple-in-design, double blind, placebo controlled, pragmatic, multicentre randomised trial.
Hypothesis to be tested
That, in infants receiving antibiotics for clinical sepsis, the addition of non-specific, polyclonal intravenous immunoglobulin IgG (IVIG) therapy reduces mortality and major morbidity compared with antibiotics alone.
Neonatal sepsis is a major cause of mortality and morbidity and has been implicated in the causation of perinatal brain damage and cerebral palsy, both in term and preterm infants [1, 2]. Although antibiotics are the mainstay of therapy, increasing numbers of bacteria are resistant to them [3, 4]. Effective adjunctive strategies are therefore needed.
Incidence, potential impact on mortality and problems in diagnosis
In a prospective study in seven Australian neonatal intensive care units (NICUs), Isaacs and colleagues reported an annual incidence of sepsis of 6.6 per 1000 live births, of which 75% were late onset (more than 48 hours after birth). Overall hospital mortality for sepsis was 10% . In a cohort of 54 UK neonatal units in 1998, 204 (5%) of 3,963 consecutive admissions to neonatal units had a positive blood culture . Of these, 16 (8%) died. Of 3,759 (95%) babies with negative blood cultures, 95 babies died (2.5%). For very low birthweight (VLBW) infants with positive blood cultures, mortality was 14%. In a North American cohort, mortality in VLBW infants with septicaemia was 21% . However, these figures may underestimate the true incidence of neonatal sepsis. Blood cultures may often be negative if less than 1 ml of blood is sampled . Furthermore, while sepsis was the primary cause of death in most infants under 1000 g at autopsy, it was clinically undiagnosed in 61% of cases . Sepsis-specific mortality rates should therefore be interpreted with caution, as the diagnosis may often be inaccurate. More reliable evidence would be provided by randomised comparisons of the effects of specific interventions on mortality from all causes.
Potential impact of sepsis on the perinatal brain
Recent evidence suggests that sepsis is also important in the pathogenesis of neuro-developmental impairment of perinatal origin. In a case-control study of 424 births, Grether and Nelson found an association between maternal infection in labour and cerebral palsy in infants with birthweight of at least 2500 g (OR 9.3, 95% CI 3.7, 23.0).
In another case-control study of 96 term infants, levels of cytokines in neonatal blood spots were consistently higher in children diagnosed with cerebral palsy at 3 years of age than in controls, suggesting that an inflammatory response may be important in the aetiology of cerebral impairment . In preterm infants, sepsis is also associated with subsequent adverse neuro-developmental outcome . Dammann and Leviton have suggested that infection remote from the preterm brain may predispose to cerebral white matter damage with disruption of oligodendroglial myelination and disordered migration of precursors . The damage could result partly from inadequate endogenous protection from developmentally regulated factors such as oligotrophins . As antenatal and postnatal sepsis may predispose to neuro-developmental impairment and disability in term and preterm infants, these are essential measures of outcome.
Possible adjunctive treatments
Newborn infants, particularly those who are very low birthweight or preterm, are deficient in IgG, which binds to cell surface receptors, provides opsonic activity, activates fixation of complement, promotes antibody dependent cytotoxicity, improves neutrophil chemiluminescence and phagocytosis and can improve neutropenia by enhancing the release of stored neutrophils [13–16]. Intravenous immunoglobulin (IVIG) is therefore a theoretically attractive strategy, with multiple mechanisms of action. Its potential clinical relevance is confirmed by recent evidence from randomised controlled trials (see below).
In animal models of sepsis, pentoxifylline, a methylxanthine deriviative, inhibits production of Tumor Necrosis Factor (TNF), preserves micro-vascular blood flow, prevents circulatory failure and intestinal vaso-constriction and improves survival [17, 18]. It is well tolerated and decreases TNF production in adults and preterm infants with sepsis [19–21]. Two randomised controlled trials (RCTs) of pentoxifylline recruited 140 preterm infants with clinical sepsis [21, 22]. Among the 107 with positive blood cultures, pentoxifylline was associated with an 86% reduction in risk of mortality (RR 0.14, 95% CI 0.03, 0.76). Outcomes for the 33 infants not included in the analyses are not available. Pentoxifylline may be a promising therapy in neonatal sepsis.
Other adjunctive strategies for prophylaxis or treatment of neonatal sepsis are also attractive, such as use of the recombinant cytokines Granulocyte Colony Stimulating Factor (G-CSF) or Granulocyte-Macrophage Colony Stimulating Factor (GM-CSF) to prevent neutropenia . However, no systematic reviews of RCTs of these agents are yet available. In four RCTs of G-CSF therapy which recruited 125 infants with neonatal sepsis, there was a trend to reduced mortality which was not statistically significant (OR 0.6, 95% CI 0.2,1.8) [24–27].
Blood products other than immunoglobulin
White cell (granulocyte) transfusions are also a logical approach. Although preliminary clinical evidence is encouraging, there are potential risks from transmission of infection (e.g. HIV or hepatitis) or from graft-versus-host disease, and the technology is not widely available . Exchange transfusion with fresh whole adult blood appeared effective in one RCT of 22 septicaemic infants, but may also transmit infection . In another RCT, in 776 infants of less than 32 weeks' gestation, there was no evidence that prophylactic fresh frozen plasma reduced the risks of mortality from all causes or of disability in survivors at 2 years .
Overall, therefore, the evidence suggests that IVIG therapy is one of the most promising strategies in neonatal sepsis and should be assessed in a definitive RCT.
Non-specific versus specific immunoglobulin
This trial will use non-specific, polyclonal IVIG (normal human IgG immunoglobulin) produced from plasma from non-UK donors. It was decided that specific IVIG would not be used and that there was no necessity to characterise the specific antibacterial profile of the non-specific IVIG for the following reasons:
1) Previous RCTs of non-specific, polyclonal IVIG in neonates and adults did not characterise any specific aspects of antibacterial function in the products used. There is therefore no reference laboratory data against which to judge the possible antibacterial efficacy of polyclonal IVIG.
2) As the mechanism of action of IVIG is likely to be multifactorial, the precise aspects of antibacterial function which should be assessed are speculative.
3) Despite the production of monoclonal antibodies with demonstrable in vitro and in vivo antibacterial function in laboratory studies, they have not been associated with reductions in mortality in RCTs. There is therefore no evidence that laboratory studies which attempt to characterise specific aspects of antibacterial function in IVIG products would be more predictive of clinical efficacy than the existing clinical evidence from RCTs in support of non-specific, polyclonal IVIG therapy.
Results of previous randomised controlled trials
A Cochrane systematic review of the prophylactic use of non-specific IVIG in 15 RCTs with a total of 5,054 preterm or low birthweight infants has demonstrated that prophylactic, non-specific IVIG reduced sepsis (RR 0.85, 95% CI 0.74, 0.98) and was safe, with no major adverse effects, but did not demonstrate a definite reduction in mortality (RR 0.89, 95% CI 0.75, 1.05) .
Mortality in trials of IVIG for suspected infection in neonates
Relative Risk (95% CI Fixed)
0.80 [0.34, 1.86]
0.17 [0.02, 1.30]
0.62 [0.23, 1.69]
1.00 [0.41, 2.43]
0.50 [0.16, 1.53]
0.63 [0.40, 1.00]
Mortality in trials of IVIG for proven infection in neonates
Relative Risk (95% CI Fixed)
2.00 [0.19, 20.82]
0.76 [0.31, 1.89]
0.27 [0.03, 2.26]
0.95 [0.15, 6.03]
0.15 [0.01, 2.46]
0.38 [0.08, 1.78]
0.49 [0.11, 2.13]
0.55 [0.31, 0.98]
The reviewers concluded that: "The reduced mortality following treatment with IVIG for subsequently proven infection, the imprecise estimate of the effect size (number needed to treat 11, 95% CI 5.6, 100) and the borderline statistical significance for the outcome of mortality in neonates with suspected infection justify further research. Researchers should be encouraged to undertake well-designed trials to confirm or refute the effectiveness of IVIG" .
Using slightly different selection criteria and methods for analysis, Jenson and Pollock have published a systematic review of three RCTs of IVIG in neonatal sepsis in which 55 infants received IVIG and 55 received placebo or no intervention . The odds ratio for mortality in treated versus control infants was 0.173 (95% CI 0.03, 0.75). These authors reached a conclusion which many would consider premature, namely that "IVIG should be considered as part of the routine therapy of neonatal sepsis". However, it remains true that, among all the interventions currently reviewed in the Cochrane Library, IVIG therapy in neonatal sepsis is associated with one of the largest reductions in the odds of death. A further RCT of IVIG in neonatal sepsis in Brazilian neonatal units is being conducted. One of the applicants (K Haque) is an investigator of this trial. Its results will be incorporated into the current meta-analysis as soon as they are available.
Mortality in trials of IVIG for proven sepsis & septic shock in adults and children
Relative Risk (95% CI Fixed)
Standard IVIG vs placebo or no intervention, ACM
2.00 [0.19, 20.82]
De Simone 1988
0.78 [0.44, 1.39]
0.57 [0.34, 0.96]
0.72 [0.51, 1.03]
0.82 [0.40, 1.70]
1.00 [0.41, 2.43]
0.49 [0.11, 2.13]
0.73 [0.57, 0.93]
IgM-enriched IVIG vs placebo or no intervention, ACM
0.80 [0.34, 1.86]
0.17 [0.02, 1.30]
0.23 [0.05, 0.97]
0.58 [0.33, 1.04]
0.48 [0.30, 0.76]
0.64 [0.51, 0.80]
A sensitivity analysis was also conducted, including only seven trials of good and fair quality. This also suggested a decreased risk of mortality (RR 0.55, 95% CI 0.40, 0.79).
The same Cochrane review also explored the effect of monoclonal antibodies; the relative risks for both anti-endotoxins and anti-cytokines were similar and were of borderline statistical significance (anti-endotoxins RR 0.93, 95% CI 0.85, 1.02; anti-cytokines RR 0.92, 95% CI 0.86, 0.99). The authors concluded that although there was evidence that non-specific IVIG appears to be beneficial "large, multi-centre studies are needed to confirm the effectiveness of polyclonal IVIGs in reducing mortality in patients with sepsis. These are particularly indicated for neonatal sepsis, where evidence for benefit is still conflicting".
Safety: No evidence of transmission of blood borne viruses or prion disease
The risk of transmissible infection by blood products remains a potent source of anxiety for many clinicians and patients. However, IVIG produced to modern standards of quality control is one of the safest blood products available. There have been no reports of transmission of viruses or prions by the IVIG to be used in this study.
Methods of production including ethanol fractionation, and the use of the pH4/pepsin virus inactivation procedure, in addition to the use of plasma originating from non-UK donors have reduced the risk of transmission of infection to an absolute minimum [46, 47].
In particular, for prion disease, leucocytes represent the main source of infectivity in Creutzfeld-Jacob disease. Owing to the physico-chemical characteristics of the abnormal prion protein, the process of partitioning and filtration during fractionation further reduces the risk of transmission in IVIG . This theoretical risk must be considered in the context of the significantly increased risk of mortality and morbidity in infants eligible for the study.
As a further safeguard, fractionation pools of IVIG are tested with PCR (polymerase chain reaction) for known blood borne viruses.
Safety: No evidence of haemolysis related to T activation of red cells
Bacteria such as Clostridia can strip neuraminic acid residues from the red cell membrane, exposing the T antigen (T activation). Adult plasma contains anti-T antibodies, so transfusing newborn infants whose red cells are T activated with whole blood, unwashed red cells or unselected plasma may lead to polyagglutination and haemolysis . However, anti-T antibodies are predominantly IgM immunoglobulins a fraction which is removed from the IVIG used in this study . T activation is not a contra-indication to its use in neonatal sepsis.
Although neonatal haemolysis has been noted in association with IVIG, it was not clinically significant . The UK Committee on Safety of Medicines has received no reports of neonatal haemolysis or other adverse reactions in association with IVIG over a 30 year period until the present (personal communication, September 1999).
IVIG is not currently widely used for prophylaxis or treatment of neonatal sepsis in UK NICUs. In 1997, a postal survey of all paediatricians who were members of the British Association of Perinatal Medicine was undertaken into practice in the investigation and treatment of neonatal sepsis . Of the 181 (66%) who responded, only 13 (7%) used IVIG routinely as adjuvant therapy alongside antibiotics.
There is good preliminary evidence that IVIG therapy may reduce mortality in severe neonatal sepsis. However, there is no information on longer term quality of survival, the number of babies included in the existing systematic reviews is small and the effect size seems larger than would be anticipated. As a consequence a reliable multicentre trial is needed to provide definitive evidence that IVIG therapy for severe neonatal sepsis is or is not of benefit, with mortality or major morbidity as the outcome. IVIG is not yet widely used as routine therapy. There remains, therefore, a window of opportunity to perform such a trial before an intervention which has been inadequately assessed begins to be incorporated into routine practice.
Methods and design
Hospitals will be eligible for entry if they can provide neonatal intensive or special care, can achieve satisfactory rates of follow up at two years and would be able to institute the routine use of adjuvant IVIG for babies with sepsis if the trial demonstrates evidence of benefit.
Babies are eligible if:
1. They are receiving antibiotics and have proven or suspected serious infection
2. They have at least one of following:
• birthweight less than 1500 g OR
• evidence of infection in blood culture, CSF or usually sterile body fluid OR
• respiratory support via an endotracheal tube
3. There is substantial uncertainty that IVIG is indicated
Exclusion criteria are:
1. IVIG has already been given
2. IVIG is thought to be needed or contra-indicated (e.g. because of severe congenital abnormality or contra-indications in the manufacturer's licensed product information sheet)
Recruitment and Trial Entry
Recruitment will depend on good teamwork, knowledge and confidence among all clinicians, particularly front line nursing and medical staff, so that parents receive appropriate information about the study before entry and throughout their baby's stay. The ORACLE study recruited over 11,000 infants in 161 centres [53, 54]. Experience from that trial suggests that it is helpful if nurses and doctors understand the study background, see clinical research as an integral part of neonatal care contributing to future quality of care, and if a named nurse is appointed and trained as a local Trial Co-ordinator. If those caring for the baby are well informed about the study, they can discuss it without transmitting anxiety. Indeed, parents are likely to feel less anxious if given the opportunity to discuss the options for their baby's treatment in the context of the study with knowledgeable staff.
The named nursing and medical representative in each unit will therefore receive opportunities for training, regular information and support to enable them to orientate and update new and established nursing and medical staff. The protocol, printed materials and relevant new research will be widely available and staff will be kept informed by newsletters, personal visits and the study website http://www.npeu.ox.ac.uk/INIS.
Many babies receive antibiotics for sepsis during their stay in a neonatal unit. Some babies may be given antibiotics more for prophylaxis than for treatment of suspected sepsis. The threshold for considering entry into INIS must be that there is a clinical suspicion that this baby has sepsis. Once an infant is considered eligible, it is important that enrolment takes place as soon as practically possible.
All parents should routinely be given an information leaflet about INIS by the nursing staff when their baby is admitted to the neonatal unit. This will include details of their local medical and nursing contact with whom they can discuss the study . If their baby becomes eligible they will be asked for consent to participate in the study, and later follow up, by the most appropriate member of staff available, in person or by telephone. If they consent in person a copy of the signed consent form will be given to the parent(s). If telephone consent is considered necessary and appropriate by the recruiting clinician, a 'Telephone Consent' form will be completed. This form should then be read and signed by the parent(s) at their next visit to the hospital. Once this has happened, a copy of the consent form will be given to the parent(s).
The practical arrangements for random allocation to trial groups will be as simple as possible, based on that used in the ORACLE study [53, 54]. Staff will open the next sequentially numbered study pack kept in the neonatal unit, which contains all the materials necessary to give a course of study drug.
an intravenous infusion of IVIG of 10 ml per kg, repeated after 48 hours.
an intravenous infusion of 10 ml per kg of 0.2% albumin solution (placebo) repeated after 48 hours.
Both infusions are of identical appearance: they are colourless and froth on agitation [see Additional File 1].
Administration of treatment
The IVIG or placebo infusion will be given according to the manufacturer's instructions, over about 4–6 hours. A second, similar, dose will be given at or around 48 hours after the first dose. No further IVIG or placebo should be given, in this or any subsequent episodes of sepsis.
All other aspects of neonatal management will be left to the discretion of the paediatrician responsible for care. No special investigations and no delays of discharge will be required.
Measurement of Outcome
Primary outcome measure
1. Mortality or major disability at two years, corrected for gestational age
Secondary short term outcomes
2. Mortality, chronic lung disease or major cerebral abnormality before hospital discharge, significant positive culture after trial entry, pneumonia, necrotising enterocolitis, duration of respiratory support
Secondary long term outcomes
3. Mortality before two years, major disability at 2 years, non-major disability at 2 years
Health service utilisation
4. Length of hospital stay and number of hospital admissions
Hospital mortality, chronic lung disease, major cerebral abnormality and length of stay will be assessed from case notes. Major disability at two years will be assessed by questionnaires sent to the child's parents and health care professionals. Major disability will be defined according to the criteria set out in the National Perinatal Epidemiology Unit (NPEU) and Oxford Regional Health Authority document and will include any major disability in the following domains: neuromotor function, seizures, auditory function, communication, visual function, cognitive function and other physical disability [56, 57].
The parental questionnaire incorporates the parent report component of the Parent Report of Children's Abilities (PARCA), which was used in the 2 year follow up of the MRC funded UKOS trial . This shortened version of the PARCA was acceptable to parents, with a high response rate in UKOS, and is currently being validated by the UKOS team. The PARCA score (both parent report and parent administered components) has been validated and was found to predict performance on the Mental Development Index of the Bayley Scales of Infant Development II [58, 59]. The overall score for the modified PARCA will give a measure of verbal and non-verbal cognitive abilities. The parents' questionnaire also includes questions about temperament, which may give an early indication of behavioural and attentional difficulties, and also includes questions about respiratory function, hearing, vision, hospital admissions, relevant diagnoses and current function in a number of domains, allowing categorisation of disability as major or non-major.
Experience of other trials in this area at the NPEU, Oxford and elsewhere, suggests that it is possible to determine early neonatal events for all babies recruited. Loss to follow-up after hospital discharge of the child is more problematic.
There are likely to be few children who cannot be traced in the UK either through the hospital of recruitment or the NHS Central Register. Similar high rates of follow-up will be expected in countries participating outside the UK.
An intention to treat analysis will be performed comparing the outcome of all children allocated IVIG with all those allocated placebo, regardless of what treatment was received, or how complete that treatment was. Statistical analysis will calculate the relative risk of an outcome in the IVIG group compared with the placebo group along with a 95% confidence interval. For subgroup analyses, 99% confidence intervals will be calculated to take account of the number of comparisons.
Ten subgroup analyses will also be undertaken, stratifying by the factors described below.
1. Birth weight. Infants of very low birth weight (VLBW: < 1500 g) v infants with birth weight ≥ 1500 g
2. Small for gestational age infants (< 10th centile) v infants ≥ 10th centile
Gestational age at birth: < 26 weeks, 26+0 to 27+6 weeks, 28+0 to 29+6 weeks, 30+0 weeks or more
4. Gender: male vs female
5. Maternal chorioamnionitis: infants born at < 30 weeks gestation to women with clinical chorioamnionitis v infants born at < 30 weeks gestation with no clinical chorioamnionitis v infants born at ≥ 30 weeks
6. Elevated maternal CRP: infants born at < 30 weeks gestation to women with elevated CRP (> 80 mg/l) v infants born at < 30 weeks gestation with no elevated maternal CRP v infants born at ≥ 30 weeks
7. Preterm birth and duration of membrane rupture: Born at < 37 weeks and membranes ruptured for < 24 hours, 24–48 hours or > 48 hours versus born at ≥ 37 weeks
- (i)Clinical evidence of high mortality risk: looking seriously ill or inactive, and has:
capillary refill time > 3 seconds OR
bowel perforation or definite necrotising enterocolitis OR
prolonged bleeding from puncture sites OR
ventilated, SaO2/FiO2 ratio or PaO2/FiO2 ratio consistent with > 15% mortality risk for gestation OR
pH consistent with > 15% mortality risk for gestation. [SaO2/FiO2 ratio, PaO2/FiO2 ratio and pH consistent with > 15% mortality risk will be extrapolated from oxygenation and pH data in a prospective cohort of 14,000 infants (UK Neonatal Staffing Study) by methods similar to that used in the development of the MRC funded CRIB score ]
- (ii)Intermediate mortality risk: not satisfying criteria for high risk, but has:
Total white cell count < 5 × 109/l OR
CRP above 15 mg/l OR
platelet count < 50 × 109/l OR
organism(s) isolated in blood or usually sterile site OR
pneumonia on chest X-ray OR
CSF consistent with bacterial meningitis
Other: not satisfying criteria for high or intermediate risk
- (i)Early onset infection (non-contaminant organisms isolated from culture sent before 48 hours):
group B streptococcal disease
- (ii)Late onset infection (non-contaminant organism isolated from culture sent after 48 hours):
gram positive organisms except coagulase negative staphylococcus
coagulase negative staphylococcus
gram negative organisms
10. Type of IVIG. This subgroup analysis will analyse separately babies recruited in hospitals using the different IVIG products included in INIS. This subgroup analysis will include baseline characteristics and treatments after randomisation as well as outcomes.
Interim analyses: the Data Monitoring and Ethics Committee
For the trial, a Data Monitoring and Ethics Committee (DMEC) has been established. This is independent of the trial organisers and meets at least once per year. During the period of recruitment to the trial, interim analyses are supplied, in strict confidence, to the DMEC, together with any other analyses the DMEC may request. In the light of interim data, and other evidence from relevant studies (including updated overviews of the relevant randomised controlled trials), the DMEC will inform the Trial Steering Committee, if in their view: i) there is proof beyond reasonable doubt that the data indicate that any part of the protocol under investigation is either clearly indicated or contra-indicated, either for all infants or for a particular subgroup of trial participants, or ii) it is evident that no clear outcome will be obtained. Decision to inform the Trial Steering Committee in either of these circumstances will in part be based on statistical considerations.
Appropriate criteria for proof beyond reasonable doubt cannot be specified precisely. A difference of at least 3 standard deviations in the interim analysis of a major endpoint may be needed to justify halting, or modifying, such a study prematurely. If this criterion were to be adopted, it would have the practical advantage that the exact number of interim analyses would be of little importance, and so no fixed schedule is proposed .
Unless modification or cessation of the protocol is recommended by the DMEC, the Trial Steering Committee, collaborators and administrative staff (except those who supply the confidential information) will remain ignorant of the results of the interim analysis. Collaborators and all others associated with the study may write through the trial office to the DMEC to draw attention to any concern they may have about the possibility of harm arising from the treatment under study, or about any other matters that may be relevant.
The membership of the Data Monitoring and Ethics Committee is :
Professor Adrian Grant (Chair)
Professor Forrester Cockburn
Professor Deborah Ashby
Mrs Hazel Thornton
Professor Neena Modi
Dr Brian McClelland
Sample Size and Feasibility
Positive blood culture rates
Number with positive cultures
Mortality (all causes)
Number with negative cultures
Mortality (all causes)
Range of estimated sample sizes
Mortality or major morbidity in control group
Mortality or major morbidity in IVIG group
Relative risk reduction
Total sample size required to detect difference with 95% confidence
About 5,000 infants will be needed to demonstrate moderate reductions in mortality or survival with major developmental delay with adequate power. Over a three year recruitment period, assuming that 7–10% of all admissions are diagnosed with clinical sepsis and considered eligible for recruitment, 150 NICUs with an average of 300 admissions per year will be required to achieve the recruitment target, assuming a 40–50% rate of recruitment of eligible infants. Neonatal units will initially be recruited in the UK, Australia and New Zealand. A census of all 186 UK neonatal intensive care units (NICUs) and 60 special care baby units in 1996 (which is a 100% response rate) found that the median number of admissions per year per NICU was 317 . If a broadly representative sample of about half of all UK NICUs and SCBUs participate, then over 50% of the projected recruitment rate for the trial will be possible within the UK, leaving the additional 50% to be recruited from the rest of the world.
This study reflects the philosophy that the only practicable way to achieve comparisons which are sufficiently large to minimise the risk of being seriously misled by the play of chance is to design trials that are extremely simple and flexible .
Experience in the OSIRIS and ORACLE studies suggests that a large, simple trial of this scale of a potentially important intervention is feasible [63, 53, 54]. Furthermore, systematic reviews of RCTs of IVIG therapy in neonatal sepsis suggest a substantial reduction in mortality. This contrasts with the systematic reviews of RCTs of antibiotics in threatened preterm birth which led to the ORACLE study, as these showed no evidence of a difference in neonatal mortality. This preliminary evidence that IVIG may reduce mortality may further enhance the appeal of the study.
The estimate of the incidence of the outcome (the event rate) for the trial is imprecise, particularly as the threshold at which clinicians will enter patients cannot be estimated. If clinicians enter babies where the likelihood of serious sepsis is lower then the event rate will also be lower. If clinicians restrict entry to only those babies who are very sick, then the event rate will be high. Either of these two scenarios is reasonable because it will define a population to which the trial result can be generalised. However, it does mean that until the trial has recruited sufficient numbers of babies it will not be possible to determine the optimum trial sample size with any certainty. As a consequence the trial sample size currently represents the minimum size desirable. Assuming the trial recruits for three years, the maximum number of babies which can be recruited during this time will be recruited and it is possible that this number may exceed 5,000. During recruitment to the trial the accumulating data will be seen by an independent Data Monitoring and Ethics Committee at least once per year (see above) and they will advise the Trial Steering Committee whether the trial has answered the clinical question being addressed. If not, the trial will continue to recruit until 5,000 babies have been recruited, or until funding is exhausted.
To safeguard the scientific integrity of the trial, data from this study should not be presented in public or submitted for publication without requesting consent from the Trial Steering Committee (see Organisation below). The success of the trial depends on the collaboration of a large number of doctors and nurses. For this reason, chief credit for the results will be given not to the committees or central organisers but to all who have collaborated in the study. Acknowledgement will include all members of the trial committees, the data co-ordinating centre, trial staff, and local co-ordinators at all collaborating centres. Authorship at the head of the paper will take the form; "The INIS Study Collaborative Group". This is the preferred option, as it avoids giving undue prominence to any individuals. All contributors to the study will be listed at the end of the report, with their contribution to the study identified.
The protocol received Multi-Centre Research Ethics Committee (MREC) approval from the Anglia and Oxford MREC on 16 October 2000 (Reference number: 00/05/53). Each country will obtain ethics approval prior to recruitment commencing.
Trial Steering Committee (see Table 6)
The Trial Steering Committee (TSC) provides overall supervision of the trial on behalf of the Medical Research Council. Its terms of reference are:
To monitor and supervise the progress of the trial towards its interim and overall objectives
2. To review at regular intervals relevant information from other sources (e.g. related trials)
3. To consider the recommendations of the Data Monitoring and Ethics Committee
4. In the light of 1, 2 and 3 above, to inform the MRC Council and relevant MRC Research Boards on the progress of the trial
The membership of the Trial Steering Committee is:
Professor Richard Cooke
Professor of Neonatal Medicine
Dr Tim Neal
Consultant in Medical Microbiology
Dr Gorm Greisen
Professor Douglas G Altman
Dr William Tarnow-Mordi
Professor Peter Brocklehurst
Meetings of the TSC will take place at least once per year.
Study Investigators' Group
The Investigators' Group will consist of the trial investigators, representatives of specific groups whose expertise is necessary for the trial, and investigators of any ancillary studies. This group will supervise the practical aspects of the trial's conduct. It will resolve problems brought to it by the Project Management Group (see below) and will be responsible for organising reporting and dissemination of the trial's results.
Project Management Group
The Project Management Group will oversee the day-to-day running of the trial.
The responsibilities of the Project Management Group include:
i) Recruitment of participating centres
ii) Distribution and supply of data collection forms and other appropriate documentation for the trial
iii) Data collection and management
iv) Organisation of the distribution system for the treatment packs
v) Organisation of the follow-up of children at 2 years of age, including the distribution of questionnaires, follow-up of non-responders and liaison with local 'follow-up' personnel
vi) Data entry and cleaning
vii) Data analysis
viii) Collection of adverse event data
ix) Organising and servicing the Data Monitoring and Ethics Committee
Each participating centre will identify a local medical co-ordinator and a local neonatal nurse co-ordinator (as necessary). The responsibility of the local co-ordinators will be to:
i) Be familiar with the trial
ii) Liaise with the INIS Co-ordinating Centre
iii) Ensure that all staff involved in the care of babies on the neonatal unit are informed about the trial
iv) Ensure that mechanisms for recruitment of eligible babies (including information material) are in place, monitor their effectiveness, and discuss reasons for the non-recruitment of any eligible babies with relevant staff
v) Ensure that supplies of data collection forms are always available, that they are completed and returned to the INIS Co-ordinating Centre promptly, and to deal with any queries arising
vi) Notify the INIS Co-ordinating Centre of any serious adverse events
vii) Facilitate other aspects of local collaboration as appropriate
viii) Make all data available for verification, audit and inspection purposes as necessary
ix) Ensure that the confidentiality of all information about trial participants is respected by all persons
The INIS Study has been approved and is funded by the UK Medical Research Council and the Australian National Health and Medical Research Council.
- Grether JK, Nelson KB: Maternal infection and cerebral palsy in infants of normal birth weight. JAMA. 1997, 278 (3): 207-11. 10.1001/jama.278.3.207.View ArticlePubMedGoogle Scholar
- Murphy DJ, Sellers S, MacKenzie IZ, Yudkin PL, Johnson AM: Case-control study of antenatal and intrapartum risk factors for cerebral palsy in very preterm singleton babies. Lancet. 1995, 346 (8988): 1449-54. 10.1016/S0140-6736(95)92471-X.View ArticlePubMedGoogle Scholar
- Parliamentary Select Committee: House of Lords inquiry into antimicrobial resistance. Commun Dis Rep CDR Wkly. 1998, 8 (17): 147-150.Google Scholar
- Levy SB: The challenge of antibiotic resistance. Sci Am. 1998, 278 (3): 46-53.View ArticlePubMedGoogle Scholar
- Isaacs D, Barfield CP, Grimwood K, McPhee AJ, Minutillo C, Tudehope DI: Systemic bacterial and fungal infections in infants in Australian neonatal units. Australian Study Group for Neonatal Infections. Med J Aust. 1995, 162 (4): 198-201.PubMedGoogle Scholar
- UK Neonatal Staffing study group: Patient volume, staffing and workload in relation to risk-adjusted outcomes in a stratified sample of UK neonatal intensive care units: a prospective evaluation. Lancet. 2002, 359: 99-107. 10.1016/S0140-6736(02)07366-X.View ArticleGoogle Scholar
- Fanaroff AA, Korones SB, Wright LL, Verter J, Poland RL, Bauer CR, Tyson JE, Philips JB, Edwards W, Lucey JF, Catz CS, Shankaran S, Oh W: Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J. 1998, 17 (7): 593-8. 10.1097/00006454-199807000-00004.View ArticlePubMedGoogle Scholar
- Schelonka RL, Chai MK, Yoder BA, Hensley D, Brockett RM, Ascher DP: Volume of blood required to detect common neonatal pathogens. J Pediatr. 1996, 129 (2): 275-8. 10.1016/S0022-3476(96)70254-8.View ArticlePubMedGoogle Scholar
- Barton L, Hodgman JE, Pavlova Z: Causes of death in the extremely low birth weight infant. Pediatrics. 1999, 103 (2): 446-51. 10.1542/peds.103.2.446.View ArticlePubMedGoogle Scholar
- Nelson KB, Dambrosia JM, Grether JK, Phillips TM: Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol. 1998, 44 (4): 665-75. 10.1002/ana.410440413.View ArticlePubMedGoogle Scholar
- Dammann O, Leviton A: Infection remote from the brain, neonatal white matter damage, and cerebral palsy in the preterm infant. Semin Pediatr Neurol. 1998, 5 (3): 190-201. 10.1016/S1071-9091(98)80034-X.View ArticlePubMedGoogle Scholar
- Dammann O, Leviton A: Brain damage in preterm newborns: Might enhancement of developmentally regulated endogenous protection open a door for prevention?. Pediatrics. 1999, 104 (3): 541-50. 10.1542/peds.104.3.541.View ArticlePubMedGoogle Scholar
- Baley JE: Neonatal sepsis: the potential for immunotherapy. Clin Perinatol. 1988, 15 (4): 755-71.PubMedGoogle Scholar
- Fujiwara T, Taniuchi S, Hattori K, Kobayashi T, Kinoshita Y, Kobayashi Y: Effect of immunoglobulin therapy on phagocytosis by polymorphonuclear leucocytes in whole blood of neonates. Clin Exp Immunol. 1997, 107 (3): 435-9. 10.1046/j.1365-2249.1997.d01-957.x.View ArticlePubMedPubMed CentralGoogle Scholar
- Christensen RD, Brown MS, Hall DC, Lassiter HA, Hill HR: Effect on neutrophil kinetics and serum opsonic capacity of intravenous administration of immune globulin to neonates with clinical signs of early-onset sepsis. J Pediatr. 1991, 118 (4 Pt 1): 606-14.View ArticlePubMedGoogle Scholar
- Harper TE, Christensen RD, Rothstein G, Hill HR: Effect of intravenous immunoglobulin G on neutrophil kinetics during experimental group B streptococcal infection in neonatal rats. Rev Infect Dis. 1986, 8 (Suppl 4): S401-8.View ArticlePubMedGoogle Scholar
- Masouye P, Gunning K, Pittet JF, Suter PM, Morel DR: Xanthine derivative HWA 138 attenuates hemodynamic and oxygen uptake dysfunction secondary to severe endotoxin shock in sheep. Circ Shock. 1992, 36 (2): 113-19.PubMedGoogle Scholar
- Lundblad R, Ekstrom P, Giercksky KE: Pentoxifylline improves survival and reduces tumor necrosis factor, interleukin-6, and endothelin-1 in fulminant intra-abdominal sepsis in rats. Shock. 1995, 3 (3): 210-15. 10.1097/00024382-199503000-00009.View ArticlePubMedGoogle Scholar
- Zeni F, Pain P, Vindimian M, Gay JP, Gery P, Bertrand M, Page Y, Page D, Bertrand JC: Effects of pentoxifylline on circulating cytokine concentrations and hemodynamics in patients with septic shock: results from a double-blind, randomized, placebo-controlled study. Crit Care Med. 1996, 24 (2): 207-14. 10.1097/00003246-199602000-00005.View ArticlePubMedGoogle Scholar
- Staudinger T, Presterl E, Graninger W, Locker GJ, Knapp S, Laczika K, Klappacher G, Stoiser B, Wagner A, Tesinsky P, Kordova H, Frass M: Influence of pentoxifylline on cytokine levels and inflammatory parameters in septic shock. Intensive Care Med. 1996, 22 (9): 888-93. 10.1007/BF02044112.View ArticlePubMedGoogle Scholar
- Lauterbach R, Zembala M: Pentoxifylline reduces plasma tumour necrosis factor-alpha concentration in premature infants with sepsis. Eur J Pediatr. 1996, 155 (5): 404-9. 10.1007/BF01955273.View ArticlePubMedGoogle Scholar
- Lauterbach R, Pawlik D, Kowalczyk D, Ksycinski W, Helwich E, Zembala M: Effect of the immunomodulating agent, pentoxifylline, in the treatment of sepsis in prematurely delivered infants: a placebo-controlled, double-blind trial. Crit Care Med. 1999, 27 (4): 807-14. 10.1097/00003246-199904000-00042.View ArticlePubMedGoogle Scholar
- Perez EM, Weisman LE: Novel approaches to the prevention and therapy of neonatal bacterial sepsis. Clin Perinatol. 1997, 24 (1): 213-29.PubMedGoogle Scholar
- Gillan ER, Christensen RD, Suen Y, Ellis R, Ven van de C, Cairo MS: A randomized, placebo-controlled trial of recombinant human granulocyte colony-stimulating factor administration in newborn infants with presumed sepsis: significant induction of peripheral and bone marrow neutrophilia. Blood. 1994, 84 (5): 1427-33.PubMedGoogle Scholar
- Drossou-Agakidou V, Kanakoudi-Tsakalidou F, Sarafidis K, Taparkou A, Tzimouli V, Tsandali H, Kremenopoulos G: Administration of recombinant human granulocyte colony-stimulating factor to septic neonates induces neutrophilia and enhances the neutrophil respiratory burst and beta2 integrin expression. Results of a randomized controlled trial. Eur J Pediatr. 1998, 157 (7): 583-8. 10.1007/s004310050884.View ArticlePubMedGoogle Scholar
- Schibler KR, Osborne KA, Leung LY, Le TV, Baker SI, Thompson DD: A randomized, placebo-controlled trial of granulocyte colony-stimulating factor administration to newborn infants with neutropenia and clinical signs of early-onset sepsis. Pediatrics. 1998, 102 (1 Pt 1): 6-13. 10.1542/peds.102.1.6.View ArticlePubMedGoogle Scholar
- Bedford Russell A, Emmerson A, Wilkinson N, Holland B, Sweet D, Davies E: UK multicentre prospective, randomised double blind placebo controlled trial of rhG-CSF (granulocyte colony-stimulating factor) in neonatal sepsis. Proceedings of the Royal College of Paediatrics and Child Health Spring meeting. 1999, G69-Google Scholar
- Carr R, Modi N, Doré J, El-Rifai R, Lindo Dl: A randomized, controlled trial of prophylactic granulocyte-macrophage colony-stimulating factor in human newborns less than 32 weeks gestation. Pediatrics. 1999, 103: 796-802. 10.1542/peds.103.4.796.View ArticlePubMedGoogle Scholar
- Cairo MS, Agosti J, Ellis R, Laver JJ, Puppala B, deLemos R, Givner L, Nesin M, Wheeler JG, Seth T, Ven van de C, Fanaroff A: A randomized, double-blind, placebo-controlled trial of prophylactic recombinant human granulocyte-macrophage colony-stimulating factor to reduce nosocomial infections in very low birth weight neonates. J Pediatr. 1999, 134 (1): 64-70. 10.1016/S0022-3476(99)70373-2.View ArticlePubMedGoogle Scholar
- Tarnow-Mordi W, Healy M: Distinguishing between "no evidence of effect" and "evidence of no effect" in randomised controlled trials and other comparisons. Arch Dis Child. 1999, 80 (1): 210-13.View ArticlePubMedPubMed CentralGoogle Scholar
- Cairo MS, Worcester CC, Rucker RW, Hanten S, Amlie RN, Sender L, Hicks DA: Randomized trial of granulocyte transfusions versus intravenous immune globulin therapy for neonatal neutropenia and sepsis. J Pediatr. 1992, 120 (2 Pt 1): 281-5.PubMedGoogle Scholar
- Tollner U, Pohlandt F, Heinze F, Henrichs I: Treatment of septicaemia in the newborn infant: choice of initial antimicrobial drugs and the role of exchange transfusion. Acta Paediatr Scand. 1977, 66 (5): 605-10. 10.1111/j.1651-2227.1977.tb07955.x.View ArticlePubMedGoogle Scholar
- Northern Neonatal Network Nursing Initiative: Randomised trial of prophylactic early fresh-frozen plasma or gelatin or glucose in preterm babies: outcome at 2 years. Northern Neonatal Nursing Initiative Trial Group. Lancet. 1996, 348 (9022): 229-32. 10.1016/S0140-6736(95)12506-X.View ArticleGoogle Scholar
- Ohlsson A, Lacy JB: Intravenous immunoglobulin for preventing infection in preterm and/or low-birthweight infants (Cochrane Review). The Cochrane Libary. 2003, Oxford: Update Software, 2Google Scholar
- Ohlsson A, Lacy JB: Intravenous immunoglobulin for suspected or subsequently proven infection in neonates (Cochrane Review). The Cochrane Libary. 2003, Oxford: Update Software, 2Google Scholar
- Samatha S, Jalalu MP, Hegde RK, Vishwanath D, Maiya PP: Role of IgM enriched intravenous immunoglobulin as an adjuvant to antibiotics in neonatal sepsis. Karnataka Pediatr J. 1997, 11: 1-6.Google Scholar
- Shenoi A, Nagesh NK, Maiya PP, Bhat SR, Rao SDS: Multicentre randomized placebo controlled trial of therapy with intravenous immunoglobulin in decreasing mortality due to neonatal sepsis. Indian Pediatr. 1999, 36: 1113-18.PubMedGoogle Scholar
- Mancila Ramirez J, Gonzales R, Castellanos C, Garcia P, Santos JI: Immunoglobulina intravenosa en el tratamiento de septicemia neonatal. Bol Med Hosp Infant Mex. 1992, 49: 4-11.Google Scholar
- Weisman LE, Stoll B, Kueser T, Rubio TT, Frank CG, Heiman HS, Subramanian KN, Hankins CT, Anthony BF, Cruess DF: Intravenous immune globulin therapy for early-onset sepsis in premature neonates. J Pediatr. 1992, 121 (3): 434-43. 10.1016/S0022-3476(05)81802-5.View ArticlePubMedGoogle Scholar
- Chen JY: Intravenous immunoglobulin in the treatment of full-term and premature newborns with sepsis. J Formosa Med Assoc. 1996, 95: 839-44.Google Scholar
- Haque KN, Zaidi MH, Bahakim H: IgM-enriched intravenous immunoglobulin therapy in neonatal sepsis. Am J Dis Child. 1988, 142 (12): 1293-6.PubMedGoogle Scholar
- Erdem G, Yurdakok M, Tekinalp G, Ersoy F: The use of IgM-enriched intravenous immunoglobulin for the treatment of neonatal sepsis in preterm infants. Turk J Pediatr. 1993, 35 (4): 277-81.PubMedGoogle Scholar
- Sidiropoulos D, Boehme U, von Muralt G, Morell A, Barandun S: Immunoglobulin supplementation in prevention or treatment of neonatal sepsis. Pediatr Infect Dis. 1986, 5 (3 Suppl): S193-4.PubMedGoogle Scholar
- Jenson HB, Pollock BH: Meta-analyses of the effectiveness of intravenous immune globulin for prevention and treatment of neonatal sepsis. Pediatrics. 1997, 99 (2): E2-10.1542/peds.99.2.e2.View ArticlePubMedGoogle Scholar
- Alejandria MM, Lansang MA, Dans LF, Mantaring JBV: Intravenous immunoglobulin for treating sepsis and shock (Cochrane Review). The Cochrane Libary. 2003, Oxford: Update Software, 2Google Scholar
- Turner ML, Ironside JW: New-variant Creutzfeldt-Jakob disease: the risk of transmission by blood transfusion. Blood Rev. 1998, 12 (4): 255-68. 10.1016/S0268-960X(98)90007-8.View ArticlePubMedGoogle Scholar
- Turner M: The impact of new variant Creutzfeldt-Jakob disease on blood transfusion practice. Br J Haematol. 1999, 106: 842-50. 10.1046/j.1365-2141.1999.01457.x.View ArticlePubMedGoogle Scholar
- Foster PR: Assessment of the potential of plasma fractionation processes to remove causative agents of transmissible spongiform encephalopathy. Transfus Med. 1999, 9 (1): 3-14. 10.1046/j.1365-3148.1999.009001003.x.View ArticlePubMedGoogle Scholar
- Osborn DA, Lui K, Pussell P, Jana AK, Desai AS, Cole M: T and Tk antigen activation in necrotising enterocolitis: manifestations, severity of illness, and effectiveness of testing. Arch Dis Child Fetal Neonatal Ed. 1999, 80 (3): F192-F197.View ArticlePubMedPubMed CentralGoogle Scholar
- Mollison P, Engelfriet C, Contreras M: Blood Transfusion in Clinical Medicine. 1997, Oxford: Blackwell, 10Google Scholar
- Haque KN, Remo C, Bahakim H: Comparison of two types of intravenous immunoglobulins in the treatment of neonatal sepsis. Clin Exp Immunol. 1995, 101 (2): 328-33.View ArticlePubMedPubMed CentralGoogle Scholar
- Haque K: Management of neonatal sepsis in the UK: a national audit of practice. J of Clinical Excellence. 2000, 2: 81-85.Google Scholar
- Taylor D, Kenyon S, Peto R, Tarnow-Mordi W: Preterm premature rupture of membranes. Lancet. 1996, 347 (8995): 204-10.1016/S0140-6736(96)90391-8.View ArticlePubMedGoogle Scholar
- Taylor D, Kenyon S, Tarnow-Mordi W: Infection and preterm labour [comment]. Br J Obstet Gynaecol. 1997, 104 (12): 1338-40.View ArticlePubMedGoogle Scholar
- Tarnow-Mordi W: Consent, information, and neonatal research [letter]. Lancet. 1999, 353 (9163): 1530-1. 10.1016/S0140-6736(05)67227-3.View ArticlePubMedGoogle Scholar
- Johnson A: Follow up studies: a case for a standard minimum data set. Arch Dis Child Fetal Neonatal Ed. 1997, 76 (1): F61-F63.View ArticlePubMedPubMed CentralGoogle Scholar
- Disability and perinatal care: measurement of health status at two years. A report of two working groups convened by the National Perinatal Epidemiology Unit and the former Oxford Regional Health Authority. 1994, Oxford: NPEUGoogle Scholar
- Saudino KJ, Dale PS, Oliver B, Petrill SA, Richardson V, Rutter M, Simonoff E, Stevenson J, Plomin R: The validity of parent-based assessment of the cognitive abilities of two-year olds. British Journal of Development Psychology. 1198, 16: 349-63.View ArticleGoogle Scholar
- Bayley N: Bayley Scales of Infant Development. 1993, San Antonio TX Psychological Corporation, 2Google Scholar
- Peto R, Pike M, Armitage P, Breslow N, Cox D, Howard S, Mantel N, McPherson K, Peto J, Smith DG: Design and analysis of randomized clinical trials requiring prolonged observation of each patient. 1. Introduction and design. Br J Cancer. 1976, 34: 585-612.View ArticlePubMedPubMed CentralGoogle Scholar
- Tucker J, Tarnow-Mordi W, Gould C, Parry G, Marlow N: UK neonatal intensive care services in 1996. Arch Dis Child Fetal Neonatal Ed. 1999, 80 (3): F233-F234.View ArticlePubMedPubMed CentralGoogle Scholar
- Peto R, Baigent C: Trials: the next 50 years. Large scale randomised evidence of moderate benefits. BMJ. 1998, 317 (7167): 1170-1.View ArticlePubMedPubMed CentralGoogle Scholar
- Early versus delayed neonatal administration of a synthetic surfactant – the judgment of OSIRIS. The OSIRIS (open study of infants at high risk of or with respiratory insufficiency – the role of surfactant) Collaborative Group. Lancet. 1992, 340 (8832): 1363-9. 10.1016/0140-6736(92)92557-V.Google Scholar
- Human Immunoglobulin for Intravenous Administration. 2000, Information Pamphlet, Scottish National Blood Transfusion ServiceGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2393/8/52/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.