Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Higher versus lower sodium intake for preterm infants

Authors' declarations of interest

Version published: 24 April 2017 Version history

https://doi.org/10.1002/14651858.CD012642

This is not the most recent version

view the current version 12 October 2023
Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To determine the effects of higher versus lower sodium supplementation in preterm infants. We will undertake three comparisons.

  1. Higher (commencing ≥ 2 mmol/kg/day) versus lower (commencing < 2 mmol/kg/day) sodium supplementation in preterm infants less than 7 days of age.

  2. Higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) or intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    1. Higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    2. Higher (≥ 5 mmol/kg/day) versus intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    3. Intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

  3. Early and late higher versus lower sodium intake.

    1. Higher (commencing ≥ 2 mmol/kg/day) or lower (commencing < 2 mmol/kg/day) sodium supplementation in preterm infants less than 7 days of age, and either:

      1. higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age; or

      2. higher (≥ 5 mmol/kg/day) versus intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age; or

      3. intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

Background

Description of the condition

Sodium imbalances may impact on neonatal outcomes including intraventricular haemorrhage, patent ductus arteriosus, respiratory distress, sensorineural hearing impairment and motor development (Moritz 2009; Oh 2012). Preterm infants are particularly at risk of early hypernatraemia (Gawlowski 2006; Jones 1976; Rees 1984; Wilkins 1992b) and later hyponatraemia (Al‐Dahhan 1983; Day 1976; Mannan 2012; Rees 1984). Hypernatraemia has been variably defined as a serum sodium > 145 mmol/L (Balasubramanian 2012) or ≥ 150 mmol/L (Baraton 2009; Harkavy 1983; Takahashi 1994); and hyponatraemia as a serum sodium < 135 mmol/L (Kloiber 1996) or < 130 mml/L (Baraton 2009; Day 1976; Takahashi 1994).

Serum sodium levels reflect water and sodium balances which change over time after birth. Initially after birth, the extracellular fluid space contracts in association with a net negative sodium and water balance with accompanying weight loss (Bauer 1989; Bauer 1991; Bauer 1993; Lorenz 1995; Shaffer 1989). A substantial portion of the loss is transepidermal water loss, particularly in extremely preterm infants (Agren 1998; Lorenz 1995; Maurer 1984). Excess water loss is associated with early hypernatraemia (Jones 1976; Rees 1984; Wilkins 1992b). Higher renal fractional excretion of sodium may ameliorate the tendency to hypernatraemia (Ross 1977), although this is limited by the relatively low glomerular filtration rate (GFR) in the first few days in extremely preterm infants (Al‐Dahhan 1983; Aperia 1981; Mannan 2012; Wilkins 1992a). Clinicians typically modify water intakes of infants in response to changes in serum sodium and other indicators of hydration, particularly in the first week.

Hyponatraemia by the end of the first week is common in very preterm infants if strategies for its prevention are not applied. A negative sodium balance was reported in 100% of infants born at less than 30 weeks' gestation, 70% at 30 to 32 weeks, 46% at 33 to 35 weeks, and 0% after 36 weeks (Al‐Dahhan 1983). Reported incidences of hyponatraemia < 130 mmol/L in very low birth weight infants increases from 25% in the first week to 65% thereafter (Baraton 2009; Day 1976; Kloiber 1996; Takahashi 1994). Various mechanisms are reported to contribute to the development of hyponatraemia in preterm infants including renal losses and needs for growth (Agostoni 2010). Preterm infants have high fractional excretion of sodium (FENa) secondary to impaired tubular sodium reabsorption (Al‐Dahhan 1983; Aperia 1981; Mannan 2012; Wilkins 1992b). Despite a relatively low GFR, it is thought the increase in GFR over time frequently exceeds the limited tubular sodium reabsorption capacity in very preterm infants (Al‐Dahhan 1983; Aperia 1981). Reported risk factors for hyponatraemia include increasing prematurity (Mannan 2012), birth weight of less than 1000 grams, feedings of fortified human milk and occurrence of an intraventricular haemorrhage (Kloiber 1996). Additional treatments including diuretics and indomethacin are also associated with hyponatraemia (Kim 2015).

Description of the intervention

Prevention of sodium imbalance in preterm infants requires specific approaches to both water and sodium intake. The effect of restricted versus liberal water intake in preterm infants is assessed in a separate Cochrane review (Bell 2014). Reflecting variations in clinical practice, water intakes were variable in the five included trials but typically commenced between 50 to 80 mL/kg/day increasing to 140 to 200 mL/kg/day, with liberal groups receiving 20 to 30 mL/kg/day higher than restricted groups. The review found restricted water intake significantly increases postnatal weight loss and reduces the risk of patent ductus arteriosus and necrotising enterocolitis, but with a trend towards increased dehydration, although this was not statistically significant.

Sodium intake in preterm infants is from a combination of parenteral and enteral sources. The European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) recommended parenteral sodium intakes specific for the phase of adaptation (Koletzko 2005). During the initial transition phase (hours to days), ESPGHAN recommends allowing for a negative net sodium balance of 2 to 5 mmol/kg; during the second intermediate phase to replete the body for electrolyte losses, replace actual water and electrolytes; and during the third stable growth phase to provide 3.0 to 5.0 mmol/kg/day (69 to 113 mg/kg/day), or higher if required, to replace losses and provide for building new tissue at intrauterine growth rates. A recent Australasian group consensus for standardised parenteral nutrition (PN) recommended minimal sodium intake of approximately 1 mmol/kg/day on day 1 using a starter PN formulation, with an increase to a maximum 4.5 mmol/kg/d in preterm and 3.5 mmol/kg/day in term infants (Bolisetty 2014). ESPGHAN recommendations for enteral intakes are 3 to 5 mmol/kg/day (Agostoni 2010). For the purposes of this review, early higher sodium supplementation is defined as commencement parenteral or enteral intakes of ≥ 2 mmol/kg/day; and later higher sodium supplementation is defined as parenteral or enteral intakes of ≥ 5 mmol/kg/day, and lower sodium supplementation as < 3 mmol/kg/day.

How the intervention might work

Although early hypernatraemia > 145 mmol/L is common in very preterm infants and may not be associated with increased morbidity (Gawlowski 2006), severe hypernatraemia has been associated with intraventricular haemorrhage in preterm infants (Lee 2015; Li 2007; Lim 2011). Severe hypernatraemia and rapid sodium correction is associated with increased mortality and convulsions in newborn infants (Bolat 2013). Early sodium restriction, particularly in parenteral fluids, may reduce the incidence and severity of hypernatraemia and its associated morbidities. Given early hypernatraemia is associated with substantial extracellular water loss, sodium requirements may be substantially influenced by water intake.

In addition, later hyponatraemia may not be a benign condition (Moritz 2009). Preterm neonates receiving restricted sodium have an increased incidence of hyponatraemia, impaired neonatal growth (Al‐Dahhan 1984; Day 1976) and worse neurodevelopment at 10 to 13 years of age compared to sodium‐supplemented infants (Al‐Dahhan 2002). Hyponatremia has also been reported to be a risk factor for sensorineural hearing loss (Ertl 2001) and cerebral palsy (Murphy 1997). After the initial period of adaptation, providing sufficient sodium intake to replace losses and provide for growth may reduce postnatal growth failure and influence development.

Why it is important to do this review

Sodium is an integral part of fluid and electrolyte therapy in newborn infants. Preterm infants have a significant burden of mortality and morbidity (Bolisetty 2015). There are concerns that imbalances of sodium intake may impact on neonatal outcomes including intraventricular haemorrhage, patent ductus arteriosus, respiratory distress, sensorineural hearing impairment and motor development (Moritz 2009; Oh 2012). In addition, extremely preterm infants continue to have high rates of postnatal growth failure (Ofek Shlomai 2014; Stoltz Sjostrom 2013). Optimising sodium intake has the potential to reduce morbidity and improve growth and long‐term outcomes of infants born preterm. There are no current systematic reviews assessing the benefits and harms of higher versus lower sodium intakes in preterm infants.

Objectives

To determine the effects of higher versus lower sodium supplementation in preterm infants. We will undertake three comparisons.

  1. Higher (commencing ≥ 2 mmol/kg/day) versus lower (commencing < 2 mmol/kg/day) sodium supplementation in preterm infants less than 7 days of age.

  2. Higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) or intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    1. Higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    2. Higher (≥ 5 mmol/kg/day) versus intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

    3. Intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

  3. Early and late higher versus lower sodium intake.

    1. Higher (commencing ≥ 2 mmol/kg/day) or lower (commencing < 2 mmol/kg/day) sodium supplementation in preterm infants less than 7 days of age, and either:

      1. higher (≥ 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age; or

      2. higher (≥ 5 mmol/kg/day) versus intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age; or

      3. intermediate (≥ 3mmol/kg/day to < 5 mmol/kg/day) versus lower (< 3mmol/kg/day) sodium supplementation in preterm infants ≥ 7 days of age.

Methods

Criteria for considering studies for this review

Types of studies

Randomised, quasi‐randomised or cluster‐randomised controlled trials.

Types of participants

Preterm infants born before 37 weeks of gestational age, and/or low birthweight infants (< 2500 grams).

Types of interventions

Nutritional supplementation that includes higher versus lower sodium supplementation in parenteral and/or enteral intake. We will exclude studies that had prespecified differential water intakes between groups.

Separate comparisons will be performed of studies that compare:

  1. early (< 7 days following birth) higher versus lower sodium intakes;

  2. late (≥ 7 days following birth) higher versus lower sodium intakes; and

  3. early and late higher versus lower sodium intake.

Definitions of sodium intake for early comparisons will be:

  • higher sodium intake: commencing ≥ 2 mmol/kg/day; and

  • lower sodium intake: commencing < 2 mmol/kg/day.

Definitions of sodium intake for late comparisons will be:

  • higher sodium intake: ≥ 5mmol/kg/day;

  • intermediate sodium intake: ≥ 3mmol/kg/day to < 5 mmol/kg/day (as per ESPGHAN recommendations: Agostoni 2010);

  • lower sodium intake: < 3mmol/kg/day.

Types of outcome measures

Primary outcomes

  1. Mortality (latest time reported up to hospital discharge)

  2. Neurodevelopmental disability at least 18 months of postnatal age (defined as neurological abnormality including cerebral palsy on clinical examination, developmental delay more than two standard deviations below population mean on a standardised test of development, blindness (visual acuity less than 6/60), or deafness (any hearing impairment requiring amplification) at any time after term corrected)

Secondary outcomes

Reported at latest time reported up to hospital discharge unless otherwise specified. For each comparison, only outcomes that are chronologically appropriate will be reported.

  1. Hyponatraemia (defined as < 130 mmol/L); early (< 7 days); late ≥ 7 days.

  2. Hypernatraemia (defined as ≥ 150 mmol/L); early (< 7 days); late ≥ 7 days.

  3. Postnatal growth failure (weight < 10th percentile) at 28 days; near term corrected age (36 to 40 weeks) or at discharge; and follow up (at least 1 year age).

  4. Late onset sepsis (≥ 1 episodes of positive bacterial culture in cerebrospinal fluid (CSF), sterile urine or blood after 48 hours age).

  5. Necrotising enterocolitis (Bell's grade ≥ 2) (Neu 2011).

  6. Patent ductus arteriosus (either haemodynamically significant (Skelton 1994) or treated with a cyclo‐oxygenase inhibitor or ligation).

  7. Chronic lung disease (respiratory support or oxygen requirement at or beyond 36 weeks' postmenstrual age (PMA)) (Shennan 1988).

  8. Bronchopulmonary dysplasia (BPD) (Jobe 2001):

    1. mild BPD, defined as a need for supplemental oxygen for ≥ 28 days but not at 36 weeks' PMA or discharge;

    2. moderate BPD defined as oxygen for ≥ 28 days plus treatment with < 30% oxygen at 36 weeks' PMA;

    3. severe BPD defined as oxygen for ≥ 28 days plus ≥ 30% oxygen and/or positive pressure at 36 weeks' PMA.

  9. Intraventricular haemorrhage (IVH) (any IVH or severe IVH (grade 3 or 4)) (Papile 1978).

  10. Periventricular leukomalacia (cystic).

  11. Term magnetic resonance imaging (MRI) brain abnormalities graded as normal, mild, moderate or severe (e.g. Inder 2003).

  12. Retinopathy of prematurity (ROP) (any ROP or severe ROP (stage 3 or 4)) (International Committee 2005).

  13. Individual components of neurodevelopment at least 18 months' postnatal age: cerebral palsy on clinical examination, developmental delay more than two standard deviations below population mean on a standardised test of development, blindness (visual acuity less than 6/60), deafness (any hearing impairment requiring amplification) at any time after term corrected age.

  14. Growth of infant:

  • days to regain birth weight;

  • maximal weight loss: gram; per cent;

  • weight gain: up to age 1 month (g/kg/day); at latest time measured (g/kg/day) (definition = from 1 month to time of discharge); to follow up beyond 12 months (kg/year);

  • linear growth:up to age 1 month (cm/week); at latest time measured (cm/week); to follow up beyond 12 months (cm/year);

  • head circumference: up to age 1 month (cm/week); at latest time measured (cm/week); to follow up beyond 12 months (cm/year);

  • change in weight z‐score: up to age 1 month; at latest time measured; to follow up beyond 12 months;

  • change in length z‐score: up to age 1 month; at latest time measured; to follow up beyond 12 months;

  • change in head circumference z‐score: up to age 1 month; at latest time measured; to follow up beyond 12 months.

Search methods for identification of studies

Electronic searches

We will search the following databases for relevant studies, using the search strategies as noted in the Appendices.

  • Cochrane Central Register of Controlled Trials (CENTRAL,Cochrane Library, current) (Appendix 1).

  • MEDLINE or PubMed (1946 to current) (Appendix 2).

  • Embase (1980 to current) (Appendix 3).

  • CINAHL (1982 to current) (Appendix 4).

Searching other resources

We will search relevant clinical trials registers such as the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/en/), Neonatal Trials Register on the Cochrane Library (neonatal.cochrane.org/), and ClinicalTrials.gov (clinicaltrials.gov/) for completed but unpublished and ongoing trials. We will search reference lists of included trials and review articles, abstracts and proceedings from the European Society of Pediatric Research, the American Academy of Pediatrics, and the Perinatal Society of Australia and New Zealand. We intend to send written enquiries to the authors of major relevant studies and experts in the field for further information or to obtain access to any unpublished trial data.

Data collection and analysis

Selection of studies

Two review authors (WC, MC) will assess titles and abstracts retrieved from the search to determine their relevance concerning the Objectives and Criteria for considering studies for this review. We will manage disagreements through discussion and/or a deciding arbiter (PB or DO). We will enter all search results into reference manager software and record studies assessed for eligibility in Review Manager 5.3 (RevMan 2014).

Data extraction and management

Two review authors (WC, MC) will design a data extraction sheet for study reports, which will be pilot tested using sample studies and revised by the other authors (ET, PB). Into this form, two authors (WC, MC) will independently extract and record key features of each study including the following details.

  • Authors.

  • Date and place of publication.

  • Study design.

  • Inclusion and exclusion criteria.

  • Setting.

  • Summary of study participant characteristics.

  • Summary of intervention and control conditions.

  • Number of participants in each arm (including dropouts).

  • Adverse events.

  • Outcome measurement and assessment time points.

  • Risk of bias (as per the domains specified in Assessment of risk of bias in included studies).

  • Any relevant additional comments reported by the study authors.

We will manage disagreements through discussion and/or a deciding arbiter (PB or DO). We will enter and present the data for each included study in a table in Review Manager 5.3 (RevMan 2014).

Assessment of risk of bias in included studies

Two review authors (WC, MC) will independently analyse each study in conjunction with the Cochrane tool for assessing risk of bias (Higgins 2011) on the following domains.

  • Selection bias.

  • Performance bias.

  • Attrition bias.

  • Reporting bias.

  • Any other bias.

We will manage any disagreements through discussion, a deciding arbiter (PB or DO) or both. We will present our assessment of risk of bias for the included studies in the 'Risk of bias' summary tables and graphs as generated through input into Review Manager 5.3 (RevMan 2014). See Appendix 5 for a more detailed description of risk of bias for each domain.

Measures of treatment effect

  1. We will present dichotomous (binary) data as a measure of relative risk by using a risk ratio (RR) with 95% confidence intervals (CIs). Where possible, we will calculate the absolute risk difference (ARD), and if statistically significant we will calculate the number needed to treat for an additional beneficial outcome (NNTB) or number needed to treat for an additional harmful outcome (NNTH) for comparison against other treatments or non‐treatment (Higgins 2011).

  2. We will present continuous data as a mean difference (MD) if the same scale is used. Alternatively, a standardised mean difference (SMD) will be calculated (that is, an average of the combined standard deviations (SDs)) in the event that each study uses a different scale, where we will assess the impact of using the highest versus the lowest of the available SDs on the overall estimate of effect. If SDs are not reported we will estimate the SD based on similar studies and use this in the meta‐analysis (Higgins 2011).

  3. We will present data reported as rates as hazard ratios. We will use rates to express outcomes over time. This may be expressed as a rate ratio, which may demonstrate outcomes more clearly than a risk ratio as it accounts for the likelihood that some participants may experience multiple events (Higgins 2011).

Unit of analysis issues

We envisage that the unit of analysis in our review will be the participant. Nonetheless, if the unit of analysis is not the same as the unit of randomisation, such as in cluster‐randomised trials, we plan to adjust for clustering by using the guidance given in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). If there are multiple measurements for the same participant (e.g. multiple reported instances of hyponatraemia or hypernatraemia or sepsis in the one patient) this will be counted as an event for each outcome for the participant.

Dealing with missing data

We intend to contact the original study authors if the information provided in the published article does not provide sufficient data or results to address the outcomes of our review.

We will perform analyses based on intention‐to‐treat (ITT) principles, whereby the missing data for randomised participants will be assumed to be treatment failures in this review. This approach of ITT analysis (that is, assuming drop‐outs as failures) may underestimate the effect of the intervention, therefore we may perform both ITT and on‐treatment (that is, non‐ITT) analyses to explore the impact of missing data on the overall outcome (Higgins 2011). Furthermore, for continuous data we will assess the impact of missing data on the overall estimate of effect by imputing missing data in the following ways: best case scenario where the missing data are considered 2 SDs greater in the intervention arm than in the control arm and worst case scenario where the missing data are considered 2 SDs less than in the control arm.

Assessment of heterogeneity

We will assess the included studies for heterogeneity through three successive steps to determine if they should be pooled with the rest of the included studies or considered separately.

  1. Two review authors (WC, MC) will independently analyse the included studies for their 'face‐value' similarities; that is, for the extent of clinical diversity (participants, interventions and outcomes), and for methodological diversity (study design and risk of bias) (Higgins 2011).

  2. We then intend to assess the included studies for statistical heterogeneity using the Chi² test with a P value of less than 0.10 being statistically significant (Higgins 2011).

  3. We then intend to calculate the I² statistic as instructed in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011); where less than 25% is likely to indicate no heterogeneity, 25% to 49% may represent minimal heterogeneity, 50% to 74% may represent moderate heterogeneity, and greater than 75% may represent substantial or high heterogeneity. The importance of the observed value of I² does depend on the magnitude and direction of the treatment effects, and strength of evidence for heterogeneity (that is, the P value from the Chi² test or the confidence interval for I²).

Assessment of reporting biases

If a sufficient number of studies have been pooled (that is, greater than 10) we plan to use a funnel plot to inspect visually the risk of publication bias, whereby more pronounced asymmetry of the funnel plot may be indicative of a substantial overestimation of the intervention effect (Higgins 2011).

Data synthesis

We will synthesise the data in accordance with the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) using one of the following methods.

  1. We will use the fixed‐effect model in the absence of statistical heterogeneity according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), where the analysis will produce an estimate of the true effect.

  2. Where cluster RCTs are included, we will use the generic inverse variance method (Higgins 2011).

If studies are clinically heterogeneous, we will not pool them in a meta‐analysis.

Quality of evidence

We will use the GRADE approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the quality of evidence for the following (clinically‐relevant) outcomes for the three comparisons between higher and lower sodium intake (early and late; early; late):

  • hyponatraemia;

  • hypernatraemia;

  • neurodevelopmental disability;

  • mortality;

  • postnatal growth failure;

  • severe intraventricular haemorrhage or periventricular leukomalacia;

  • chronic neonatal lung disease; and

  • necrotising enterocolitis.

Two authors will independently assess the quality of the evidence for each of the outcomes above. We will consider evidence from randomised controlled trials as high quality but downgrade the evidence one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates and presence of publication bias. We will use the GRADEpro 2014 Guideline Development Tool to create a ‘Summary of findings’ table to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence in one of four grades.

  1. High: we are very confident that the true effect lies close to that of the estimate of the effect.

  2. Moderate: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

  3. Low: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.

  4. Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Where significant subgroup effects are found, these will be incorporated into the 'Summary of findings' table as appropriate.

Subgroup analysis and investigation of heterogeneity

If sufficient data are available or heterogeneity exists, we intend to explore further the treatment effect in specific subgroups (restricted to the primary outcome measures), including the following.

  • Gestation age of infant: extremely preterm (born < 28 weeks gestational age); very preterm (< 32 weeks); preterm (< 37 weeks).

  • Birth weight: extremely low birth weight (≤ 1000 grams); very low birth weight (≤ 1500 grams); low birth weight (≤ 2500 grams); appropriate for gestational age (birth weight percentile > 10th centile); small for gestational age (birth weight percentile ≤ 10th centile).

  • Fluid intake: restricted water intake (≤ 60 ml/kg day 1 and ≤ 150 ml/kg day 7) versus liberal water intake (≥ 80 ml/kg day 1 and > 150 ml/kg after day 7) (Tammela 1992).

  • Stable (no significant system related comorbidity) versus sick infants (included infants with comorbidity).

  • Sodium intake varied in parenteral intake only versus varied in enteral intake only versus varied in parenteral and enteral intake.

Sensitivity analysis

If sufficient data are available, we intend to explore methodological heterogeneity through sensitivity analyses.

We will perform these by including only low risk studies with adequate allocation concealment, randomisation or blinding of treatment, and less than 10% loss to follow‐up. In order to determine the impact of risk of bias on the overall effect estimate, we plan to compare high risk of bias studies to low risk of bias studies and test for subgroup differences using the Chi² and I² statistics (Higgins 2011).