Infant isolation and cohorting for preventing or reducing transmission of healthcare‐associated infections in neonatal units
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
Primary objectives
1. To determine the effectiveness of isolation for preventing transmission of HAI or colonization with HAI‐causing pathogens in the neonatal unit.
2. To determine the effectiveness of cohorting for preventing transmission of HAI or colonization with HAI‐causing pathogens in the neonatal unit.
Secondary objectives
1. To determine the effects of isolation on neonatal mortality.
2. To determine the effects of cohorting on neonatal mortality.
3. To determine perceived or documented adverse effects of isolation or cohorting infection control measures.
We will perform subgroup analyses based on: gestational age; birth weight; patient subgroups on whether colonized or infected; cluster type; infection type; setting type in low‐ versus middle‐ versus high‐income countries; type of neonatal unit; and type of HAI.
Background
Description of the condition
Neonatal infections are the leading cause of death in neonates (Lawn 2006). Healthcare‐associated infections (HAIs) are infections that patients acquire within a healthcare setting which were not present or incubating at the time of admission. These include perinatally‐acquired infections in hospitalized neonates (CDC 2015). Due to variations in definitions and reporting, the exact global burden of HAI cannot be accurately estimated. Approximately 1.7 million HAI occur annually in hospitals in the USA and are responsible for 99,000 deaths and severe morbidity (Klevens 2007; Stone 2009; Hooven 2014). HAIs account for up to USD 45 billion annually in healthcare costs (Scott 2009). The attributable cost of one healthcare‐associated bloodstream infection ranges from USD 6000 to USD 39,000 (Payne 2004; Elward 2005). Nearly 33,000 infants are diagnosed with HAI each year in neonatal intensive care units (NICUs) in the USA (Klevens 2007). The incidence of HAI is higher in NICUs than in other intensive care units (Ford‐Jones 1989; Saiman 2006), and may be attributed to host factors including prematurity, immaturity of the immune system and impaired skin barrier, immature neonatal microbiome and the use of invasive procedures. Environmental risk factors in the NICU that influence the incidence of HAI include poor design of the unit, overcrowding, staffing inadequacy and poor infection control policies (Saiman 2006; Polin 2012; Hooven 2014)
The most frequent HAIs in the NICU are central line‐associated bloodstream infections (CLABSI), pneumonia and infections of the skin and soft tissues, urinary tract and the central nervous system (Polin 2012; Hooven 2014). HAIs, even after early diagnosis and treatment, increase the risk of mortality, adverse neurodevelopment, length of hospital stay and health resource utilization (Adams‐Chapman 2006; Bassler 2009; Schulman 2009; AAP 2012; Hocevar 2012; Manzoni 2013). Hence prevention or reduction of HAI is essential to improve clinical outcomes. Many hospitals and health care systems have implemented targets for preventing HAI as part of patient safety measures.
Description of the intervention
Infection control measures in the NICU often include transmission‐based precautions to decrease horizontal transmission from infants who are infected or colonized with infectious agents to those who are not infected or colonized. These include contact precautions, droplet precautions and airborne infection isolation precautions, determined by the mode of transmission of different infectious agents. One component of transmission‐based precautions, along with use of personal protective equipment (PPE), is isolation or cohorting of patients (Siegal 2007).
Isolation refers to the care of infants in separate rooms or other confined areas that are physically separated from other infants (Landelle 2013). Infants may be isolated because they are known or suspected to be colonized or infected with a pathogen based on clinical diagnosis, microbiologic confirmation or epidemiology, or because they are particularly at risk of acquiring a HAI (protective isolation). Isolation or cohorting may be effective in controlling nosocomial infections by preventing horizontal spread from patient to patient (e.g. for Candida and MRSA infections) (Gastmeier 2004). Centers for Disease Control and Prevention (CDC) guidelines recommend single room isolation for infections transmitted by direct contact or indirectly via equipment or other surfaces (Garner 1986). Single room isolation coupled with adherence to hand hygiene has been shown to decrease transmission of MRSA infections in the intensive care unit (Cheng 2010). However, NICUs are often faced with lack of isolation rooms and/or an inadequate number of health care personnel devoted to the care of infants in these isolation rooms. Isolation requirements may depend on factors including severity of illness in the infant, available resources and the transmissibility of the infection. Guidelines of the American Academy of Pediatrics state that it may be unnecessary to isolate a neonate (except in the case of neonatal varicella‐zoster or an epidemic of bacterial infection) under the following conditions: i) there is sufficient medical and nursing staff on duty, ii) sufficient space between stations, iii) two or more sinks for hand washing available in each nursery area and iv) continuing instruction is provided regarding the way infections spread (AAP 2012).
Cohorting is the physical segregation of infants in separate areas where newborns with similar exposures, colonization, or infections are cared for by designated staff assigned exclusively to these infants. Cohorts are created based on clinical and microbiological diagnoses, epidemiology, and mode of transmission of the infectious agent (Siegal 2007). Cohorting of infants may be useful in reducing horizontal transmission of infection or colonization by pathogens that are transmissible by contact (Koch 2003; Rosenberger 2011; Rosenberger 2012). In a study of MRSA surveillance and cohorting of colonized neonates, cohorting decreased MRSA colonization in non‐colonized neonates and decreased MRSA‐associated bloodstream infections (Kaushik 2015). Cohorting may be more feasible than single room isolation during outbreaks in the NICU setting: Cohorting does not require isolation rooms and may reduce resource utilization compared to single room isolation. However, the success of this system depends on the strict adherence of the healthcare staff to the cohort system (AAP 2012). Assigning or cohorting healthcare workers to care for patients infected or colonized with a single target pathogen aims to prevent transmission of the pathogen to uninfected patients, but is difficult to achieve in the face of staffing shortages in hospitals (Siegal 2007).
How the intervention might work
Isolation and cohorting infection control measures decrease HAI in the NICU by preventing horizontal spread of infection or colonization from infant to infant. Most data regarding efficacy and safety of infection control measures are based on studies of infection control of MRSA infections predominantly from adult and pediatric units (Karchmer 2002; Tawney 2015; Kullar 2016). Isolation and cohorting physically separates infants from potential sources of pathogens, which include other infants in the unit, healthcare staff or visitors, to decrease colonization and infection.
Isolation and cohorting may introduce potential risks to patients, which may include decreased frequency of care, decreased patient observation and parental anxiety (Landelle 2013). In adults, other investigators have reported decreased patient‐health care worker contact, changes in systems of care leading to delays, increased non‐infectious adverse events, increased symptoms of depression and anxiety, and decreased patient satisfaction (Morgan 2009; Abad 2010; Bearman 2012). Although no adverse effects due to infection control measures have been reported in neonates, potential adverse consequences in the NICU relate to decreased provider–patient interactions that could lead to inadvertent extubations, intravenous extravasations, medication errors and parent dissatisfaction.
Risks due to infection control measures need to be balanced against the benefits of decreasing horizontal transmission in the neonatal unit for optimal neonatal outcomes.
Why it is important to do this review
Healthcare‐associated infections have major implications for the preterm infant by increasing mortality, morbidity and healthcare costs (Scott 2009; Moran 2010; Donovan 2013). Infection control measures are expensive and consume valuable resources. Neonatal data indicating mortality or morbidity outcomes or cost‐effectiveness of implementing infection control measures are lacking and potential adverse effects due to infection control measures have been reported in adult studies. The evidence for the effectiveness and safety of infection control measures (cohorting or isolation) in preventing HAI in neonatal units has not been reviewed and hence this review.
Objectives
Primary objectives
1. To determine the effectiveness of isolation for preventing transmission of HAI or colonization with HAI‐causing pathogens in the neonatal unit.
2. To determine the effectiveness of cohorting for preventing transmission of HAI or colonization with HAI‐causing pathogens in the neonatal unit.
Secondary objectives
1. To determine the effects of isolation on neonatal mortality.
2. To determine the effects of cohorting on neonatal mortality.
3. To determine perceived or documented adverse effects of isolation or cohorting infection control measures.
We will perform subgroup analyses based on: gestational age; birth weight; patient subgroups on whether colonized or infected; cluster type; infection type; setting type in low‐ versus middle‐ versus high‐income countries; type of neonatal unit; and type of HAI.
Methods
Criteria for considering studies for this review
Types of studies
Trials that are randomized or quasi‐randomized at the level of the cluster (where clusters may be defined by NICU, hospital, ward or other subunits of the hospital) will be included. We will include cross‐over trials with a washout period of more than 4 months (arbitrarily defined).
Types of participants
Neonates in neonatal units which implement isolation or cohorting as infection control measures to prevent HAI.
Types of interventions
Isolation and cohorting infection control measures in the NICU performed to decrease infection transmission.
Comparisons
1. Isolation compared to no isolation performed to decrease infection transmission in the neonatal unit.
2. Cohorting compared to no cohorting performed to decrease infection transmission in the neonatal unit.
3. Isolation compared to cohorting performed to decrease infection transmission in the neonatal unit.
Operational definitions
Isolation: Isolation refers to the nursing of infants in single rooms or other confined areas that are physically separated from other patients (Landelle 2013). CDC guidelines recommend single room isolation for infections transmitted by direct contact or indirectly via equipment or other surfaces (Garner 1986). Guidelines of the American Academy of Pediatrics recommend isolation in the case of neonatal varicella‐zoster or an epidemic of bacterial infection (AAP 2012). AAP guidelines state that it may otherwise be unnecessary to isolate a neonate under the following conditions: i) there is sufficient medical and nursing staff on duty, ii) sufficient space between stations, iii) two or more sinks for hand washing available in each nursery area and iv) continuing instruction is provided regarding the way infections spread (AAP 2012).
Cohorting: Physical segregation of infants with colonization or infection in a separate area in the NICU with infants with colonization or infection with the same organism, with nursing personnel designated to looking after these cohorts until discharge or at least two negative surveillance cultures. The need for surveillance cultures on the patient or the care‐giver should be based on local infection control policy and dependent on the organism.
Types of outcome measures
Primary outcomes
Our primary outcomes will be HAI, and colonization with HAI‐causing organisms and will be measured as infection and colonization rates.
Healthcare‐associated infections (HAI) are infections people acquire within a healthcare setting that were not present or incubating at the time of admission. These include perinatally‐acquired infections in hospitalized neonates (CDC 2015).
Definitions of outcome measures
1. Infection rate will be presented as number of infections per 1000 patient days
2. Colonization rate will be presented as number of colonizations per 1000 patient days.
For both infection and colonization rates, the denominator will be all infants in the neonatal unit irrespective of colonization or infection status and colonization and infection rates will be tabulated separately.
Bloodstream infections (BSI): Microbiologically confirmed bloodstream infection (bacterial or fungal) with growth in blood cultures or positive by a molecular test including polymerase chain reaction, evaluated during hospital stay and evaluated as positive or negative.
Meningitis: Microbiologically confirmed meningitis (bacterial or fungal) with detection of pathogen in cerebrospinal fluid cultures or positive by a molecular test including polymerase chain reaction, evaluated during hospital stay and evaluated as positive or negative.
Urinary tract infection (UTI): The diagnosis is based on positive culture from a urine specimen collected by catheterization or suprapubic aspiration (urine collected in sterile bags has a high rate of false positives and hence is not recommended). Any growth in a suprapubic aspirate specimen (one colony is equivalent to 1000 CFU/ml) and a colony count of more than 50,000 CFU/ml or between 10,000 to 50,000 with associated pyuria will be defined as UTI (O'Donovan 2016).
Skin and soft tissue infections (SSI) as defined by CDC 2015: ''Infection occurs within 30 days after any operative procedure AND involves only skin and subcutaneous tissue of the incision AND patient has at least one of the following: a. purulent drainage from the superficial incision, b. organisms identified from an aseptically‐obtained specimen from the superficial incision or subcutaneous tissue by a culture or non‐culture based microbiologic testing method which is performed for purposes of clinical diagnosis or treatment, c. superficial incision that is deliberately opened by a surgeon, attending physician or other designee and culture or non‐culture based testing is not performed. AND patient has at least one of the following signs or symptoms: pain or tenderness; localized swelling; erythema; or heat. A culture or non‐culture based test that has a negative finding does not meet this criterion. d. diagnosis of a superficial incisional SSI by the surgeon or attending physician or other designee.'' (www.cdc.gov/nhsn/xls/icd10‐pcs‐pcm‐nhsn‐opc.xlsx).
Ventilator‐associated pneumonia (VAP) defined as per CDC 2015 by using a combination of imaging, clinical and laboratory criteria: ''Worsening gas exchange (e.g. oxygen desaturations, increased oxygen requirements, or increased ventilator demand)AND at least three of the following:• Temperature instability • Leukopenia (≤4000 WBC/mm3) or leukocytosis (> 15,000 WBC/mm³) and left shift (> 10% band forms) • Increased respiratory secretions or increased suctioning requirements • Apnea, tachypnea, nasal flaring with retraction of chest wall or nasal flaring with grunting • Wheezing, rales, or rhonchi • Cough • Bradycardia (<100 beats/min) or tachycardia (> 170 beats/min).''
(www.cdc.gov/nhsn/PDFs/pscManual/6pscVAPcurrent.pdf)
Central line‐associated BSI (CLABSI) defined as per CDC 2015. ''A laboratory‐confirmed bloodstream infection (LCBI) where central line (CL) or umbilical catheter (UC) was in place for >2 calendar days on the date of event AND the line was also in place on the date of event or the day before. If a CL or UC was in place for >2 calendar days and then removed, the date of event of the LCBI must be the day of discontinuation or the next day to be a CLABSI.''
(www.cdc.gov/nhsn/pdfs/pscmanual/4psc_clabscurrent.pdf).
Secondary outcomes
-
All‐cause mortality during hospital stay at 28 days of age.
-
Length of hospital stay measured in days.
-
Potential adverse effects of isolation and cohorting infection control measures include:
a. Decreased patient‐provider interactions that are measured by observation during the study period, measured as number of episodes or decreased time of interaction.
b. Parent dissatisfaction measured by a validated survey instrument (Latour 2012).
c. Other non‐infectious adverse events such as medication errors (measured in number of errors reported by the healthcare staff or the hospital patient safety/risk staff).
d. Extubations measured as numbers during the study period or intravenous infiltrations or complications measured by number of events during study period.
Search methods for identification of studies
Electronic searches
Please refer to Cochrane Neonatal Review Group's search strategy. Relevant trials in any language will be searched in the following databases:
1. Cochrane Central Register of Controlled Trials (CENTRAL, in the Cochrane Library)
2. Electronic journal reference databases:
MEDLINE via PubMed (1966 to present)
Embase (1980 to present)
CINAHL (1982 to present)
The detailed search strategy for MEDLINE via PubMed is given in Appendix 1. This will be adapted to suit Embase, CINAHL and CENTRAL.
Searching other resources
1. Abstracts of conferences: proceedings of Pediatric Academic Societies (American Pediatric Society, Society for Pediatric Research and European Society for Paediatric Research) will be searched from 1990 in Pediatric Research and www.abstracts2view.com/pas/ (2000 to present).
2. Ongoing trials will be searched with the search engines provided at the web sites www.clinicaltrials.gov, www.controlled‐trials.com, www.who.int/ictrp and www.anzctr.org.au/TrialSearch.aspx.
3. Authors who published in this field will be contacted for possible unpublished studies.
4. Additional searches will be made from the reference lists of identified clinical trials and in the review authors' personal files.
Data collection and analysis
We will use the standardized method of Cochrane Neonatal for conducting a systematic review (neonatal.cochrane.org/en/index.html).
Selection of studies
Two review authors (MP and RD) will independently assess the titles and the abstracts of studies identified by the search strategy for eligibility for inclusion in this review. If we cannot do this reliably by title and abstract, then we will obtain the full text version assessment. We will resolve any differences by mutual discussion. We will obtain a full text version of all available studies for quality assessment.
Data extraction and management
We will use pre‐designed forms for trial inclusion and exclusion, data extraction and for requesting additional published information from authors of the original reports. We will independently extract data using specifically designed paper forms for identified eligible trials. We will compare the extracted data for differences which will then be resolved by discussion.
Assessment of risk of bias in included studies
We will use the standardized review methods of the Cochrane Neonatal Review Group (CNRG) to assess the methodological quality of the studies. Review authors will independently assess the quality of the included studies using the standard criteria developed by Cochrane and the CNRG. We will assess the risk of bias for cluster randomized trials (section 16.3.2) and for individual randomized trials (Table 8.5a) as recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). In addition to risk of bias, quality of evidence will be assessed by the GRADE method as evidence of high, moderate, low and very low quality based on imprecision, inconsistency, indirectness and publication bias (Higgins 2011). We will resolve all discrepancies by discussion and consensus. We will obtain study protocols for risk of bias assessment.
Cluster‐randomized trials
For cluster‐randomized trials, we will assess the following: recruitment bias, baseline imbalance, loss of clusters or individuals from the cluster, whether analyzed taking clustering into account (unit of analysis error) and contamination between clusters.
(i) Recruitment bias can occur when individuals are recruited to the trial after the clusters have been randomized, as the knowledge of whether each cluster is an ‘intervention’ or ‘control’ cluster could affect the types of participants recruited.
(ii) In cluster‐randomized trials, there is a possibility of chance baseline imbalance between the randomized groups, in terms of either the clusters or the individuals. Although not a form of bias as such, the risk of baseline differences can be reduced by using stratified or pair‐matched randomization of clusters. Reporting of the baseline comparability of clusters, or statistical adjustment for baseline characteristics, can help reduce concern about the effects of baseline imbalance.
(iii) Attrition bias: Loss of clusters from a trial or missing outcomes for individuals within clusters may also lead to a risk of bias in cluster‐randomized trials.
(iv) Cluster‐randomized trials that are analyzed by incorrect statistical methods, not taking the clustering into account, create a ‘unit of analysis error’ and produce over‐precise results (the standard error of the estimated intervention effect is too small) and P values that are too small. Approximate methods will be used to correct trial results that do not allow for clustering.
(v) In a meta‐analysis including both cluster and individually randomized trials, or including cluster‐randomized trials with different types of clusters, possible differences between the intervention effects being estimated will be considered.
Cross‐over trials
We will assess risk of bias in cross‐over trials as follows (Higgins 2011):
(i) Whether the cross‐over design is suitable: The cross‐over design is suitable to study a condition that is (reasonably) stable, and where long‐term follow‐up is not required.
(ii) Whether there is a carry‐over effect: A carry‐over effect means that the observed difference between the treatments depends upon the order in which they were received, hence the estimated overall treatment effect will be affected (usually underestimated, leading to a bias towards the null).
(iii) Whether only first period data are available: Cross‐over trials for which only first‐period data are available will be considered to be at risk of bias, especially when the investigators explicitly used the two‐stage strategy.
(iv) Incorrect analysis: The analysis of a cross‐over trial should take advantage of the within‐person design, and use some form of paired analysis.
(v) Comparability of results with those from parallel‐group trials in the absence of carry‐over.
Measures of treatment effect
We will report relative risk (RR) and risk difference (RD) for dichotomous outcomes and mean differences (MDs) for continuous outcomes with 95% confidence intervals (CIs) when eligible trials are identified. The number needed to treat for an additional beneficial outcome (NNTB) or number needed to treat for an additional harmful outcome (NNTH) will be calculated with 95% CIs if there is a statistically significant reduction or increase in RD.
Unit of analysis issues
The unit of analysis will be the cluster (for example neonatal unit or subunit) in cluster‐randomized trials and we will report on whether the sample size was estimated based on the intra‐cluster correlation co‐efficient (ICC) and whether the trial had been analyzed at the cluster level (the unit of randomization) or at the level of the individual. The unit of analysis in cross‐over trials will be the individual patient and will analyzed as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Dealing with missing data
We will contact the authors of published studies if clarifications are required, or to provide additional information. In the case of missing data, the number of participants with missing data will be described in the 'Results' section and the 'Characteristics of included studies' table. The results will only be presented for the available participants. We will discuss the implications of the missing data in the discussion of the review.
Assessment of heterogeneity
We plan to estimate the treatment effects of individual trials and examine heterogeneity between trials by inspecting the forest plots and by using the Chi² test which assesses whether observed differences in results are compatible with chance alone (Higgins 2011). A low P value (or a large Chi² statistic relative to its degree of freedom) provides evidence of heterogeneity of intervention effects (variation in effect estimates beyond chance). However the Chi² statistic has low power when meta‐analyzed studies have small sample size or are few in number. We will also quantify the impact of heterogeneity using the I² statistic (which incorporates the Chi² statistic). We will grade the degree of heterogeneity as none if less than 25%, low if between 25% and 49%, moderate if between 50% and 74% and high if greater than 75%. If we detect statistical heterogeneity, we will explore the possible causes (for example, differences in study quality, participants, intervention regimens, or outcome assessments) using post hoc subgroup analyses. We plan to use a fixed‐effect model for meta‐analysis.
Assessment of reporting biases
We will attempt to obtain study protocols of all included studies and compare outcomes reported in the protocols to those reported in the included studies. We will investigate reporting and publication bias by examining the degree of asymmetry of a funnel plot if at least 10 studies are included in the meta‐analysis. Where we suspect reporting bias we will attempt to contact study authors, asking them to provide missing outcome data. Where this is not possible, and the missing data are thought to introduce serious bias, we will explore the impact of excluding such studies in the overall assessment of results by sensitivity analyses.
Data synthesis
We will use Review Manager 5 (RevMan) software for statistical analysis (Review Manager 2014), and intend to use a fixed‐effect model for meta‐analysis. We will perform statistical analyses according to the recommendations of the CNRG.
Cluster randomized trials
For cluster randomized trials, if analyzed appropriately at the level of the cluster and if summary estimates are available, we will synthesize data using the generic inverse variance method. If summary estimates are unavailable or the trials were not analyzed at the cluster level, we will adjust the sample size by using the intracluster co‐efficient (ICC) and design effect (approximate analyses) (Higgins 2011). Analysis of cluster‐randomized trials will be performed as recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). In cluster‐randomized trials that were analyzed appropriately at the cluster level using the ICC, we will use the summary estimate to generate natural log of the RR and standard error of the log RR, entered in RevMan and meta‐analyzed using the generic inverse variance method (Review Manager 2014). In cluster‐randomized trials that were not analyzed at the cluster level, where the ICC is available we will calculate the design effect using the ICC and adjust the sample size for analyses. If the ICC is not available, we will use an assumed ICC from similar trials or perform approximate analysis as recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). When ICC could not be assumed due to variability in the study design or outcome, we will summarise the results without meta‐analysis.
Cross‐over trials
Methods of analysis of cross‐over trials (Higgins 2011): If neither carry‐over nor period effects are thought to be a problem, then an appropriate analysis of continuous data from a two‐period, two‐intervention cross‐over trial is a paired t‐test. This evaluates the value of ‘measurement on experimental intervention (E)’ minus ‘measurement on control intervention (C)’ separately for each participant. The mean and standard error of these difference measures are the building blocks of an effect estimate and a statistical test. The effect estimate will be included in a meta‐analysis using the generic inverse‐variance method in Review Manager 2014.
A paired analysis is possible if the data in any one of the following is available: a. individual participant data from the paper or by correspondence with the trialist; b. the mean and standard deviation (or standard error) of the participant‐specific differences between experimental intervention (E) and control intervention (C) measurements; c. the mean difference and one of the following: (i) a t‐statistic from a paired t‐test; (ii) a P value from a paired t‐test; (iii) a confidence interval from a paired analysis; d. a graph of measurements on experimental intervention (E) and control intervention (C) from which individual data values can be extracted, as long as matched measurements for each individual can be identified as such.
Quality of evidence
We will use the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach, as outlined in the GRADE Handbook (Schünemann 20133), to assess the quality of evidence for the main comparison at the outcome level.
Two authors will independently assess the quality of the evidence for each of the outcomes above. We will consider evidence from randomized 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 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:
-
High: We are very confident that the true effect lies close to that of the estimate of the effect.
-
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.
-
Low: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
-
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.
Subgroup analysis and investigation of heterogeneity
We will perform the following subgroup analyses if data are available:
Gestational age
Term if ≥ 37 weeks
Late Preterm: 34 to 36 weeks and 6 days
< 34 weeks
< 28 weeks
Birth weight
≥ 2500 grams
< 2500 grams
< 1500 grams
< 1000 grams
Infant subgroups based on type of colonization or infection
Infants colonized with an infectious agent
Infants infected with an infectious agent
Type of unit/cluster
Hospital, unit or a part of the unit.
Type of infection
Bacterial (Gram‐positive or Gram‐negative or specific organisms), viral or fungal (Candida or non‐Candida fungal infections).
Type based on availability of healthcare resources
Low‐, middle‐ and high‐income countries, with bands based on World Bank data and classification (World Bank 2016 accessed 8/8/2014). For the 2016 fiscal year, low‐income economies are defined as those with a GNI per capita, calculated using the World Bank Atlas method, of USD 1045 or less in 2014; lower‐middle‐income economies are those with a GNI per capita between USD 1046 and USD 4125; upper‐middle‐income economies are those with a GNI per capita between USD 4126 and USD 12,735; high‐income economies are those with a GNI per capita of USD 12,736 or more.
Clusters or neonatal units will be subgrouped by low‐, middle‐ and high‐income countries based on the World Bank classification.
Type of neonatal unit
Newborn nursery, level II or level III/IV neonatal units
Type of healthcare associated infection
Bloodstream infections, meningitis, ventilator‐associated pneumonia, skin and soft tissue infections and CLABSI.
Sensitivity analysis
We will explore methodological heterogeneity through the use of sensitivity analyses based on identified trials during the review process by excluding studies with high risk of bias.
