NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Allen MC, Donohue P, Gilmore M, et al. Inhaled Nitric Oxide in Preterm Infants. Rockville (MD): Agency for Healthcare Research and Quality (US); 2010 Oct. (Evidence Reports/Technology Assessments, No. 195.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Inhaled Nitric Oxide in Preterm Infants

Inhaled Nitric Oxide in Preterm Infants.

Show details


Our objective was to review and synthesize the evidence on the use of inhaled nitric oxide (iNO) in preterm infants born at or before 34 weeks gestation age who require respiratory support. This review addresses the short term outcomes bronchopulmonary dysplasia, cardiopulmonary risks, infectious risks, neurological risks, as well as short term survival and death. Long term outcomes including pulmonary outcomes, neurodevelopmental outcomes, growth, chronic medical conditions, and survival to childhood were also assessed. The results of this report will be presented to an NIH Consensus Panel in October 2010.

Topic Development

The core team worked with technical experts, the NIH Consensus Panel Chair, to develop and refine the Key Questions that are presented in Chapter 1 (Introduction). Prior to searching for literature, we clarified the definitions of these key questions and the types of evidence which we would include in our review. Topic development was facilitated by the results of preliminary searches, discussions among team members, and input from our Technical Expert Panel.

Key Questions 1 and 2 address short term impact of iNO use on preterm infants. Key Question 1 addresses the impact of iNO on survival and/or bronchoplulmonary dyplasia. Key Question 2 addresses short term risks to preterm infants receiving iNO therapy. Based on discussion with our experts, we decided to limit our review to randomized controlled trials for these two questions.

Key Question 3 addresses long term outcomes of iNO use in preterm infants. This question focuses on pulmonary and neurodevelopmental outcomes. We did not limit the consideration of studies by study design. We identified, and abstracted separately case reports and case series. However, we ultimately chose not to include case reports and case series in our formal review as the level of detail in these reports was generally insufficient.

The impact of iNO therapy on bronchopulmonary dyplasia, death, and/or neurodevelopmental outcomes across subpopulations of premature infants is addressed in Key Question 4, and influence of the timing of initiation, mode of delivery, dose and duration, or concurrent therapies is addressed in Key Question 5. Included studies for this question were not limited by study design except for the exclusion of case reports and case series, as for Key Question 3.

Analytic Framework

We developed a framework (Figure 1) to illustrate the components of the key questions, including the population, intervention and outcomes. The framework also delineates the subgroups; treatment characteristics, such as dose of iNO; and specific short and long term outcomes of interest. Short term outcomes were defined as adverse events and clinical outcomes associated with iNO treatment that occur during the initial hospitalization after birth. Long term outcomes were defined as the effects of iNO treatment on infant health and functional outcome in early childhood and include measures of chronic pulmonary disease, growth, developmental delay and disability, and survival.

This figure describes the characteristics of the patients included in this study, short term risks and outcomes (Key Questions 1 and 2), the patient subgroups under investigation (Key Question 4), treatment characteristics (Key Question 5) and long term health outcomes (Key Question 3) that may occur after use of iNO in preterm infants.

Figure 1

Analytic framework.

Search Strategy

Searching the literature involved identifying reference sources, formulating a search strategy for each source, and executing and documenting each search. For the searching of electronic databases we used controlled vocabulary terms (i.e., MeSH, EMTREE), combined with text words for iNO (see Appendix B, Detailed Search Strategies) We also looked for eligible studies by reviewing the references in pertinent reviews, by scanning conference proceedings, by querying our experts, and through knowledge shared at core team meetings.


Our search included electronic and hand searching. On November 9, 2009, we ran searches of MEDLINE® (using PubMed), EMBASE®, the Cochrane Central Register of Controlled Studies (CENTRAL), and PsycInfo databases. These searches were run again on June 23, 2010. We also searched the references of articles included in this study and those tagged as of interest during the screening process. As information on long term outcomes for infants treated with iNO is just emerging, we also scanned the proceedings of the Pediatric Academic Societies Meetings in 2009 and 2010. was searched for ongoing or completed trials. Investigators of ongoing trials were not contacted for information. We decided that the investigators conducting ongoing trials would only be contacted if they were studying outcomes with no published information. There were no limits used in the searches, including any based on publication date.

Search Terms and Strategies

We developed a strategy for MEDLINE, accessed via PubMed, based on an analysis of the MeSH terms and text words of key articles identified a priori. The PubMed strategy formed the basis for the strategies developed for the other electronic databases (see Appendix B).

Organization and Tracking of the Literature Search

The results of the searches were downloaded into ProCite® version 5.0.3 (ISI ResearchSoft, Carlsbad, CA). Duplicate articles retrieved from the multiple databases were removed prior to initiating the review. From ProCite, the articles were uploaded to Distiller SR © (Evidence Partners, Ottawa, Ontario). We used this software to store full articles in portable document format (PDF) and to track the results of the abstract screen, article screen, and data abstraction.

Study Selection

Abstract Screen

Each abstract was independently screened by two reviewers. An abstract was excluded at this level if it did not reporting any original data, did not include human data, did not include infants born at less than or equal to 34 weeks of gestation, did not include preterm infants requiring respiratory support, did not include preterm infants treated with inhaled nitric oxide, did not address any of the key questions, or addressed Key Question 1 and or 2 but was not a randomized controlled trial. An option was provided for reviewers to indicate other reasons for exclusion. Articles tagged as non English were reviewed by individuals fluent in the language of publication to determine eligibility. (Appendix C, Abstract Review Form).

Abstracts were promoted to be screened using full text article if both reviewers agreed that the abstract could apply to one or more of the key questions. An abstract could be excluded for different reasons by the two reviewers. Disagreements about the eligibility of an abstract were resolved by discussion between the two reviewers or by adjudication of a third reviewer.

Article Screen

Full text articles underwent another independent review by paired investigators to determine whether they should be included in the full data abstraction (see Appendix C, Article Inclusion/Exclusion Form). If articles were deemed to have applicable information, they were included in the data abstraction. Articles could be excluded at this level for the same set of reasons used at the abstract screen level with an additional exclusion criterion of no abstractable data. Articles that had English language abstracts that were promoted to this level but were tagged for exclusion as “not English language” were reviewed by investigators fluent in the specific language for eligibility.

Articles were promoted to data abstraction if both reviewers agreed. An article could be excluded for different reasons by the two reviewers. Disagreements about the eligibility of an article were resolved by discussion between the two reviewers or by adjudication of a third reviewer.

Data Abstraction

We used an independent review process to abstract data from the included articles. In this process, both a clinical expert and a research assistant completed all relevant data abstraction forms independently. Reviewers were not masked to the articles’ authors, institutions, or journal.48 Disagreements that could not be resolved between the reviewers were resolved through consensus adjudication at team meetings.

For all articles, reviewers extracted information on general study characteristics: study design, whether the study was a followup or additional analysis of another study, location, recruitment start and end dates, inclusion and exclusion criteria, description of the study intervention, iNO dose and duration, and length of followup (see Appendix C, Study Characteristics Form). Participant characteristics were also abstracted: number of participants, gestational age, birth weight, participant age, sex, and relevant background data such as disease severity, mode of ventilation, and concurrent medications. Maternal characteristics were also collected on this form (see Appendix C, Participant Characteristics Form).

Reviewers abstracted data, for all study arms and subgroups, on a predefined set of outcomes (see Appendix C, All Outcomes). Case reports were abstracted separately to identify whether they included data relevant to this study (see Appendix C, Case report form). These data were ultimately not included as the level of detail in these reports was generally insufficient.

Quality Assessment of Individual Studies (Risk of Bias Assessment)

In order to assess the risk of bias in randomized controlled trials, we used the Cochrane Collaboration Tool for Assessing Risk of Bias from the Cochrane Handbook for Systematic Reviews of Interventions.49 This tool was used to assess six categories of potential bias; (1) sequence generation, (2) allocation concealment, (3) blinding, (4) incomplete data reported, (5) selective reporting bias as well as (6) other sources of bias. For each bias category reviewers answered one or more questions and entered “Yes” for a low risk of bias, “No” for a high risk of bias or “Unclear.”

For the observational studies we adapted the Newcastle-Ottawa Scale in order to determine the risk of bias of the reported data in both cohort and case control studies.50 This form assessed possible sources of bias including (1) representativeness of the study cohort, (2) selection of the control cohort (if applicable), (3) selection of treated patients, (4) presence of the outcome of interest at the start of the study, (5) comparability of the cohorts, (6) reporting bias, (7) whether the followup was long enough for outcomes to occur, and (8) incomplete data reported. Similar to the risk of bias forms for randomized control trials, we used question based forms where reviewers entered “Yes” for a low risk of bias, “No” for a high risk of bias or “Unclear” for questions about each source of bias.

The risk of bias forms were completed independently by paired reviewers. In the case of a disagreement, the two original reviewers conferred and agreed upon a single answer. These assessment instruments are included in Appendix C, Risk of Bias Forms.

Grading of the Body of Evidence

At the completion of our review, we assessed the quantity, quality and consistency of the body of available evidence addressing Key Questions 1 through 5. We used an evidence grading scheme recommended by the GRADE Working Group, and adapted by AHRQ in their Draft Methods Guide,51 and recently published in the Journal of Clinical Epidemiology.52 We considered the strength of the study designs with randomized controlled trials as the highest level of evidence, followed by observational studies. If an outcome was evaluated by at least one randomized controlled trial as well as observational studies our evidence grade was based on the randomized controlled trials and followed by the quality of the cohort studies. If an outcome was evaluated by one or no randomized controlled trials, our evidence grade was based on the single randomized controlled trial in addition to the best available observational study.

We assessed the quality and consistency of the best available evidence, including assessment of the risk of bias in relevant studies, as well as aspects of consistency, directness, and precision as described in the Draft Methods Guide51 and Owens, 2010.52

For each outcome of interest, two investigators graded the major outcomes for each Key Question and then the entire team discussed their recommendations and reached consensus.

Data Synthesis

We created a set of detailed evidence tables containing information extracted from eligible studies. We stratified the tables according to applicable key question. Once evidence tables were created, we rechecked selected data elements against the original articles. If there was a discrepancy between the data abstracted and the data appearing in the article, this discrepancy was brought to the attention of the investigator in charge of the specific data set and the data were corrected in the final evidence tables.

Meta-analyses were completed using MetaAnalyst. The program was developed by the Tufts Evidence-based Practice Center under contract with AHRQ.53 The analyses were performed using a Der-Simonian Laird random effects model.54 In this program, a Woolf-Haldane continuity correction of 0.5 was used when a cell contained zero events.55 In all analyses, we examined the risk ratio for each outcome. Sensitivity analyses were performed to determine stability of the results. In general, meta-analyses were completed for outcomes reported across more than one study where the definition and measurement of the outcome was determined to be similar. Where relevant, further details regarding the decision to conduct or not conduct meta-analyses, the inclusion and exclusion of articles from the meta-analysis, and any sensitivity analyses, are provided in the results section.

Peer Review

We recruited external technical experts from diverse professional backgrounds, including neonatology, pulmonology, cardiology, and neurodevelopment. The technical experts were asked for input regarding key steps of the process, including development of the analytic framework, outcomes, and search strategies. In addition to the technical experts, three peer reviewers were recruited from various clinical and methodological settings.

Throughout the project, the core team sought feedback from the external technical experts and the NIH Panel Chair. A draft of the report was sent to the technical experts and peer reviewers, as well as to representatives of AHRQ, and the NIH Office of Medical Applications Research Panel Chair for this project. In response to the comments from the technical experts and peer reviewers, we revised the evidence report and submitted a summary of the comments and their disposition.


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...