EBNEO Commentary: A network meta‐analysis of postnatal corticosteroids for bronchopulmonary dysplasia: Has the most appropriate treatment been revealed?
Bibliographic record
Abstract
Ramaswamy VV, Bandyopadhyay T, Nanda D, Bandiya P, Ahmed J, Garg A, Roehr CC, Nangia S. Assessment of Postnatal Corticosteroids for the Prevention of Bronchopulmonary Dysplasia in Preterm Neonates: A Systematic Review and Network Meta-analysis. JAMA Pediatr 2021; 175(6): e206826. PMID 33720274. Systemic postnatal corticosteroids (PNCs) reduce the risk of bronchopulmonary dysplasia (BPD) in preterm infants but may contribute to long-term neurodevelopmental harm. The need to identify a treatment regimen with a reassuring risk-benefit balance remains.1 In an extraordinary feat of data aggregation, Ramaswamy and colleagues include 62 randomised clinical trials (RCTs) enrolling 5559 neonates in a systematic review and network meta-analysis (NMA) to determine which of 14 distinct PNC exposures (Table) may be most appropriate.2 The authors suggest moderately early-initiated medium cumulative dose systemic dexamethasone is the best regimen. Unlike traditional pairwise meta-analyses, NMAs compare multiple interventions, adding relevance to clinical scenarios with various alternative treatment options. NMAs also allow indirect comparisons between interventions, linking them through a common comparator.3 For example, no RCTs have compared early systemic hydrocortisone (EHC) to late-initiated low-cumulative dose systemic dexamethasone (LaLdDx), but a shared comparison to placebo in separate RCTs allows a relative risk estimation. The validity of NMA results relies on specific assumptions. A critical one is that the data linked to produce indirect evidence display ‘transitivity’.3 In essence, that aggregation may be justified by similarities in key population characteristics. Tools that guide this subjective judgement include assessment of whether the populations are similar on important effect modifiers and whether the indirect comparisons could be assessed in head-to-head trials. Extending the example above, we consider the tools against the PREMILOC (EHC) and DART (LaLdDx) trials.4, 5 A meta-regression identified the baseline risk of developing BPD as an important effect modifier for the effect of PNCs on death or cerebral palsy.6 The rates of death or BPD in the control groups of PREMILOC and DART were 49% and 91%, respectively, questioning their similarity.4, 5 In turn, a hypothetical trial of EHC versus LaLdDx would presumably compare prophylactic EHC to LaLdDx only amongst infants remaining on mechanical ventilation at 14 days. Exposing subjects to LaLdDx irrespective of respiratory status would raise ethical concerns. However, the indirect comparison generated by the NMA is not of prophylactic EHC versus expectant management with rescue LaLdDx. Though we focus on EHC versus LaLdDx as an example, these concerns extend to various comparisons. One could argue that transitivity is generally violated when comparing interventions applied in distinct stages of disease progression, and that use of PNCs as prophylaxis versus treatment of evolving severe BPD are distinct indications best suited to distinct comparisons.3 Lastly, we should avoid labelling the PNC regimen most effective against BPD as ‘most appropriate’. Although a reduction in BPD with various PNC regimens is probable, the avoidance of risk for adverse long-term neurodevelopmental outcomes is uncertain, and the latter will guide appropriateness. Further, as the authors acknowledge, the quality of evidence supporting the conclusion is low. Despite these points of caution, Ramaswamy and colleagues add a rich contribution to the literature, shining a helpful light on the path forward. Conclusively arriving at best practice will require additional high-quality research, followed by cautious data synthesis that prioritises long-term outcomes of importance to patients and their families. URL: https://ebneo.org/postnatal-corticosteroids-meta-analysis. None.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.038 | 0.219 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.008 | 0.006 |
| Bibliometrics | 0.003 | 0.004 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.008 | 0.002 |
| Research integrity | 0.021 | 0.017 |
| Insufficient payload (model declined to judge) | 0.017 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".