OBJECTIVES: Dysregulated pulmonary coagulation and inflammation is a hallmark of respiratory failure in various etiologies. Excessive fibrin deposition contributes to alveolar collapse, impaired gas exchange, and progression to pulmonary fibrosis. Nebulized heparin can mitigate these coagulation and inflammation disturbances. Although several randomized controlled trials have explored its effects, results remain inconsistent and limited by small patient populations. We conducted a random-effects meta-analysis to calculate the risk ratio (RR) and 95% CIs.
DATA SOURCES: We systematically searched PubMed, Embase, and Cochrane Central Register of Controlled Trials for randomized controlled trials comparing nebulized unfractionated heparin to standard care or placebo in adult patients with respiratory failure either invasively mechanical ventilated or not. The primary outcome was all-cause mortality at the longest follow-up.
STUDY SELECTION: We included randomized clinical trials enrolling adult patients with respiratory failure, comparing nebulized heparin vs. standard care or placebo, and reporting at least one clinical outcome, including all-cause mortality.
DATA EXTRACTION: Two independent investigators extracted data on trial design, setting, etiology of respiratory failure, heparin dosing regimens, follow-up duration, and outcomes. Discrepancies were resolved by consensus.
DATA SYNTHESIS: We identified 16 studies (787 receiving nebulized heparin, 833 control). Six (38%) were multicenter, five focused on COVID-19, 12 enrolled ICU patients, and dosing clustered around 25,000 international units (IUs) three times a day (~75,000 IU/d for ~10 d). At the longest follow-up, nebulized heparin reduced all-cause mortality vs. control (110/645 [17.1%] vs. 157/711 [22.1%]; RR, 0.79; 95% CI, 0.66-0.95; with ten studies included). Nebulized heparin was also associated with more ventilation-free days by day 28 (mean difference, +4.85; 95% CI, 1.47-8.24). Major bleeding was rare (1.1 vs. 0.7%; RR, 1.48; 95% CI, 0.42-5.18), while no minor bleeding or heparin-induced thrombocytopenia was reported.
CONCLUSIONS: Nebulized unfractionated heparin may improve survival in patients with respiratory failure without increasing adverse events.
| Discipline Area | Score |
|---|---|
| Hospital Doctor/Hospitalists |  |
| Internal Medicine | |
| Respirology/Pulmonology | |
| Intensivist/Critical Care |  |
A meta-nalysis of 17 small trials of nebulized heparin in respiratory failure with disparate causes. The effect size here is likely inflated (NNT is reported as 20). This is provocative for future efficacy trials but should not inform clinical practice, in my opinion.
Well done but not practice informing.
Several studies have suggested that heparin may be therapeutic for hypoxemic respiratory failure / ARDS but the individual studies are heterogeneous in patient types, dosing and administration (IV, inhaled) and, therefore, are not conclusive or practice changing. This aggregates the data for nebulized heparin and suggests it may be efficacious and perhaps safe, although bleeding remains a notable risk in this analysis.
As a hospitalist, this is the first I am hearing of this. In this meta-analysis, it looked as though inhaled heparin had a significant benefit in a randomized designs. Although there seemed to be some heterogeneity in the different trials that included floor patients - ICU, COVID, non-COVID - the heterogeneity measures were not significant. I was surprised that the benefits looked more significant in floor patients than in the ICU. This does not seem to be accepted in guidelines anywhere yet. Given that we do not have anything else in ARDS, this does look promising; however, it does seem to work in smoke inhalation patients.
This meta-analysis suggests that inhaled heparin may be beneficial in patients with respiratory failure. The findings are limited by a weak level of evidence, a positive effect seen in non-intubated patients, and variability of the patient populations. May have potential as an area for future study.
The 10 included trials were all relatively small and only 1 had reached statistical significance on its own, so the meta-analysis adds new information. Perhaps a bit surprising that this would work. Potential sources of bias are a concern as no effect was seen in trials at low risk of bias.
