Medical News Stories
Inhaled Nitric Oxide and the Risk of Renal Dysfunction in Patients with Acute Respiratory Distress Syndrome: a Propensity-Matched Cohort Study
Ruan SY, Wu, HY, Lin HH, et al. Critical Care. 2016;20:389.
Study Question: In patients with acute respiratory distress syndrome (ARDS), does inhaled nitric oxide (iNO) therapy increase the risk of requiring renal replacement therapy (RRT), compared to non-users?
Study Description: This single-center, retrospective study evaluated the risk of RRT in adults with moderate/severe ARDS (PaO2/FiO2 <200 mmHg) receiving iNO, compared to a propensity-matched cohort of non-users. Patients previously on RRT or extracorporeal membrane oxygenation were excluded. Patients were ventilated using tidal volumes and PEEP recommended by the ARDS network. No standard protocol for iNO was used and patients receiving any dose were defined as iNO-users. Hazard modeling was used to calculate the hazard ratio (HR). The propensity-matched cohort was analyzed using cumulative incidence functions to account for competing risks of death. The primary outcome was incident RRT until the first occurrence of the following: death from any cause, transfer to another hospital, or at 30 days.
Results: A total of 302 patients were included in the propensity-matched cohort (151 in each group). Fifty-two percent had moderate ARDS and pneumonia was the primary cause (80%). The median duration of iNO was 4 days (IQR 2-7) and median initial dose was 20 ppm (IQR 15-20). Overall 30-day mortality for matched patients was 43%. The propensity matching model found a significantly increased risk of RRT with iNO (34 vs. 23%; HR 1.59; 95% CI 1.08-2.34; p = 0.02). In a subgroup analysis, age ≥65 years was associated with higher risk of iNO-associated RRT (p = 0.05).
Conclusion(s): ARDS patients exposed to iNO had increased risk for requiring RRT and age ≥65 years were at greatest risk, compared to patients that were not exposed.
Perspective: Similar to RCTs, this study found that iNO therapy led to an increased risk of RRT. However, results should be interpreted cautiously due to the study’s single-center, retrospective nature. While propensity-matching attempted to adjust for confounders, nephrotoxic medications weren’t considered in matching and may have resulted in unmeasured confounding, especially in this critically ill patient population. Additionally, iNO and RRT was managed at provider discretion and it is unclear if this influenced outcomes. Confounding by indication may also be a remaining source of bias, although propensity-matching attempted to control for this. Despite these limitations, clinicians should monitor renal function and avoid the concurrent use of nephrotoxic medications with iNO.