Purpose: Congenital diaphragmatic hernia (CDH) is associated with significant in-hospital mortality, pulmonary morbidity and length-of-stay (LOS). We hypothesized that oxygen requirements on hospital day 30 may predict subsequent in-hospital mortality, LOS, and oxygen needs at discharge and as such could be useful for risk-stratification and counseling.
Methods: Patients in the CDH Study Group registry with a LOS>=30 days were analyzed (2007-2010). Oxygen requirements at 30-days were defined as (1) room-air (2) noninvasive supplementation (3) mechanical ventilation and (4) extracorporeal membrane oxygenation (ECMO). Crude analysis was used to compare oxygen requirements with associated risks and outcomes. Using multivariate analysis, the association of day-30 oxygen requirements with subsequent mortality and discharge oxygen needs was adjusted for confounders (including APGAR, birthweight, ECMO duration and defect size).
Results: 862 patients in the registry had a LOS>=30 days. 542(62.9%) needed oxygen supplementation at 30 days. Of those, 244(45.0%) were on noninvasive supplementation, 279(51.5%) were on mechanical ventilation, and 19(3.5%) were on ECMO. Survival to discharge was 99.7, 98.0, 74.9 and 26.3% in patients on room-air, non-invasive supplementation, ventilator, and ECMO at 30-days respectively (p<.001, figure-1). The proportion of patients on oxygen at discharge was 0.6%, 43.2%, and 63.3% respectively in those on room air, noninvasive supplementation and mechanical ventilation at 30-days (p<.001). Of those on room-air on day-30, only 13.2% required a stay over 60 days, whereas 47.3% on noninvasive supplementation, 85.2% on a ventilator, and 100% of survivors on ECMO at 30-days required a long-stay (p<.001). Patients requiring mechanical ventilation or ECMO at 30-days had significantly greater mortality on survival analysis, (p<.001, figure-2), however there was no difference between those on room-air and those on noninvasive oxygen (p=.31). On multivariate analysis, need for noninvasive oxygen on day-30 was not associated with an increased risk of mortality compared to those on room-air (Hazard Ratio 0.91 95%CI .21-4.51, p=.90), whereas use of ventilator (HR 5.07 95%CI 1.65-22.3, p=.003) or ECMO (HR 19.57, 95%CI 5.07-97.72, p=<.001) were significant predictors of in-patient mortality. Need for non-invasive supplementation or mechanical ventilation on day-30 was associated with a respective 21.94 (95%CI 4.26-402.99, p<.001) and 43.24 (95%CI 8.51-790.81, p<.001) increased odds of oxygen use at discharge in long-stay survivors compared to those on room-air.
Conclusions: Patients on room-air or noninvasive oxygen on day-30 had a shorter total LOS, decreased need for oxygen at discharge and greatly decreased risk of mortality compared to those with greater oxygen requirements. The day-30 oxygen requirement is the strongest predictor of oxygen needs at discharge and in-patient mortality and may be used as a simple prognostic indicator for family counseling, discharge planning, and identification of high-risk infants.
Figure 1. Flow diagram
Figure 2. Survival Analysis of 862 CDH patients grouped by oxygen needs at day 30.