Risk of recurrent respiratory failure subsequent to successful weaning from invasive mechanical ventilation initiated for non-surgical reasons: A nationwide cohort study

[[abstract]]Background and Aims: Numerous patients survive respiratory failure and subsequently separate from mechanical ventilation every year. Nevertheless, it remains elusive how the clinical course evolves after successful ventilator liberation. This study aimed to estimate the risk levels of recurrent respiratory failure and death after weaning success. Methods: This nationwide cohort study used the entire population database of Taiwan’s National Health Insurance (NHI). We included patients who were eventually separated from invasive mechanical ventilation successfully among those were newly placed on such care for at least two days for non-surgical reasons from March 1, 2010, through September 30, 2011. We estimated the probabilities of reinstituting invasive mechanical ventilation by the Kaplan–Meier estimator, and depicted the risk distribution over time on the basis of the hazard function. Results: There were 56,394 patients newly initiating invasive mechanical ventilation for non-surgical reasons during the study period. Of them, 34,550 patients (61.3%) were successfully weaned from ventilators. The median age of the study cohort was 74 years (IQR, 60–82 years), and 36.3% were women. The median duration for the index mechanical ventilation episode was 8 days (IQR, 5–16 days). After a successful weaning, the subsequent risks of reinstituting invasive mechanical ventilation were 29.5% (95% CI, 29–30%) by six months and 36.8% (95% CI, 36.2–37.4%) by 12 months. The risks for a composite outcome combining reinstitution of invasive mechanical ventilation and death were 45.2% (95% CI, 44.7–45.7%) by six months and 55.0% (95% CI, 54.5–55.6%) by 12 months. The risk of reinstituting invasive mechanical ventilation was trending downwards over time and levelled off on the 30th day after ventilator liberation. Conclusion: The patients just recovering from respiratory failure were at high risks for recurrent respiratory failure and death. The risks were particularly remarkable within 30 days after ventilator liberation