This study demonstrates that influenza-induced ARDS and PMV have independent risk factors, including BMI > 25, combined bacterial pneumonia, neuromuscular blockade use over 48 hours during treatment, and ECMO support. Additionally, we found that risk factors for unsuccessful weaning after PMV are ECMO support and presence of bacteremia.
In our cohort, an average BMI > 25 (overweight and obese) was observed, with a higher prevalence of patients with BMI > 25 in the PMV group. Recent research describing the epidemiological shifts since the emergence of H1N1 influenza has reported severe outcomes in adults with obesity. Individuals who have obesity and contract H1N1 influenza are more likely to require mechanical ventilation and experience longer stays in the intensive care unit (ICU), along with an increased risk of mortality[11, 12].
Several clinical studies on ARDS have noted higher BMI in ARDS patients than in non-ARDS patients. Development of ARDS increases significantly with higher body weight [13].
Obesity is a common comorbidity among hospitalized patients with pandemic H1N1 influenza infection. [14] Recent experimental studies in mice have suggested that obesity impairs dendritic cell function and recruitment of monocytes into the lungs during influenza infection [15]. A meta-analysis demonstrated a significant association between obesity in critically ill patients and prolonged mechanical ventilation duration [16].
Over the past two years, the literature has begun to highlight that the proportion of patients with BMI > 25 of COVID-19 is nearly twice that of those with influenza [17].
BMI > 25 correlates positively with risk of severe events related to COVID-19, increasing significantly from 8.1–20% and 30.6%. Indeed, overweight status stands as the sole predictive factor for severe events associated with SARS-CoV-2 infection[18].
A COVID-19 study also indicated that patients with higher BMI spent more days on ventilators than those with normal weight, aligning with our cases and the results of this study [19].
The combination of influenza virus and bacterial pneumonia can exacerbate the severity of ARDS, infectious shock, and multiorgan failure. [20] In patients requiring hospitalization, bacterial pneumonia is more commonly seen in influenza than in COVID-19. [21, 22]。
During influenza virus infection, changes in respiratory epithelial cells and host immune responses lead to exposure of the epithelial surface. As the infection progresses, bacteria can adhere, and respiratory bacteria can accumulate in the airway epithelial mucus, thereby promoting secondary bacterial pneumonia.[23]
In the 2009 H1N1 pandemic, bacterial pneumonia as a pulmonary complication was associated with prolonged duration of mechanical ventilation. [24]
In our cohort, the presence of bacterial pneumonia prolonged the duration of mechanical ventilation, and in another study conducted by our team, we found that severe influenza-related ARDS hospital-acquired lower respiratory tract infections are associated with prolonged mechanical ventilation and worse prognosis.[25] Another multicenter Italian study on COVID-19 patient weaning from ventilators also identified late-onset ventilator-associated pneumonia as one of the factors influencing ventilator liberation. [26]。In another study on COVID-19 with superinfections, bacteremia accounted for 47.4% of cases, and patients with secondary infections had a prolonged mechanical ventilation time of up to 37 days. [27] Bacteremia can progress to systemic inflammatory response syndrome (SIRS), sepsis, septic shock, and multiple organ dysfunction syndrome (MODS).[28, 29]
The occurrence of sepsis can affect diaphragmatic stability and lead to failed ventilator liberation. The presence of sepsis is associated with evident diaphragmatic weakness. [30–34]
Currently, diaphragm dysfunction in critically ill patients is believed to occur primarily through two mechanisms: ventilator-induced diaphragmatic dysfunction (VIDD) [35] and sepsis-induced dysfunction. [36]
Sepsis typically impairs oxygen consumption, increases anaerobic metabolism, and leads to metabolic acidosis. The need to compensate for acidemia increases ventilation requirements and may result in failed ventilator liberation [37].
In recent years, research has indicated that spontaneous ventilation can help to improve hypoxemia and lung compliance and reduce diaphragm atrophy in patients with acute respiratory distress syndrome (ARDS).
[38–40] However, spontaneous breathing can lead to elevated respiratory drive and vigorous inspiratory efforts, causing uneven pressure distribution and potentially resulting in patient self-inflicted lung injury (P-SILI) [41]. Multiple animal studies have demonstrated detrimental cycles, such as asynchrony with the ventilator, elevated transpulmonary pressure, and double triggering, which further worsen lung injury. [31, 42–44]
A trial in 2010 reported improved 90-day survival rates in severe ARDS patients receiving neuromuscular blockade (NMBAs) [45], but a larger trial in 2019 contradicted these findings. [46] Therefore, use of NMBAs in ARDS patients remains controversial. Although current evidence does not support routine early use of neuromuscular blockade in all adult patients with moderate to severe ARDS, utilizing NMBAs in the early stages of ARDS to ensure good synchrony with the ventilator and promote lung-protective strategies remains a reasonable treatment option. There are formal guidelines that recommend continuous infusion of NMBAs for less 48 hours [47] and evaluation daily by specialized physicians for ongoing use. [48, 49] Clinical data indicate that using neuromuscular blocking agents can reduce barotrauma and improve physiological and clinical outcomes. [50] However, there are potential adverse effects on diaphragmatic contractile function and delayed extubation. [51] Several studies and a meta-analysis on acute respiratory distress syndrome (ARDS) have confirmed that neuromuscular blocking agents (NMBAs) do not improve mortality rates, ventilator-free days (VFDs), or the duration of mechanical ventilation. [52–54]. However, it is important to note that the incongruences in research methodologies lead to an inability to reach a definitive consensus on this matter.
In contrast, our study findings indicate a significant impact of utilizing NMBAs for more than 48 hours on the extension of mechanical ventilation duration, known as prolonged mechanical ventilation (PMV).
ECMO provides circulatory or respiratory support in cases of refractory cardiogenic shock or ARDS. Multiple studies have indicated that ECMO is feasible and effective for ARDS patients caused by H1N1 infection in 2009[55–57].
One study assessing ECMO-related mortality risk reported that the mortality rate for influenza-induced ARDS patients receiving ECMO support was the lowest observed thus far, despite an average duration of mechanical ventilation support of up to 40 days. [58]。
In a multicenter retrospective study conducted in Italy, COVID-19 and influenza-related ARDS receiving ECMO patients had longer durations of invasive mechanical ventilation than influenza patients, with durations of 33 days and 25 days[59], respectively. However, the mortality rate during COVID-19 was higher than that during the 2009 H1N1 period, potentially due to more frequent use of noninvasive ventilation (HFNC) forms before endotracheal intubation, leading to more severe self-inflicted lung injury. [41, 60]。
In studies involving ECMO usage for H1N1 influenza patients, the duration of mechanical ventilation before ECMO initiation was identified as an important prognostic factor. [57] A multicenter study conducted in COVID-19 patients yielded similar results. [61]
Numerous international multicenter studies have substantiated the elevated mortality rates associated with ARDS, which are estimated to be approximately 40%.[62–65] In another study focusing on the association of higher tidal volumes with increased mortality conducted by our group, a 30-day mortality rate of 23.2% was observed[66], but the estimated mortality rate for patients after PMV increased to 39.7%. This similarity with large international studies confirms the association between PMV and a higher mortality rate.
Our study has several limitations. First, it is a retrospective study, which may have resulted in missing statistical and medical data, leading to variations. Second, there was a lack of consistency in the treatment strategies for influenza-related ARDS among the different study sites. The treatment policies of participating centers were not standardized, which increased potential confounding factors. Last, our study focused on patients with ARDS caused by influenza. Therefore, whether the findings apply to ARDS caused by other factors needs to be confirmed through future, more rigorous prospective clinical studies.
Despite these limitations, to the best of our knowledge, this is the first multicenter study that elucidates risk factors associated with prolonged mechanical ventilation in patients with ARDS caused by influenza. Our study may aid clinicians with regard to treatment directions and decision-making for critically ill patients.