Ultrasound-based evaluation of respiratory muscle effort in patients with type 2 respiratory failure secondary to acute exacerbation of chronic obstructive pulmonary disease

ABSTRACT Introduction Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) represent one of the leading causes of type 2 (hypercapnic) respiratory failure and often require non-invasive respiratory support strategies such as high-flow nasal cannula (HFNC) or non-invasive ventilation (NIV). Diaphragmatic dysfunction plays a central role in the pathophysiology of respiratory decompensation, reflecting excessive or ineffective respiratory effort. While the predictive value of diaphragmatic ultrasound is well documented, limited evidence exists regarding the role and compensatory activation of accessory respiratory muscles in this setting. Identifying non-invasive indices of respiratory muscle load may improve early recognition of patients at risk of non-invasive support failure. Methods This prospective observational study was conducted at the Respiratory Intermediate Care Unit of the University Hospital of Modena. Twenty-five patients admitted with AECOPD and PaO₂/FiO₂ < 300 mmHg were enrolled and underwent a 2-hour HFNC trial as a first attempt at non-invasive respiratory support per our institution protocol, consistent with mounting evidence of comparable effectiveness to NIV. Ultrasound was performed at admission (T0) and after 2 hours (T1) to assess thickening fractions of the diaphragm (TFdi), parasternal intercostals (TFic), and sternocleidomastoid (TFscm). Inspiratory effort was concurrently quantified using nasal pressure swings (Pnose). Blood gases and clinical severity scores were also recorded. Correlations between muscle thickening, Pnose, gas exchange, acid-base status, and HFNC outcomes were analyzed. Results At baseline, median TFdi was 27% [IQR 18–38], while TFic and TFscm were 0% [0–39] and 0% [0–24], respectively. Intercostal activation was observed in 40% and SCM activation in 52% of patients. TFdi inversely correlated with TFic (r = –0.34; p = 0.01) and TFscm (r = –0.41; p = 0.03), while TFic and TFscm were directly correlated (r = 0.39; p = 0.005). TFdi was negatively correlated with Pnose on both hemidiaphragms (r = –0.49 right, –0.52 left; p < 0.001), while both TFic and TFscm were positively correlated with Pnose (r = 0.52 and 0.66, respectively). Additionally, TFdi inversely correlated with PaCO₂ (r = –0.29; p = 0.04) and positively with PaO₂/FiO₂, whereas TFic and TFscm showed positive correlations with PaCO₂ and negative correlations with oxygenation and pH. Patients who experienced HFNC failure exhibited significantly lower TFdi (24% vs 32.5%; p = 0.03) and markedly higher TFic (30% vs 0%; p = 0.003) and TFscm (19% vs 0%; p = 0.004) compared to those who responded successfully. Conclusions In patients with AECOPD, diaphragmatic dysfunction is strongly associated with increased inspiratory effort and compensatory recruitment of accessory muscles. The integration of ultrasound-derived thickening fractions and nasal pressure swings offers a comprehensive and non-invasive approach to monitor respiratory effort. This multimodal evaluation may enhance early identification of patients at high risk of HFNC failure and guide timely therapeutic escalation, potentially improving clinical outcomes and avoiding delayed intubation.