Introduction Data from the HOT-HMV UK trial showed an improvement in admission-free survival with the addition of home mechanical non-invasive ventilation (HMV) to home oxygen therapy (HOT) in patients with persistent hypercapnia following an acute exacerbation of COPD [JAMA;317:2177]. A post-hoc analysis was conducted to investigate (1) which baseline patient characteristics predict 12 month outcome and (2) the difference of these characteristics between treatment groups.

Method Patients were randomised to HOT or HOT-HMV if they had persistent hypercapnia (PaCO₂ >7 kPa) 2 weeks following resolution of respiratory acidosis (pH >7.30) secondary to an acute exacerbation of COPD. Non-invasive ventilation was titrated to treat nocturnal hypoventilation and patients were followed up for 1 years after discharge. Between group comparison of readmission and death were assessed in terms of baseline demographics, anthropometrics, lung function, gas exchange, quality of life and dyspnoea level.

Results 116 patients were enrolled and randomised to HOT (n=59) or HOT-HMV (n=57) with (mean ±sd or median [IQR]) age 67±10 years, BMI 22 [18–26] kg/m², FEV₁0.6±0.2 L, PaCO₂59±7 mmHg, SRI-SS 46±15, SGRQ-SS 74 [63–80], MRC dyspnoea score 5 [4–5]. Patients allocated HOT were less likely to be admitted with increasing BMI (25±5 vs 22±6 kg/m²; p=0.044). Patients allocated to HOT-HMV are were less likely to be admitted if they had higher FEV₁ (0.66±0.24 vs 0.54±0.20 L; p=0.042), lower levels of dyspnoea (MRC dyspnoea score 4±1.0 vs 5±0.5; p=0.002) or higher levels of specific measures of quality of life (SGRQ-AC 85±13 vs 92±7; p=0.015) (Table 1). Baseline severity of respiratory failure did not predict 12 month outcome in either group.

Conclusion Factors influencing outcome in patients with COPD and persistent hypercapnia receiving HOT-HMV treatment were airways obstruction and level of dyspnoea. However, in the patients receiving HOT alone, BMI was the only factor. There was no between group difference in with the exception of a sub-scale of the SGRQ. Interestingly, the severity of respiratory failure at baseline does not influence risk of readmission or death within 12 months as the patients all demonstrated severe chronic respiratory failure