Ward-Based Noninvasive Ventilation for Acute Hypercapnic Respiratory Failure Unrelated to Chronic Obstructive Pulmonary Disease

Table 3

Multivariable logistic regression demonstrating the predictors of in-hospital mortality.

Variable		Odds ratio
Total non-COPD AHRF
Pre-NIV pH > 7.25	pH > 7.15	2.223^a (1.130–4.375)	0.021
	pH 7.15–7.25	1.865^a (1.091–3.187)	0.023
Pre-NIV pH	0.004 (0.000–0.062)	<0.001
Pneumonia on admission	5.313^b (2.326–12.131)	<0.001
Bronchiectasis on admission	4.236^b (1.680–10.679)	0.002
NMD on admission	4.038^b (1.888–8.636)	0.003
Fluid overload on admission	3.735^b (1.556–8.966)	0.021
Age	1.034 (1.017–1.051)	>0.001
Subgroup (BHH): domiciliary NIV	0.065 (0.009–0.488)	0.008
Subgroup: obesity-related AHRF
Pre-NIV pH > 7.25	pH > 7.15	7.800^a (1.843–33.013)	0.005
	pH 7.15–7.25	2.035^a (0.523–7.915)	0.305
Age	1.030 (0.981–1.082)	0.231
Subgroup: non-obesity-related AHRF
Pre-NIV pH > 7.25	pH > 7.15	4.538^a (1.694–9.354)	0.002
	pH 7.15–7.25	1.843^a (1.038–3.272)	0.037
Age	1.030 (1.010–1.050)	0.005

^aThe odds ratio is against the reference group: pH < 7.25, ^bThe odds ratio is against the reference group: obesity-related AHRF. The pre-NIV pH was analysed as the continuous variable; COPD: chronic obstructive pulmonary disease; AHRF: acute hypercapnic respiratory failure; NIV: noninvasive ventilation; NMD: neuromuscular disease; BHH: Birmingham Heartlands Hospital.