. Author manuscript; available in PMC: 2022 Jan 21.

Published in final edited form as: Eur Respir J. 2009 Dec 23;35(5):1079–1087. doi: 10.1183/09031936.00072709

TABLE 5.

Multivariate predictors of death on Cox’s proportional hazards analysis in idiopathic, familial and anorexigen-associated pulmonary arterial hypertension

	Hazard ratio
	Regression 1^#	Regression 2^¶	Regression 3⁺

Age at diagnosis per decade increase	1.21 (1.04–1.39)	1.17 (1.01–1.34)	1.21 (1.05–1.39)
Functional class per 1-class increase	1.17 (0.92–1.51)	1.41 (1.10–1.80)	1.12 (0.86–1.45)
Right atrial pressure per 5-mmHg increase	1.42 (1.23–1.65)
Mean pulmonary arterial pressure per 5-mmHg increase		1.01 (0.94–1.08)
Cardiac index per 1-L·min⁻¹·m⁻² increase			0.54 (0.38–0.77)

Data are presented with 95% confidence intervals in parentheses. Likelihood ratio (LR) analysis was performed versus the model with age and functional class alone. The model with age, functional class and right atrial pressure is the best model. Adding cardiac index is marginally additive in this primary pulmonary hypertension subgroup (LR test for model 1 with cardiac index versus model 1 without cardiac index: p=0.063).

^#:

independent variables: age at diagnosis, functional class and right atrial pressure (p<0.0001 on LR analysis)

^¶:

independent variables: age at diagnosis, functional class and mean pulmonary arterial pressure (p=0.78 on LR analysis)

^+:

independent variables: age at diagnosis, functional class and cardiac index (p=0.0004 on LR analysis).