. 2019 May 15;4(6):556–563. doi: 10.1001/jamacardio.2019.1384

Table 3. Comparative Effectiveness of Pulmonary Artery Pressure Sensor-Based Management on Clinical Event Rates and Days Lost.

Outcome	Mean (SD)		Hazard Ratio^a or Risk Ratio^b (95% CI)	Absolute Difference, d	P Value
Outcome	Treatment Cohort (n = 1087)	Control Cohort (n = 1087)	Hazard Ratio^a or Risk Ratio^b (95% CI)	Absolute Difference, d	P Value
Clinical event rates per patient per y
Heart failure hospitalization^c	0.65	0.88	0.76 (0.65-0.89)	NA	<.001
Mortality, death per y	0.23	0.30	0.70 (0.59-0.83)	NA	<.001
Heart failure or death	0.90	1.23	0.73 (0.64-0.84)	NA	<.001
Days lost per patient^a
To death	46.2	64.2	0.72 (0.62-0.84)	−17.9	<.001
To heart failure hospitalization or death	50.0	68.4	0.73 (0.63-0.85)	−18.5	<.001
To any-cause hospitalization or death	56.9	74.5	0.77 (0.68-0.88)	−17.5	<.001

Abbreviation: NA, not applicable.

^{^a}

Hazard ratios, 95% CIs, and P values for events were derived using the Andersen-Gill extension of the Cox proportional hazards model. Mortality rates are the Kaplan-Meier estimates of mortality.

^{^b}

Mean, 95% CI, hazard ratio, and P values for comparing days lost were derived from nonparametric bootstrap model.

^{^c}

Inpatient heart failure hospitalization; identified in primary diagnosis code per Centers for Medicare and Medicaid Services.