Table 2. Key Secondary Outcome Measures at 12 Months.
| Measure | Median (Interquartile Range) | ||
|---|---|---|---|
| Participants With Interatrial Shunt Device (n = 21) | Control Participants (n = 22) | P Value | |
| Cardiovascular death | |||
| Available data, No. (%) [95% CI]a,b | 1 (4.8) [0.1-23.8] | 1 (4.5) [0.1-22.8] | >.99 |
| Kaplan-Meier cumulative incidence, % (95% CI)c | 4.8 (0.0-19.2) | 5.0 (0.0-17.6) | .99 |
| Total heart failure–associated admissions/visits, rate per patient-year (95% CI)d | 0.22 (0.08-0.58) | 0.63 (0.33-1.21) | .06 |
| Days alive and without hospitalization | 353 (339-363) | 340.5 (330-353) | .16 |
| Days alive without heart failure–associated hospitalization | 359 (351-365) | 351 (331-365) | .17 |
| Hospitalizations for a heart failure–associated event per patient, No. (%) | .09 | ||
| 0 | 18 (85.7) | 14 (63.6) | .13 |
| 1 | 1 (4.8) | 4 (18.2) | |
| 2 | 0 (0.0) | 1 (4.5) | |
| ≥3 | 2 (9.5) | 3 (13.6) | |
| Change from baseline values at 12 mo | |||
| Surviving participants, No. (%) | 20 (95) | 19 (86) | |
| New York Heart Asssociation class | −1 (−1 to 0) | 0 (−1 to 0) | .08 |
| 6-min Walk time distance | 16 (−57 to 30) | 13.6 (−10 to 72) | .31 |
| Kansas City Cardiomyopathy Questionnaire | |||
| Overall summary score | 10.5 (0.7 to 18.8) | 8.1 (−5.7 to 20.6) | .57e/.78f |
| Clinical summary score | 10.4 (−6.5 to 26.0) | 3.1 (−4.2 to 18.8) | .83e/.89f |
| EuroQol 5-dimension scale score | 0.0 (−0.2 to 0.1) | 0.0 (−0.1 to 0.2) | .81e/.25f |
Includes patients with at least 320 days of follow-up or a cardiovascular death within 365 days postprocedure; P values were calculated using the Fisher exact test.
The 95% CI for available data are based on the 2-sided exact CI of the percentage based on the binomial distribution for each treatment group.
Patients not experiencing cardiovascular death were censored at 12 months or last known follow-up examination; P values were calculated using the log-rank test.
Calculated as the total number of heart failure–associated admissions or visits requiring intravenous diuretics (including all admissions or visits to intensive care units, non–intensive care units, emergency departments, acute care facilities, and outpatient physician or nurse visits for heart failure in which the patient received diuretics) for all participants, divided by the total follow-up in years for all participants; Poisson regression was used to compare rates per patient-year.
The P value comparing treatment arms for changes at month 12 was computed using analysis of covariance adjusted for baseline value.
A mixed-effect repeated-measures model with baseline quality of life as a covariate was used to compare treatments simultaneously across all points through month 12; the underlying within-patient covariance matrix was assumed to be unstructured.