Skip to main content
. 2019 Jun 11;321(22):2193–2202. doi: 10.1001/jama.2019.7108

Figure 2. One-Year Cumulative Event Rates of All-Cause Mortality or Stroke Among Patients With Bicuspid and Tricuspid Aortic Stenosis in Unadjusted and Propensity-Matched Cohorts.

Figure 2.

The P values were obtained from Cox proportional hazards models. In the unadjusted cohort, the median follow-up for the bicuspid group was 44 days (interquartile range [IQR], 31-365 days) and for the tricuspid group, 55 days (IQR, 32-365 days). In the propensity score–matched cohort, the median follow-up for the bicuspid group was 44 days (IQR, 31-365 days) and for the tricuspid, 53 days (IQR, 32-365 days). This is a continuous registry in which all patients will not have reached the 1-year follow-up at any given time point. At 1 year, there were missing data from 1586 patients in the bicuspid group (660 had not completed their first year of follow-up at the time of the analysis; 926, unknown) and 1514 patients in the tricuspid group (684 had not completed their first year of follow-up at the time of the analysis; 830, unknown), which were further assessed in the Centers for Medicare & Medicaid Services linked sensitivity analyses (eFigure 3 in the Supplement).