Twenty-Four-Hour Mean Arterial Pressure and Pulse Pressure Are Associated with Hospitalization Duration at Delivery in Pregnant Women Referred for Cardiovascular Risk Assessment

Figures

• 

  Figure 1. Comparison of selected clinical and blood pressure-derived variables according to hospitalization duration. Women were classified as having ≤4 days or >4 days of hospitalization after delivery, according to the operational cohort-specific definition of above-median hospitalization duration. Blue and green boxes correspond to the ≤4 days and >4 days groups, respectively. In the boxplots, circles indicate outliers located more than 1.5 times the interquartile range from the box, whereas asterisks indicate extreme values located more than 3 times the interquartile range from the box. (a) Office and out-of-office blood pressure indices; and (b) nighttime blood pressure dipping. Results marked with ** and * reached a p-value of less than 0.05 and 0.1. The numerical values of all indices are shown in Table 2. BP—blood pressure; SBP—office systolic BP; 24-hSBP—24 h SBP; dSBP—daytime SBP; nSBP—nighttime SBP; DBP—office diastolic BP; 24-hDBP—24 h DBP; dDBP—daytime DBP; nDBP—nighttime DBP; mmHg—millimeter of mercury; %—percentage.

• 

  Figure 2. Receiver operating characteristic (ROC) curves for above-median hospitalization (>4 days). ROC curves are presented as exploratory analyses of apparent discriminative performance within the study cohort. (a) ROC curves for three multivariable logistic regression models: a clinical model without blood pressure indices (Model 1, blue line), a model including 24 h mean arterial pressure (24-hMAP; Model 2, green line), and a model including 24 h pulse pressure (24-hPP; Model 3, red line). The area under the curve (AUC) was 0.836 (SE 0.036; 95% CI 0.765–0.907) for Model 1, 0.866 (SE 0.033; 95% CI 0.801–0.931) for Model 2, and 0.859 (SE 0.035; 95% CI 0.791–0.926) for Model 3. At a predefined false-positive rate of 30%, the 24-hMAP model achieved a sensitivity of approximately 90%. In contrast, at a predefined false-negative rate of 30%, the 24-hPP model achieved a specificity approaching 90%. (b) ROC curves for individual office and ambulatory blood pressure-derived indices. The AUC values were 0.637 (SE 0.049; 95% CI 0.542–0.733; p = 0.010) for office MAP, 0.484 (SE 0.054; 95% CI 0.379–0.588; p = 0.759) for office PP, 0.743 (SE 0.048; 95% CI 0.648–0.837; p < 0.001) for 24-hMAP, and…

• 

  Figure 3. Kaplan–Meier analysis of time to hospital discharge according to 24 h mean arterial pressure. Kaplan–Meier curves showing (a) survival probability (i.e., probability of remaining hospitalized) and (b) cumulative risk of hospital discharge according to quartiles of 24 h mean arterial pressure (MAP) measured during pregnancy. Patients in higher MAP quartiles exhibited a lower probability of discharge over time and a slower increase in cumulative discharge rates, indicating longer hospital stays. A progressive separation of curves across quartiles suggests a dose–response relationship between increasing MAP and delayed discharge. Differences between groups were assessed using the log-rank test. Median length of stay increased from 3.0 days in the lowest quartile (SE 0.23; 95% CI 2.55–3.45) to 6.0 days in the highest quartile (SE 0.70; 95% CI 4.63–7.38), suggesting a dose–response relationship. Differences between groups were statistically significant (log-rank p < 0.05).