Predictors of Use of Pharmacomechanical Therapies versus Standard Therapy in Intermediate-Risk Pulmonary Embolism: A Retrospective Study from Two Medical Centers

Abstract Pharmacomechanical therapy (PMT) is increasingly used in patients with intermediate-risk pulmonary embolism (PE), but the clinical and anatomic factors associated with operator selection of PMT rather than standard anticoagulation remain incompletely characterized. We retrospectively analyzed 202 consecutive patients with intermediate-risk PE defined for this study by right ventricular dysfunction (right ventricle [RV]/left ventricle [LV] ratio > 0.9) and positive troponin. Ninety-seven patients underwent PMT with mechanical thrombectomy or catheter-directed thrombolysis, and 105 received standard therapy. Baseline characteristics, hemodynamics, PE distribution, and 30-day events were extracted. A multivariable binary logistic regression model was developed to identify factors associated with PMT selection, with internal validation using a 70/30 training–test split. Patients selected for PMT differed substantially from controls at baseline, including higher admission heart rate, higher pulmonary artery pressure, higher RV/LV ratio, lower systolic blood pressure, and a greater frequency of saddle PE. Concurrent deep vein thrombosis was documented in 0/97 PMT patients and 44/105 controls (p < 0.0001). In multivariable analysis, admission heart rate (odds ratio [OR] 1.0315, 95% confidence interval [CI] 1.0051–1.0585; p = 0.019), pulmonary artery pressure (OR 1.0438, 95% CI 1.0103–1.0785; p = 0.010), and PE distribution as a multilevel factor (overall p = 0.019) were associated with PMT selection. Training and test areas under the curve were 0.8049 and 0.8307, respectively. Thirty-day mortality was 5/97 (5.2%) in the PMT group and 5/105 (4.8%) in controls; serious bleeding from procedure to discharge occurred in 7/97 (7.2%) and 2/105 (1.9%), respectively. These event data are descriptive and were not adjusted for treatment selection bias. In this retrospective two-center cohort, higher admission heart rate, higher pulmonary artery pressure, and PE distribution were associated with operator selection of PMT. The model showed good discrimination in internal training and test datasets. Because treatment was not randomized and important baseline imbalances were present, observed clinical outcomes should be considered descriptive and should not be interpreted as evidence of comparative efficacy or safety of PMT versus standard therapy.