Machine-Learning-Based In-Hospital Mortality Prediction for Transcatheter Mitral Valve Repair in the United States.

BACKGROUND: Transcatheter mitral valve repair (TMVR) utilization has increased significantly in the United States over the last years. Yet, a risk-prediction tool for adverse events has not been developed. We aimed to generate a machine-learning-based algorithm to predict in-hospital mortality after TMVR. METHODS: Patients who underwent TMVR from 2012 through 2015 were identified using the National Inpatient Sample database. The study population was randomly divided into a training set (n = 636) and a testing set (n = 213). Prediction models for in-hospital mortality were obtained using five supervised machine-learning classifiers. RESULTS: A total of 849 TMVRs were analyzed in our study. The overall in-hospital mortality was 3.1%. A naïve Bayes (NB) model had the best discrimination for fifteen variables, with an area under the receiver-operating curve (AUC) of 0.83 (95% CI, 0.80-0.87), compared to 0.77 for logistic regression (95% CI, 0.58-0.95), 0.73 for an artificial neural network (95% CI, 0.55-0.91), and 0.67 for both a random forest and a support-vector machine (95% CI, 0.47-0.87). History of coronary artery disease, of chronic kidney disease, and smoking were the three most significant predictors of in-hospital mortality. CONCLUSIONS: We developed a robust machine-learning-derived model to predict in-hospital mortality in patients undergoing TMVR. This model is promising for decision-making and deserves further clinical validation.

Pressor response to initial blood pressure monotherapy is associated with cardiovascular mortality.

BACKGROUND: A paradoxical pressor systolic response to initial antihypertensive monotherapy has been observed in 8% of hypertensive patients. The long-term consequences of this finding are unknown. METHODS: We included 945 hypertensive patients with baseline systolic blood pressure (SBP) ≥140mm Hg. A 4-week washout period free of antihypertensive drugs was allowed for those already on treatment at entry. Mortality outcomes were ascertained from the National Death Index. Subjects were categorized by SBP response into depressor (≥10mm Hg fall), nonresponder, and pressor (≥10mm Hg rise) categories. RESULTS: There were 268 fatalities. Of these, 100 (37%) were from cardiovascular disease (CVD), of which 70 (70%) were due to coronary artery disease (CAD). A pressor response was associated with higher SBP at 1 year compared with the nonresponder or depressor response (141 vs. 136 vs. 136mm Hg). CVD mortality was greater in pressors than depressors (hazard ratio (HR) = 3.0; 95% confidence interval (CI) = 1.4-6.4; P = 0.004], as was CAD (HR = 3.1; 95% CI = 1.4-6.8; P < 0.01) and all-cause mortality (HR = 1.7; 95% CI = 1.1-2.6; P = 0.02), after adjusting for 1-year SBP and other possible confounders. CONCLUSIONS: We found the incidence of a pressor response to monotherapy at 3 months was significantly, specifically, and independently associated with higher subsequent cardiovascular mortality.

Enduring direct association of baseline plasma renin activity with all-cause and cardiovascular mortality in hypertensive patients.

BACKGROUND: Plasma renin activity (PRA) has been associated with cardiovascular disease mortality (CVD) events among hypertensive patients. We now report a long-term follow-up to assess the enduring association of PRA to CVD and all-cause mortality. METHODS: Participants (3,791) in a systematic hypertension treatment study had entry systolic blood pressure (BP) ≥140 mm Hg and mean age 52. CVD and all-cause mortality was ascertained for mean of 16 years. Pretreatment PRA was analyzed as a continuous variable, and by tertiles. The 10-year Framingham score was similarly examined. Hazard ratios (HRs) were estimated from multivariate Cox proportional hazard models. RESULTS: There were 804 deaths, and 360 (45%) were CVD. PRA was associated with all-cause mortality and CVD, but not cancer or non-CVD. Although T3 had lower mean baseline and follow-up systolic BP than T1, (146 vs. 152 mm Hg (P < 0.001) and 135 vs. 139 mm Hg (P < 0.001), respectively), T3 had 37% higher all-cause mortality (HR: 1.37, 95% confidence interval (CI): 1.15-1.63, P < 0.001) and 70% higher CVD mortality (HR: 1.70, 95% CI: 1.29-2.23, P < 0.001) after adjustment. The difference between T3 and T1 in mortality from coronary artery disease and myocardial infarction was more pronounced than for all CVD. PRA also significantly improved CVD risk estimation provided by Framingham. CONCLUSIONS: These findings extend and reinforce previous evidence that pretreatment PRA has a significant, independent, specific, and direct long-term association with CVD mortality. Moreover, PRA adds significantly to risk identified by the Framingham score.