Identification of patients at risk of new onset heart failure: Utilizing a large statewide health information exchange to train and validate a risk prediction model.

BACKGROUND: New-onset heart failure (HF) is associated with poor prognosis and high healthcare utilization. Early identification of patients at increased risk incident-HF may allow for focused allocation of preventative care resources. Health information exchange (HIE) data span the entire spectrum of clinical care, but there are no HIE-based clinical decision support tools for diagnosis of incident-HF. We applied machine-learning methods to model the one-year risk of incident-HF from the Maine statewide-HIE. METHODS AND RESULTS: We included subjects aged ≥ 40 years without prior HF ICD9/10 codes during a three-year period from 2015 to 2018, and incident-HF defined as assignment of two outpatient or one inpatient code in a year. A tree-boosting algorithm was used to model the probability of incident-HF in year two from data collected in year one, and then validated in year three. 5,668 of 521,347 patients (1.09%) developed incident-HF in the validation cohort. In the validation cohort, the model c-statistic was 0.824 and at a clinically predetermined risk threshold, 10% of patients identified by the model developed incident-HF and 29% of all incident-HF cases in the state of Maine were identified. CONCLUSIONS: Utilizing machine learning modeling techniques on passively collected clinical HIE data, we developed and validated an incident-HF prediction tool that performs on par with other models that require proactively collected clinical data. Our algorithm could be integrated into other HIEs to leverage the EMR resources to provide individuals, systems, and payors with a risk stratification tool to allow for targeted resource allocation to reduce incident-HF disease burden on individuals and health care systems.

Incremental Value of Deformation Imaging and Hemodynamics Following Heart Transplantation: Insights From Graft Function Profiling.

OBJECTIVES: This study investigated to define graft dysfunction and to determine its incremental association with long-term outcome after heart transplantation (HT). BACKGROUND: Although graft failure is an established cause of late mortality after HT, few studies have analyzed the prognostic value of graft dysfunction at 1- and 5-year follow-up of HT. METHODS: Patients who underwent HT and completed their first annual evaluation with right heart catheterization and echocardiography at Stanford University between January 1999 and December 2011 were included in the study. Hierarchical clustering was used to identify modules to capture independent features of graft dysfunction at 1 year. The primary endpoint for analysis consisted of the composite of cardiovascular mortality, re-transplantation, or heart failure hospitalization within 5 years of HT. The study further explored whether changes in graft dysfunction between 1 and 5 years were associated with 10-year all-cause mortality. RESULTS: A total of 215 HT recipients were included in the study. Using hierarchical clustering, 3 functional modules were identified; among them, left ventricular global longitudinal strain (LVGLS), stroke volume index, and right atrial pressure (RAP) or pulmonary capillary wedge pressure (PCWP) captured key features of graft function. Graft dysfunction based on pre defined LVGLS in absolute value <14%, stroke volume index <35 ml/m2, RAP >10 mm Hg, or PCWP >15 mm Hg were present in 41%, 36%, and 27%, respectively. The primary endpoint at 5 years occurred in 52 patients (24%), whereas 10-year all-cause mortality occurred in 30 (27%) of 110 patients alive at 5 years. On multivariate analysis, RAP (standardized hazard ratio: 1.63), LVGLS (standardized hazard ratio: 1.39), and a history of hemodynamically compromising rejection within 1 year (hazard ratio: 2.18) were independent predictors of 5-year outcome. RAP at 5 years, as well as change in RAP from 1 to 5 years, was predictive of 10-year all-cause mortality. CONCLUSIONS: RAP and LVGLS at the first annual evaluation provide complementary prognostic information in predicting 5-year outcome after HT.

Perioperative mortality is the Achilles heel for cardiac transplantation in adults with congenital heart disease: Evidence from analysis of the UNOS registry.

BACKGROUND: Adults with congenital heart disease may present with end-stage heart failure necessitating orthotopic heart transplant (OHT). We sought to review the United Network for Organ Sharing (UNOS) experience with this unique cohort focusing on surgical outcomes and survival. METHODS: From the UNOS registry, 737 adult congenital heart disease recipients (ACHDR) out of 26,993 OHT patients (2.7%) who underwent OHT were studied to analyze early and late outcomes and compared to non-congenital recipients (NCR) over a 15-year period (2000-2014). RESULTS: More ACHDR underwent OHT in the recent-era (3.1%; 2010-2014) as compared to the initial-era (2.5%; 2000-2004; p = 0.03). ACHDR were more likely female (40% vs. 24%; p < 0.01), younger (35 vs. 53 years; p < 0.01), less likely to have ventricular assist devices (4.2% vs. 19.3%; p < 0.01), more likely to have class-II panel-reactive antibody >10%; p < 0.01, and were listed for a longer time (249 vs. 181 days; p < 0.01). When compared to the NCR in the same period, the ACHDR cohort had longer postoperative length of stay (27 vs. 20 days; p < 0.01), higher operative mortality (11.5% vs. 4.6% p < 0.001), higher incidence of primary graft dysfunction (4.3% vs. 2.6%; p < 0.01), and higher need for dialysis (20% vs. 9%; p < 0.01). Primary graft dysfunction is the most common cause of death in (5.8%) ACHDR. Although short-term survival is poorer, long-term survival of ACHDR was found to be equivalent or better than NCR in long term. CONCLUSIONS: Perioperative morbidity and mortality adversely affects short-term survival in ACHDR. ACHDR who survive the first post-transplant year have equivalent or better long-term survival than NCR.