Data Service

The Cardiovascular Medicine Analytics Center connects cardiovascular investigators with the data expertise needed to develop feasible, rigorous research. We support projects using clinical trial data and approved electronic health record resources, from cohort and variable definitions through data preparation, statistical analysis, and publication. Access to each resource follows its institutional or study-specific approval process. Contact us to discuss your research question, available data, and collaboration needs.

Comparison of Surgical and Medical Treatment for Congestive Heart Failure and Coronary Artery Disease(STICH, 2002-2007; STICHES till 2015) is a multicenter, international, randomized trial that addresses two specific primary hypotheses in patients with clinical heart failure (HF) and left ventricular (LV) dysfunction who have coronary artery disease amenable to surgical revascularization.

The first hypothesis (N=1,212) is that restoration of blood flow by means of coronary revascularization recovers chronic LV dysfunction and improves survival, as compared to intensive medical therapy alone. The second hypothesis (N=1,000) is that surgical ventricular restoration (SVR) to a more normal LV size improves survival free of subsequent hospitalization for cardiac cause compared to CABG alone. Patients eligible for either medical therapy or CABG, but not eligible for the SVR procedure (Stratum A), were randomized in equal proportions to medical therapy alone versus CABG plus medical therapy. Patients eligible for all three therapies (Stratum B) were be randomized in equal proportions to medical therapy alone, CABG plus medical therapy, and CABG plus SVR plus medical therapy. Patients whose severity of angina or CAD makes them inappropriate for medical therapy alone (Stratum C) were randomized in equal proportions to CABG plus medical therapy versus CABG plus SVR plus medical therapy.

More details about STICH [and STICHES].

ToRsemide compArisoN With furoSemide FORManagement of Heart Failure (TRANSFORM-HF, 2018-2022) is a large-scale, pragmatic, randomized, unblinded clinical effectiveness study comparing torsemide versus furosemide as treatment for heart failure. A total of N=2,859 hospitalized patients with heart failure were recruited at 60 hospitals in the United States. Heart failure need not be the reason for hospitalization. Patients were randomized 1:1 to either oral torsemide OR oral furosemide prior to hospital discharge. Oral dosing of torsemide compared to furosemide was 1mg:2mg. The specific loop diuretic dose was at the discretion of the treating physician with the above noted conversion.

The primary objective of the TRANSFORM-HF study is to compare the treatment strategy of torsemide versus furosemide on clinical outcomes over 12 months in patients with heart failure who were hospitalized. Patients had 30-day, 6-month and 12-month follow-up phone contacts for assessments of vital status, interval hospitalizations, adherence, and quality of life. Subsets of patients enrolled early in the study had additional phone contacts beyond 12 months, up to 30 months, at six-month intervals, to document vital status.

More details about the study.

CMAC helps investigators turn a clinical question into a well-defined, analysis-ready study using Yale New Haven Health electronic health record data. Working with Yale’s Joint Data Analytics Team (JDAT) and Research Informatics Office, we support cohort definition, eligibility criteria, variable specification, and preparation of research data requests. For approved projects, we develop reproducible extraction and analysis workflows using provisioned data, including OMOP common data model resources in secure research environments such as Kamino. Our work includes data cleaning, code and terminology mapping, longitudinal data organization, missing-data assessment, and clinical review of study variables and outcomes. We then support statistical analyses, including survival and longitudinal modeling, comparative effectiveness research, and prediction model development. Project-specific support can also include machine learning and natural language processing to identify information in clinical notes, with appropriate approvals and validation. Data availability and access depend on the study protocol, institutional review, JDAT authorization, and the approved project scope.

Yale clinical data resources and research requests.

REVIVED-BCIS2 (Revascularization for Ischemic Ventricular Dysfunction) was a multicenter, randomized trial that enrolled 700 patients in the United Kingdom with severe ischemic left ventricular dysfunction, extensive coronary artery disease, and demonstrable myocardial viability. The trial compared percutaneous coronary intervention (PCI) plus optimal medical therapy with optimal medical therapy alone. Its primary outcome was death from any cause or hospitalization for heart failure; additional assessments included ventricular function and patient-reported health status. CMAC investigators have contributed to collaborative analyses of REVIVED-BCIS2 and to harmonized individual participant data analyses alongside STICH/STICHES. This work supports research on prognosis, clinical outcomes, and treatment differences across patient groups. CMAC can help collaborators develop secondary research questions, harmonize variables and outcomes, and design rigorous analyses that account for differences between trial populations and treatment eras. Collaboration and data access are subject to the trial investigators’ review, applicable agreements, and project approvals.

REVIVED-BCIS2 trial publication. Collaborative STICHES and REVIVED analysis

• EPIC Cosmos
• LUMEDX
• PROMISE data
• ECG Warehouse
• Cedars-Sinai Collaboration

The Cardiovascular Medicine Analytics Center offers computational solutions for research projects. We handle projects that involve large datasets or require sophisticated computational methods, such as Bayesian statistics, among other complex computational needs. Our center is equipped to consult and provide tailored solutions that will drive your research forward. Whether you're dealing with vast amounts of data or navigating the intricacies of advanced statistical models, our expertise is at your service. We encourage researchers facing computational challenges in their projects to leverage this unique opportunity. Please reach out to us for consultation for computationally intensive projects.

• Yale Center for Research Computing