# Power of Bayesian Methods, RCTs, and Decision Models to Inform CRT Uncertainties

> **NIH NIH R01** · DUKE UNIVERSITY · 2020 · $704,628

## Abstract

Heart failure (HF) affects nearly 6 million individuals in the US, one of every eight death certificates lists HF
as a primary or contributing cause of death, and HF is associated with a significant reduction in quality of life
for patients and caregivers alike. Cardiac resynchronization therapy (CRT) has represented one of the most
important advances in the care of select patients with HF. Although CRT has substantially improved outcomes
among HF patients, it is widely recognized that approximately 1/3 of patients who undergo this invasive
procedure do not derive clinical improvements. Research on causes of CRT non-response and approaches to
mitigating non-response is a priority in the fields of HF and electrophysiology. Although many important
questions can be answered through national registries and secondary analyses of landmark trials, lack of data
granularity and adjudicated outcomes in registries, and small numbers of patients in key subgroups in trials,
limit these approaches. Our group has successfully used Bayesian statistical methods to combine primary
patient-level data from RCTs of the implantable cardioverter defibrillator. We propose that a similar approach
for CRT trials, combined with stakeholder prioritization and a decision modeling framework, have the power to
overcome many important limitations of existing platforms for CRT research.
 Our long-term goal is to enhance the ability of the NHLBI to provide evidence-based decisionmaking tools
to aid patients, providers, and policymakers in the use of CRT for the treatment of HF. To achieve this overall
goal, we have four specific aims (SA): (1) To work with diverse stakeholders to identify and prioritize timely
clinical and policy questions regarding the comparative effectiveness of CRT; (2) To develop a generalizable
decision modeling framework for the treatment of HF; (3) To use Bayesian statistical techniques to devise a
model for predicting patient and population health and economic outcomes; and (4) To combine the framework
from SA#2, the Bayesian model from SA#3, and patient level data from existing RCTs and registries, to
explore high-priority questions identified in SA#1. Our research will build off our team's expertise and
experience in research prioritization, evidence synthesis, chronic disease modeling, Bayesian statistical
techniques, stakeholder engagement, and methods of disseminating evidence-based decision models to
patients, providers, and policymakers. We will collaborate with principal investigators from existing trials of
CRT to harness the power of patient-level data from over ten years of clinical trials representing nearly 10,000
patients and use registry data to explore whether the outcomes observed in the community are predicted by
available RCT evidence. In an era in which great importance is placed on defending clinical practice with
rigorous supporting evidence, our research brings together stakeholder engagement, decision analytic
methods, Bayesian st...

## Key facts

- **NIH application ID:** 9838292
- **Project number:** 5R01HL131754-04
- **Recipient organization:** DUKE UNIVERSITY
- **Principal Investigator:** Gillian Sanders Schmidler
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2020
- **Award amount:** $704,628
- **Award type:** 5
- **Project period:** 2016-12-15 → 2023-06-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/9838292

## Citation

> US National Institutes of Health, RePORTER application 9838292, Power of Bayesian Methods, RCTs, and Decision Models to Inform CRT Uncertainties (5R01HL131754-04). Retrieved via AI Analytics 2026-05-23 from https://api.ai-analytics.org/grant/nih/9838292. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*