Active Projects
Research projects currently underway at WIRE, spanning clinical outcomes, quality improvement, observational research, and research infrastructure.
EUS-Guided FNB Without Rapid On-Site Cytologic Examination (ROSE)
PI: Matthew T. Peller, MD
An institutional quality study evaluating the performance and outcomes of EUS-guided fine needle biopsy (FNB) performed without rapid on-site cytologic examination (ROSE). This study examines whether high-quality tissue acquisition for cytopathologic diagnosis can be reliably achieved without ROSE, which has significant implications for resource utilization and procedural efficiency in endoscopic ultrasound.
- Institutional quality study
- EUS-guided tissue acquisition
- Resource utilization
- Cytopathology outcomes
PANC-GLP: Acute Pancreatitis Risk with GLP-1 Agonist Therapy
PI: Zachary L. Smith, DO, MSc
A target trial emulation examining the risk of acute pancreatitis in patients with obesity who initiate GLP-1 receptor agonist therapy compared with those who do not. This study leverages large-scale real-world data through WIRE's institutional data partnerships to address an emerging clinical question as GLP-1 agonist prescribing continues to expand rapidly.
- Target trial emulation
- GLP-1 receptor agonists
- Acute pancreatitis risk
- Real-world evidence
Target Trial Emulation of NSAIDs and Statins for Recurrent Acute Pancreatitis Prevention
PI: Zachary L. Smith, DO, MSc
A target trial emulation using TriNetX Linked Claims data from 143,742 acute pancreatitis patients across 29 healthcare organizations, comparing NSAID therapy, statin therapy, and no pharmacotherapy for the prevention of recurrent acute pancreatitis. This study applies cloning-censoring-weighting methodology to emulate a three-arm randomized trial, generating preliminary data for a planned large-scale VA Cooperative Studies Program RCT of celecoxib and atorvastatin for pancreatitis prevention.
- Target trial emulation
- TriNetX
- Pancreatitis prevention
- Claims-based research
Colorectal Cancer Screening Outcomes After Positive Multitarget Stool DNA Testing
PI: Zachary L. Smith, DO, MSc
A multi-paper TriNetX Linked Claims research program examining follow-up colonoscopy completion and downstream outcomes after positive multitarget stool DNA (mt-sDNA/Cologuard) testing. In a cohort of 8,419 patients, the central finding challenges the prevailing narrative of patient non-compliance: 72.1% of apparent non-completers had fewer than 90 days of observable claims after their test result, indicating insurance discontinuity rather than refusal. Among patients with adequate observation, the defensible follow-through rate is 78.4%. Additional studies will examine screening trajectory patterns, diagnostic cascade costs, interval colorectal cancer incidence, and comparative effectiveness across screening modalities.
- Colorectal cancer screening
- mt-sDNA testing
- Insurance discontinuity
- Follow-up colonoscopy
- Claims-based research
Reporting Patterns and P-Value Distributions in TriNetX and Nationwide Inpatient Sample Research
PI: Zachary L. Smith, DO, MSc
A preregistered scoping review analyzing p-value distributions and reporting bias indicators across 4,323 TriNetX publications and 2,537 Nationwide Inpatient Sample publications (2017–2026). Using p-curve, z-curve, caliper tests, and density discontinuity analyses, the study finds that structural reporting bias indicators are shared across both corpora and appear to be a general feature of large observational database research. However, TriNetX cohort-level analytics showed significantly greater threshold-specific asymmetry (8:1 vs. 3.9:1 caliper ratio), suggesting platform-based tools may amplify analytic flexibility at the significance boundary.
- Meta-research
- P-value distribution
- Reporting bias
- TriNetX
- Observational research methodology
Redundancy Mapping in Procedural Trials: A Cross-Specialty Meta-Research Analysis
PI: Zachary L. Smith, DO, MSc
A cross-specialty meta-research analysis quantifying trial redundancy in procedural medicine across two domains: endoscopic mucosal resection for colorectal polyps (47 RCTs) and radial versus femoral vascular access for coronary intervention (46+ RCTs). The study measures temporal overlap, duplicated enrollment, and endpoint fragmentation, then models the efficiency gains achievable through hypothetical platform trial architectures with shared controls, adaptive stopping, and harmonized endpoints.
- Trial redundancy
- Platform trials
- Cross-specialty analysis
- Research efficiency
Research Waste in Post-ERCP Pancreatitis Prevention Trials
PI: Zachary L. Smith, DO, MSc
A multi-dimensional analysis of research waste in post-ERCP pancreatitis prevention trials, applying the Chalmers and Glasziou framework to evaluate waste across four stages: question selection, study design, publication practices, and reporting quality. This study quantifies the extent to which PEP prevention research has suffered from redundant trial questions, inconsistent outcome definitions, selective reporting, and poor design quality, despite being one of the most heavily studied topics in therapeutic endoscopy.
- Research waste
- Post-ERCP pancreatitis
- Chalmers-Glasziou framework
- Publication practices
ERCP Outcomes During the Nationwide Intravenous Fluid Shortage
PI: Hassaan Zia, MD
An observational study examining whether the nationwide intravenous fluid shortage impacted clinical outcomes following ERCP procedures. This timely investigation leverages a natural experiment created by the supply chain disruption to evaluate the effect of modified fluid resuscitation strategies on post-ERCP complications, including post-ERCP pancreatitis.
- Natural experiment design
- IV fluid shortage impact
- Post-ERCP pancreatitis
- Supply chain disruption
Project Harmonize
PI: Zachary L. Smith, DO, MSc & Gokulakrishnan Balasubramanian, MD
Project Harmonize is an initiative to create standardized institutional procedural repositories for endoscopic data. By harmonizing data collection across endoscopic procedures, this project establishes the infrastructure needed for rigorous observational research, quality improvement, and future pragmatic trials. The project addresses a critical gap in how procedural data is captured, stored, and made available for research at the institutional level. A central aim of Harmonize is the development of novel, HIPAA-compliant de-identification software and automated pipelines capable of rapidly converting unstructured clinical procedure reports into structured, research-ready datasets — transforming raw procedural documentation into scalable institutional repositories without compromising patient privacy or data integrity.
- Procedural data standardization
- HIPAA-compliant de-identification software
- Automated clinical-to-research data pipelines
- Institutional repositories
- Research infrastructure
- Cross-procedural harmonization