High (expert-level):
Administrative data (HCUP-NIS, HCUP-SID), logistic regression, ICD-10 coding, SAS programming, use of large language models for coding and editing
Moderate (working proficiency, including unrefreshed past expertise):
Python, SQL, ML models, generalized estimating equations (GEEs), mixed models, SEER-Medicare data, R programming
Aspirational / Actively learning:
Snowflake for Epic data, reinforcement learning (RL)
Below is a selection of major research projects as of December 2025. My current work is self-funded, and I welcome opportunities for collaboration. Please feel free to reach out. At this time, I do not have openings for research assistants or coordinators.
Classification of Frailty in Surgical Patients
Role: Principal Investigator
Status: In analysis and manuscript development
This project focuses on identifying limitations in current frailty models used in administrative datasets and developing validated, robust frailty classification systems. The goal is to create risk-adjustment tools that enable more accurate assessment of frailty’s impact on surgical outcomes across state and national patient cohorts and improve hospital quality comparisons.
Key methods and data sources include:
Impact of Frailty on In-Hospital Outcomes of Major Thoracic Surgery
Role: Principal Investigator
Status: Paper 1 in second-round review at Journal of Geriatric Oncology; additional analyses planned
This research examines how frailty influences in-hospital outcomes among patients with lung cancer undergoing major thoracic surgery. Much of the prior literature is dated and limited by small cohorts, heterogeneous populations, and methodological shortcomings. Our work addresses these gaps using contemporary, representative datasets and improved outcome classification.
Innovations include:
Impact of Surgical Urgency on In-Hospital Mortality
Role: Principal Investigator
Status: Paper in second-round review at Anesthesiology
A recent study (Anesthesiology, July 2025) highlighted that the majority of perioperative mortality is associated with non-elective procedures. While the finding is correct, the downstream recommendation—prioritizing urgency over procedural type for resource allocation—oversimplifies clinical reality.
This project contextualizes those findings by demonstrating that urgency alone does not fully characterize perioperative risk. For example, a non-elective laparoscopic appendectomy is not comparable to an emergent aortic root replacement for acute dissection. Our work clarifies how urgency interacts with procedure type, acuity, and case complexity, providing a more nuanced framework for perioperative risk prediction and resource planning.
ICU Utilization Trends for Non-Cardiac Surgical Patients Over Time
Role: Principal Investigator
Status: In analysis
Clinicians working in surgical ICUs have long observed a shift in case mix: fewer routine post-operative admissions and more admissions from the wards, ED, and outside hospital transfers. However, these trends have not been systematically quantified.
This project characterizes:
The goal is to provide empirical evidence for what many clinicians experience daily and to inform planning for ICU staffing, bed allocation, and capacity management.
Device Design: Improving the Ergonomics of Handheld Ultrasound Probes (2020–2023)
Role: Principal Investigator
Status: Not active at this time
Sponsor: MGH Department of Anesthesia, Critical Care, and Pain Medicine Innovation Grants
Funding: $6,000 (2021); $10,000 (2022)
This project involved the design and early testing of an ergonomic accessory to improve stability, needle alignment, and infection control during ultrasound-guided vascular access. The device aimed to reduce operator strain, improve image acquisition, and support consistent needle tip visibility—common challenges in both central and peripheral access procedures.
Although the project is not currently active, it laid the foundation for ongoing interests in procedural ergonomics and device innovation.