Reducing Adverse Events Following Thoracic Surgery

Thoracic surgery is high-stakes, high-risk surgery, with 30-60% incidence of postoperative adverse events (AEs) depending on surgical procedure. Dr. Seely and team defined all thoracic surgery AEs and their severity (see https://ottawatmm.org/) (Seely et al., Ann Thor Surg, 2010) and have long monitored all AEs after all thoracic surgery in Ottawa. They built software to enable AE data collection, harmonized with international organizations (Sigler et al., JTCVS Open, 2021) and helped demonstrate how AEs augment the risk of mortality, worsen length of stay, readmissions, and oncologic outcomes, impair patient experience, and raise hospital costs (Grigor et al., Ann Thor Surg, 2017).  To help utilize this data to improve care, Dr. Seely led the initiation of positive deviance (PD) quality improvement (QI) interventions, which combine best experience (i.e. PD) with best evidence to generate recommendations to improve care. This collegial and novel QI approach demonstrated improvements in AEs in Thoracic surgery (Ivanovic et al., Ann Thor Surg, 2015; Ivanovic et al., J Healthc Qual, 2018) and OR times in Orthopedics (Gold et al., Can J Surg, 2023), in Ottawa. Embracing the importance of a national learning network (Seely et al., Can J Respir Crit Care Sleep Med, 2021), 14 Canadian thoracic centres spanning 7 provinces (two-thirds of Canadian thoracic surgeons) now collect identical AE data using Ottawa-developed definitions and software, with 5 more centers working to join (see https://www.canadianthoracicsurgeons.ca/cats-qipdb/). Dr. Seely and team led the implementation of national PD QI interventions to utilize this data to improve care, and results from multiple studies to date demonstrate reductions in AEs (Seely et al., BMJ Open Qual, 2023; Jones et al., BMJ Open Qual, 2023) Dr. Seely and Co-PIs Drs. Jamie Brehaut and Monica Taljaard were awarded a CIHR Project Grant to evaluate this national PD QI program on care and further improve it.

Variability-derived Predictive Clinical Decision Support

Dr. Seely and team and many others have shown that preserved degree and complexity of heart and respiratory rate variability (HRV & RRV) are associated with health, and illness is characterized by a reduction in variability. Dr. Seely initiated his research to utilize commonly discarded monitored waveform data to improve care, by developing variability-derived predictive models and clinical decision support (CDS) tools (Ahmad et al., PLoS ONE, 2009; Bravi et al., Biomed Eng Online, 2011; Bravi et al., PLoS One, 2012; Seely et al., Crit Care, 2014). Dr. Seely and his team have further developed and implemented and evaluated three variability-derived CDS tools, helping to establish a roadmap for the evolution of monitoring (Seely et al., Crit Care, 2024).  

Timely and safe extubation (i.e. endotracheal tube removal) in critically ill patients is vitally important, as prolonged mechanical ventilation and failed extubation (i.e. re-intubation<48 hrs; 15% incidence) are associated with increased, morbidity, mortality, costs, intensive care unit (ICU) stays. To address this challenge, Dr. Seely and his team developed Extubation Advisor (EA), a CDS software tool which provides a standardized and improved assessment of extubation readiness based on RRV (Seely et al., Crit Care, 2014). Dr. Seely, with co-PI Dr. Karen Burns, completed an observational study (n=117; 2 ICUs) and an interventional Phase I study (n=29; 3 ICUs) (Sarti et al., BMJ Open, 2021; Hryciw et al., J Intensive Care Med, 2024) and they are currently conducting a CIHR-funded multicentre pilot RCT (10 sites, n=110) of Extubation Advisor nearing completion. EA became a patented regulatory approved commercial product in Europe and Health Canada in 2023, commercialized by the company Dr. Seely founded, Therapeutic Monitoring Systems, with patent licensing agreements with the OHRI.  

Donation after circulatory determined death (DCD) allows for organ retrieval after death determination after withdrawal of life sustaining measures (WLSM), with subsequent transplantation; however, uncertainty in predicting time to death leads to decreased efficiency, interest and acceptability of DCD programs, distress for donor families and waste healthcare resources. In 2022, Dr. Seely and team derived and validated a variability-derived predictive model to provide a probability of dying within 30, 60 and 120 min after WLSM (Scales et al., Crit Care Explor, 2022). Dr. Seely and team, along with co-PI Dr. Sonny Dhanani, developed Donation Advisor (DA), which identifies DCD candidates likely to become successful organ donors and objectively characterizes organ ischemia during DCD. They are currently performing a multicenter feasibility RCT of DA.  

Given infection is a common presentation in emergency departments (EDs), and given its potential to develop into life-threatening sepsis, ED and ICU physicians have the difficult task of prognosticating these patients daily. A critical decision for physicians is whether to admit or transfer a patient to a high-volume ICU with improved outcomes, yet unnecessary hospital transport is both costly and harmful. Based on HRV and bloodwork in the ED, Dr. Seely and team developed and implemented a novel disposition CDS tool called Sepsis Advisor (SA) which quantifies the risk of future deterioration in patients with infection based on HRV and bloodwork (Barnaby et al., Shock, 2018). This SA tool has been submitted to Health Canada in 2025 for regulatory approval, and is proposed for hospital and region wide implementation in Ottawa.

Understanding Variability

Dr. Seely has long been interested in theoretical research, initially hypothesizing the importance of variability monitoring in critical care within the framework in complex systems science (Seely et al., Crit Care Med, 2000), then focussing on the importance of embracing uncertainty (Seely, Perspect Biol Med, 2013), and more recently, on the vital importance of human entropy production (Brodeur et al., Entropy, 2023; Seely, Entropy 2020).  Dr. Seely has hypothesized that multiscale self-similar fractal structure contained in HRV & RRV is related to its capacity to maximize entropy production, and that health is characterized by robust entropy production at rest combined with the ability to elevate it in response to exercise or heat stress (i.e. both function & adaptability). Supported by a New Frontiers Research Fund Exploration grant, Dr. Seely and co-PIs Drs. Glen Kenny (physiologist) and Andre Longtin (physicist) are pioneering groundbreaking research.