Projects

Completed and ongoing research

Completed Projects
Published, presented, and launched works
SORC Tool Launched
SORCTracks: A Visualization of Temporary Health Service Disruptions Across Alberta
Tung A
Visualization Tool. Systematic Outcomes Research Collaborative. Version 1.5. 2026 Oct 14.
Abstract
Background
Temporary changes in health service availability are posted in many places and are difficult to compare across Alberta.
Methods
SORCTracks reconstructs Alberta Health Services temporary service-disruption postings from archived service-listing snapshots and related news or notice records. Source records are standardized by site, service, and date, duplicate captures are reconciled, recurring schedules are expanded, and overlapping intervals are merged within the site-level summaries. The archive begins in August 2021 and is refreshed through the latest completed month across the analyzed Emergency Department, Obstetrics, Acute Care, Surgery or Operative Capability, and Other Services layers. A separate daily snapshot view shows current public postings.
Results
SORCTracks presents searchable site-level records and summaries of publicly posted disruption hours, episodes, timing, stated reasons, and geographic context. Users can inspect individual source-supported intervals and compare the documented burden across sites, service layers, years, and time-of-week patterns.
Conclusion
SORCTracks puts these records in one searchable place. It describes what Alberta Health Services posted and does not show whether a service was actually unavailable, how far patients travelled, or what happened to them. The historical archive and daily Live view are independent public-data tools and are not Alberta Health Services products.
SORC Tool Launched
SORCShortages: A View of Canadian Drug Shortage and Discontinuation Reports
Tung A
Visualization Tool. Systematic Outcomes Research Collaborative. Version 1.0. 2026.
Abstract
Background
Canadian drug shortage and discontinuation reports are public, but they are not organized for easy comparison across products, report types, and time.
Methods
SORCShortages uses the public Health Product Shortages Canada source and organizes individual reports into comparable episodes. Each report is retained as one episode. Start dates use the actual date when available and otherwise the anticipated date. Observed end dates are used when available, ongoing episodes are right-censored at the dated source snapshot, and estimated end dates are not treated as observed resolution. Shortage and discontinuation reports remain separate, and source-reported Tier 3 status is not reclassified by SORC.
Results
Users can search and compare reports by report type, active status, product, duration, tier, and time. Source-date flags and missing optional fields are retained so that data limitations remain visible.
Conclusion
SORCShortages provides a monthly way to inspect Canadian shortage and discontinuation reporting for research and public understanding. Reported episodes are not a census of unmet patient need, patient harm, or local inventory status. The tool is not affiliated with Health Canada and is not intended for clinical or operational decision-making.
SORC Published
Debiasing and Educational Interventions in Medical Diagnosis: A Systematic Review
Tung A, Melchiorre M.
Systematic Review. University of Toronto Medical Journal. 2023; 100(1).
Abstract
Background
Cognitive bias contributes to diagnostic error, and educational or other debiasing interventions have been proposed to reduce its effect. This review characterized the strategies used, the biases targeted, and their reported effect on diagnostic accuracy.
Methods
MEDLINE, EMBASE, Healthstar, and PsycInfo were searched on April 25, 2022. Eligible studies evaluated an intervention intended to improve diagnostic accuracy. Screening was completed in Rayyan, quality was assessed with the JBI Critical Appraisal Tools, and extraction and synthesis were recorded in Excel.
Results
Of 2,232 records identified, 17 studies were included. The interventions fell into three broad groups: decision-support or other tool use, education about cognitive biases, and education about debiasing strategies. Results were mixed across all three groups. Confirmation, availability, and search-satisfying biases were among the most frequently targeted.
Conclusion
The included interventions showed some evidence of benefit, but findings were inconsistent. No study evaluated diagnostic decision-making in a clinical setting, and none reported long-term follow-up. Future work should examine why effects vary and whether findings from vignettes or physician trainees transfer to routine clinical practice. PROSPERO: CRD42022331128.
SORC Presented
Geographic Context of Publicly Posted Complete Obstetric Service Unavailability in Alberta: A Retrospective Archive-Based Analysis
Tung A, Nurmambetova E, Zhang M.
Original Study. Poster Presentation. Presented at the Federation of Medical Women of Canada 2026 Annual Meeting.
Abstract
Background
Complete local obstetric service unavailability may require patients to seek care outside their community. We characterized the duration of publicly posted complete obstetric service unavailability in Alberta and the proximity of affected sites to another currently listed maternity-related service.
Methods
Using Alberta Health Services archived disruption snapshots and news or notice postings, we reconstructed intervals from August 1, 2021, to June 30, 2026, during which all local obstetric services or obstetric coverage were reported as unavailable. Repeated captures were reconciled, recurring schedules were expanded, and overlapping intervals were unioned within each site and calendar year. For each affected site, road distance and estimated driving time to the nearest other facility matching predefined maternity-related terms were calculated using a local OSRM automobile-routing matrix based on the Geofabrik Alberta OpenStreetMap extract and an AHS directory reviewed July 26, 2026.
Results
Thirty-six sites contributed 262,232 cumulative site-hours, representing 67.2% of reconstructed maternity-disruption site-hours. The median road distance to another currently directory-listed maternity-related facility was 77.7 km (IQR 41.2–89.7), corresponding to 61.5 driving minutes (IQR 34.2–72.7). Six sites were at least 100 road km away, and Slave Lake and Wabasca were at least 120 driving minutes away. Sundre, Rimbey, Three Hills, and Westlock contributed 157,896 site-hours (60.2%) but were 38.0–62.4 road km from another currently listed maternity-related facility.
Conclusion
Duration burden and geographic separation were distinct dimensions of publicly documented obstetric-service instability. Most cumulative hours were concentrated at four sites with shorter estimated driving times, whereas several lower-burden communities were more geographically isolated. Distances based on the current AHS directory provide geographic context only and do not establish historical service availability, actual patient travel, or clinical suitability.
SORC Presented
Prevention Messaging for Micromobility Lithium-Ion Battery Fires: A Comparison of Clinical and Grey Literature
Nurmambetova E, Tung A, Wolfman D, Wong JN, Rosenfield D, Rowe BH.
Original Study. Presented, 11th Annual Canadian Burn Conference 2026.
Abstract
Background and Objectives
Lithium-ion battery fires involving electric micromobility devices can cause severe burn and inhalation injury. Clinical reports and public safety organizations both make prevention recommendations, but their emphasis may differ. This study compared prevention messages in the clinical and grey literature.
Methods
Clinical recommendations were drawn from a prospectively registered systematic review (PROSPERO CRD420261341238). A targeted grey-literature review examined 16 source groups from consumer-safety agencies, fire services, regulators, standards organizations, and manufacturers. Key source pages were rechecked through August 7, 2026. Recommendations were grouped by theme and compared descriptively.
Results
Eleven clinical records were included. Clinical and grey literature overlapped on supervised charging, manufacturer-approved batteries and chargers, damaged-battery warning signs, safe storage, avoiding battery modification, product certification, regulation, injury surveillance, public education, and charging infrastructure. Grey literature additionally addressed professional e-bike conversions, disposal and recycling, product registration and recalls, incident reporting, producer and distributor responsibilities, and building-level fire-safety controls. Health-system preparedness for multiple-casualty incidents was identified in the clinical literature but not as an explicit recommendation in the reviewed grey sources.
Discussion and Conclusion
Clinical reports and public guidance generally reinforce the same immediate safety messages, while grey literature extends prevention into regulatory, product-life-cycle, and building-level measures. Together, they provide complementary targets for prevention messaging. This comparison describes recommendation content and does not assess intervention effectiveness.
Ongoing Projects
Accepted and submitted works
SORC Submitted
Thrombotic Microangiopathy Following Systemic Adeno-Associated Virus Gene Therapy: A Systematic Review of Clinical Presentation and Management
Tung A, Wolfman D, Acharya S, Villarreal Andrade D, Melchiorre M, Kolski H, Sun HL.
Systematic Review. Submitted to European Journal of Haematology. 2026.
Abstract
Objective
To synthesize patient-level evidence on adeno-associated virus-associated thrombotic microangiopathy (AAV-TMA), describing presentation, timing, investigations, management, outcomes, and supporting an initial approach.
Methods
We conducted a systematic review (PROSPERO CRD420251150401) in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and reported in accordance with the updated PRISMA 2020 statement. On 16 September 2025, searches were conducted in Medline, Embase, Global Health, Scopus, Web of Science, and CINAHL. We included reports of post-AAV patients with TMA-like presentations. Two reviewers independently screened records and extracted study- and patient-level data. Study quality was appraised using the Joanna Briggs Institute checklists.
Results
We identified 656 records and included 12 records reporting 22 patients. Exposures were primarily onasemnogene abeparvovec (17/22) and fordadistrogene movaparvovec (3/22). Symptom onset clustered within the 5 to 10 days post-infusion window. Renal-predominant features were common (among cases with reporting: hypertension 15/17, oliguria/anuria 13/17, proteinuria 15/16, hematuria 9/9). Schistocytes were present in 16/16 cases, and thrombocytopenia was severe (median 17×10³/µL). ADAMTS13 activity was reported in 9/22, and none were severely deficient. Most patients received corticosteroids. PLEX was used in 6/17 cases and C5 inhibitors in 10/21 cases. ICU admission occurred in 8/10 cases, renal replacement therapy in 7/22, and mortality in 2/22.
Conclusion
Post-systemic AAV-TMA onset clustered in a narrow period with prominent renal involvement. These findings support phenotype- and ADAMTS13-informed triage, balancing thrombotic thrombocytopenic purpura (TTP) coverage with consideration of complement-directed therapy in severe presentations.
SORC Submitted
Immersive Virtual Reality for Surgical Skills Training: A Review of Evidence Maturity and Effectiveness
Zhu H, Nurmambetova E, Tung A, Zheng B, Armour A.
Evidence Review. Poster Presentation. Submitted to the Canadian Conference for the Advancement of Surgical Education (C-CASE) 2026.
Abstract
Background
Immersive virtual reality (iVR) is increasingly used for technical surgical-skills training. As iVR technology and applications have evolved, it remains unclear how evidence has progressed in methodological rigor, assessment beyond the virtual environment, and relevant outcomes. We characterized the contemporary evidence base and identified evidence gaps.
Methods
We included studies evaluating interactive iVR delivered through a head-mounted display for technical surgical-skills training in medical students, surgical trainees, fellows, or surgeons. Studies published before January 2021 were compared with those from January 2021 onward. We extracted study design, sample size, assessor blinding, assessment setting, retention, implementation reporting, study-quality characteristics, and outcome domains. Findings were synthesized descriptively to characterize evidence maturity.
Results
Thirty-six studies met eligibility criteria, including 16 published before January 2021 and 20 thereafter. Contemporary studies were more frequently randomized (19/20, 95% vs 10/16, 63%) and assessed performance beyond the iVR environment (19/20, 95% vs 12/16, 75%). However, assessment in cadaveric or operating-room settings remained uncommon (5/20, 25% vs 6/16, 38%). Delayed retention was assessed in one study and no contemporary studies, while no study evaluated patient outcomes. Contemporary studies assessed technical performance, accuracy, errors, or completion time.
Discussion
The evidence base has expanded and methodological rigor has improved, but relevant validation remains limited.
Conclusion
Contemporary iVR research demonstrates progress in study design and assessment beyond the virtual environment, while evidence for durable skill retention, operative transfer, implementation, and patient benefit remains insufficient. Future studies should prioritize rigorous comparative designs and relevant outcomes.
SORC Submitted
Burn and Inhalation Injuries Associated With Lithium-Ion Battery Fires in Electric Mobility Devices: A Systematic Review
Tung A, Thornton J, Melchiorre M, Wolfman D, Wong JN, Rosenfield D, Rowe BH.
Systematic Review. Submitted to Journal of Emergency Medicine. 2026.
Abstract
Background
Lithium-ion battery fires involving electric mobility devices have been reported, but their frequency and clinical characteristics have not been well summarized. Burn and inhalation injuries associated with these events, along with their incident characteristics, acute management, and outcomes, were captured in this review.
Methods
A systematic review was conducted in accordance with the Cochrane Handbook and PRISMA 2020 and registered prospectively in PROSPERO (CRD420261341238). MEDLINE, Embase, Scopus, Web of Science Core Collection, and CINAHL were searched up to 2026. Backward and forward citation searching and targeted Google Scholar and Google web searches were also performed. Human studies reporting extractable original data on injuries associated with lithium-ion battery fires or explosions involving electric mobility devices were eligible.
Results
Eleven records comprising 334 patients or casualties from seven countries were included. Most were case reports, case series, or observational reports in adults from burn centres. The most severe cases usually followed indoor micromobility fires, often during charging. Across the most directly comparable hospital-based reports, inhalation injury ranged from 33% to 88%, operative management from 40% to 78%, and mortality from 0% to 11%. Reported burn severity was substantial, with mean total body surface area values ranging from 14.5% to 27.5% in the major cohorts. In contrast, one pediatric short-circuit series described uniformly minor hand burns managed non-operatively without mortality.
Conclusions
Lithium-ion battery fires involving electric mobility devices occur predominantly in adults following an indoor battery fire, often during charging. Fires result in combined burn and inhalation injuries, frequent surgeries, prolonged hospitalizations, and mortality. Prevention efforts should focus on certified devices and chargers and on reducing indoor overnight charging.
Active projects
SORC In Progress
Temporary emergency department service disruptions in Alberta from 2021 to 2026
Tung A, Nguyen V, Zhang M, Beamish L, Cross E.
Original Study. In preparation.
Abstract
Objectives
To characterize the burden, timing, and geographic context of temporary emergency department (ED) service disruptions in Alberta using public service disruption postings.
Methods
We conducted a retrospective descriptive study of Alberta Health Services temporary service disruption listings and public notices from August 1, 2021 through July 31, 2026. Posted disruptions were reconstructed into site-level disruption periods, with overlapping periods combined to avoid double counting. We summarized disruption hours by site, year, time of week, timing structure, and posted reason. Road distance and estimated driving time to the nearest other listed ED were modelled, and concurrent disruption at that ED was measured.
Results
Fifty EDs had 162,721 posted disruption hours. Burden was concentrated at a small number of sites, with the ten highest-burden EDs accounting for 78.2% of all hours. Night accounted for 52.6% of disruption hours despite representing 37.5% of calendar time. Scheduled and Sustained patterns accounted for 83.9% of all disruption hours. Workforce shortages were mentioned in notices covering 69.5% of hours. The median road distance to the nearest other listed ED was 49.4 km, and 16 of 50 affected EDs had at least some concurrent disruption at that site.
Conclusions
Temporary ED service disruption in Alberta was concentrated at a small number of sites and frequently recurred, particularly overnight. Time-based monitoring can capture recurring service loss and concurrent disruption at the nearest listed ED that closure counts alone do not.
SORC In Progress
Rural Maternity Service Instability in Alberta: An Archive-Based Analysis of Publicly Posted Hospital Service Disruptions
Tung A, Zhang M, Nurmambetova E, Kornelsen J.
Original Study. In preparation.
Abstract
Background
Temporary hospital disruptions can affect local delivery care, caesarean capability, epidural access, and maternity backup. Alberta Health Services (AHS) publicly posts these changes, but their cumulative burden is not available in a single system-level dataset. We characterized publicly documented maternity-related hospital capability disruptions across Alberta using archived records.
Methods
We conducted a retrospective descriptive reconstruction of AHS temporary service-disruption postings from August 1, 2021, through July 31, 2026. We included postings explicitly affecting maternity care, labour and delivery, caesarean or operative maternity capability, epidural access, or maternity backup. Snapshot and notice records were standardized by site, deduplicated, and reconstructed into dated intervals. Reopening notices and subsequent postings were used only when they provided source-supported evidence of a changed endpoint. Overlapping intervals at the same site were unioned to estimate unique site-level disruption hours. Capability layers overlap and are therefore not additive.
Results
Across 45 sites, 520 reconstructed disruption periods contributed 394,216.2 unioned disruption hours. Annual burden was highest in 2022 at 92,570.5 h and remained substantial in 2023 (85,798.5 h), 2024 (76,257.5 h), and 2025 (74,182.8 h). 38,536.0 h were recorded through July 31, 2026. The highest cumulative burdens occurred in Sundre, Rimbey, Three Hills, Slave Lake, and Westlock. Disruptions affecting all obstetric services and caesarean, epidural, or surgical backup capability were common, but layer-specific hours are non-additive.
Conclusion
Publicly posted AHS records show recurrent and sustained loss of maternity-related hospital capability across Alberta. Several high-burden sites reflect long posted suspensions, including intervals already active at the archive start and therefore left-censored. Capability-specific reporting better reflects operational changes that may affect local delivery access, while these public records do not establish real-time availability, patient outcomes, or a formal rurality classification.
SORC In Progress
Maintaining Acute Care Access During Wildfire-Related Hospital Disruptions: A Systematic Review
Systematic Review. In preparation.
Abstract
Background
Wildfires can disrupt hospital operations through direct fire exposure, smoke intrusion, infrastructure strain, evacuation risk, and regional patient redistribution. Existing reviews have focused more broadly on wildfire-related emergency department impacts, emergency medical services response, or hospital evacuation across hazards, rather than specifically on how hospitals and acute-care systems preserved service continuity during wildfire-related disruption. This review synthesised peer-reviewed and selected gray-literature evidence on evacuation, transfer, surge response, and early recovery strategies used to maintain acute-care access during wildfire events.
Methods
We conducted a systematic review of studies and operational reports describing wildfire- or bushfire-related disruption to hospitals and acute-care services. We searched MEDLINE, Embase, CINAHL, and Scopus using controlled vocabulary and text words related to wildfires, hospital and critical care settings, and evacuation, transfer, continuity, and recovery processes. We also undertook targeted gray-literature searching using structured Google web searches and searches of relevant hospital, health-system, emergency-preparedness, and government websites. Eligible records included hospital- or system-level reports of evacuation, transfer, service continuity, operational response, or early recovery during real wildfire events. Two reviewers independently screened records and extracted data on setting, disruption type, operational strategies, and reported outcomes. Given expected methodological heterogeneity, findings were synthesised narratively.