Perioperative Data Cluster

Vision and
Description

The vision of TRIPLE-A is to establish a data-driven platform for research and quality projects within perioperative medicine.

The platform utilizes structured perioperative data from the Electronic Patient Record Sundhedsplatformen to enable comparisons, monitoring, and continuous improvement of local guidelines, as well as testing new interventions and perioperative workflows.

The platform serves as a foundation for high-quality epidemiological research and automated data extraction in clinical studies.

The project is based on the OPIAID database. Data constitutes a complete perioperative dataset ±7 days around surgery with additional relevant variables from one year before to one year after surgery. The database holds 1.1 million surgical procedures.

 

TRIPLE-A was initiated by CEPRA members from the Capital Region and Region Zealand and is run from Bispebjerg and Frederiksberg Hospital. TRIPLE-A collaborates via CEPRA through the perioperative data cluster.

Steering committee: Anders Peder Højer Karlsen, Markus Harboe Olsen, Anders Kehlet Nørskov, Janus Jacobsen, Ole Mathiesen.

Clinician /
Data Scientist Collaboration

TRIPLE-A benefits from data being structured and coded into relevant variables by anesthesiologists with domain knowledge, research experience, and database expertise.

Example 1: TRIPLE-A defines acute kidney injury based on creatinine increases, hourly diuresis, and/or dialysis requirements per KDIGO guidelines.

Example 2: Cumulative postoperative opioid consumption (0-24 hours) is calculated from all opioid administrations automatically converted into morphine equivalents.

Prediction

Together with The IT University of Copenhagen, TRIPLE-A develops predictive models for implementation in EPR systems. Initially, this involves the OPIAID algorithm for individualized opioid dosing in surgical procedures (opiaid.dk). Over the coming years, additional predictive models will be developed for mortality, intraoperative blood pressure thresholds and fluid strategies, expected long-term surgical outcomes, and early detection of patients at high risk for poor surgical outcomes.

aCRF

TRIPLE-A develops an automated CRF for direct data extraction from the EPR, based on coded variables where various data sources can be integrated into relevant variables. This automated CRF format is expected to support the majority of future anesthesiological clinical trials conducted in Region East. Additionally, researchers can associate variables for long-term follow-up directly extracted via the EPR, eliminating the need for manual journal review.

Quality Assurance

The TRIPLE-A model can generate detailed reports summarizing a multitude of perioperative questions, such as:

  1. Generating overview of pain management and pain intensity following specific surgical procedures.

  2. Comparing the risk of organ failure, ICU admission and death between two surgical techniques.

  3. Comparing length of stay and risk of reoperation before and after implementation of epidural catheter treatment as a standard for laparoscopic colon resection.

Expansion to a National Platform

We are developing the model in Region East based on EPR data, after which we will offer to support other regions in deploying this framework nationally, provided their EPR systems are compatible with a TRIPLE-A solution.

Cost Savings

We estimate that TRIPLE-A can reduce costs through efficiencies in both research and quality assurance. Execution of research and quality assurance projects will also become significantly more resource-efficient—economically and timewise—by utilizing TRIPLE-A's automated processes.