In the fast-paced, high-pressure environment of an emergency department (ED), it’s easy to assume that any case requiring imaging is of high priority. However, not every imaging order from the ED needs to be considered “STAT,” and distinguishing between high and low acuity patients can be challenging, especially for the radiology teams reviewing these requests. Effectively prioritizing cases is critical to ensuring that truly “STAT” or urgent cases receive the immediate attention they require, while less critical cases are handled as appropriately without overwhelming radiologists.
This is where AI and workflow automation have the potential to make a significant impact. By integrating advanced imaging algorithms and automation rules into hospital workflows, it can offer a new level of prioritization for EDs–allowing healthcare professionals to identify and expedite the most critical cases to relevant subspecialties without unnecessary delays.
The ED Conundrum: Not Every Case is “STAT”
Imagine the scene: a patient arrives at the ED, suspected of having a stroke. The case is, without a doubt, a top priority for the ED, radiology and neurovascular care teams. However, not every patient coming through the doors is having a stroke, and requires that level of immediate attention. The sheer volume of imaging orders placed in the ED can overwhelm radiology departments, especially when many, if not most, orders are marked as “STAT” or urgent, regardless of the patient’s actual acuity level.
On top of that, radiology teams are not always aware of the granular priorities within the ED. For example, a patient needing to be discharged may be prioritized for imaging because they’re taking up valuable bed space. In this case, the priority isn’t medical acuity but rather an operational need to free up resources. The lack of transparency between the ED and radiology creates inefficiencies and leads to misaligned priorities.
Introducing AI and Workflow Automation-Driven Prioritization
Now, imagine if exams could be automatically categorized in the worklist, not solely based on whether it was an ED request, but on the actual acuity of the patient. Instead of the standard blanket “STAT” label, the intelligent worklist could flag high-priority cases and re-categorize lower-priority cases based on predefined rules agreed upon by both the ED and radiology.
Here’s how AI-enhanced workflow automation works:
- Configurable Triage Rules: The worklist can be set to follow customized rules based on the facility’s needs. For example, a facility’s system could use three levels of priority: urgent, high and STAT. Urgent reads are all ED imaging requests, high priority reads are ED escalated exams and STAT is reserved for true emergencies like strokes.
- Enhanced Transparency: Radiology teams can see exactly why a case has been escalated. Meaning it does not just flag the case as high priority–it provides context, such as change in patient condition or critical finding.
- Automated Findings: Clinical AI imaging algorithms can also quickly rule out negatives or alert on suspected positive findings, allowing care teams to facilitate next steps of care for both high and low acuity patients.
The Importance of Ruling Out Negatives
pathologies can be confidently dismissed. These “non-emergent” patients, while not needing immediate critical intervention, still require quick turnaround for discharge or further non-emergency treatment.
This is where AI shines in the emergency department.
Consider this scenario: if AI can quickly rule out a fracture on a chest X-ray, the emergency medicine team can more confidently and rapidly discharge these patients. This speeds up patient flow, reduces bed occupancy and allows the radiology department to focus more on the higher acuity cases. By identifying cases that can be safely deprioritized, AI gives radiology the breathing room to focus on what’s truly critical.
Patient Journey in the ED: A Stroke Example
To bring all this into context, let’s walk through the journey of an ED patient at a health system utilizing AI-enhanced workflow automation.
A 47-year-old female presents to the ED with slurred speech, and a NCCT, CTA and CTP exam is ordered. Suspected stroke patient orders are typically considered STAT however this ED is also interested in whether or not the patient has a bleed, and if not, what actions can be taken.
- Phase 1: Immediate Escalation and Worklist Prioritization
The emergency physician expedites the NCCT exam as a “candidate for thrombolytics” which then prioritizes that read to the top of the worklist. This is in addition to the STAT level of prioritization assigned to the patient.
- Phase 2: AI-Driven Results
AI also rapidly processes the scans, rules out the bleed, and flags a suspected positive finding for an LVO. The care team is notified instantly either with the AI result or through the Radiology report, setting off the next phase of treatment–rapid coordination for thrombolysis and intervention.
- Phase 3: Streamlined Coordination
With AI’s ability to quickly escalate positive LVO cases and notify the ED, radiology and neurology care teams, they can more efficiently communicate and remain in sync. After timely TNK administration, the patient is then quickly transferred and admitted for thrombectomy treatment without any further delays.
Radiology and AI: A Symbiotic Relationship
For radiology, the introduction of AI doesn’t mean a loss of control–it means clarity. Radiologists are still responsible for reviewing and interpreting images, but AI helps them more quickly triage and identify critical findings or rule out non-critical findings.
Moreover, workflow automation ensures that Radiologists are not unnecessarily interrupted or communicated to in an untimely manner. By automatically flagging cases with clear justification, such as the need for discharge, it offers ease and transparency that was previously lacking, fostering better collaboration between the ED and radiology.
The Future of ED Prioritization
AI-enhanced workflow automation can revolutionize how we triage ED patients. By clearly distinguishing between high and low priority cases, it brings a new level of clinical efficiency to radiology departments and speeds up patient flow in the ED. The ability to quickly rule out negatives and prioritize critical cases allows for better resource management and ultimately better patient outcomes.
