Example Use Case
How early risk identification enables proactive care and reduces avoidable hospitalizations.
Oncology Care
In oncology, avoidable hospitalizations often occur between treatment cycles.
Patients may begin to deteriorate days or weeks before a hospitalization, but early warning signs are not always identified in time.
SmithAI’s P2A framework helps surface those signals earlier and connects them to a structured outreach workflow.
The Challenge
Cancer patients frequently experience:
- Subtle symptom progression
- Lab value changes
- Complications related to treatment
These signals often exist in routine clinical data, but without a structured system, they may not trigger timely intervention.
As a result:
- Patients deteriorate at home
- Emergency department visits increase
- Avoidable hospitalizations occur
How P2A Applies
Using routine clinical and laboratory data, P2A identifies patients at elevated risk of hospitalization in the near term.
Each day, high-risk patients are surfaced in a nurse outreach queue.
Care teams can then:
Contact Patients Earlier
Assess Symptoms Before Escalation
Adjust Medications if Needed
Coordinate Care With Oncology Providers
The workflow fits within existing care management processes.
What Changes
Without P2A, care is often reactive.
With P2A, care becomes proactive.
Instead of responding after deterioration, teams can intervene earlier when patients are still stable enough to manage outside the hospital.
Example Workflow
Day
0
Patient risk score increases based on lab and clinical data
Day
1
Patient appears in nurse outreach queue
Day
1-2
Nurse contacts patient and identifies early symptoms
Day
2-3
Care plan adjusted or follow-up scheduled
Potential hospitalization avoided
Measurable Impact
Early intervention can lead to:
- Fewer avoidable hospitalizations
- Reduced emergency department utilization
- Improved patient experience
- Better alignment with quality metrics
This includes improvements in measures such as oncology-related hospitalization rates and overall care quality performance.
Designed for Real Clinical Use
This is not a theoretical model.
P2A is designed to operate within real-world clinical environments:
- Uses existing data
- Fits current workflows
- Supports, not replaces, clinical judgment
- Includes governance and monitoring
The result is a practical, scalable approach to improving patient outcomes.
Explore What This Could Look Like in Your System
We work with health systems to define use cases, validate data, and implement P2A in a controlled, measurable way.