AI-Driven Referral Intake: How a 65-Census Hospice Scaled Admissions Without Adding Staff
Business Type
Hospice Agencies
Location
Thousand Oaks, CA
Size
Mid-sized agency with a 65-patient average daily census
Challenge
Losing time-sensitive hospital referrals due to manual data entry and slow clinical triage.
The Challenge
The agency was struggling with 'referral leakage.' When local hospitals or SNFs sent referral packets via fax or encrypted email, the intake coordinator spent 3-5 hours manually entering demographics, medications, and history into MatrixCare. In the hospice industry, the first agency to respond usually wins the admission; by the time the clinical director reviewed the file, the family had often already signed with a competitor.
Additionally, weekend and on-call coordination was fragmented. On-call nurses were receiving raw PDF attachments on their mobile devices, making it difficult to quickly assess patient acuity or medication needs (like morphine or oxygen requirements) before arriving at the home. This led to delays in the first nursing visit and increased stress for grieving families during the critical 24-hour admission window.
The Solution
Services Used
- • Intelligent Document Processing (IDP)
- • Custom LLM Workflow Automation
- • MatrixCare API Integration
Timeline
6 weeks
Integrations
- • MatrixCare
- • Twilio (SMS Alerts)
- • OpenAI GPT-4o (Clinical Summarization)
- • RightFax
The Results
18 hours/week
Time Saved
$4,200/month
Cost Saved
22% increase in admission conversion
Revenue Impact
12 minutes
Average Response Time
99.2%
Data Entry Accuracy
"Read Laboratories transformed our intake process. Instead of digging through 40-page PDFs, our clinical directors get a concise 1-paragraph summary of the patient's status and immediate needs within minutes of the referral arriving."
— Clinical Director, Southern California Hospice Group
Implementation Timeline
During the first two weeks, Read Laboratories mapped the agency's MatrixCare custom fields and analyzed common referral formats from local health systems. By week four, we deployed an AI-powered 'Intake Assistant' that extracted ICD-10 codes and patient demographics directly into a staging area. The final two weeks focused on configuring Twilio alerts that notify on-call nurses via SMS with a summarized 'cheat sheet' of the patient's clinical status as soon as the referral is processed.
FAQ
Is the AI HIPAA compliant?
Yes. We utilize enterprise-grade AI instances with Business Associate Agreements (BAAs) in place, ensuring that PHI is encrypted and never used to train public models.
Does this work with handwritten physician notes?
Our Intelligent Document Processing (IDP) engine uses advanced OCR that can accurately transcribe most handwritten physician signatures and clinical notes found in standard hospice referral packets.
Do we still need an intake coordinator?
Yes. The AI acts as a co-pilot. It handles the 'grunt work' of data entry and summarization, allowing your intake coordinator to focus on family communication and clinical validation rather than typing.
Can this integrate with WellSky or Axxess instead of MatrixCare?
Absolutely. We can build custom connectors for any EMR that offers an API or allows for structured data imports via HL7 or SFTP.
Want results like these?
Free consultation. We'll look at your specific situation and tell you exactly what's possible.
Book a Call →Serving Hospice Agencies businesses nationwide. Based in Westlake Village, CA.