Case Study: Heart Specialists Queensland

How a Cardiology Practice Achieved Specialty-Grade AI Documentation

Practice Type: Specialist Cardiology
Location: Brisbane, Queensland
Providers: 4 Cardiologists
Patient Volume: 60+ patients/day
Implementation: March 2025

---

Executive Summary

Heart Specialists Queensland implemented Regenemm Voice to address complex specialist documentation needs that generic AI scribes couldn't handle. The cardiology-specific AI agents delivered:

• 75% reduction in documentation time
• Clinical-grade notes with appropriate cardiology terminology
• Automated risk scoring (CHA₂DS₂-VASc, HEART Score)
• 95% first-draft acceptance rate (minimal editing)
• Same-day letters to referring GPs

---

The Challenge

Specialist Documentation Complexity

Cardiology documentation is among the most complex in medicine. Heart Specialists Queensland faced unique challenges:

"We tried three different AI documentation tools before Regenemm. They all failed on cardiology-specific content. The notes sounded like a medical student wrote them—generic and missing critical clinical nuance."

— Dr Andrew Mitchell, Director

Key Pain Points:

Issue	Impact
Generic AI outputs	Required 15-20 min editing per note
Missing specialty terminology	CHA₂DS₂-VASc, SYNTAX scores not captured
Complex medication regimens	Cardiac medications frequently misunderstood
ECG/imaging interpretation	Not integrated into documentation
Lengthy referral letters	20+ minutes per specialist letter

Why Generic AI Failed

Previous AI documentation attempts failed because:

1. No cardiology training: AI didn't understand cardiac terminology
2. No risk calculators: Manual entry of risk scores still required
3. Poor medication handling: Complex cardiac drug regimens confused the AI
4. No clinical reasoning: Just transcription, no clinical intelligence

---

The Solution

Cardiology-Specific AI Agent

Regenemm's specialty agent architecture was the differentiator:

"When Regenemm demonstrated their cardiology agent, we saw immediate differences. It understood what a 'positive ETT with 2mm ST depression in leads V4-V6' meant clinically. Other systems just transcribed the words."

— Dr Lisa Park, Interventional Cardiologist

Specialty Features:

Feature	Capability
Cardiology vocabulary	5,000+ cardiac terms, procedures, devices
Risk calculators	Automatic CHA₂DS₂-VASc, HEART, GRACE scoring
Medication expertise	Understands cardiac drug interactions and protocols
Investigation integration	Structures ECG, echo, cath findings appropriately
Clinical reasoning	Suggests differentials based on presentation

---

Sample Output: Before vs After

Before (Generic AI):

The patient is a 68 year old man with chest pain. He has some risk factors. ECG showed changes. Will do more tests and follow up.

After (Regenemm Cardiology Agent):

CARDIOLOGY CONSULTATION NOTE

PRESENTING COMPLAINT
68-year-old male presenting with exertional chest tightness, 
onset 3 weeks ago, progressive. Character: substernal pressure, 
radiating to left arm. Triggered by climbing stairs (2 flights), 
relieved by rest within 5 minutes. Associated with dyspnoea. 
No rest pain, syncope, or palpitations.

CARDIOVASCULAR RISK FACTORS
• Hypertension (on Ramipril 10mg, controlled)
• Type 2 Diabetes (HbA1c 7.2%, on Metformin 1g BD)
• Dyslipidaemia (LDL 2.8 on Atorvastatin 40mg)
• Ex-smoker (20 pack-year history, quit 2019)
• Family history: Father MI age 55

EXAMINATION
• BP 138/82, HR 72 regular, SpO2 98% RA
• JVP not elevated
• Heart sounds dual, no murmurs
• Chest clear
• No peripheral oedema
• Peripheral pulses palpable

INVESTIGATIONS
ECG: Sinus rhythm, rate 70. Minor T wave flattening V5-V6. 
No acute ischaemic changes.

ASSESSMENT
Likely stable angina on background of multiple cardiovascular 
risk factors. Clinical presentation consistent with flow-limiting 
coronary artery disease.

HEART Score: 5 (Intermediate risk)
• History: Moderately suspicious (1)
• ECG: Non-specific changes (1)  
• Age: >65 (2)
• Risk factors: ≥3 (1)
• Troponin: Pending (0)

PLAN
1. Urgent CT coronary angiogram
2. Commence Aspirin 100mg daily (if not contraindicated)
3. GTN spray PRN for symptom relief
4. Optimise risk factor management
5. Review in 1 week with CTCA results
6. If significant disease, proceed to invasive angiography

---

Results

Documentation Efficiency

Metric	Before	After	Improvement
Documentation time per patient	18 min	4.5 min	75% reduction
Editing time per note	12 min	2 min	83% reduction
First-draft acceptance rate	40%	95%	+55 percentage points
GP letter turnaround	5-7 days	Same day	85% faster

Clinical Quality

Metric	Before	After
Risk scores documented	60%	100%
Complete medication list	75%	98%
Investigation results structured	50%	100%
Clinical reasoning documented	40%	95%

Financial Impact

Time Value Recovered:

• Documentation time saved: 13.5 min/patient
• Average patients/cardiologist/day: 15
• Daily time saved per cardiologist: 3.4 hours
• 4 cardiologists × 3.4 hours × 200 days = 2,720 hours/year

ROI:

• Specialist hourly rate: $300
• Annual value recovered: $816,000
• Annual subscription (Specialist tier): $36,000
• Implementation + cardiology customisation: $8,000
• Total Year 1 cost: $44,000
• Year 1 net benefit: $772,000
• ROI: 1,755%

---

Referrer Feedback

GP Response to AI-Generated Letters

The practice surveyed referring GPs about the new letter format:

Feedback Metric	Result
"Letters more useful than before"	94% agree
"Risk scores help my management"	89% agree
"Turnaround time improved"	100% agree
"Would refer more patients"	78% agree

GP Testimonials:

"The letters now include CHA₂DS₂-VASc scores and clear anticoagulation recommendations. It makes my job so much easier when the patient comes back to me."

— Dr Emma Watson, Referring GP

---

Cardiologist Testimonials

"The AI actually thinks like a cardiologist. When I mention atrial fibrillation, it automatically calculates CHA₂DS₂-VASc. When I discuss chest pain, it stratifies with HEART score. This is clinical intelligence, not just transcription."

— Dr Andrew Mitchell, Director

"For the first time, I'm finishing my notes before my next patient arrives. And the quality is better than what I used to write myself—more structured, more complete."

— Dr Lisa Park, Interventional Cardiologist

"The GP letters are a revelation. They used to take 20 minutes each. Now they're generated automatically with all the relevant information, risk scores, and recommendations. GPs are telling me they're the best letters they receive."

— Dr Raj Patel, Electrophysiologist

---

Keys to Success

1. Specialty-specific AI: Generic AI fails for complex specialties
2. Risk calculator automation: Manual scoring wastes specialist time
3. Investigation structure: Cardiology requires structured results capture
4. GP letter quality: Referrers judge you by your letters
5. Clinical reasoning documentation: Audit and medico-legal protection

Lessons Learned

"Don't try to make generic AI work for specialty practice. The time you waste editing defeats the purpose. Invest in a solution that understands your specialty from day one."

— Dr Andrew Mitchell, Director

---

Practice Profile

Attribute	Detail
Practice Name	Heart Specialists Queensland
Location	Brisbane, Queensland
Type	Private specialist cardiology
Subspecialties	Interventional, electrophysiology, heart failure
Providers	4 Cardiologists
Support Staff	6 clinical + 4 admin
Patient Base	15,000+ patients
PMS	Genie Solutions
Regenemm Start Date	March 2025

---

Get Similar Results

Ready to transform specialist documentation at your practice?

1. Request a Specialty Demo — See your specialty's AI agent in action
2. Calculate Your ROI — Estimate your potential savings
3. Start a Pilot — Begin your 30-day specialist trial

---

Related Resources:

• FHIR Integration Guide
• Implementation Checklist
• More Case Studies

---

Case study published with permission. Some details anonymised for privacy.