RONKE
CLINICAL EVIDENCE

We hold ourselves to clinical standards, not software standards.

Every question RONKE asks traces to evidence. Every claim we make will be validated the way medicine is validated: measured, peer-reviewed, published.

THE EVIDENCE BASE

Structured telephone support works. RONKE makes it scale.

Systematic reviews of structured telephone support in heart failure consistently show reduced readmissions and mortality. The barrier has never been the evidence, it's been the workforce to deliver it. One nurse can call 40 patients. RONKE calls all of them.

20–25%
reduction in readmissions
£3–5
saved per £1 spent
CUMULATIVE READMISSIONS · 90 DAYS · ILLUSTRATIVE
Usual care (30% at day 30) Structured telephone monitoring
VALIDATION PATHWAY

Shadow first. Prove safety before touching care.

RONKE's scoring runs in shadow mode alongside clinical judgement before it ever influences a decision, validated step by step, in partnership with Manchester Metropolitan University.

PHASE 1 · NOW
Shadow scoring
Implemented today: the deterministic score runs silently alongside nurse assessments, logging only, unable to affect any decision. Agreement is measured and thresholds tuned with clinicians before anything can go live.
PHASE 2
NHS pilot
Supervised deployment with a heart failure service. Readmission, escalation accuracy and patient experience measured.
PHASE 3
Controlled trial
RCT-level evaluation with academic partners. Primary endpoint: 30-day readmission. Findings submitted for peer review.
PHASE 4
Publication & scale
Peer-reviewed evidence base supporting national adoption, and extension to COPD, diabetes and kidney disease.
CLINICAL GOVERNANCE

A nurse-led reporting tool. Never an autonomous clinician.

RONKE screens and reports. Every clinical decision, every escalation, every medication change, every reassurance, is made by a qualified professional. That boundary is architectural, not aspirational.

Built on NICE guidance
Assessment protocols mapped to NICE Guideline NG106. Every change to the clinical logic and scoring thresholds is reviewed and signed off by our Clinical Safety Officer before it can be used live.
Explainable scoring
Every Green / Amber / Red decision traces to specific patient responses. Clinicians can always see why.
Fail-safe defaults
Anything uncertain escalates to a human, and every call ends with a scripted 999 safety signpost, regardless of risk level.
Full audit trail
Every call recorded, every score logged, every action attributable. Governance-ready from day one.
Safety enforced in code, not policy
Patient data is isolated by clinical site on every query, identifiers are masked before anything is logged, and the clinical logic is a protected module that cannot change without Clinical Safety Officer sign-off.
Named Clinical Safety Officer
Dr. Moslem Abdelghafar, Cardiothoracic Surgeon, SCTS Innovation Committee, reviews and signs off all clinical logic before live deployment.
DCB0129 & data protection
Developed to the NHS clinical-risk standard DCB0129. UK-only data storage. A DPIA and DSPT submission are in preparation, and the regulatory classification is being confirmed with an independent MHRA consultant.
RONKE AI DOES
Call patients on any standard telephone, landline or mobile
Conduct a structured 5–7 minute clinical conversation
Generate a structured report for the reviewing nurse
Alert the nurse when a patient requires urgent review
Close every call with a scripted 999 safety signpost
RONKE AI NEVER DOES
Diagnose any medical condition
Give medical advice of any kind
Recommend changes to medication
Make any clinical decision
Replace the clinical judgement of the nurse
"Every question in RONKE's framework is derived from clinical evidence. We don't ask how you're feeling, we ask the questions that predict readmission."
Discuss a pilot with us →
RONKE © 2026 RONKE AI Ltd · Reg. No. 16865109 · Manchester, UK
Home Technology About hello@ronkeai.com