Clinical Governance Statement

 

 

 

Last Updated: June 2026

1. Introduction

Clinical governance is the framework through which Physio From Home is accountable for continuously improving the quality of our services and safeguarding high standards of care. This statement outlines how we ensure the safety, effectiveness, and quality of our AI-assisted physiotherapy assessment service.

2. Our Clinical Team

Qualified Physiotherapists

All physiotherapists who review assessments and create rehabilitation programs for Physio From Home are:

• Registered with the Health and Care Professions Council (HCPC)
• Hold a recognised physiotherapy degree qualification
• Maintain up-to-date Continuing Professional Development (CPD)
• Have appropriate professional indemnity insurance
• Undergo regular performance reviews

3. The Role of AI in Our Service

AI-Assisted Assessment

Our service uses advanced AI technology (powered by Anthropic’s Claude) to conduct initial patient assessments. The AI is programmed to:

• Gather relevant clinical information through structured conversational questions
• Screen for red flag symptoms requiring urgent medical attention (see Section 4)
• Generate a preliminary provisional diagnosis with confidence rating
• Generate an AI clinical summary for the reviewing physiotherapist
• Identify when conditions are outside our scope (e.g., chronic inflammatory conditions)
• Enforce single-injury-per-assessment limits

Optional Camera-Based Movement Analysis

Patients may optionally use their device camera to measure joint range of motion during the assessment. This feature:

• Provides an estimated range of motion (in degrees) to support the reviewing physiotherapist’s judgement
• Is a guided estimate, not a clinical goniometric measurement, and has the inherent accuracy limits of single-camera analysis
• Processes all video on the patient’s own device — no image or video is stored or transmitted; only the resulting measurement is saved (see our Privacy Policy, Section 2.5)
• Is always optional, with a self-report alternative available for every movement
• Does not make any clinical decision — the physiotherapist reviews the measurement alongside all other assessment information

Human Oversight

Importantly, the AI does not work in isolation. Every AI-generated assessment is reviewed by a qualified, HCPC-registered physiotherapist before any rehabilitation program is sent to the patient. The physiotherapist:

• Reviews the AI-generated clinical summary
• Reviews the full assessment transcript for verification
• Reviews any uploaded medical images
• Reviews the body chart pain/Pins & Needles markers
• Reviews any camera-measured range-of-motion values, where the patient chose to use the optional movement measurement feature
• Validates or adjusts the provisional diagnosis
• Creates a personalised rehabilitation program
• Identifies any cases requiring escalation or referral
• Identifies any safeguarding concerns from the transcript

Ongoing Support: The Patient Portal AI Assistant

After a rehabilitation program has been issued, patients can raise questions or concerns about their rehabilitation through a clinical AI assistant within their secure patient portal. This assistant provides general, rehabilitation-related guidance and is governed by the same principles as the rest of our service:

• The AI assistant supports, and does not replace, the judgement of a qualified physiotherapist
• It does not provide emergency or urgent care, and patients presenting potential red-flag or worsening symptoms are directed to seek appropriate medical attention
• Where a patient does not feel their concern has been answered, they can forward it to a qualified, HCPC-registered physiotherapist, who may review and, where appropriate, adjust the rehabilitation program
• A physiotherapist response to an escalated concern is provided within a maximum of 5 working days

This ensures that patients always retain a route to a human clinician, and that any clinical decision — such as changing a program — remains with a qualified physiotherapist.

Data the AI Does Not Access

To maintain clinical appropriateness and data minimisation, the following data is collected but not shared with the AI: height, weight, BMI, and uploaded medical images. These are only available to the reviewing physiotherapist.

4. Patient Safety

Red Flag Screening

Patient safety is screened by a deterministic red-flag safety layer that operates independently of the AI. Every message a patient submits during the assessment is automatically checked against a fixed set of clinically-approved rules before it is processed by the AI. Because this layer does not rely on the AI’s interpretation, red-flag screening does not depend on the AI behaving correctly — it provides a consistent safety net that runs on every assessment.

The rules and patient-facing wording for this layer have been reviewed and signed off by an HCPC-registered physiotherapist. The screening includes:

General red flags (all assessments):

• Cauda equina syndrome symptoms (saddle numbness, bladder/bowel dysfunction)
• Signs of fracture after significant trauma
• Signs of infection (fever with severe pain, night sweats, chills)
• Neurological deterioration (sudden severe weakness, inability to walk)
• Cardiac symptoms (chest pain, difficulty breathing)
• Unexplained weight loss with pain

Region-specific screening:

• Lower back pain:Structured 5-question cauda equina screening protocol, specifically asking about new or recent onset of each symptom to distinguish from pre-existing conditions
• Neck pain:5D’s and 3N’s screening for vertebrobasilar insufficiency (dizziness, diplopia, drop attacks, dysarthria, dysphagia, nystagmus, nausea, numbness around face/lips)
• Knee pain:Mechanical symptom screening (clicking/clunking, locking, giving way)

When red flags are detected, patients are immediately advised to seek urgent medical attention and the assessment is terminated. The physiotherapy team is notified via urgent email alert.

Pregnancy and Postnatal Screening

Patients who indicate they are currently pregnant, or within three months of giving birth, are screened separately:

• Where their symptoms are in the lumbar spine, thoracic spine, hip, sacroiliac joint or pelvis, they are signposted to a women’s health physiotherapist rather than assessed through the standard pathway, as these areas are best assessed by a specialist during and shortly after pregnancy. Where the patient has paid, a full refund is arranged.
• Pregnancy-specific red flags — including signs of pre-eclampsia (severe headache, visual disturbance, sudden swelling, severe upper abdominal pain), possible deep vein thrombosis, vaginal bleeding, reduced or absent fetal movements, or leaking fluid — prompt the patient to contact their midwife, maternity unit, or emergency services as appropriate.
• Where symptoms are in other regions, the assessment proceeds with pregnancy-safe adaptations and an appropriate caution flag for the reviewing physiotherapist.

The pregnancy and postnatal screening rules and wording have also been reviewed and signed off by an HCPC-registered physiotherapist.

Scope Limitations

We clearly communicate that our service:

• Is designed to complement, not replace, face-to-face physiotherapy
• Is not a substitute for emergency medical care
• Is not suitable for chronic multi-joint inflammatory conditions (fibromyalgia, rheumatoid arthritis, psoriatic arthritis)
• Provides guidance based on information provided by the patient
• Recommends face-to-face assessment if symptoms persist or worsen

5. Quality Assurance

We maintain quality through:

Clinical Audit: Regular review of assessments and outcomes

AI Output Verification: All AI-generated clinical summaries are labelled “AI-generated — please verify against full transcript” to ensure physiotherapists cross-reference before acting

Patient Feedback: Systematic collection and analysis of patient feedback

Incident Reporting: Robust system for reporting and learning from incidents

Complaints Analysis: All complaints are reviewed to identify improvement opportunities (see our Complaints Policy)

CPD: Ensuring all physiotherapists maintain their professional development

6. Evidence-Based Practice

Our rehabilitation programs are based on current evidence and best practice guidelines. We regularly review and update our protocols in line with the latest research and guidance from bodies such as NICE (National Institute for Health and Care Excellence) and the Chartered Society of Physiotherapy.

7. Data Security and Confidentiality

We take patient confidentiality seriously:

• All data is encrypted in transit and at rest
• Hosting is on UK/EU-based servers (Vercel London, Upstash EU, Cliniko EU)
• Assessment data is stored securely and transferred to our clinical management system (Cliniko) for physiotherapist review
• No patient health data is included in email communications — only reference numbers and secure links
• We comply with UK GDPR and the Data Protection Act 2018
• See our Privacy Policy and Data Protection Impact Assessment for full details

8. Professional Standards

All physiotherapists working with Physio From Home adhere to:

• HCPC Standards of Conduct, Performance and Ethics
• HCPC Standards of Proficiency for Physiotherapists
• CSP Code of Members’ Professional Values and Behaviour (where applicable)
• Our own internal policies and procedures

9. Continuous Improvement

We are committed to continuously improving our service. We regularly review patient outcomes, feedback, and complaints to identify areas for enhancement. Our AI system prompt and screening protocols are periodically reviewed and updated to improve accuracy and safety.

10. Transparency

We believe in transparency. We clearly inform patients that their initial assessment is conducted by AI and subsequently reviewed by a qualified physiotherapist. Patients are provided with a reference number and can request a copy of their assessment data at any time by contacting help@physiofromhome.com.

11. Contact

For clinical governance queries, please contact:

Email: help@physiofromhome.com

Website: www.physiofromhome.com