What The Physio Assistant Is and Why It Exists

The Physio Assistant (AIPT Guru) is a clinical education and documentation helper for physiotherapists and rehab clinicians in hospital and primary care. It synthesises peer-reviewed evidence (e.g., RCTs, systematic reviews) and recognised guidelines (NICE/CSP, APTA, APA when requested), highlights red flags and referral pathways, and outputs copy-ready clinical notes and teaching aids. It does not give medical advice; it supports reasoning and documentation.\n\nHow it’s designed to work:\n• Evidence-first: Summaries are anchored to study type and evidence strength (GRADE/PEDro when available) and can include PubMed-style links in Auto-Cite mode.\n• Context-aware: Tailors to region-specific guidance if you say “Use NICE” or “Follow APTA”.\n• Decision support: Logic-tree walkthroughs for differential diagnosis and triage (e.g., Ottawa Ankle Rules, Canadian C-Spine Rule, vestibular red flags).\n• Outcomes & documentation: Interprets common measures (ODI, SPADI, TUG, NDI) with typical MCID/MDC values and produces SOIER/SOAP/ISOBAR notes.\n\nIllustrative examples:\n• ED ankle injury: Runs an ankle inversion logic tree → applies Ottawa Ankle Rules → flags tenderness at posterior edge distal fibula + inability to weight-Physio Assistant functionsbear → recommends imaging threshold (as per decision rule) and lists red flags requiring urgent medical assessment.\n• MSK low back pain (LBP): Produces an evidence map summarising exercise therapy (moderate–high evidence), manual therapy (adjunct; mixed evidence), routine imaging (not indicated without red flags), and low-value modalities (e.g., passive electrotherapy) with rationale and citations.\n• Post-op shoulder on surgical ward: Generates a SOIER note, interprets SPADI change versus typical MCID (~13 points on 0–100), and drafts an MDT handover to OT and nursing with precautions and mobility plan.

Core Functions and How They Are Used

  • Evidence & Guideline Synthesiser

    Example

    Sciatica with leg-dominant pain in primary care: Produces a concise evidence brief comparing exercise, education, manual therapy as adjuncts, and epidural steroid injections; labels each by study type (e.g., SR/RCT), effect direction, and evidence strength (e.g., Moderate by GRADE); aligns with a requested guideline set (e.g., NICE NG59 for LBP/sciatica, APTA CPG). It explicitly marks outdated/low-value options (e.g., routine imaging without red flags, passive modalities with minimal effect) and notes when evidence is contested.

    Scenario

    A First Contact Practitioner builds a pathway update: The Assistant auto-lists key recommendations (e.g., keep active, screen red flags, safety-netting), contraindications, and referral thresholds; provides a one-page summary with PubMed-style links for the MDT governance pack and highlights where local policy should mirror national guidance.

  • Clinical Reasoning & Logic Tree Decision Support

    Example

    Dizziness triage: The Assistant runs a vestibular logic tree differentiating peripheral (e.g., BPPV: brief positional vertigo, fatigable horizontal/rotatory nystagmus) from central features (e.g., direction-changing nystagmus, severe ataxia, focal neuro signs). It proposes appropriate bedside tests (e.g., Dix–Hallpike, HINTS+ overview), flags red flags (neurologic deficits, new severe headache), and outlines urgent referral triggers.

    Scenario

    Same-day access MSK clinic receives a whiplash case: The Assistant applies the Canadian C-Spine Rule (age, dangerous mechanism, paresthesias; low-risk factors; active rotation) to determine if imaging is indicated, documents the reasoning, and drafts safety-net advice. In an ankle sprain, it applies the Ottawa Ankle/Foot Rules to reduce unnecessary radiographs and standardise triage across clinicians.

  • Rehab Planning, Documentation & Outcomes

    Example

    ACL reconstruction: Generates a week-by-week clinical timeline with criteria-based progression (e.g., effusion control → ROM restoration → strength symmetry benchmarks → running/readiness testing), load management cues, and red flags (e.g., calf pain/swelling → DVT rule-out). It pairs this with an EMR-ready SOIER or SOAP note and an outcomes tracker (e.g., KOOS subscales, hop tests). For shoulder rehab, it interprets SPADI improvement (e.g., +15 points exceeds a commonly cited MCID ~13) and for LBP interprets ODI change (e.g., −12 points exceeds a typical MCID ~10), noting population/context caveats.

    Scenario

    On a surgical ward (day 2 post–rotator cuff repair), the Assistant drafts: (1) SOIER assessment with precautions and mobility status; (2) an MDT handover highlighting sling use, sleeping position, and opioid weaning plan; (3) a patient-friendly exercise sheet with stage-appropriate dosing; (4) an outcome plan with review intervals and MCID targets to guide discharge planning.

Who Benefits Most

  • Frontline physiotherapists and MDT clinicians in hospital and primary care (ED/FCP/ACP, inpatient rehab, outpatient MSK, community)

    They need rapid, defensible reasoning and documentation under time pressure. The Assistant standardises triage (e.g., Ottawa/Canadian rules), highlights red flags and referral thresholds, aligns care with regional guidance (NICE/APTA/APA on request), and outputs EMR-ready notes and safety-netting language. This improves consistency across shifts and sites, supports audit readiness, and reduces low-value care.

  • Physiotherapy educators, students, and service leads/quality improvement teams

    Educators and students use logic trees and teaching mode to unpack differential diagnosis, compare interventions by study type and evidence strength, and rehearse exam-style cases. Service leads use Auto-Cite evidence summaries and guideline alignment to build/update protocols, create pathway toolkits, and track outcomes against MCID/MDC thresholds for governance and service evaluation.

How to Use The Physio Assistant

  • Visit aichatonline.org for a free trial without login, also no need for ChatGPT Plus.

    Open the site and launch The Physio Assistant instantly—no account, subscription, or credit card required.

  • Set context & prerequisites

    State your role (student, MSK, neuro, senior clinician), region or guideline authority (NICE, APTA, APA, CSP), and desired output format (SOIER, SOAP, ISOBAR). Prerequisites: internet access; anonymize all patient data; educational use only—this tool does not provide medical advice.

  • Describe the task clearly

    Provide concise clinical context (e.g., condition, acuity, key findings, goals). Examples: differential for heel pain; ACL rehab week-by-week; vestibular screening; outcome measure interpretation (e.g., ODI, SPADI with MCID/MDC); guideline-aligned referral thresholds; MDT handover text.

  • Use the built-in tools

    Quick commands: “start logic tree for shoulder pain,” “Walk through ACL rehab,” “Track post-op recovery for rotator cuff,” “Why did you suggest that?,” “What if the symptom changed?” Enable Auto-Cite (APA/VHow to use Physio Assistantancouver with PubMed links), evidence grading (GRADE/PEDro), red-flag alerts, and region-specific guidance.

  • Optimize your workflow

    Choose modes: clinic (concise), teaching (expanded), quick (one-line), or time mode (30s/2m/5m). Ask for copy-paste SOIER/SOAP/ISOBAR notes, outcome score interpretations, or rehab toolkits/timelines. Keep prompts specific, avoid identifiers, and note boundaries: informational only, no diagnosis, no background processing.

  • Case Triage
  • Guideline Review
  • Rehab Planning
  • Outcome Tracking
  • MDT Handover

Five In-Depth Q&A About The Physio Assistant

  • What is The Physio Assistant and how is it different from generic AI tools?

    It is an AI physiotherapy assistant focused on clinical education. It summarizes peer-reviewed evidence (with PubMed-linked citations), aligns to regional guidance (NICE, APTA, APA, CSP), flags red-flags/referral pathways, and generates copy-ready clinical notes (SOIER/SOAP/ISOBAR). It also offers structured logic trees for differential diagnosis and rehab planning—features tailored to hospital and primary care physio workflows.

  • How does it ensure evidence quality and transparency?

    It prioritizes systematic reviews, RCTs, and reputable guidelines, labels study types, and uses strength-of-evidence conventions (e.g., GRADE, PEDro where applicable). It highlights where evidence is strong, mixed, or outdated, and includes Auto-Cite (APA/Vancouver) with PubMed links for verification and further reading.

  • Can it help with triage and red-flag screening?

    Yes. Logic Trees guide region-by-region triage (e.g., shoulder, thoracic, heel, low back, vestibular) and incorporate tools like Ottawa Ankle Rules, Canadian C-Spine Rule, STarT Back, and vestibular central vs peripheral screening. It surfaces red-flag indicators and suggests guideline-aligned referral thresholds while remaining educational (not medical advice).

  • What documentation and handover outputs can I generate?

    Request SOIER, SOAP, ISOBAR, MDT handover, or discharge summaries. Outputs can include diagnosis differentials, goals, test findings, exercise dose (frequency/intensity/volume), precautions/contraindications, red-flag status, and follow-up plans—formatted for quick EMR copy-paste and aligned to your selected guideline region.

  • Can it build rehab plans and support teaching?

    Yes. Use the Toolkit Builder and Clinical Timeline Generator for condition-specific programs (e.g., ACL, rotator cuff, OA knee). You can add milestones, progressions, and outcome tracking (ODI, SPADI, TUG with MCID/MDC). Teaching mode expands reasoning, provides flashcards/quizzes for CPD, and supports 'Why did you suggest that?' explanations for transparent clinical logic.

cover