Open patient consultation demonstration ↗
Browse expanded register: 116 condition-based interview drafts across 24 domains ↗Detailed clinical decision pathways and language review remain pending. No patient demographics are assigned.
Open physician evaluation workspace v0.9 ↗Fictional cases only · Clinical protocols and source validation remain under review.
Draft interview prompts are available in Urdu, Pakistani Sindhi, Pakistani Punjabi (Shahmukhi) and Pakistani Pashto alongside English. Native Pakistani clinicians must review each language and dialect. This does not translate the whole interface or provide live multilingual AI or speech.
The GP reviews, edits and owns every clinical decision. Questions below are structured prompts, not an autonomous diagnosis.
Children, pregnancy / postpartum and complex emergencies need dedicated, validated pathways. This initial adult prototype does not supply those algorithms.
Demonstration profile only. Entering a registration number does not verify it or create a secure account.
For a fictional adult case, generate a draft history summary, missing-information list and follow-up questions for the GP. This is not a diagnostic or prescribing service. The draft does not change the encounter, release a patient plan or send a referral.
Regional-language AI drafts are not enabled in this release.
Leave unmeasured observations blank. Record method and context. These entries do not generate an automated risk score or clearance.
This status is a prototype annotation, not a legal signature or prescription.
Review v0.7 with fictional cases. No clinical approval or integration with the existing Humsafar service is claimed.
Record observations against the case. These are reviewer judgements, not automated validation scores.
No completed case reviews recorded in this tab.
These sources inform design principles; they do not validate this implementation or replace locally approved protocols.
Purpose: fictional adult-case evaluation of a proposed physician-facing clinical documentation and decision-support workflow. The present software is a deterministic interface, not an AI diagnostic system. Sponsor: Digital Landscape (Pvt) Ltd. Clinical lead and regulatory classification: to be agreed.
This is a proposed pilot scope, not a national ranking of visit frequency. GP management depends on examination, severity, available facilities and reliable follow-up. Necessary hospital referral must never be discouraged.
Local evidence and service framework: Pakistan EPHS (Ministry / WHO / DCP3); 2024 PHC quality study in Pakistan; WHO Pakistan family practice approach. This proposed list requires review by Pakistani family physicians before becoming a clinical protocol.
Agreement by clinical lead and authorities is pending. For this interface: all agreed scenarios completed; unresolved critical or major issues prevent progression; missing data preserved; exports recover the correct case and version; changes are retested. No clinical accuracy, sensitivity, specificity or HIPAA compliance result is established.
The journey uses first-contact safety assessment, history and examination, targeted investigations, shared decisions, medication safety and follow-up. WHO guidance is intended for international adaptation; NICE recommendations are UK guidance and require review for Pakistan. A framework reference does not establish that this prototype fully implements or is endorsed by its publisher.
Other symptom pathways remain draft structured assessment prompts pending condition-specific guideline selection and Pakistani clinician approval. Guideline references checked 23 October 2026. English and regional-language drafts need separate clinical review.
Nothing is saved automatically. Closing this tab loses unsaved entries. You can download an encrypted review file and restore it on another device with your chosen passphrase. This is file encryption, not an account or a clinical record service.
Share fictional assessment reports through an agreed review process. Downloads remain your responsibility. There is no central review inbox or automated referral/alert function in this pilot.
Fictional demonstration directory only. Your supplied national doctor database is excluded from this circulation package. Search functions demonstrate the interface; these clinicians and clinics do not exist.
Offline demonstration fixture: three invented records, not real clinicians. The separate owner-private hosted pilot contains the provisional imported directory. No real doctor data is included here.
20 review drafts based on your supplied issue categories and international reference frameworks. GP assessment is essential. These are prompts, not validated diagnostic algorithms or autonomous treatment. Warning-sign examples are not exhaustive. No ranking or prevalence estimate is claimed.
For regulatory testing, document limitations and corrections in Regulatory review. These fields are included in encounter exports. Referral and result tracking is manual, without alerts or automatic transmission.
Fictional testing only. Exchange encrypted files between the patient demonstration and this GP workspace. No live messaging, authenticated identity, notifications or real consultation is connected.
Patient home measurements are retained as reported history, not verified GP observations. Review them independently.
Release requires fictional consent, GP name, patient ID, plan, patient explanation, safety-net and follow-up date. Only this patient-facing subset is exported; private assessment notes are excluded. The file does not verify the sender’s identity.
Download regulatory review package — fictional data only (.zip)
Submission purpose: preliminary assessment of intended use, usability, clinical workflow and a proposed evaluation protocol. This is not a request to certify unimplemented AI functionality or approve use with patients.
Existing internet-based fictional conversation review. Record its displayed version separately. No unified patient record or shared GP sign-on is implemented.
This v0.7 workspace: fictional cases, observations, predefined prompts, draft regional language text, references and reviewer reports. Optional AI documentation drafting is available in Patient & history. No AI diagnosis, prescribing service or autonomous consultation is provided.
Do not progress to a real-patient pilot while critical or major findings remain unresolved, required security and identity controls are absent, or the agreed clinical / regulatory approvals are outstanding. Interface checks do not measure clinical diagnostic accuracy.
Email fictional review reports and the case log to shahab.rizvi@outlook.com. Automated submission is not connected. Review logs remain in this tab until exported; encrypted review files include the log.
Humsafar GP v0.7 · 23 October 2026. Mental wellbeing version and hosting release must be identified independently. Registration, certificate number, approving authority and approval dates: not issued / not claimed.