Standard 22: Telemedicine (Version 11)

Why it matters. A telemedicine service that works only for patients with a fast internet connection and a smartphone leaves behind those who most need care. A route for patients who cannot connect keeps the service fair.

Sample. Review 10 telemedicine bookings where the connection failed or access was limited, and the routes taken.

1. Patients who cannot connect are referred or redirected to an alternative.

Method. Check each sampled failed connection for the alternative recorded.

Evidence to show. Digital access check and alternative route record (TEL-FRM-001).

Pass when. Every failed connection has an alternative route recorded.

2. Access is checked at booking.

Method. Check each sampled booking for the access check.

Evidence to show. Digital access check and alternative route record (TEL-FRM-001).

Pass when. At least 95% of bookings show the access check.

3. No patient is left with no next step.

Method. Check each sampled case for the next step and its completion.

Evidence to show. Referral tracking register (PHC-REG-002).

Pass when. Every case has a next step completed or in progress.

Immediate fail (any one means the criterion is not met). A rural patient turned away because the video link failed and no alternative was offered.

Why it matters. Artificial intelligence can help with diagnosis and triage, but it can also make errors that look confident. WHO’s guidance on AI in health asks for safety, transparency and human oversight, and a facility that introduces AI must follow it.

Sample. Review the AI systems in use for diagnosis, triage or care decisions, and their introduction records.

1. Each AI system is introduced in line with WHO AI governance guidance.

Method. Check each system for the assessment against the WHO guidance.

Evidence to show. AI System Assessment (AI-FRM-001).

Pass when. Every AI system has an assessment recorded.

2. The responsible use policy is in place.

Method. Check the policy for the WHO principles applied.

Evidence to show. Responsible Use of AI in Care Policy (AI-POL-001).

Pass when. The policy is in force and covers the WHO principles.

3. For telemedicine, the country of each remote patient is checked.

Method. Check the assessment for the country check.

Evidence to show. AI System Assessment (AI-FRM-001).

Pass when. Every assessment for a remote service includes the country check.

Immediate fail (any one means the criterion is not met). An AI triage tool used with no assessment.

Why it matters. An AI system can work well at first and then drift, missing diagnoses or over-diagnosing as the patients change. Auditing its results regularly is the only way to notice.

Sample. Review the AI monitoring register and the audit records for each AI system in use.

1. AI results are audited for over-diagnosis and missed diagnosis.

Method. Check each audit for the over-diagnosis and missed-diagnosis review.

Evidence to show. AI Monitoring Register (AI-REG-001).

Pass when. Every AI system has an audit recorded in the period.

2. Patient complaints related to AI-assisted care are tracked.

Method. Check the complaints log for AI-related complaints and their tracking.

Evidence to show. AI Monitoring Register (AI-REG-001).

Pass when. Every AI-related complaint is tracked to closure.

3. Monitoring reflects the limits of remote assessment.

Method. Check the audit for the remote-assessment limits.

Evidence to show. AI Performance and Safety Monitoring (AI-SOP-001).

Pass when. Remote audits include the limits.

Immediate fail (any one means the criterion is not met). An AI tool never checked after it went live.

Why it matters. Staff who will use an AI system every day know the practical problems that a manager may not. Consulting them before it is introduced finds those problems early.

Sample. Review the staff consultation records for each AI system introduced in the last 12 months.

1. Staff who will use the AI system are consulted before introduction.

Method. Check each introduction for the consultation before go-live.

Evidence to show. Staff Consultation Record (AI-FRM-002).

Pass when. Every introduction has a consultation before go-live.

2. Training needs are identified.

Method. Check each consultation record for the training need recorded.

Evidence to show. Staff Consultation Record (AI-FRM-002).

Pass when. Every consultation identifies the training need.

3. Remote clinicians across time zones are consulted.

Method. Check the consultation attendance for remote clinicians.

Evidence to show. Staff Consultation Before AI Introduction (AI-SOP-002).

Pass when. Remote clinicians are included in each consultation.

Immediate fail (any one means the criterion is not met). An AI system introduced with no staff input.

Why it matters. When an AI system contributes to a wrong decision, it is not clear who is responsible: the clinician, the vendor or the facility. A documented accountability arrangement answers that question before there is an incident.

Sample. Review the accountability arrangements for each AI system, and any AI-assisted decision reviewed in the last 12 months.

1. Accountability for AI-assisted care is documented.

Method. Check each AI system for the accountability arrangement.

Evidence to show. Accountability for AI-Assisted Care (AI-POL-002).

Pass when. Every AI system has accountability documented.

2. A named party is responsible for an AI-assisted decision.

Method. Check each reviewed decision for the responsible party named.

Evidence to show. Accountability for AI-Assisted Care (AI-POL-002).

Pass when. Every reviewed decision names the responsible party.

3. Accountability is defined across borders for remote care.

Method. Check the policy for the cross-border arrangement.

Evidence to show. Accountability for AI-Assisted Care (AI-POL-002).

Pass when. The policy covers cross-border accountability.

Immediate fail (any one means the criterion is not met). An AI-assisted decision with no one named as responsible.

Why it matters. Staff who cannot learn new skills fall behind the care they are asked to give. Giving every staff member access to ongoing education keeps the workforce able to meet the standards.

Sample. Review the training records for 10 workforce members, including remote staff.

1. Each workforce member has access to ongoing education.

Method. Check each sampled record for the training offered in the last year.

Evidence to show. Training matrix register (WRK-REG-001).

Pass when. Every sampled member has training offered in the last year.

2. The access route reaches providers who work alone.

Method. Check that remote staff are on the training list and have access.

Evidence to show. Competency and training procedure (WRK-SOP-001).

Pass when. Remote staff are on the list with access.

3. Education is maintained against the role requirement.

Method. Check each record for the role-required training completed.

Evidence to show. Training matrix register (WRK-REG-001).

Pass when. At least 95% of records show the role-required training.

Immediate fail (any one means the criterion is not met). No training offered for three years.

Why it matters. Clinicians who work alone from a screen can become isolated and burn out, and the strain shows in their care. Monitoring wellbeing and offering support keeps the workforce able to practise safely.

Sample. Review the wellbeing and isolation check records for remote clinicians in the last 12 months.

1. Remote clinicians’ wellbeing and isolation are checked.

Method. Check each remote clinician for the quarterly check.

Evidence to show. Remote clinician wellbeing check record (TEL-FRM-002).

Pass when. Every remote clinician has a check each quarter.

2. Each clinician has regular team contact and a support route.

Method. Check each record for the team contact and the support route agreed.

Evidence to show. Remote clinician wellbeing check record (TEL-FRM-002).

Pass when. Every record shows team contact and a support route.

3. Warning signs are acted on.

Method. Check each warning sign for the action and its date.

Evidence to show. Staff support after distressing events (WRK-SOP-004).

Pass when. Every warning sign has an action within the week.

Immediate fail (any one means the criterion is not met). A clinician working alone for a year with no contact with colleagues and no check.

Why it matters. A carbon footprint shows where a facility’s greenhouse emissions come from. It does not need to be perfect; a basic baseline from energy and travel data is enough to start measuring change.

Sample. Review the carbon footprint baseline for digital infrastructure and the energy use records.

1. A carbon footprint baseline is carried out, including digital infrastructure.

Method. Check the baseline record for the digital infrastructure included.

Evidence to show. Carbon Footprint Baseline Record (ENV-FRM-009).

Pass when. The baseline includes the digital infrastructure.

2. The baseline method is documented.

Method. Check the method against the procedure.

Evidence to show. Carbon Footprint Baseline Method (ENV-SOP-007).

Pass when. The method is documented and followed.

3. Energy use is monitored to support the baseline.

Method. Check the monthly energy record for the period.

Evidence to show. Monthly Energy Use Record (ENV-FRM-008).

Pass when. Monthly energy use is recorded for the period.

Immediate fail (any one means the criterion is not met). No assessment of emissions ever carried out.

Why it matters. Purchases carry environmental and social costs: wasteful packaging, disposable products, and suppliers who exploit workers. Procurement that considers these reduces both.

Sample. Review the supplier evaluation records for 5 technology or hosting purchases in the last 12 months.

1. Procurement considers environmental and social responsibility.

Method. Check each sampled purchase for the environmental criterion in the evaluation.

Evidence to show. Supplier evaluation record (PRC-FRM-001).

Pass when. Every sampled purchase has the criterion applied.

2. Hosting and technology choices weigh environmental impact.

Method. Check each hosting or technology choice for the impact considered.

Evidence to show. Supplier evaluation record (PRC-FRM-001).

Pass when. Every hosting or technology choice considers the impact.

3. Single-use decisions are reviewed against safety requirements.

Method. Check each single-use decision for the safety review.

Evidence to show. Energy Use Monitoring and Reduction (ENV-SOP-006).

Pass when. Every single-use decision is reviewed.

Immediate fail (any one means the criterion is not met). All purchasing on lowest price regardless of waste.

Why it matters. Telemedicine can save patients long journeys, and that is an environmental and personal benefit. Where it is meaningful, the provider should measure and record it.

Sample. Review the patient travel avoided record for the last 12 months.

1. The benefit of reduced patient travel is recorded where meaningful.

Method. Check the record for the journeys avoided.

Evidence to show. Patient travel avoided record (TEL-FRM-003).

Pass when. The record shows the journeys avoided each quarter.

2. Any claim about the benefit is supported by the record.

Method. Check each published claim for the record behind it.

Evidence to show. Patient travel avoided record (TEL-FRM-003).

Pass when. Every claim is supported by the record.

3. The benefit is reported in the sustainability progress report.

Method. Check the sustainability progress report for the travel benefit.

Evidence to show. Sustainability Progress Report (ENV-FRM-003).

Pass when. The travel benefit appears in the report.

Immediate fail (any one means the criterion is not met). A claim that telemedicine is green, with no evidence.

Why it matters. Staff and patients can help a sustainability goal only if they know what it is. Communicating the goals in plain words makes the approach something people do, not something on paper.

Sample. Review the sustainability communication record and the staff and patient notices.

1. The sustainability approach and goals are communicated to staff.

Method. Check the communication record for staff notice.

Evidence to show. Sustainability Communication Record (ENV-REG-001).

Pass when. The goals are communicated to staff.

2. Patients are informed of the approach.

Method. Check the notice for patient information.

Evidence to show. Staff and Patient Sustainability Notice (ENV-FRM-007).

Pass when. The notice reaches patients.

3. Remote staff are reached as well as office-based staff.

Method. Check the communication channels for remote staff.

Evidence to show. Sustainability Communication Record (ENV-REG-001).

Pass when. Remote staff are reached.

Immediate fail (any one means the criterion is not met). Goals never shared with staff.

Why it matters. Goals that are never reported to the people who govern the facility are easily dropped. A set reporting schedule to governance keeps progress visible and acted on.

Sample. Review the governance reports on sustainability progress for the last 12 months.

1. Progress is reported to governance on a defined schedule.

Method. Check each governance report against the schedule.

Evidence to show. Governing Reporting of Sustainability Progress (ENV-SOP-005).

Pass when. Every scheduled report is made.

2. Goals are reviewed on the same schedule for telemedicine.

Method. Check the review for the telemedicine goals.

Evidence to show. Governing Reporting of Sustainability Progress (ENV-SOP-005).

Pass when. The telemedicine goals are reviewed on schedule.

3. Governance acts on the report.

Method. Check the governance minutes for the decision taken.

Evidence to show. Sustainability Progress Report (ENV-FRM-003).

Pass when. Each report has a decision recorded.

Immediate fail (any one means the criterion is not met). No report to governance on sustainability ever given.

Why it matters. In a remote consultation, the clinician cannot see who is on the other end. Confirming the patient’s identity at the start protects against the wrong person receiving care, or a record being made for the wrong patient.

Sample. Review 10 remote consultation records for the identity and privacy checks.

1. Patient identity is confirmed at the start of every remote consultation.

Method. Check each sampled consultation for the identity check.

Evidence to show. Remote consultation identity and privacy check (TEL-FRM-004).

Pass when. Every sampled consultation shows the identity check.

2. The clinician confirms who else is present.

Method. Check each record for the others-present entry.

Evidence to show. Remote consultation identity and privacy check (TEL-FRM-004).

Pass when. Every record shows who was present.

3. The clinician confirms whether the consultation is private.

Method. Check each record for the privacy entry.

Evidence to show. Remote consultation identity and privacy check (TEL-FRM-004).

Pass when. Every record shows the privacy check.

Immediate fail (any one means the criterion is not met). A consultation held with a family member pretending to be the patient.

Why it matters. A patient seen remotely must know that a video or phone call cannot always replace an examination. Telling them the limits, and when they must be seen in person, is part of informed consent.

Sample. Review 10 remote consultation records for consent and the limits explained.

1. The patient is told the limits of remote care.

Method. Check each sampled record for the limits explained.

Evidence to show. Remote care consent and limits record (TEL-FRM-005).

Pass when. Every sampled record shows the limits explained.

2. Consent to remote care is recorded before the consultation.

Method. Check each record for the consent time against the consultation time.

Evidence to show. Consent Form (CON-FRM-001).

Pass when. Every record shows consent before the consultation.

3. The patient is advised when an in-person examination is needed.

Method. Check each record for the in-person advice where it applies.

Evidence to show. Remote care consent and limits record (TEL-FRM-005).

Pass when. Every case where it applies shows the in-person advice.

Immediate fail (any one means the criterion is not met). A patient with a suspected fracture advised remotely with no warning about the limit.

Why it matters. A prescription written without seeing the patient can miss an allergy or an interaction. A safe process for remote prescribing limits what can be prescribed without an examination.

Sample. Review 10 remote prescriptions, including any controlled medicine.

1. A remote prescription is issued only where there is enough information.

Method. Check each sampled prescription for the sufficiency decision.

Evidence to show. Prescribing during remote consultation (TEL-POL-002).

Pass when. Every sampled prescription shows the sufficiency decision.

2. Allergies, interactions and current medicines are checked.

Method. Check each prescription for the three checks.

Evidence to show. Prescription record (PHM-FRM-001).

Pass when. Every prescription shows the three checks.

3. No controlled medicine is prescribed remotely to a patient never seen.

Method. Check each controlled medicine for a prior in-person review.

Evidence to show. Prescribing during remote consultation (TEL-POL-002).

Pass when. No controlled medicine is prescribed to a patient never seen.

Immediate fail (any one means the criterion is not met). A controlled medicine prescribed remotely to a patient never seen.

Why it matters. A remote consultation that produces no clinical record leaves the next clinician blind to what was decided. A complete record, available to the treating clinician, keeps care continuous.

Sample. Review the clinical records for 10 remote consultations.

1. Each remote consultation produces a clinical record.

Method. Check each sampled consultation for a record.

Evidence to show. Remote consultation clinical record check (TEL-FRM-006).

Pass when. Every sampled consultation has a record.

2. The record is complete.

Method. Check each record against the required fields.

Evidence to show. Remote consultation clinical record check (TEL-FRM-006).

Pass when. Every record is complete.

3. The treating clinician can access the record.

Method. Check access for the treating clinician.

Evidence to show. Remote consultation clinical record check (TEL-FRM-006).

Pass when. Every record is accessible to the treating clinician.

Immediate fail (any one means the criterion is not met). A patient’s remote consultation with no notes on file.

Why it matters. A video call that drops in the middle of a consultation, especially one where the patient is unwell, must not leave the patient without help. A clear route to reach them by another method keeps the patient safe.

Sample. Review the failed remote consultations in the last 12 months, and the follow-up records.

1. Each failed consultation has a call-back by another method.

Method. Check each failure for the call-back recorded.

Evidence to show. Technical failure and follow-up record (TEL-FRM-007).

Pass when. Every failure has a call-back recorded.

2. The call-back is made the same day.

Method. Check each failure for the time of the call-back.

Evidence to show. Technical failure and follow-up record (TEL-FRM-007).

Pass when. At least 95% of call-backs are the same day.

3. Urgent cases are escalated.

Method. Check each urgent case for the escalation recorded.

Evidence to show. Technical failure during a remote consultation (TEL-SOP-003).

Pass when. Every urgent case has an escalation.

Immediate fail (any one means the criterion is not met). A patient with chest pain whose call dropped and who was not called back.