NMC Telemedicine Practice Guidelines: A Hospital Checklist

Only a Registered Medical Practitioner enrolled on the State or National Medical Register may provide a teleconsultation, and the registration number has to appear on the prescription and on the practitioner-facing communications the patient receives. That sounds trivial until you look at how most hospitals onboard a visiting consultant: a scanned degree, a bank mandate, a slot template, and no verified registration number anywhere in the system. If your credentialing file cannot produce a current registration number for every doctor on the tele-roster, you have a gap that a council complaint will find immediately.

The guidelines also anticipated a mandatory training course for practitioners offering telemedicine. Whether that requirement is currently live and in what form is a question for your medical administrator to confirm against the present NMC position rather than something to assume from a 2020 reading. What is not in doubt is that a hospital which can show a dated internal training record for each tele-consulting doctor is in a far stronger position than one which cannot, regardless of the statutory status of any external course.

There is a second category the roster usually misses: nurses, counsellors, and health workers who facilitate a consultation at the patient end or at a spoke site. They are not prescribing, but they are part of the clinical chain, and their scope needs to be written down. A health worker who relays a clinical instruction without a practitioner in the loop is an incident waiting to be reconstructed by someone else.

Almost every prescribing restriction in the guidelines hangs on whether this is a first consultation or a follow-up, so your system has to answer that question mechanically rather than leaving it to the doctor to remember. A follow-up, in the sense the guidelines use it, means the same practitioner and the same patient continuing on the same health condition within a defined window, and not simply any second visit to your hospital. A cardiology follow-up does not make a dermatology complaint a follow-up.

The practical implication is that your teleconsultation screen needs a condition-linked episode, not just a patient identifier. If the doctor opens a consultation and the interface shows nothing more than the patient name and a blank note, they will classify from memory and the classification will drift. Show the previous encounter for that condition, its date, the practitioner who saw it, and let the system propose first-or-follow-up with the doctor able to override and state why.

The trade-off is friction. Consultants object, reasonably, that they do not want to spend the first ninety seconds of a ten-minute slot on classification. The answer is to default correctly and make the override a single click with a mandatory one-line reason. If you make it a form, they will pick whichever option closes the form fastest, and your prescribing controls will be running on garbage input.

The guidelines recognise video, audio, and text-based or asynchronous exchange as legitimate modes, and they do not rank them as good, better, best. What they do is tie certain actions to certain modes. The clearest example is prescribing: some medicines may only be prescribed at a first consultation if that consultation was conducted over video. That single linkage is why the mode has to be a recorded field on the encounter and not an incidental fact about how the call happened to connect.

Mode also has to be able to change mid-consultation and be recorded as having changed. Rural bandwidth being what it is, a video consult that degrades into an audio call is a daily event. If the doctor started on video, verified the patient visually, then dropped to audio, the record should say so. A system that stores only the mode the session was booked in will misdescribe a large share of your consultations, and the misdescription will always be discovered at the worst moment.

Text and asynchronous consultation deserve a separate policy decision. A WhatsApp exchange between a consultant and a known patient is a teleconsultation under these guidelines whether or not the hospital acknowledges it, and it carries the same record-keeping and prescribing obligations. Hospitals that pretend this is not happening simply have an unmanaged channel. Either bring it into a governed messaging route or say explicitly, in writing, that clinical advice on personal numbers is outside scope.

Identity verification in telemedicine runs in two directions and hospitals usually build only one. The patient has to be established as the person the record belongs to, and the practitioner has to be identifiable to the patient by name and registration number. The second half is the one that gets skipped, and it is the cheaper of the two to fix: a persistent on-screen band carrying the doctor name, speciality, and registration number, plus the same details on every prescription and message, closes it permanently.

For the patient side, the guidelines allow reasonable means rather than prescribing one method. In an Indian hospital the workable ladder is: an existing UHID matched at booking, a one-time password to the registered mobile at session start, and a visual check against the record for video consults. Where an ABHA Number is already on the profile, it strengthens matching across facilities but it is not by itself a session-level authentication. Do not conflate the two.

Caregiver consultations need their own path. When the patient is a minor, is incapacitated, or is simply not the person holding the phone, the record must name the caregiver, state the relationship, and note the authorisation basis. This is where most identity trails collapse, because the consultation was booked under the patient and conducted with someone else entirely, and nothing in the note says so.

A teleconsultation record is held to the same standard as an in-person one, and the guidelines are explicit that the practitioner should maintain the notes, images, and prescriptions exchanged. In practice this means the consultation cannot live inside the video platform. Whatever tool carries the call, the clinical content has to land in the medical record with the same encounter identifier, the same clinician attribution, and the same retention rules as an OPD visit. If the video vendor is your system of record, you have outsourced your medico-legal position to a contract.

Fees are permitted and should be charged and receipted exactly as a physical consultation would be, which in India also means the receipt behaves consistently for GST purposes with the rest of your outpatient billing. The commercially awkward case is the consultation that fails on connectivity after two minutes. Decide your refund and reschedule rule in advance, publish it, and build it into the billing screen, because deciding it case by case at the counter produces both revenue leakage and complaints.

Storage should be deliberate rather than accidental. Decide what is retained: the note always, the prescription always, the exchanged images usually, the call recording only if you have a specific reason and a consent basis for it. HealUDoc can hold the teleconsultation note, the mode, and the prescription against the same encounter as the patient's in-person visits, which is what makes a longitudinal record readable rather than split across two systems.

A compliance document that lives in a shared drive changes nothing. What changes behaviour is a short SOP written in the sequence the doctor actually experiences: joining, identifying, classifying, consulting, prescribing, closing, and documenting. Two pages, with the prescribing restrictions on the second page as a table the doctor can glance at. Anything longer will be signed and never read, and you will have a training record with no training behind it.

Train in the specialities, not in an auditorium. Physicians, dermatologists, psychiatrists, and paediatricians hit completely different edges of these guidelines, and a combined session teaches everyone the parts they will never use. Run four short department-level sessions using your own recent consultations, anonymised, and let the consultants argue about the borderline cases. The disagreements are the useful part, and they tell you which defaults to set in the software.

Then audit monthly against a small, fixed sample. Pull twenty teleconsultations at random and check five things: registration number present, mode recorded, first-or-follow-up classified, consent captured, prescription consistent with the permitted list for that classification. Report the counts to the medical advisory committee alongside your other quality indicators. A programme that is measured survives a change of medical superintendent. One that is only documented does not.